Convergent synthesis of homogeneous Glc1Man 9GlcNAc2-protein and derivatives as ligands of molecular chaperones in protein quality control

Article abstract of DOI:10.1021/ja204831z

A detailed understanding of the molecular mechanism of chaperone-assisted protein quality control is often hampered by the lack of well-defined homogeneous glycoprotein probes. We describe here a highly convergent chemoenzymatic synthesis of the monoglucosylated glycoforms of bovine ribonuclease (RNase) as specific ligands of lectin-like chaperones calnexin (CNX) and calreticulin (CRT) that are known to recognize the monoglucosylated high-mannose oligosaccharide component of glycoproteins in protein folding. The synthesis of a selectively modified glycoform Gal₁Glc ₁Man₉GlcNAc₂-RNase was accomplished by chemical synthesis of a large N-glycan oxazoline and its subsequent enzymatic ligation to GlcNAc-RNase under the catalysis of a glycosynthase. Selective removal of the terminal galactose by a β-galactosidase gave the Glc₁Man ₉GlcNAc₂-RNase glycoform in excellent yield. CD spectroscopic analysis and RNA-hydrolyzing assay indicated that the synthetic RNase glycoforms maintained essentially the same global conformations and were fully active as the natural bovine ribonuclease B. SPR binding studies revealed that the Glc₁Man₉GlcNAc₂-RNase had high affinity to lectin CRT, while the synthetic Man₉GlcNAc ₂-RNase glycoform and natural RNase B did not show CRT-binding activity. These results confirmed the essential role of the glucose moiety in the chaperone molecular recognition. Interestingly, the galactose-masked glycoform Gal₁Glc₁Man₉GlcNAc₂-RNase also showed significant affinity to lectin CRT, suggesting that a galactose β-1,4-linked to the key glucose moiety does not significantly block the lectin binding. These synthetic homogeneous glycoprotein probes should be valuable for a detailed mechanistic study on how molecular chaperones work in concert to distinguish between misfolded and folded glycoproteins in the protein quality control cycle.

Full text of DOI:10.1021/ja204831z

ARTICLE
pubs.acs.org/JACS
Convergent Synthesis of Homogeneous Glc1Man9GlcNAc2-Protein
and Derivatives as Ligands of Molecular Chaperones in Protein
Quality Control
Mohammed N. Amin, Wei Huang, Rahman M. Mizanur, and Lai-Xi Wang*
Institute of Human Virology, Department of Biochemistry & Molecular Biology, University of Maryland School of Medicine, Baltimore,
Maryland 21201, United States
S Supporting Information
b
ABSTRACT: A detailed understanding of the molecular mechanism of chaperone-
assisted protein quality control is often hampered by the lack of well-defined homo-
geneous glycoprotein probes. We describe here a highly convergent chemoenzymatic
synthesis of the monoglucosylated glycoforms of bovine ribonuclease (RNase) as specific
ligands of lectin-like chaperones calnexin (CNX) and calreticulin (CRT) that are known
to recognize the monoglucosylated high-mannose oligosaccharide component of glyco-
proteins in protein folding. The synthesis of a selectively modified glycoform Gal1Glc1-
Man9GlcNAc2-RNase was accomplished by chemical synthesis of a large N-glycan
oxazoline and its subsequent enzymatic ligation to GlcNAc-RNase under the catalysis
of a glycosynthase. Selective removal of the terminal galactose by a β-galactosidase gave
the Glc1Man9GlcNAc2-RNase glycoform in excellent yield. CD spectroscopic analysis and RNA-hydrolyzing assay indicated that
the synthetic RNase glycoforms maintained essentially the same global conformations and were fully active as the natural bovine
ribonuclease B. SPR binding studies revealed that the Glc1Man9GlcNAc2-RNase had high affinity to lectin CRT, while the synthetic
Man9GlcNAc2-RNase glycoform and natural RNase B did not show CRT-binding activity. These results confirmed the essential role
of the glucose moiety in the chaperone molecular recognition. Interestingly, the galactose-masked glycoform Gal1Glc1Man9Glc-
NAc2-RNase also showed significant affinity to lectin CRT, suggesting that a galactose β-1,4-linked to the key glucose moiety does
not significantly block the lectin binding. These synthetic homogeneous glycoprotein probes should be valuable for a detailed
mechanistic study on how molecular chaperones work in concert to distinguish between misfolded and folded glycoproteins in the
protein quality control cycle.
’ INTRODUCTION
folding. Once the last Glc residue is removed by R-glucosidase II
(G-II), the resulting Man9GlcNAc2-protein will be released from
the CNX/CRT folding cycle. At this point, an ER-residing UDP-
glucose:glycoprotein glucosyltransferase (UGGT) serves as a
“folding sensor” to detect and reglucosylate the misfolded or
partially folded protein to generate the monoglucosylated glyco-
form, which enters into the CNX/CRT cycle again for refolding
(Figure 1).1,2 Despite tremendous efforts in this field, the
molecular mechanism of the protein quality control process that
involves a number of specific proteinꢀcarbohydrate interactions
is still not fully understood. A major obstacle is the difficulty
to obtain homogeneous, structurally well-defined glucosyla-
ted N-glycoproteins for detailed mechanistic and functional
studies.3ꢀ5 As an elegant chemical approach, Ito and co-workers
have synthesized an array of monoglucosylated high-mannose
type N-glycans and derivatives, and used them as probes to study
their interactions with lectin-like chaperones.6,7 Nevertheless,
preparation of related homogeneous natural glycoproteins,
which are highly desirable as probes for understanding how a
Asparagine (N)-linked glycosylation is an important co- and
post-translational modification of proteins in eukaryotic cells.
Recent studies have suggested that the attached high-mannose
type N-glycans with subtle structural difference play a pivotal role
in various steps of the protein quality control process, including
the calnexin (CNX)/calreticulin (CRT)-mediated protein fold-
ing and the endoplasmic reticulum (ER)-associated protein
degradation pathways.1 N-Glycosylation occurs when a tetra-
decasaccharide precursor, Glc3Man9GlcNAc2, is transferred by
an oligosaccharyltransferase (OST) from a sugar-lipid precursor
to an asparagine side chain in the consensus Asn-X-Ser/Thr
sequon of a nascent polypeptide (where X can be any amino acid
except proline). The precursor is then processed by ER R-
glucosidases I and II (G-I and G-II) to di- and monoglucosylated
glycoforms, respectively. The monoglucosylated form, Glc1-
Man9GlcNAc2 (Glc, glucose; Man, mannose; GlcNAc, N-acet-
ylglucosamine), is an essential ligand recognized by molecular
chaperones calnexin (CNX) and calreticulin (CRT). Upon
binding to the monoglucosylated glycoform, CNX and CRT
recruit other molecular chaperones such as protein disulfide
isomerase-like protein ERp57 to form a complex for appropriate
Received: May 26, 2011
Published: August 06, 2011
r
2011 American Chemical Society
14404
dx.doi.org/10.1021/ja204831z J. Am. Chem. Soc. 2011, 133, 14404–14417
|

Journal of the American Chemical Society
ARTICLE
Mucor hiemalis endo-β-N-acetylglucosaminidases (Endo-A and
Endo-M), respectively, were shown to be particularly useful for
synthesizing glycoproteins carrying large natural high-mannose
and complex type N-glycans.11 A key hypothesis in the present
synthetic design is that a large synthetic N-glycan oxazoline (4) with
a lactose moiety attached at the outer mannose residue of the
natural high-mannose core could also serve as a donor substrate
of the glycosynthase mutant EndoA-N171A or EndoM-N175A
for transglycosylation. Thus, the synthesis of glycoprotein tar-
gets 1 and 2 could be reduced to the preparation of GlcNAc-
RNase (3) and the key oligosaccharide oxazoline (4) (Figure 2).
Retrosynthetic disconnection of 4 led to three major building
blocks, the D2D3 arm pentasaccharide fragment (5), the lactose-
containing D1 arm fragment (6), and the core disaccharide
glycosyl acceptor (7) (Figure 2). The retrosynthetic design of
the large N-glycan oxazoline (4) was similar to the chemical
synthesis of the monoglucosylated N-glycan as reported
previously,6a which also used a convergent coupling of the D1
and D2D3 building blocks with a core β-mannoside moiety. But
different protecting groups and different glycosyl donors and
acceptors were used.
Chemical Synthesis of the Selectively Modified and
Monoglucosylated High-Mannose Type N-Glycan Oxazoline
(4). The synthesis of the target N-glycan oxazoline (4) required
the preparation of three key building blocks 5, 6, and 7. The
preparation of the D2D3 arm pentasaccharide fragment (5) was
summarized in Scheme 1. Briefly, selective de-O-acetylation of 8
with acetyl chloride in MeOH gave the 2-OH derivative (9),
which was glycosylated with the known glycosyl imidate 1015 to
provide disaccharide 11. Compound 11 was converted into the
corresponding glycosyl trichloroacetimidate 12,16 which was
then used to glycosylate diol 1317 under the catalysis of TMSOTf
to afford the pentasaccharide derivative (14) in 97% yield. De-O-
acylation followed by O-benzylation gave the per-O-benzylated
thioglycoside (15). Finally, the thioglycoside was converted into
the glycosyl fluoride (5) by treatment with NBS and then with
DAST (Scheme 1).
For the synthesis of lactose-containing pentasaccharide inter-
mediate (6), the known thioglycoside (16)18 was treated with
NBS, followed by DAST, to give a mixture of R- and β-lactosyl
fluoride (β/R, 12:1), from which the β-lactosyl fluoride (17) was
isolated in 88% yield. Glycosylation between 17 and acceptor
18,19 via activation with Cp2Zr(OTf)2 and DTBMP,20 afforded
single R-stereoisomer 19, which was isolated in 68% yield. A
relatively small coupling constant of the H-1 of Glc (J1,2 = 3.5
Hz) suggested that the newly formed glycosidic bond between
the Glc and Man moieties is in an R-configuration. Compound
19 was then converted into glycosyl fluoride 20, and the latter
was coupled with disaccharide 21 under the promotion of
Cp2HfCl2/AgOTf to provide pentasaccharide 22 in 69% yield.
Changing the benzoyl groups to benzyl groups gave the corre-
sponding thioglycoside (23), which was then converted to
glycosyl fluoride 6 by NBS-promoted hydrolysis and subsequent
treatment of the resulting hemiacetal with DAST (Scheme 2).
On the other hand, the core disaccharide (7) was prepared
following our previously reported procedures.10
Figure 1. The clnexin/calreticulin-assisted folding cycle and the struc-
tures of synthetic glycoprotein probes (1) and (2).
specific oligosaccharide and the misfolded protein domain act in
concert to provide the right signal for folding, is much more
challenging. Some of the glycoforms may be only transiently
present in the quality control process and are extremely difficult
to isolate from natural source. For example, in a classic study on
calnexin-mediated protein folding using bovine ribonuclease B
(RNase B) as the model glycoprotein, the monoglucosylated
RNase B was prepared through in vitro glucosylation of a mixture
of denatured RNase B glycoforms (ranging from Man5GlcNAc2-
RNase to Man9GlcNAc2-RNase) using radiolabeled UDP-glu-
cose and enzyme UGGT, which gave only about 2% mono-
glucosylated RNase B with uncharacterized ratio of the
glycoforms.3 The use of undefined glycoform mixtures as probes
complicated the interpretation of the experimental results.
We report here a convergent chemoenzymatic synthesis of
homogeneous glycoforms of bovine ribonuclease (RNase),
including the monoglucosylated glycoform Glc1Man9GlcNAc2-
RNase (1) and a selectively modified derivative Gal1Glc1-
Man9GlcNAc2-RNase (2) (Gal represents a galactose moiety).
Glycoprotein (2), in which the terminal Glc residue is selectively
masked by a β-1,4-linked Gal residue, is an unnatural glycoform
designed for probing how site-specific modification on the Glc
moiety perturbs its recognition by lectin-like molecular chaper-
ones. We also report here the binding of the synthetic glycoforms
with molecular chaperone CRT. Our experimental data showed
that while nonglucosylated glycoforms, including the natural
bovine RNase B (Man5ꢀ9GlcNAc2-RNase) and a synthetic high-
mannose glycoform Man9GlcNAc2-RNase, either in the folded
or denatured forms, were not recognized by CRT, the synthetic
monoglucosylated glycoform (1) could bind to CRT with high
affinity. Interestingly, the modified monoglucosylated glycoform
(2) also demonstrated significant binding to lectin CRT.
’ RESULTS AND DISCUSSION
Retrosynthetic Design of Glycoproteins 1 and 2. Our
synthesis was based on a chemoenzymatic approach that was
developed by us and others, which employs endoglycosidase-
catalyzed transglycosylation for regio- and stereoselective liga-
tion of a synthetic N-glycan (in the form of the activated sugar
oxazoline) to a GlcNAc-protein.8ꢀ13 A major advance in this
chemoenzymatic method is the discovery of novel endoglyco-
sidase mutants (glycosynthases) that can use highly activa-
ted sugar oxazoline as donor substrate for transglycosylation,
but lack the activity to hydrolyze the ground state product
because of the site-directed mutation at the critical catalytic
residue.11,13,14 Two glycosynthase mutants, the EndoA-N171A and
the EndoM-N175A originated from Arthrobacter protophormiae and
With the key synthetic building blocks in hand, we next turned
our attention to the assembly of the dodecasaccharide oxazoline
(4). As outlined in Scheme 3, glycosylation of glycosyl fluoride 6
with disaccharide 7 under the promotion of Cp2HfCl2/AgOTf
gave heptasaccharide 24 as a single isomer in 62% yield. It should
be noted that an initial attempt to use thioglycoside 23 directly as
14405
dx.doi.org/10.1021/ja204831z |J. Am. Chem. Soc. 2011, 133, 14404–14417

Journal of the American Chemical Society
ARTICLE
Figure 2. Retrosynthetic analysis of glycoproteins 1 and 2.
Scheme 1. Synthesis of the D2D3 Arm Pentasaccharide 5a
a Reagents and yields: (a) 5% AcCl in MeOH, 82%. (b) TMSOTf, 84%. (c) (1) NBS, H2O, quant.; (2) CCl3CN, DBU, 86%. (d) TMSOTf, 97%. (e) (1)
NaOMe, MeOH; (2) NaH, BnBr, 86%. (f) (1) NBS, H2O; (2) DAST, 70%.
glycosyl donor for coupling with disaccharide acceptor 7 via the
(BrC6H4)3NSbCl6-promoted radical glycosylation21 gave only a
very low yield of corresponding heptasaccharide 24, even when a
large amount of the promoter was applied (data not shown). This
was probably due to the low reactivity of thioglycoside 23. The
benzylidene acetal group in 24 was removed by mild acidic
14406
dx.doi.org/10.1021/ja204831z |J. Am. Chem. Soc. 2011, 133, 14404–14417

Journal of the American Chemical Society
ARTICLE
Scheme 2. Synthesis of Pentasaccharide 6a
a Reagents and yields: (a) (1) NBS, H2O, quant.; (2) DAST, 88%. (b) Cp2Zr(OTf)2, DTBMP, 68%. (c) (1) NBS, H2O, quant.; (2) DAST, 91%.
(d) Cp2HfCl2, AgOTf, 69%. (e) (1) NaOMe, MeOH; (2) NaH, BnBr, 88% in two steps. (f) (1) NBS, H2O; (2) DAST, 94% in two steps.
Scheme 3. Synthesis of Dodecasaccharide Oxazoline 4a
a Reagents and yields: (a) Cp2HfCl2, AgOTf, 62%. (b) AcOH-EtOH-H2O, 91%. (c)Cp2HfCl2, AgOTf, 61%. (d) AcSH, 2,6-lutidine, 86%.
(e) H2, Pd(OH)2/C, 92%. (f) 2-Chloro-1,3-dimethylimidazolinium chloride (DMC), TEA, quant.
hydrolysis to afford diol 25, which was then selectively glyco-
sylated at the primary hydroxyl group with pentasaccharide
fragment 5 to provide the dodecasaccharide skeleton (26) in
61% yield. The R-mannoside linkage for the newly generated
glycosidic bond in 26 was evident by the relatively large
coupling constants (over 170 Hz) between the C-1 and H-1
of all the R-mannosides (the coupling constants (1JCH) for the 8
R-mannosides and one R-glucoside appeared at 174.7, 174.0,
173.9, 173.8, 173.8, 173.3, 172.3, 171.8, and 170.3 Hz, re-
spectively). Conversion of the 2-azido group into the 2-acet-
amido group was achieved by treatment with AcSH in the
presence of 2,6-lutidine to give 27 in 86% yield. De-O-benzyla-
tion was performed via hydrogenolysis of 27 in the presence of
Pd(OH)2/C to afford the free dodecasaccharide 28. The
removal of the multiple O-Bn groups by Pd-catalyzed hydro-
genolysis went very well to give the free oligosaccharide (28) in
almost quantitative yield. Finally, oxazoline formation was
achieved in a single step by treatment of free oligosaccharide
28 with an excess amount of 2-chloro-1,3-dimethylimidazoli-
nium chloride (DMC)/TEA12,22 to furnish the dodecasacchar-
ide oxazoline (4), which was isolated by gel filtration in a
quantitative yield (Scheme 3).
14407
dx.doi.org/10.1021/ja204831z |J. Am. Chem. Soc. 2011, 133, 14404–14417

Journal of the American Chemical Society
ARTICLE
Scheme 4. Synthesis of Homogeneous Glycoproteins 1, 2, and 30
Synthesis of the Target Glycoproteins through Enzymatic
Transglycosylation. We have previously shown that a glyco-
synthase mutant, EndoA-N171A, was able to transfer sugar
oxazoline of natural high-mannose type N-glycans to a GlcNAc-
protein acceptor to form a homogeneous glycoprotein with the
natural β-1,4-linkage for the newly generated glycosidic bond.11
The use of the glycosynthase mutant is important for the
synthesis, as wild-type Endo-A leads to quick hydrolysis of the
high-mannose type glycoprotein product. To test whether the
lactosemodified high-mannose N-glycan oxazoline (4) serves as
a donor substrate for the enzyme, we performed the EndoA-
N171A catalyzed transglycosylation between oxazoline 4 and
GlcNAc-RNase 3 (donor/acceptor, 7:1) in a phosphate buffer
(pH 7.0) (Scheme 4). The GlcNAc-RNase was prepared by
which was recycled for the transglycosylation by conversion into the
activated sugar oxazoline (4), using a single-step transformation in
water with 2-chloro-1,3-dimethylimidazolinium chloride (DMC)/
TEA.12,22 The homogeneity of the product Gal1Glc1Man9GlcNAc2-
RNase (2) was evident from its SDSꢀPAGE and ESIꢀMS analysis
(Figure 3). Moreover, Dionex HPAEC-PAD monosaccharide com-
position analysis of 2 gave the expected molar ratios of monosacchar-
ides: Gal/Glc/Man/GlcN = 1:1:9:2 (Figure S2, Supporting Informa-
tion). To obtain the monoglucosylated glycoform (1), the synthetic
glycoprotein (2) was treated with a β-galactosidase from Streptococcus
pneumoniae in a phosphate buffer (pH 5.5). It was found that the
Streptococcus β-galactosidase could selectively and efficiently remove
the terminal galactose to give the monoglucosylated glycoform
Glc1Man9GlcNAc2-RNase (1) in essentially quantitative yield, with-
out any side reactions. The SDSꢀPAGE and ESIꢀMS analysis of
glycoproteins 1 and 2 were shown in Figure 3. The purified Gal1Glc1-
Man9GlcNAc2-RNase (2) (lane 3, Figure 3A) and Glc1Man9Glc-
NAc2-RNase (1) (lane 4, Figure 3A) appeared as a single band at ca.
16 kDa, which is about 2 kDa larger than the precursor GlcNAc-RB
(3) that had an apparent molecular weight of 14 kDa (lane 2,
Figure 3A). The data indicates the attachment of a single large
N-glycan to the precursor after transglycosylation. Deconvolu-
tion of the MS data of glycoprotein 2 gave a molecular mass of
15 872 Da, which matched well with the calculated molecular
mass (M = 15 870 Da). On the other hand, deconvolution of the
ESIꢀMS data of glycoprotein 1 gave a molecular mass of 15 710
Da, which was in good agreement with the calculated molecular
mass (M = 15 708 Da). For comparison, we also prepared the
nonglucosylated glycoform, Man9GlcNAc2-RNase (30), using
Man9GlcNAc-oxazoline 2911 as the donor substrate. The En-
doA-N171A catalyzed reaction gave a 78% yield of glycoprotein
30 under the same conditions as that for the synthesis of gly-
coprotein 2. The transglycosylation yield for the synthesis of
deglycosylation of natural bovine ribonuclease B (Man5ꢀ9
-
GlcNAc2-RNase) with Endo-H, following our previously re-
ported procedure.9 We found that the selectively modified
high-mannose type N-glycan oxazoline (4) acted as a good
substrate for EndoA-N171A, and the enzymatic reaction led to
a gradual formation of glycoprotein product 2 that appeared
slightly earlier than the acceptor GlcNAc-RNase (3) under a
reverse-phase HPLC condition (for an HPLC monitoring profile
of the enzymatic transglycosylation, see Figure S1, Supporting
Information). After 5 h, the reaction was stopped and the product
(2) was readily isolated by HPLC in 80% yield (calculated on the
basis of the used GlcNAc-RNase). The transformation of
GlcNAc-RNase (3) to the product could be driven to completion
when additional sugar oxazoline was added to the reaction
mixture. The enzymatic reaction was readily performed on a
milligram scale of the GlcNAc-protein acceptor, but could be
scaled up when needed. Moreover, upon the completion of the
enzymatic reaction, we were able to efficiently recover the excess sugar
oxazoline by HPLC in the form of free reducing oligosaccharide,
14408
dx.doi.org/10.1021/ja204831z |J. Am. Chem. Soc. 2011, 133, 14404–14417

Journal of the American Chemical Society
ARTICLE
Figure 4. The CD spectra of different glycoforms of ribonuclease. GN-
RB, the GlcNAc-RB; LacM9GN2-RB, Gal1Glc1Man9GlcNAc2-RB;
M9GN2-RB, Man9GlcNAc2-RB; GlcM9GN2-RB, Glc1Man9GlcNAc2-
RB; M5ꢀ9GN2-RB, Man5ꢀ9GlcNAc2-RB (the natural bovine RNase
B). Except the extensively trimmed glycoform GlcNAc-RB, all other
synthetic glycoforms that bear a large N-glycan as well as the natural
RNase B showed very similar CD spectroscopic profiles.
Figure 3. Analysis of the synthetic glycoproteins. (A) SDSꢀPAGE:
lane 1, natural RNase B; lane 2, GlcNAc-RNase; lane 3, glycoprotein 2;
lane 4, glycoprotein 1. (B) ESIꢀMS of glycoprotein 2. (C) ESIꢀMS of
glycoprotein 1.
Figure 5. RNA-hydrolyzing activity of different glycoforms of ribonu-
clease B (RB). A solution of yeast RNA (0.5 mg/mL) was incubated with
each of the RNase glycoforms (4 μg/mL) in a buffer (pH 5.0), and the
UV absorbance (300 nm) of the solution was recorded. This figure
shows a representative profile of the RNA hydrolysis by the individual
glycoforms of RNase B (RB). RB-DA stands for the heat-deactivated
natural RNase B.
glycoproteins 2 and 30 was comparable, suggesting that the
glycosynthase mutant EndoA-N171A was equally efficient for
transferring the sugar oxazolines corresponding to both the
natural high-mannose type N-glycan and its derivatives with
selective modifications on the outer residues of the high-man-
nose core.
removed to leave only the innermost GlcNAc residue at the
N-glycosylation site, seemed to have an enhanced R-helical
content. The estimated percentages of the secondary structures
for the GlcNAc-RNase (3) were 31 ( 9.9% R-helix, 27 ( 4.5% β-
strands, 12 ( 2.0% β-turns, and 30 ( 5.6% unordered structures.
The CD spectroscopy of GlcNAc-RNase (3) was very similar to
that of the glycan-truncated RNase B carrying heterogeneous
Man2ꢀ4GlcNAc2 glycans as reported previously.24 These results
suggest that the chemoenzymatic manipulations including the
enzymatic deglycosylation of RNase B and subsequent enzymatic
transglycosylation with sugar oxazoline did not denature the
glycoproteins. In addition, the experimental data also indicate that
as long as the high-mannose core was retained, modification at the
outer residues of the N-glycan had less impact on the protein’s
conformations.
Circular Dichroism and RNA-Hydrolyzing Activity of the
Synthetic RNase Glycoforms. To evaluate the structural integ-
rity of the synthetic glycoproteins, we measured their circular
dichroism (CD) spectra and compared them to the natural
RNase B (31) that carries heterogeneous high-mannose type
N-glycans (Man5ꢀ9GlcNAc2). The CD spectroscopic analysis
was performed in the far UV wavelength region (190ꢀ260 nm)
that allows the monitoring of the changes in secondary structures
of the glycoproteins. As shown in Figure 4, all the synthetic
glycoforms with a large synthetic high-mannose type N-glycan,
including glycoforms 1, 2, and 30, had almost identical spectro-
scopic profiles similar to that of the natural RNase B. An analysis
using the web-based DICHROWEB server23 gave essentially the
same estimate (within the error) on the percentages of secondary
structures for glycoforms 1, 2, and 30, as well as the natural
RNase B (31): 21 ( 8.8% R-helix, 33 ( 7.5% β-strands, 13 (
1.8% β-turns, and 33 ( 0.9% unordered structures. In comparison,
the truncated form GlcNAc-RNase (3), in which the N-glycan was
To probe the functions of the synthetic glycoforms, we
measured the enzymatic activity of the respective synthetic
glycoforms as well as the natural RNase B (Sigma). RNase B is
14409
dx.doi.org/10.1021/ja204831z |J. Am. Chem. Soc. 2011, 133, 14404–14417

Journal of the American Chemical Society
ARTICLE
Figure 6. SPR sensorgrams of the binding between immobilized lectin CRT and different RNase glycoforms. The binding analysis was performed at
10 °C, (A) with Glc1Man9GlcNAc2-RNase (1), (B) with nonglucosylated Man9GlcNAc2-RNase (30) and natural Man5ꢀ9GlcNAc2-RNase (31), (C)
with Gal1Glc1Man9GlcNAc2-RNase (2), (D) with denatured RNase glycoproteins.
known to catalyze the hydrolysis of the 30,50-phosphodiester
bonds of RNA. It was previously reported that degradation of
RNA by ribonuclease was accompanied by a shift in the UV
absorbance of the RNA substrate toward the shorter wave-
lengths.25 Thus, measurement of the decrease of absorbance at
300 nm constitutes a simple and very straightforward method for
assessing the RNA-hydrolyzing activity of RNase.25 Using a
commercially available yeast RNA extract as the substrate for
testing hydrolytic activity, we measured the enzymatic hydrolysis
of the yeast RNA by the individual RNase glycoforms. A
representative enzymatic hydrolysis profile was shown in
Figure 5. We found that the synthetic glycoforms (1, 2, and
30) that carry a large N-glycan had similar enzymatic activity as
that of the natural RNase B, while the glycoform (GlcNAc-RB)
that carries only a monosaccharide moiety at the glycosylation
site showed a slightly higher enzymatic activity than the others.
Dwek and co-workers previously performed a comparative study
on the hydrolytic activity of RNase A (nonglycosylated form)
with several truncated glycoforms of RNase B, including the
Man5GlcNAc2, Man1GlcNAc2, and GlcNAc2-glycoforms.26
Their studies indicated that the attachment of an N-glycan at
Asn-34 in the bovine ribonuclease resulted in 2- to 4-fold
decrease in the enzymatic activity when compared with the
nonglycosylated RNase A. The difference in enzymatic activity
between the nonglycosylated and glycosylated RNases (RNase A
and B, respectively) was mainly attributed to steric hindrance
exerted by the attachment of a large N-glycan. They also found
that natural RNase B that contains a mixture of Man5ꢀ9GlcNAc2
glycoforms had the same activity as that of Man5GlcNAc2-RNase
B, suggesting that attachment of additional residues beyond the
high-mannose core did not further affect the enzymatic activity.
Our experimental data with the synthetic glycoforms are consistent
with the previously reported observations. Together with the CD
analysis, our RNA-hydrolyzing activity data provides strong evi-
dence indicating that the RNase glycoforms synthesized through
our chemoenzymatic approach are properly folded and are fully
functional with almost identical enzymatic activity as the natural
RNase B.
Binding Studies with Lectin Calreticulin (CRT). With the
successful synthesis of the glycan-defined RNase glycoforms, we
performed binding studies with the molecular chaperone calre-
ticulin (CRT) using surface plasmon resonance (SPR) technol-
ogy. The lectin CRT was immobilized on the chips and the
different glycoforms were probed as analytes at 10 °C. The
results are shown in Figure 6. As expected, the synthetic
Glc1Man9GlcNAc2-RNase (1) was efficiently recognized by
the immobilized CRT (Figure 6A), with a deduced KD of 38 nM
at 10 °C (kon = 1.28 ꢁ 104 Mꢀ1 sꢀ1; koff = 4.82 ꢁ 10ꢀ4 sꢀ1). In
contrast, the nonglucosylated Man9GlcNAc2-RNase (30), as
well as the native RNase B (Man5ꢀ9GlcNAc2-RNase) (31),
did not show any binding to CRT within the tested concentra-
tion range (30ꢀ2000 nM) (Figure 6B). These results confirm
the essential role of the R-1,3-linked glucose moiety in the
Glc1Man9GlcNAc2 glycan for CRT recognition.
The SPR analysis of the interaction between CRT and the
selectively modified glycoform (2) yielded an interesting result
(Figure 6C). We found that CRT showed significant affinity to
the Gal1Glc1Man9GlcNAc2-RNase (2), in which the terminal
glucose was masked by a β-1,4-linked galactose. Fitting the
binding data to a 1:1 Langmuir binding model gave a KD of
110 nMfor the interaction (kon =1.81ꢁ 104 Mꢀ1 sꢀ1;koff =1.99ꢁ
10ꢀ3 ꢀ1). A comparison of the KD values indicates that
s
Glc1Man9GlcNAc2-RNase (1) binds to CRT about 2-fold more
efficiently than Gal1Glc1Man9GlcNAc2-RNase (2) binds to
14410
dx.doi.org/10.1021/ja204831z |J. Am. Chem. Soc. 2011, 133, 14404–14417

Journal of the American Chemical Society
ARTICLE
CRT. The kinetic data suggests that this was attributed mainly to
the relatively fast dissociation rate (koff) of the modified mono-
glucosylated derivative. Nevertheless, the high affinity of glyco-
form 2 to lectin CRT (with a KD of 110 nM) indicated that
selective attachment of a galactose moiety at the 4-OH of the
terminal Glc residue did not significantly impair the interaction
between the monoglucosylated glycan and lectin CRT.
unfolded polypeptide was substrate dependent, as the affinity of
CRT to unfolded proteins was relatively weak compared with its
binding to monoglucosylated glycoprotein and the affinity varied
with different proteins. Thus, an explanation for the lack of CRT
binding to unfolded nonglucosylated RNase B was that unfolded
RNase B might lack the polypeptide elements recognized by
CRT.29 It should be also pointed out that while lectin CRT (and
CNX) may or may not have direct interactions with the
misfolded or unfolded protein domain of a given glycoprotein,
the lectins, upon binding with the monoglucosylated glycoform
of the misfolded protein, are able to recruit other molecular
chaperones such as the disulfide isomerase-like protein ERp57
that directly recognizes and binds to misfolded protein domain to
promote appropriate folding of the glycoprotein. With the
availability of the homogeneous, monoglucosylated glycoforms
of RNase B and other glycoproteins, now we have the molecular
probes to address how CRT-mediated recognition acts in con-
cert with chaperone ERp57 for promoting protein folding.
Some kinetic studies on the binding of CRT with monogluco-
sylated oligosaccharide substrates were previously reported.4,5,7,30
The binding studies with synthetic Glc1Man1ꢀ4 oligosaccharides
indicated that the whole D1 arm of Glc1Man9GlcNAc2 was
involved in CRT recognition.5,7,30 For example, an isothermal
titration calorimetry (ITC) measurement indicated that the syn-
thetic tetrasaccharide GlcR1ꢀ3ManR1ꢀ2ManR1ꢀ2ManR1-Me
was at least 50-fold more potent than the disaccharide GlcR1ꢀ
3ManR1-Me in binding to CRT.5 A SPR binding study using
monoglucosylated chicken IgG fractions as the model glycopro-
teins for CRT binding at 10 °C gave a k1 and kꢀ1 of 1.5 ꢁ 104 Mꢀ1
The high-affinity binding of the monoglucosylated high-man-
nose type glycoproteins (1 and 2), together with the observation
that the corresponding nonglucosylated Man9GlcNAc2-RNase
(30) and natural Man5ꢀ9GlcNAc2-RNase (31) did not bind to
CRT, clearly indicated the requirement of the specific R-1,3-
glucose moiety for CRT recognition. To further confirm that the
binding of Glc1Man9GlcNAc2-RNase (1) and Gal1Glc1Man9-
GlcNAc2-RNase (2) to CRT was not caused by potential
changes of conformations in the protein domains, we denatured
the natural RNase B (31) and the synthetic RNase glycoproteins
(1, 2, and 30). Partial and complete denaturations were achieved
by dissolving the glycoproteins in a Tris buffer (pH 8.6) contain-
ing 8 M urea, in the absence or presence of 0.15 M dithiothreitol
(DTT) (reducing agent), respectively, following the previously
reported methods.27 Reverse-phase HPLC (RP-HPLC) analysis
indicated that while the folded glycoprotein appeared as a single
peak, partial denaturation of RNase B led to the formation of a
new species that appeared as a broad peak with a longer retention
time as compared with the folded glycoprotein (Figure S3,
Supporting Information), suggesting the exposure of some
hydrophobic residues in the denatured species. On the other
hand, complete denaturation with the cleavage of the disulfide
bonds led to a major peak that appeared with a longer retention
time than the native species, together with a series of minor
species (Figure S3). The synthetic glycoproteins demonstrated
similar RP-HPLC patterns as the RNase B before and after
denaturation (data not shown). We performed the CRT binding
studies with the completely denatured glycoproteins (Figure 6D).
It was found that the denatured monoglucosylated glycoproteins
(1 and 2) showed almost the same CRT-binding patterns as the
corresponding folded glycoproteins (1 and 2), respectively. On
the other hand, the nonglucosylated glycoproteins, Man9-
GlcNAc2-RNase (30) and Man5ꢀ9GlcNAc2-RNase (31), still
did not bind to CRT after denaturation. The partially denatured
glycoproteins behaved the same as the corresponding completely
denatured glycoprotein in CRT-binding (data not shown).
Taken together, our binding studies with well-defined homo-
geneous RNase glycoforms that are in either folded or denatured
states clearly suggest that CRT recognizes specifically the
monoglucosylated high-mannose type N-glycan in the RNase,
independent of the protein domain and its folding states. These
results are consistent with previous observations when mixtures
of monoglycosylated and nonglycosylated RNase B glycoforms
were used for CRT recognition studies.3,28 It should be pointed
out that there were different views on whether CRT and calnexin
(CNX) have direct interactions with the protein domain in
protein folding process. Some studies suggested that the mono-
glucosylated oligosaccharides are necessary and sufficient for
association of a glycoprotein with CRT and CNX, regardless of
its folding status;3,28 other studies indicated that in addition to its
lectin property to recognize monoglucosylated oligosaccharides,
CRT could also bind to unfolded nonglycosylated protein
domain to form stable complexes that prevent protein aggrega-
tion, although folded protein domain did not bind to CRT.29 In
the latter case, it was also observed that the binding of CRT to
sꢀ1 and 2 ꢁ 10ꢀ2
s
ꢀ1, respectively.4 The CRT binding kinetic
parameters for Glc1Man9GlcNAc2 were very similar to the mono-
glucosylated chicken IgG fractions under the same binding
conditions.4 In comparison, our Glc1Man9GlcNAc2-RNase/CRT
binding study gave comparable association rates (kon = 1.28 ꢁ 104
M
ꢀ1 sꢀ1). However, our dissociation rate with Glc1Man9GlcNAc2-
RNase (koff = 4.82 ꢁ 10ꢀ4
s
ꢀ1) seemed much lower than that
reported for the monoglucosylated IgG fractions. It is not clear
whether this was due to the difference of the protein context (IgG
vs RNase) that may affect the conformations of the N-glycans
differentially or due to the heterogeneous glycosylation nature of
the IgG molecules used. A more detailed binding study using well-
characterized homogeneous monoglucosylated IgG and other
glycoproteins would be required to clarify this point.
The relatively high affinity of Gal1Glc1Man9GlcNAc2-RNase
(2) to lectin CRT (with a KD of 110 nM) suggests that the 4-OH
of the terminal Glc residue could tolerate modification without
significantly impairing its interaction with lectin CRT. Recently,
the crystal structure of CRT complexed with Glc1Man3 tetra-
saccharide was solved at a 1.95 Å resolution.31 The structure of
the complex clearly shows that the Glc1Man3 tetrasaccharide
binds to a long channel on CRT involving an extensive network
of hydrogen bonds and hydrophobic interactions. The 2-OH of
Glc was involved in hydrogen bonding with Gly124 and Lys111
of CRT; the 3-OH of Glc was engaged with the Tyr128 of CRT,
while the 4-OH of Glc was protruding toward outside the cavity
without any hydrogen bonds involved (Figure S4, Support-
ing Information). This structural arrangement suggests that the
4-OH may tolerate some modifications and has a space to accom-
modate a substituent. On the other hand, the 3-OH is involved in
a hydrogen-bond network with residue Tyr128 of CRT. The
X-ray crystal structure of the complexes also explains why CRT
does not bind to the natural diglucosylated (Glc2Man9GlcNAc2)
14411
dx.doi.org/10.1021/ja204831z |J. Am. Chem. Soc. 2011, 133, 14404–14417

Journal of the American Chemical Society
ARTICLE
intermediate glycoform, in which the second Glc residue was R-
1,3-linked to the innermost Glc residue.1,2 Our findings suggest
that the 4-OH position of the monoglucosylated ligands would
be suitable for selective modification with appropriate tags to
provide novel probes for structural or functional studies.
250 mm) at 40 °C. The column was eluted with a linear gradient of
24ꢀ33% aq MeCN containing 0.1% TFA for 30 min at a flow rate of
0.5 mL/min. Preparative HPLC was performed with a Waters 600
HPLC instrument on a Waters C18 column (5.0 μm, 10 ꢁ 250 mm).
The column was eluted with a gradient of 23ꢀ29% aq MeCN containing
0.1%TFA for 30 min at a flow rate of 4 mL/min.
Synthesis of 2,3,4,6-Tetra-O-benzyl-β-D-galactopyranosyl-
(1f4)-3,4,6-tri-O-benzyl-β-D-glucopyranosyl Fluoride (17).
To an ice-cooled solution of phenyl 2,3,4,6-tetra-O-benzyl-β-D-
galactopyranosyl)-(1f4)-2,3,6-tri-O-benzyl-1-thio-β-D-glucopyrano-
side 16 (913 mg, 0.857 mmol) in acetoneꢀwater (9:1, 10 mL) was
added NBS (458 mg, 2.57 mmol). The reaction mixture was stirred for
1 h. Monitoring of the reaction with TLC (hexaneꢀethyl acetate, 2/1)
indicated complete hydrolysis of thioglycoside. The reaction was
quenched with NaHCO3 and the mixture was extracted with ethyl
acetate. The extract was dried and concentrated to dryness. The dried
crude product was dissolved in anhydrous CH2Cl2 (18 mL) and the
solution was cooled to ꢀ20 °C. To the solution was added DAST
(225 μL, 1.71 mmol). The reaction mixture was stirred for 1 h and then
quenched with saturated NaHCO3 (aq). The mixture was washed with
brine, dried over Na2SO4, and filtered. The filtrate was concentrated and
the residue was subjected to silica gel chromatography (hexaneꢀethyl
acetate, 2/1) to give the β-lactosyl fluoride (17) (736 mg, 88%) and its
R-isomer (68 mg, 8%). Characterization of the β-lactosyl fluoride (17):
HRMS MALDI-TOF: [M + Na]+ calcd for C61H63O10FNa, 997.4303,
found 997.4303. 1H NMR (CDCl3, 500 MHz): δ 7.34ꢀ7.12 (m, 35H),
5.22 (dd, J = 53.5, 6.5 Hz, H-1, 1H), 4.99 (d, J = 11.0 Hz, PhCH2, 1H),
4.96 (d, J = 12.0 Hz, PhCH2, 1H), 4.80 (d, J = 11.5 Hz, PhCH2, 1H), 4.78
(d, J = 13.0 Hz, PhCH2, 1H), 4.73ꢀ4.68 (m, 5H), 4.55 (d, J = 11.5
Hz, PhCH2, 1H), 4.54 (d, J = 12.0 Hz, PhCH2, 1H), 4.40 (d, J = 7.5 Hz,
H-1gal, 1H), 4.39 (d, J = 12.5 Hz, PhCH2, 1H), 4.34 (d, J = 12.0 Hz,
PhCH2, 1H), 4.25 (d, J = 11.5 Hz, PhCH2, 1H), 4.04 (br t, J = 9.0 Hz,
1H), 3.90 (d, J = 2.4 Hz, H-4gal, 1H), 3.81 (dd, J = 10.5, 4.0 Hz, 1H), 3.74
(dd, J = 10.0, 7.5 Hz, 1H), 3.64 (dd, J = 9.5, 1.0 Hz, 1H), 3.59ꢀ3.50 (m,
3H), 3.47 (dd, J = 9.0, 2.0 Hz, 1H), 3.39ꢀ3.32 (m, 3H); 13C NMR
(CDCl3, 100 MHz): δ 139.13, 138.92, 138.86, 138.61, 138.19, 138.16,
138.07, 128.51, 128.45, 128.38, 128.30, 128.18, 128.01, 127.96, 127.88,
127.71, 127.57, 109.78, 102.78, 82.62, 81.82, 81.72, 81.07, 80.85, 80.03,
75.91, 75.46, 75.22, 75.15, 74.85, 74.69, 73.70, 73.56, 73.37, 73.19, 72.71,
68.24, 67.93. Characterization of the R-lactosyl fluoride: HRMS MAL-
DI-TOF: [M + Na]+ calcd for C61H63O10FNa, 997.4303, found
997.4364. 1H NMR (CDCl3, 500 MHz): δ 7.36ꢀ7.12 (m, 35H), 5.48
(dd, J = 53.5, 2.4 Hz, H-1, 1H), 5.07 (d, J = 11.0 Hz, PhCH2, 1H), 4.97
(d, J = 11.0 Hz, PhCH2, 1H), 4.83 (d, J = 12.0 Hz, PhCH2, 1H), 4.80 (d,
J = 11.0 Hz, PhCH2, 1H), 4.74ꢀ4.70 (m, 5H), 4.66 (d, J = 12.0
Hz, PhCH2, 1H), 4.56 (d, J = 11.0 Hz, PhCH2, 1H), 4.51(d, J = 12.0 Hz,
PhCH2, 1H),, 4.38 (d, J = 12.0 Hz, PhCH2, 1H), 4.33 (d, J = 7.0 Hz,
H-1gal, 1H), 4.31 (d, J = 12.0 Hz, PhCH2, 1H), 4.25 (d, J = 12.0 Hz,
PhCH2, 1H), 4.01 (br t, J = 9.0 Hz, 1H), 3.90ꢀ3.88 (m, 2H), 3.86ꢀ3.80
(m, 2H), 3.75 (dd, J = 10.0, 8.0 Hz, 1H), 3.57ꢀ3.48 (m, 2H), 3.45 (dd,
J = 9.0, 2.0 Hz, 1H), 3.39ꢀ3.32 (m, 3H); 13C NMR (CDCl3, 125 MHz):
δ 139.03, 138.72, 138.69, 138.52, 138.49, 138.33, 138.00, 137.85, 128.37,
128.22, 127.98, 127.89, 127.79, 127.72, 127.66, 127.55, 127.52, 127.40,
127.17, 105.83, 102.77, 82.46, 79.88, 79.67, 78.24, 75.59, 75.50, 75.33,
74.73, 73.88, 73.71, 734.43, 73.16, 72.60, 68.21, 67.27.
’ CONCLUSION
An efficient chemoenzymatic synthesis of homogeneous gly-
coprotein probes, including the monoglucosylated glycoform
Glc1Man9GlcNAc2-RNase and its selectively modified deriva-
tive, was achieved. The relaxed substrate specificity of the
endoglycosidase mutant and the highly convergent nature of
the chemoenzymatic method point to a potentially general
approach to the assembly of both natural and tailor-made
glycoprotein probes that are hitherto difficult to obtain for
CRT/CNX-mediated protein folding studies. The synthetic
homogeneous RNase glycoforms showed very similar global
structures and RNA-hydrolyzing activities. The binding studies
with lectin CRT indicated the essential role of the monoglucose
residue in the glycoprotein for CRT recognition, and the
unexpected high affinity of the selectively modified monogluco-
sylated glycoform implicated interesting structural features in the
molecular recognition of CRT. The availability of the well-
defined glycoprotein probes should enable a more detailed
mechanistic study on how a fine N-glycan structure functions
in concert with a misfolded or partially folded protein domain to
provide the right signal in the protein quality control process.
Among others, an immediate application of the synthetic mono-
glucosylated glycoform is to probe the mechanism and kinetics of
the CRT/CNX-mediated interactions between a misfolded
glycoprotein and other molecular chaperones such as the disul-
fide isomerase-like protein ERp57. The synthetic monoglucosy-
lated RNase glycoform (Glc1Man9GlcNAc2-RNase) can be
intentionally denatured and misfolded. Then, the complex of
misfolded Glc1Man9GlcNAc2-RNase and lectin CRT (or CNX)
can be used to decipher how differentially misfolded status of the
complex recruits and interacts with the disulfide isomerase-like
protein ERp57 for proper protein folding.
’ EXPERIMENTAL SECTION
Materials and Methods. The N171A mutant of endo-β-N-acet-
ylglucosaminidase from A. protophormiae (Endo-A) was overproduced
in Escherichia coli following the reported procedure.11 Calreticulin
(CRT) was purchased from United States Biological (Swampscott,
MA). Bovine pancreatic ribonuclease B was purchased from Sigma-
Aldrich (St. Louis, MO). Thin-layer chromatography (TLC) was
performed on silica gel coated aluminum plates (Sigma-Aldrich). Flash
column chromatography was performed on silica gel 60 (230ꢀ400
mesh). SDSꢀPAGE was performed on a gradient (4ꢀ10%) polyacry-
lamide gel. NMR spectra were recorded on a VARIAN 500 MHz
spectrometer and JEOL ECX 400 MHz spectrometer. The chemical
shifts were assigned in parts per million (ppm). TMS and acetone were
used as internal standards for solvents CDCl3 and D2O, respectively.
ESIꢀMS spectra were measured on a Micromass ZQ-4000 single-
quadrupole mass spectrometer and LXQ linear ion trap mass spectro-
meter (Thermo Scientific). High-resolution (HR) MALDI-TOF spectra
were measured on a MALDI-TOF/TOF 4800 spectrometer (Applied
Biosystems) with 2,5-dihydroxybenzoic acid as matrix and Cal 4700
standard peptide mixture was used (Applied Biosystems) as internal
standard. Analytical RP-HPLC for glycoprotein analysis was performed
on a Waters 626 HPLC instrument with a C18 column (3.5 μm, 4.6 ꢁ
Synthesis of Phenyl 2,3,4,6-Tetra-O-benzyl-β-D-galacto-
pyranosyl-(1f4)-2,3,6-tri-O-benzyl-R-glucopyranosyl-(1f3)-
2,4,6-tri-O-benzyl-1-thio-R-D-mannopyranoside (19). A mixture
of monosaccharide 18 (38 mg, 0.070 mmol), glycosyl fluoride 17 (82 mg,
0.084 mmol), and MS 4A (0.50 g) in tolueneꢀtetrahydrofuran (9/1,
18 mL) was stirred for 30 min and cooled to ꢀ20 °C. Then, DTBMP
(22 mg, 0.105 mmol) and Cp2Zr(OTf)2 (46 mg, 0.77 mmol) were added.
The resulting mixture was stirred for 5 h under argon at ꢀ20 °C to room
temperature and then filtered through a Celite pad. The filtrate was diluted
14412
dx.doi.org/10.1021/ja204831z |J. Am. Chem. Soc. 2011, 133, 14404–14417

Journal of the American Chemical Society
ARTICLE
with EtOAc and washed with saturated NaHCO3 (aq), water, and brine,
dried over Na2SO4, and filtered. The filtrate was concentrated to dryness.
The residue was subjected to flash chromatography on silica gel
(hexaneꢀethyl acetate, 4/1) to give trisaccharide 19 (72 mg, 68%).
ESIꢀMS: calcd for C94H96O15S, M = 1497.82; Found (m/z), 1520.43
[M + Na]+. 1H NMR (CDCl3, 500 MHz): δ 7.46ꢀ7.43 (m, 2H),
7.33ꢀ7.09 (m, 53H), 5.54 (s, H-1, Man, 1H), 5.12 (d, J = 4.0
Hz, H-1Glc,1H), 5.04 (d, J = 11.0 Hz, PhCH2, 1H), 4.94 (d, J = 11.0
Hz, PhCH2, 1H), 4.72 (d, J = 10.5 Hz, PhCH2, 1H), 4.69 (d, J = 12.5 Hz,
PhCH2, 1H), 4.67 (d, J = 13.0 Hz, PhCH2, 1H), 4.64 (d, J = 11.5 Hz,
PhCH2, 1H), 4.62 (d, J = 12.5 Hz, PhCH2, 1H), 4.58ꢀ4.48 (m, 10H), 4.43
(d, J = 12.0 Hz, PhCH2, 1H), 4.37ꢀ4.30 (m, 4H), 4.20 (d, J = 12.5 Hz,
PhCH2, 1H), 4.16 (br s, 1H), 4.05 (dd, J = 9.0, 2.0 Hz, 1H), 4.02ꢀ3.97 (m,
2H), 3.88 (m, 2H), 3.81ꢀ3.76 (m, 2H), 3.70 (m, 2H), 3.63 (dd, J = 11.5,
2.0 Hz, 1H) 3.49ꢀ3.45 (m, 2H), 3.35ꢀ3.32 (m, 3H); 13C NMR (CDCl3,
100 MHz): δ 139.57, 139.16, 138.83, 138.55, 138.32, 138.19, 134.74,
131.60, 129.03, 128.39, 128.29, 128.22, 128.04, 127.98, 127.87, 127.81,
127.65, 127.51, 127.27, 127.15, 127.01, 103.31, 99.73, 85.28, 82.65, 80.08,
79.87, 78.88, 77.52, 75.31, 75.07, 74.74, 73.70, 73.49, 73.38, 73.26, 73.02,
72.62, 71.55, 69.36, 68.56, 68.34.
Synthesis of 2,3,4,6-Tetra-O-benzyl-β-D-galactopyranosyl-
(1f4)-2,3,6-tri-O-benzyl-R-D-glucopyranosyl-(1f3)-2,4,6-
tri-O-benzyl-R-D-mannopyranosyl Fluoride (20). To an ice-
cold solution of compound 19 (203 mg, 0.135 mmol) in acetoneꢀwater
(9:1, 6 mL) was added NBS (72 mg, 0.406 mmol) and the mixture was
stirred for 2 h. TLC (hexaneꢀethyl acetate, 2/1) indicated complete
disappearance of starting materials. The mixture was diluted with
EtOAc. The organic layer was washed with saturated NaHCO3 (aq),
water, and brine; dried over Na2SO4; and filtered. The filtrate was
concentrated to dryness to give the hemiacetal (ESIꢀMS, calculated,
M = 1405.65; found (m/z), 1428.46 [M + Na]+). The crude product was
dissolved in anhydrous CH2Cl2 (10 mL) and cooled to ꢀ10 °C. To the
solution was added DAST (45 μL, 0.34 mmol). The mixture was stirred
for 50 min at ꢀ10 °C and the reaction was then quenched with saturated
NaHCO3 (aq). The organic layer was separated and washed with water
and brine, dried over Na2SO4, and filtered. The filtrate was concentrated
and the residue was subjected to flash chromatography on silica gel
(tolueneꢀethyl acetate, 12/1) to give compound 20 (173 mg, 91%).
HRMS MALDI-TOF: [M + Na]+ calcd for C88H91O15FNa, 1429.6239,
found 1429.6284. 1H NMR (CDCl3, 500 MHz): δ 7.32ꢀ7.08 (m, 50H),
5.48 (d, JHꢀF = 55.0 Hz, H-1, 1H), 5.15 (d, J = 4.0 Hz, H-1Glc,1H), 5.04
(d, J = 11.0 Hz, PhCH2, 1H), 5.01 (d, J = 12.0 Hz, PhCH2, 1H), 4.95 (d,
J = 11.5 Hz, PhCH2, 1H), 4.72ꢀ4.62 (m, 7H), 4.58ꢀ4.52 (m, 3H), 4.47
(d, J = 7.5 Hz, H-1Gal, overlapped, 1H), 4.48ꢀ4.44 (m, 3H, overlapped
with H-1Gal), 4.36ꢀ4.31 (m, 3H), 4.21 (d, J = 12.5 Hz, PhCH2, 1H),
4.08ꢀ4.05 (m, 2H), 4.02ꢀ3.96 (m, 3H), 3.92ꢀ3.85 (m, 3H), 3.78 (dd,
J = 11.0, 4.0 Hz, 1H), 3.71 (br t, J = 9.0 Hz, 1H), 3.68ꢀ3.62 (m, 3H),
3.50ꢀ3.46 (m, 2H), 3.36ꢀ3.34 (m, 3H); 13C NMR (CDCl3, 125 MHz):
δ 139.34, 139.03, 138.72, 138.46, 138.17, 138.07, 138.06, 128.37, 128.32,
128.15, 127.97, 127.88, 127.78, 127.68, 127.61, 127.43, 127.36, 127.27,
127.21,126.95, 105.79 (d, J = 229.0 Hz, C-1), 103.26, 99.36, 82.49, 79.99,
79.85, 78.65, 77.46, 75.20, 75.00, 74.66, 74.32, 73.85, 73.62, 73.39, 73.32,
73.18, 72.91, 72.54, 71.48, 68.70, 68.49, 68.25.
H-40, 1H), 5.87 (s, H-1, 1H), 5.83 (dd, J = 10.0, 3.0 Hz, H-3, 1H), 5.78
(dd, J = 10.0, 3.0 Hz, H-30, 1H), 5.21 (s, H-10, 1H), 4.95 (m, H-5, 1H),
4.70 (br s, H-2, 1H), 4.60 (m, H-6, 2H), 4.53 (m, H-50, 1H), 4.50ꢀ4.26
(m, H-20, H-60, 3H); 13C NMR (CDCl3, 125 MHz): δ 166.24 (PhCO ꢁ
2), 165.61 (PhCO ꢁ 2), 165.52 (PhCO), 165.41 (PhCO), 133.47,
133.41, 133.37, 133.03, 131.86, 129.90, 129.72, 129.28, 128.96, 128.42,
127.94, 101.41, 86.63, 72.17, 71.69, 69.88, 67.55, 66.87, 63.66, 63.38.
Synthesis of Phenyl 2,3,4,6-Tetra-O-benzyl-β-D-galacto-
pyranosyl-(1f4)-2,3,6-tri-O-benzyl-R-D-glucopyranosyl-(1f3)-
2,4,6-tri-O-benzyl-R-D-mannopyranosyl-(1f2)-3,4,6-tri-O-
benzoyl-R-D-mannopyranosyl-(1f2)-3,4,6-tri-O-benzoyl-1-
thio-R-D-mannopyranoside (22). To a mixture of glycosyl fluoride
20 (1.15 g, 0.817 mmol), disaccharide 21 (0.72 g, 0.681 mmol), and
dried MS (2.30 g, 4 Å) in CH2Cl2 (30 mL) were added Cp2HfCl2 (0.46
g, 1.23 mmol) and AgOTf (0.63 g, 2.451 mmol) at 0 °C. The reaction
mixture was stirred for 4 h at room temperature and then filtered
through a Celite pad. The filtrate was washed with saturated NaHCO3
(aq), water, and brine; dried over Na2SO4; and filtered. The solution was
concentrated to dryness. The residue was subjected to flash chroma-
tography on silica gel (tolueneꢀethyl acetate, 15/1) to afford 22 (1.15 g,
69%). ESIꢀMS: calcd for C148H140O31S, M = 2446.74; Found (m/z),
1
2469.51 [M + Na]+. H NMR (CDCl3, 500 MHz): δ 7.96ꢀ7.91 (m,
Ar), 7.51ꢀ6.99 (m, Ar), 6.03 (t, 1H), 5.96 (dd, J = 10.0, 9.5 Hz, 1H),
5.91 (s, 1H), 5.85 (dd, J = 9.5, 3.0 Hz, 1H) 5.77 (dd, 1H), 5.50 (br s, 1H),
5.01ꢀ4.92 (m, H-1Glc, H-1Man, PhCH2, overlapped, 5H), 4.67ꢀ4.50
(m, 13H), 4.42ꢀ4.32 (m, 8H), 4.24ꢀ4.19 (m, 3H), 4.13ꢀ4.02 (m, 4H),
3.89ꢀ3.84 (m, 6H), 3.76ꢀ3.62 (m, 3H), 3.48ꢀ3.46 (m, 3H),
3.33ꢀ3.32 (m, 4H). 13C NMR (CDCl3, 125 MHz): δ 166.22
(PhCO), 166.19 (PhCO), 165.69 (PhCO), 165.43 (PhCO), 165.33
(PhCO), 165.29 (PhCO), 139.49, 139.09, 18.71, 138.49, 1138.26,
138.03, 133.31, 132.93, 132.75, 131.53, 130.00, 129.92, 129.73,
129.59, 129.20, 129.02, 128.55, 128.34, 128.27, 128.13, 127.98,
127.84, 127.78, 127.70, 127.64, 127.43, 127.30, 127.15, 126.85,
102.91, 100.62, 100.51, 98.86, 86.39, 82.61, 79.91, 79.49, 78.97, 78.23,
75.98, 75.20, 74.77, 74.64, 73.66, 73.38, 73.09, 72.80, 72.53, 71.98, 71.48,
71.00, 69.82, 68.25, 67.89, 67.02, 63.74, 63.57.
Synthesis of Phenyl 2,3,4,6-Tetra-O-benzyl-β-D-galacto-
pyranosyl-(1f4)-2,3,6-tri-O-benzyl-R-D-glucopyranosyl-(1f3)-
2,4,6-tri-O-benzyl-R-D-mannopyranosyl-(1f2)-3,4,6-tri-O-
benzyl-R-D-mannopyranosyl-(1f2)-3,4,6-tri-O-benzyl-1-
thio-R-D-mannopyranoside (23). To a solution of compound 22
(1.21 g, 0.493 mmol) in tolueneꢀmethanol (1/1, 20 mL) was added a
solution of NaOMe in MeOH (0.5 M, 0.5 mL). The mixture was stirred
for 5 h at 50 °C when TLC and ESIꢀMS indicated the completion de-O-
acylation. Solvent was removed and the residual foam was dissolved in
anhydrous DMF (10 mL) and cooled to 0 °C. To the solution were
added NaH (177 mg, 7.39 mmol), BnBr (0.88 mL, 7.39 mmol), and
TBAI (34 mg, 0.09 mmol). The reaction mixture was stirred for 26 h at
room tmeperature, diluted with EtOAc, and washed with water and
brine. The organic layer was dried over Na2SO4, filtered, and the filtrate
was concentrated. The residue was subjected to flash chromatography
on silica gel (tolueneꢀethyl acetate, 14/1) to give 23 as a foam (1.03 g,
88%). HRMS MALDI-TOF: [M + Na]+ calcd for C148H152O25SNa,
2384.0241, found 2384.0249. 1H NMR (CDCl3, 500 MHz): δ
7.42ꢀ7.01 (m, Ar, 85H), 5.73 (s, H-1, 1H), 5.24 (s, H-1, 1H), 5,12
(s, H-1, 1H), 5.70 (d, J = 3.0 Hz, H-1Glc, 1H) 4.99 (d, J = 10.5 Hz,
PhCH2, 1H), 4.92 (d, J = 12.0 Hz, PhCH2, 1H), 4.84 (d, J = 11.0 Hz,
PhCH2,1H),4.78(d,J=10.5Hz,PhCH2,1H),4.67(d,J= 11.0 Hz, PhCH2,
1H), 4.63 (d, J = 12.0 Hz, PhCH2, 1H), 4.62ꢀ4.54 (m, 6H), 4.52ꢀ4.43 (m,
H-1Gal overlapped, 16H), 4.33ꢀ4.25 (m, 6H), 4.19 (d, J = 11.5 Hz, 1H),
4.13ꢀ4.08 (m, 4H), 3.98ꢀ3.79 (m, 11H), 3.75 (dd, J = 10.0, 9.5 Hz, 1H),
3.71ꢀ3.53 (m, 4H), 3.49ꢀ3.44 (m, 2H), 3.34ꢀ3.29 (m, 4H).
Synthesis of Phenyl 3,4,6-Tri-O-benzoyl-R-D-mannopyra-
nosyl-(1f2)-3,4,6-tri-O-benzoyl-1-thio-R-D-mannopyrano-
side (21). To a solution of compound 11 (498 mg, 0.452 mmol) in
CH2Cl2 (12 mL) was added 5% AcCl in MeOH (15 mL) at 0 °C and the
resulting mixture was stirred for 8 h. The solution was concentrated to
dryness and the residue was subjected to flash silica gel column
chromatography (tolueneꢀEtOAc, 15/1) to afford compound 21
(374 mg, 79%) as a white amorphous solid. HRMS MALDI-TOF: [M
+ Na]+ calcd for C60H50O16SNa, 1081.2717, found 1081.2711. 1H
NMR (CDCl3, 500 MHz): δ 8.04ꢀ7.91 (m, 12H), 7.52ꢀ7.15 (m,
23H), 6.00 (dd, J = 10.0, 9.5 Hz, H-4, 1H), 5.94 (dd, J = 10.0, 9.5 Hz,
Synthesis of 2,3,4,6-Tetra-O-benzyl-β-D-galactopyranosyl-
(1f4)-2,3,6-tri-O-benzyl-R-D-glucopyranosyl-(1f3)-2,4,6-
14413
dx.doi.org/10.1021/ja204831z |J. Am. Chem. Soc. 2011, 133, 14404–14417

Journal of the American Chemical Society
ARTICLE
tri-O-benzyl-R-D-mannopyranosyl-(1f2)-3,4,6-tri-O-ben-
zyl-R-D-mannopyranosyl-(1f2)-3,4,6-tri-O-benzyl-R-D-
mannopyranosyl Fluoride (6). To a solution of compound 23
(0.969 g, 0.41 mmol) in acetone (10 mL) were added NBS (0.183 g, 1.04
mmol) and H2O (0.5 mL) at ꢀ10 °C. The reaction mixture was stirred
for 2 h, when TLC (tolueneꢀEtOAc, 6/1) revealed the disappearance of
starting material. The solvent was evaporated and residue was dissolved
in EtOAc. The solution was successively washed with saturated NaH-
CO3 (aq), water, and brine; dried over Na2SO4; and filtered. The filtrate
was concentrated. The residue was dissolved in anhydrous CH2Cl2
(15 mL). To the solution was added DAST (108 μL, 0.83 mmol) at
ꢀ10 °C. The mixture was stirred for 45 min, diluted with CH2Cl2, and
washed with saturated NaHCO3 (aq), brine, and water. The organic
layer was dried over Na2SO4 and filtered. The filtrate was concentrated
and the residue was subjected to flash chromatography on silica gel
(tolueneꢀEtOAc, 15/1) to afford compound 6 (0.898 g, 93%). HRMS
MALDI-TOF: [M + Na]+ calcd for C142H147FO25Na, 2294.0113, found
2294.0135. 1H NMR (CDCl3, 500 MHz): δ 7.70ꢀ7.13 (m, Ar, 80H),
5.73 (d, J = 54.5 Hz, H-1, 1H), 5.24 (s, H-1, 1H), 5,09 (br s, H-1, H-1Glc
overlapped, 2H), 5.00 (d, J = 10.5 Hz, PhCH2, 1H), 4.92 (d, J = 11.5 Hz,
PhCH2, 1H), 4.82 (d, J = 10.5 Hz, PhCH2, 1H), 4.80 (d, J = 11.5 Hz,
PhCH2, 1H), 4.68 (d, J = 11.0 Hz, PhCH2, 1H), 4.63 (d, J = 12.0 Hz,
PhCH2, 1H), 4.62ꢀ4.54 (m, 6H), 4.51ꢀ4.39 (m, H-1Gal, overlapped,
17H), 4.35ꢀ4.26 (m, 5H), 4.23ꢀ4.19 (m, 3H), 4.13ꢀ4.08 (m, 3H),
3.97ꢀ3.91 (m, 5H), 3.84ꢀ3.79 (m, 5H), 3.77ꢀ3.66 (m, 4H),
3.60ꢀ3.54 (m, 3H), 3.50ꢀ3.47 (m, 2H), 3.33ꢀ3.29 (m, 4H). 13C
NMR (CDCl3, 125 MHz): δ 139.50, 139.10, 138.79, 138.60, 138.54,
138.48, 138.36, 138.33, 138.28, 138.27, 138.24, 138.13, 137.95, 132.46,
130.88, 128.51, 128.35, 128.32, 128.30, 128.29, 128.19, 128.14, 128.06,
128.03, 127.97, 127.90, 127.87, 127.85, 127.78, 127.74, 127.71, 127.66,
127.63, 127.60, 127.56. 127.49, 127.45, 127.42, 127.31, 127.22, 127.19,
126.88, 126.75, 106.62, 102.99, 101.19, 99.38, 90.03, 82.60, 79.99, 79.58,
79.07, 78.17, 75.25, 75.15, 75.00, 74.85, 74.64, 73.97, 73.80, 73.61, 73.38,
73.32, 73.28, 73.24, 73.11, 72.52, 72.46, 72.35, 71.99, 71.66, 71.23, 69.22,
68.65, 68.24, 68.17.
127.99, 127.93, 127.91, 127.87, 127.83, 127.76, 127.74, 127.68, 127.65,
127.59, 127.52, 127.48, 127.46, 127.41, 127.38, 127.26, 127.19, 127.16,
127.14, 126.85, 126.70, 102.92, 101.09, 101.07, 100.33, 100.03, 82.56,
81.54, 79.98, 79.71, 79.55, 79.13, 78.79, 78.21, 75.54, 75.39, 75.22, 75.13,
75.07, 74.91, 74.82, 74.60, 73.71, 73.60, 73.41, 73.35, 73.28, 73.13, 73.07,
72.90, 72.74, 72.49, 72.35, 72.26, 71.55, 71.46, 71.25, 70.74, 69.72, 69.57,
68.98, 68.31, 68.22, 68.07, 66.99, 65.69.
Synthesis of Benzyl 2,3,4,6-Tetra-O-benzyl-β-D-galacto-
pyranosyl-(1f4)-2,3,6-tri-O-benzyl-R-D-glucopyranosyl-
(1f3)-2,4,6-tri-O-benzyl-R-D-mannopyranosyl-(1f2)-3,4,6-
tri-O-benzyl-R-D-mannopyranosyl-(1f2)-3,4,6-tri-O-ben-
zyl-R-D-mannopyranosyl-(1f3)-2-O-benzyl-β-D-mannopyra-
nosyl-(1f4)-2-azido-3,6-di-O-benzyl-2-deoxy-β-D-glucopyr-
anoside (25). A solution of compound 24 (320 mg, 0.104 mmol) in
AcOHꢀEtOHꢀH2O (4/1/1) was stirred at 100 °C for 5 h. The
reaction mixture was then concentrated to dryness and the residue
was subjected directly to flash chromatography on silica gel (hexaneꢀ
EtOAc, 2/1) to afford compound 25 (284 mg, 91%) as a white foam.
HRMS MALDI-TOF: [M + Na]+ calcd for C182H191N3O35Na,
1
3001.3156, found 3001.3189. H NMR (CDCl3, 500 MHz): δ 7.39ꢀ
7.08 (m, 100H), 5.31 (s, H-1, 1H), 5.30 (s, H-1, 1H), 5.17 (s, H-1, 1H),
5.10 (d, PhCH2, overlapped, 1H), 5.08 (d, H-1Glc overlapped, 1H), 5.00
(d, J = 11.0 Hz, PhCH2, 1H), 4.99 (d, J = 11.0 Hz, PhCH2, 1H), 4.92 (d,
J = 11.5 Hz, PhCH2, 1H), 4.90 (d, J = 11.0 Hz, PhCH2, 1H), 4.81 (d, J =
10.5 Hz, PhCH2, 1H), 4.80 (d, J = 11.5 Hz, PhCH2, 1H), 4.70ꢀ4.39
(m, 25H), 4.34ꢀ4.24 (m, 6H), 4.19 (d, J = 11.5 Hz, 2H), 4.11ꢀ4.07
(m, 3H), 3.96ꢀ3.91 (m, 3H), 3.85ꢀ3.73 (m, 5H), 3.70ꢀ3.60 (m, 5H),
3.58ꢀ3.51 (m, 4H), 3.49ꢀ3.43 (m, 4H), 3.41ꢀ3.27 (m, 4H), 3.21 (br d,
1H), 3.11 (m, 1H), 2.84 (m, 1H); 13C NMR (CDCl3, 125 MHz):
δ 139.49, 139.18, 139.10 138.98, 138.79, 138.57, 138.45, 138.35, 138.25,
138.03, 137.98, 137.65, 137.52, 136.84, 128.53, 128.45, 128.35, 128.25,
128.12, 128.00, 127.94, 127.85, 127.76, 127.72, 127.68, 127.66, 127.63,
127.61, 127.58, 127.63, 127.61, 127.58, 127.54, 127.51, 127.47, 127.43,
127.37, 127.35, 127.32, 127.21, 127.09, 127.04,126.78, 102.97, 100.84,
100.31, 100.22, 99.56, 82.59, 81.39, 80.30, 79.99, 79.56, 79.10, 78.20,
75.75, 75.55, 75.38, 75.15, 74.97, 74.82, 74.69, 74.58, 74.09, 73.75, 73.48,
73.36, 73.08, 72.62, 72.49, 72.38, 71.85, 71.46, 71.36, 70.73, 70.16, 69.65,
69.01, 68.24, 68.11, 65.82, 62.44.
Synthesis of Benzyl 2,3,4,6-Tetra-O-benzyl-β-D-galacto-
pyranosyl-(1f4)-2,3,6-tri-O-benzyl-R-D-glucopyranosyl-
(1f3)-2,4,6-tri-O-benzyl-R-D-mannopyranosyl-(1f2)-3,4,6-
tri-O-benzyl-R-D-mannopyranosyl-(1f2)-3,4,6-tri-O-ben-
zyl-R-D-mannopyranosyl-(1f3)-2-O-benzyl-4,6-O-benzyli-
dene-β-D-mannopyranosyl-(1f4)-2-azido-3,6-di-O-benzyl-
2-deoxy-β-D-glucopyranoside (24). To a mixture of pentasacchar-
ide donor 6 (870 mg, 0.382 mmol), disaccharide acceptor 7 (223 mg,
0.273 mmol), and dried MS (1 g, 4 Å) in CH2Cl2 (20 mL) were added
Cp2HfCl2 (232 mg, 0.612 mmol) and AgOTf (315 mg, 1.225 mmol) at
0 °C. The mixture was stirred for 5 h at room temperature. After dilution
with CH2Cl2, the reaction mixture was filtered through a Celite pad. The
filtrate was washed with saturated NaHCO3 (aq), brine, and water; dried
over Na2SO4; and filtered. The filtrate was concentrated to dryness and
the residue was subjected to flash chromatography on silica gel
(tolueneꢀEtOAc, 6/1) to give compound 24 (517 mg, 62%). MAL-
DI-TOF MS: calcd for C189H195O35N3, M = 3068.57; Found (m/z),
Synthesis of Benzyl 2,3,4,6-Tetra-O-benzyl-β-D-galacto-
pyranosyl-(1f4)-2,3,6-tri-O-benzyl-R-D-glucopyranosyl-
(1f3)-2,4,6-tri-O-benzyl-R-D-mannopyranosyl-(1f2)-3,4,6-
tri-O-benzyl-R-D-mannopyranosyl-(1f2)-3,4,6-tri-O-benzyl-R-
D-mannopyranosyl-(1f3)-{2,3,4,6-tetra-O-benzyl-R-D-ma-
nnopyranosyl-(1f2)-3,4,6-tri-O-benzyl-R-D-mannopyrano-
syl-(1f3)-[2,3,4,6-tetra-O-benzyl-R-D-mannopyranosyl-
(1f2)-3,4,6-tri-O-benzyl-R-D-mannopyranosyl-(1f6)]-2,4-
di-O-benzyl-R-D-mannopyranosyl-(1f6)}-2-O-benzyl-β-D-
mannopyranosyl-(1f4)-2-azido-3,6-di-O-benzyl-2-deoxy-
β-D-glucopyranoside (26). To a mixture of pentasaccharide donor 5
(139 mg, 0.061 mmol), heptasaccharide acceptor 25 (181 mg, 0.061
mmol), and dried MS (1 g, 4 Å) in CH2Cl2 (10 mL) was added
Cp2HfCl2 (369 mg, 0.971 mmol) and AgOTf (499 mg, 1.94 mmol) at
0 °C. The mixture was stirred for 9 h at room temperature and then
diluted with CH2Cl2, and filtered through a Celite pad. The filtrate was
washed with saturated NaHCO3 (aq), brine, and water; dried over
Na2SO4; and filtered. The solution was concentrated and the residue
was subjected to flash chromatography on silica gel (hexaneꢀEtOAc, 2/
1) to obtain compound 26 (193 mg, 61%). MALDI-TOF MS: calcd for
C324H337O60N3, M = 5233.13; Found (m/z), 5256.12 [M + Na]+. 1H
NMR (CDCl3, 500 MHz): δ 7.43ꢀ6.92 (m, 180H), 5.31 (s, H-1, 1H),
5.28 (s, H-1, 2H), 5.20 (s, H-1, 1H), 5.17 (s, H-1, 2H), 5.09 (br s, 2H),
4.99 (m, 2H), 4.93 and 4.91 (2s, overlapped, 2H), 4.85ꢀ4.68 (m, 7H),
4.70ꢀ4.28 (m, 67H), 4.24ꢀ4.17 (m, 4H), 4.15ꢀ4.02 (m, 9H),
3.99ꢀ3.77 (m, 25H), 3.73ꢀ3.57 (m, 9H), 3.53ꢀ3.40 (m, 12H),
1
3091.56 [M + Na]+. H NMR (CDCl3, 500 MHz): δ 7.38ꢀ7.08 (m,
105H), 5.34 [s, PhCH(O)2, 1H], 5.29 (s, H-1, 1H), 5.24 (s, H-1, 1H),
5.12 (d, H-1Glc, overlapped, 1H), 5.07 (s, H-1, 1H), 5.11ꢀ5.03 (m,
PhCH2, overlapped, 2H), 4.99 (d, J = 10.5 Hz, PhCH2, 1H), 4.93ꢀ4.89
(m, 3H), 4.84ꢀ4.82 (m, 3H), 4.73 (br s, 2H), 4.69ꢀ4.53 (m, 9H),
4.51ꢀ4.40 (m, 12H), 4.36ꢀ4.16 (m, 10H), 4.13ꢀ4.06 (m, 3H),
3.98ꢀ3.95 (m, 3H), 3.93ꢀ3.85 (m, 4H), 3.80ꢀ3.77 (m, 3H),
3.73ꢀ3.67 (m, 4H), 3.62ꢀ3.45 (m, 5H), 3.43ꢀ3.29 (m, 5H), 3.15
(dd, 1H), 2.87 (m, 1H); 13C NMR (CDCl3, 100 MHz): δ 139.46,
139.07, 138.76, 138.73, 138.61, 138.58, 138.40, 138.49, 138.46, 138.40,
138.35, 138.25, 138.15, 138.02, 137.68, 137.17, 136.80, 129.00, 128.96,
128.50, 128.4 128.35, 128.31, 128.27, 128.24, 128.14, 128.10, 128.05,
14414
dx.doi.org/10.1021/ja204831z |J. Am. Chem. Soc. 2011, 133, 14404–14417

Journal of the American Chemical Society
ARTICLE
3.35ꢀ3.29 (m, 6H), 3.23ꢀ3.04 (m, 7H); 13C NMR (CDCl3, 125
MHz): δ 139.51ꢀ137.98 (Ar), 128.68ꢀ126.59 (Ar), 102.97, 102.32,
100.95, 100.30, 99.52, 99.35, 98.90, 96.76, 96.73, 82.61, 81.25, 80.25,
80.10, 79.98, 79.56, 79.45, 79.35, 78.13, 78.07, 75.54, 75.26, 75.23, 75.16,
74.98, 74.75, 74.71, 74.46, 74.13, 73.95, 73.57, 73.38, 73.19, 73.08, 72.69,
72.49, 72.28, 72.18, 72.01, 71.95, 71.75, 71.51, 70.55, 70.33, 69.40, 69.34,
68.92, 68.52, 68.41, 68.23, 67.62, 66.84, 66.62, 66.52, 65.61.
mannopyranosyl-(1f2)-r-D-mannopyranosyl-(1f3)-[r-D-
mannopyranosyl-(1f2)-r-D-mannopyranosyl-(1f6)]-r-D-
mannopyranosyl-(1f6)}-β-D-mannopyranosyl-(1f4)-1,2-
dideoxy-D-glucopyrano]-[2,1-d]-2-oxazoline (4). To a solution
of compound 28 (4.0 mg, 1.99 nmol) in H2O (0.5 mL) was added Et3N
(25 μL, 180 nmol) and 2-chloro-1,3-dimethylimidazolinium chloride
(DMC) (11 mg, 65 nmol) at 0 °C. The reaction mixture was monitored
by DIONEX HPAEC-PAD. Within 1 h, the HPAEC analysis indicated
that the free oligosaccharide was converted into a new oligosaccharide
that was eluted earlier than the reducing sugar under the HPAEC
conditions. The product was purified by gel filtration on a Sephadex
G-10 column that was eluted with 0.1% aq TEA to afford sugar oxazoline
4 (4.0 mg, quant.) as a white solid after lyophilization. MALDI-TOF MS:
calcd for C74H123O60N, M = 1986.73; Found (m/z), 2009.66 [M +
Na]+. 1H NMR (D2O, 500 MHz): δ 6.09 (d, J = 6.5 Hz, H-1, Oxazoline,
Synthesis of Benzyl 2,3,4,6-Tetra-O-benzyl-β-D-galacto-
pyranosyl-(1f4)-2,3,6-tri-O-benzyl-R-D-glucopyranosyl-
(1f3)-2,4,6-tri-O-benzyl-R-D-mannopyranosyl-(1f2)-3,4,6-
tri-O-benzyl-R-D-mannopyranosyl-(1f2)-3,4,6-tri-O-ben-
zyl-R-D-mannopyranosyl-(1f3)-{2,3,4,6-tetra-O-benzyl-R-
D-mannopyranosyl-(1f2)-3,4,6-tri-O-benzyl-R-D-mannopy-
ranosyl-(1f3)-[2,3,4,6-tetra-O-benzyl-R-D-mannopyranosyl-
(1f2)-3,4,6-tri-O-benzyl-R-D-mannopyranosyl-(1f6)]-2,4-
di-O-benzyl-R-D-mannopyranosyl-(1f6)}-2-O-benzyl-β-D-
mannopyranosyl-(1f4)-2-acetamido-3,6-di-O-benzyl-2-
deoxy-β-D-glucopyranoside (27). To a solution of compound 26
(59.0 mg, 11.27 mmol) in AcSH (3 mL) was added 2,6-lutidine (3 mL)
and the solution was stirred at room temperature for 15 h. The reaction
mixture was diluted with CH2Cl2; washed with saturated NaHCO3 (aq),
brine, and water; dried over Na2SO4 and filtered. The filtrate was
concentrated and the residue was subjected to flash chromatography on
silica gel (hexaneꢀEtOAc, 2/1) to give compound 27 (47 mg, 86%)
MALDI-TOF MS: calcd for C326H341O61N, M = 5249.16; Found (m/
z), 5272.15 [M + Na]+. 1H NMR (CDCl3, 500 MHz): δ 7.39ꢀ6.98 (m,
180H), 6.09 (d, J = 6.5 Hz, AcNH, 1H), 5.32 (s, H-1, 1H), 5.28 (s, H-1,
H), 5.20 (s, H-1, 1H), 5.17 (s, H-1, 2H), 5.09 (br s, 2H), 4.99 (m, 2H),
4.93 and 4.91 (2s, overlapped, 2H), 4.85ꢀ4.80 (m, 7H), 4.71ꢀ4.28 (m,
67H), 4.24ꢀ4.18 (m, 4H), 4.15ꢀ4.02 (m, 9H), 3.99ꢀ3.74 (m, 25H),
3.73ꢀ3.57 (m, 9H), 3.55ꢀ3.40 (m, 12H), 3.35ꢀ3.30 (m, 6H),
3.23ꢀ3.04 (m, 7H), 1.71 (s, CH3CONH). 13C NMR (CDCl3, 125
MHz): δ 170.25 (CO), 139.49ꢀ137.56 (Ar), 128.69ꢀ126.47 (Ar),
102.99, 100.86, 100.54, 100.22, 100.10, 99.45, 99.16, 96.81, 96.63, 94.76,
82.60, 80.00, 79.92, 79.87, 79.73, 79.57, 79.41, 79.28, 79.04, 78.19, 78.06,
75.39, 75.25, 75.12, 74.80, 74.62, 74.41, 74.01, 73.81, 73.59, 73.38, 73.26,
73.09, 72.94, 72.78, 72.51, 72.24, 72.00, 71.96, 71.67, 71.39, 71.22, 70.64,
70.36, 69.26, 68.98, 68.90, 68.81, 68.63, 68.25, 68.19, 67.53, 66.68, 23.09
(CH3CONH).
Synthesis of β-D-Galactopyranosyl-(1f4)-r-D-glucopyra-
nosyl-(1f3)-r-D-mannopyranosyl -(1f2)-r-D-mannopyra-
nosyl-(1f2)-r-D-mannopyranosyl-(1f3)-{r-D-mannopyra-
nosyl-(1f2)-r-D-mannopyranosyl-(1f3)-[r-D-mannopyra-
nosyl-(1f2)-r-D-mannopyranosyl-(1f6)]-r-D-mannopyra-
nosyl-(1f6)}-β-D-mannopyranosyl-(1f4)-2-acetamido-2-
deoxy-r,β-D-glucopyranose (28). A mixture of compound 27 (62
mg, 11.8 μmol) and 20% Pd(OH)2/C (41 mg) in 50% AcOH (aq.)
(12 mL) was stirred under H2 atmosphere for 21 h. The reaction mixture
was filtered through a Celite pad. The filtrate was dissolved in water and
lyophilized. The crude product was purified on a Sephadex LH-20
column by elution with H2O. Fractions containing the product were
pooled and lyophilized to give the free dodecasaccharide (28) (22 mg,
92%) as a white solid. MALDI-TOF MS: calcd for C74H125O61N, M =
2004.75; Found (m/z), 2027.67 [M + Na]+, ESIꢀMS: 1002.84 [M +
2H]2+. 1H NMR (D2O, 500 MHz): δ 5.40 (s, 1H), 5.34 (s, 1H), 5.30 (s,
1H), 5.25 (d, J = 2.5 Hz, H-1Glc, 1H), 5.23 (s, H-1, 1H), 5.14 (s, H-1, H),
5.05 (s, H-1, 1H), 5.04 (s, H-1, 2H), 5.09 (br s, 2H), 4.86, 4.76, 4.70
(overlapped with solvent H-OD) 4.45 (d, J = 8.0 Hz, H-1Gal, 1H),
4.23ꢀ3.53 (m, 72H), 2.04 (s, 3H). 13C NMR (D2O,, selected signals,
125 MHz): δ 170.21, 101.87, 101.30, 101.06, 99.88, 99.69, 99.21, 99.14,
97.22, 22.98.
1H), 5.37 (s, 1H), 5.32 (s, 1H), 5.29 (s, 1H), 5.25 (d, J = 2.5 Hz, H-1Glc
,
1H), 5.14 (s, H-1, 1H), 5.03 (s, H-1, 3H), 5.09 (s, H-1, 1H), 4.91 (s, H-1,
1H), 4.44 (d, J = 7.5 Hz, H-1Gal, 1H), 4.23ꢀ3.54 (m, 69H), 3.07ꢀ3.03
(m, 3H), 1.90 (s, 3H). 13C NMR (D2O, selected signals, 125 MHz): δ
167.8, 102.20, 101.61, 101.06, 99.85, 99.83, 99.11, 97.17, 16.75.
Synthesis of Glycoprotein Gal1Glc1Man9GlcNAc2-RNase
(2). A solution of dodecasaccharide oxazoline 4 (1.50 mg, 755 nmol)
and GlcNAc-RNase 3 (1.50 mg, 108 nmol) in a phosphate buffer
(50 mM, pH 7.0, 50 μL) was incubated with Endo-A mutant N171A
(50 μg) at 23 °C. The reaction was monitored by reverse-phase HPLC.
After 5 h, the product was purified by HPLC to afford glycoprotein 2
(1.38 mg, 86.5 nmol, 80%). The glycoprotein was quantified with a
standard solution of natural RNase by measuring the UV absorbance at
280 nm. Analytical RP-HPLC of 2, tR = 21.6 min (see the description in
general method for the conditions); ESIꢀMS of 2: calcd M = 15 870 Da;
Found (m/z), 1764.45 [M + 9H]9+, 1588.25 [M + 10H]10+, 1443.95
[M + 11H]11+, 1323.71 [M + 12H]12+, 1221.99 [M + 13H]13+ and 1134.80
[M + 14H]14+. Deconvolution of the ESIꢀMS dada, M = 15 872.0 Da.
Synthesis of Glycoprotein Man9GlcNAc2-RNase (30). A
solution of Man9GlcNAc oxazoline 29 (500 μg, 0.30 μmol) and
GlcNAc-RNase 3 (500 μg, 0.036 μmol) in a phosphate buffer
(50 mM, pH 7.0, 20 μL) was incubated with EndoA-N171A (20 μg)
at 23 °C for 5 h. The glycoprotein product was isolated by preparative
HPLC to give Man9GlcNAc2-RNase (30) as white foam after lyophi-
lization (438 μg, 78%). Analytical RP-HPLC of 30, tR = 22 min (see the
description in general method for the conditions); ESIꢀMS of 30: calcd,
M = 15547.0 Da; found, 15548.9 Da (deconvolution data).
Synthesis of Glycoprotein Glc1Man9GlcNAc2-RNase (1).
Glycoprotein 2 (535 μg, 33.7 nmol) was incubated with β-1,4-galacto-
sidase from S. pnemoniae (6 mU) in a phosphate buffer (50 mM, pH 5.5)
at 37 °C. The reaction was monitored by ESIꢀMS. After 2 h, ESIꢀMS
indicated the complete removal of the galactose residue. The product
was purified by reverse-phase HPLC to give glycoprotein 1 (530 μg,
33.7 nmol, quantitative). The glycoprotein was quantified with a
standard solution of natural RNase B by UV absorbance at 280 nm.
ESIꢀMS of 1: calcd M = 15 708 Da; Found (m/z), 1748.23 [M + 9H]9+,
1572.01 [M + 10H]10+, 1429.19 [M + 11H]11+, 1311.4 [M + 12H]12+
,
1210.72 [M + 13H]13+ and 1124.33 [M + 14H]14+. Deconvolution of
the ESIꢀMS data, M = 15710.0 Da.
RNase Activity Assay. The ribonuclease activity of the RNase
glycoforms was measured by following the method described in the
literature.25 Briefly, a solution (0.5 mg/mL) of yeast RNA (Sigma) in a
sodium acetate buffer (500 μL, 100 mM, pH 5.0) was placed in a
spectrophotometer. The enzymatic reaction was then initiated by the
addition of 2 μg of natural RNase B or each of the individual synthetic
glycoforms. The absorbance at 300 nm was recorded in every 2 min. The
control reaction was run with the addition of heat-denatured RNase B.
The decrease of absorbance at 300 nm over time indicated the progress
of degradation of the RNA by the ribonuclease activity.
Synthesis of 2-Methyl-[β-D-galactopyranosyl-(1f4)-r-D-
glucopyranosyl-(1f3)-r-D-mannopyranosyl-(1f2)-r-D-
mannopyranosyl-(1f2)-r-D-mannopyranosyl-(1f3)-{r-D-
14415
dx.doi.org/10.1021/ja204831z |J. Am. Chem. Soc. 2011, 133, 14404–14417

Journal of the American Chemical Society
ARTICLE
Circular Dichroism (CD) Spectroscopy. The CD spectroscopic
analysis of natural RNase B and the synthetic RNase glycoforms was
performed with JASCO 715 instrument at ambient temperature using a
0.1 cm cuvette. Samples were prepared in sodium phosphate buffer
(5 mM, pH 7.5) at a concentration of 5ꢀ10 μM. The samples were
scanned from 260 to 190 at 10 nm/min with a bandwidth of 1 nm and
data patching of 0.5 nm. Each spectrum was smoothed and blank
subtracted and represented the average of three scans. The data was
analyzed using the DICHROWEB server publicly available at online.23
Partial and Complete Denaturation of RNase Glycopro-
teins. The synthetic RNase glycoproteins (1, 2, and 30) and the natural
RNase B (31) were denatured following previously reported methods.27
For partial denaturation without breaking the disulfide bonds, the
glycoproteins were dissolved in 0.1 M Tris-HCl buffer (pH 8.5)
containing 8 M urea. The solution was incubated at 23 °C for 3 h,
and the pH was adjusted to 3.0 by adding HCl. The solution was then
dialyzed against 0.1 M acetic acid and the dialyzed solution was
lyophilized, which was used for reverse-HPLC analysis and for SPR
studies. For complete denaturation, the glycoproteins were incubated
with 8 M urea and 0.15 M dithiothreitol (DTT) in a 0.1 M Tris-HCl
buffer (pH 8.5) for 3 h. Then, the pH was adjusted to 3.0 and the
solution was dialyzed against 0.1 M acetic acid. After lyophilization, the
denatured glycoproteins were used for reverse-HPLC analysis and for
SPR studies.
procedures for the synthesis of pentasaccharide 5. This material is
available free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
lwang@som.umaryland.edu
’ ACKNOWLEDGMENT
We thank Dr. Shuang Jake Yang and Dr. Hui Zhang (Johns
Hopkins University) for technical assistance in measuring the
HRMS (MALDI-TOF) spectra. This work was supported by the
National Institutes of Health (NIH grants R01 GM080374 and
R01 GM080374-S1).
’ REFERENCES
(1) (a) Helenius, A.; Aebi, M. Science 2001, 291, 2364–2369. (b)
Wang, S.; Ng, D. T. Nat. Cell Biol. 2008, 10, 251–253. (c) Aebi, M.;
Bernasconi, R.; Clerc, S.; Molinari, M. Trends Biochem. Sci. 2010,
35, 74–82.
(2) Spiro, R. G. J. Biol. Chem. 2000, 275, 35657–35660.
(3) Zapun, A.; Petrescu, S. M.; Rudd, P. M.; Dwek, R. A.; Thomas,
D. Y.; Bergeron, J. J. Cell 1997, 88, 29–38.
Surface Plasmon Resonance (SPR) Measurements. SPR
measurements were performed on a Biacore T100 instrument (GE
Healthcare). Calreticulin (CRT) was purchased from United States
Biological (Swampscott, MA) and was dialyzed against a HEPES buffer
(2 mM, pH 7.4) containing 0.5 mM CaCl2 at 4 °C for 16 h to remove
Tris buffer in the original sample that would otherwise interfere with the
active-ester-based immobilization. Approximately 3000 resonance units
(RU) of CRT was immobilized on a CM5 sensor chip in a sodium
acetate buffer (10 mM, pH 5.5) at 10 °C, using amine coupling kit
provided by the manufacturer. RNase glycoproteins were determined at
10 °C under a flow rate of 30 μL/min. HBS-P+ buffer (10 mM HEPES,
150 mM NaCl, 0.05% surfactant P20, pH 7.4) containing 5 mM CaCl2
was used as sample buffer and running buffer. Association was measured
for 2 min and dissociation for 10 min at same flow rate (30 μL/min).
The surface regeneration was performed by 0.5% SDS at a flow rate of
10 μL/min for 60 s. The analytes were injected in a concentration of 2
μM, 1 μM, 500 nM, 250 nM, 125 nM, 62.5 nM, and 31.25 nM,
respectively. Kinetic analyses were performed by global fitting of the
binding data to a 1:1 Langmuir binding model using BIAcore T100
evaluation software and the equilibrium constant KD was derived. For
Glc1Man9GlcNAc2-RNase (1), kon = 1.28 ꢁ 104 Mꢀ1 sꢀ1; koff = 4.82 ꢁ
10ꢀ4 sꢀ1; and KD = 38 nM at 10 °C. For Gal1Glc1Man9GlcNAc2-RNase
(2), kon = 1.81 ꢁ 104 Mꢀ1 sꢀ1 ; koff = 1.99 ꢁ 10ꢀ3 sꢀ1; and KD = 110 nM
at 10 °C. For a comparative study, the binding of partially denatured and
completely denaturated (unfolded) RNase glycoproteins was measured
at 250 nM concentration. The denatured Glc1Man9GlcNAc2-RNase
(1) and Gal1Glc1Man9GlcNAc2-RNase (2) showed almost the same
CRT-binding profiles as those of the corresponding folded glycopro-
teins, whereas the nonglucosylated Man9GlcNAc2-RNase (30) and
Man5ꢀ9GlcNAc2-RNase (31) did not bind to CRT before and after
denaturation.
(4) Patil, A. R.; Thomas, C. J.; Surolia, A. J. Biol. Chem. 2000,
275, 24348–24356.
(5) Kapoor, M.; Srinivas, H.; Kandiah, E.; Gemma, E.; Ellgaard, L.;
Oscarson, S.; Helenius, A.; Surolia, A. J. Biol. Chem. 2003, 278, 6194–
6200.
(6) (a) Matsuo, I.; Wada, M.; Manabe, S.; Yamaguchi, Y.; Otake, K.;
Kato, K.; Ito, Y. J. Am. Chem. Soc. 2003, 125, 3402–3403. (b) Arai, M. A.;
Matsuo, I.; Hagihara, S.; Totani, K.; Maruyama, J.; Kitamoto, K.; Ito, Y.
ChemBioChem 2005, 6, 2281–2289. (c) Totani, K.; Ihara, Y.; Matsuo, I.;
Ito, Y. J. Biol. Chem. 2006, 281, 31502–31508. (d) Bosis, E.; Nachliel, E.;
Cohen, T.; Takeda, Y.; Ito, Y.; Bar-Nun, S.; Gutman, M. Biochemistry
2008, 47, 10970–10980. (e) Takeda, Y.; Totani, K.; Matsuo, I.; Ito, Y.
Curr. Opin. Chem. Biol. 2009, 13, 582–591. (f) Totani, K.; Ihara, Y.;
Tsujimoto, T.; Matsuo, I.; Ito, Y. Biochemistry 2009, 48, 2933–2940.
(7) Tatami, A.; Hon, Y. S.; Matsuo, I.; Takatani, M.; Koshino, H.; Ito,
Y. Biochem. Biophys. Res. Commun. 2007, 364, 332–337.
(8) (a) Li, B.; Zeng, Y.; Hauser, S.; Song, H.; Wang, L. X. J. Am.
Chem. Soc. 2005, 127, 9692–9693. (b) Zeng, Y.; Wang, J.; Li, B.; Hauser,
S.; Li, H.; Wang, L. X. Chem.—Eur. J. 2006, 12, 3355–3364. (c) Wei, Y.;
Li, C.; Huang, W.; Li, B.; Strome, S.; Wang, L. X. Biochemistry 2008,
47, 10294–10304. (d) Huang, W.; Wang, D.; Yamada, M.; Wang, L. X. J.
Am. Chem. Soc. 2009, 131, 17963–17971. (e) Huang, W.; Groothuys, S.;
Heredia, A.; Kuijpers, B. H.; Rutjes, F. P.; van Delft, F. L.; Wang, L. X.
ChemBioChem 2009, 10, 1234–1242. (f) Schwarz, F.; Huang, W.; Li, C.;
Schulz, B. L.; Lizak, C.; Palumbo, A.; Numao, S.; Neri, D.; Aebi, M.;
Wang, L. X. Nat. Chem. Biol. 2010, 6, 264–266. (g) Umekawa, M.;
Higashiyama, T.; Koga, Y.; Tanaka, T.; Noguchi, M.; Kobayashi, A.;
Shoda, S.; Huang, W.; Wang, L. X.; Ashida, H.; Yamamoto, K. Biochim.
Biophys. Acta 2010, 1800, 1203–1209. (h) Huang, W.; Zhang, X.; Ju, T.;
Cummings, R. D.; Wang, L. X. Org. Biomol. Chem. 2010, 8, 5224–5233.
(i) Huang, W.; Li, J.; Wang, L. X. ChemBiochem 2011, 12, 932–941. (j)
Rising, T. W.; Claridge, T. D.; Moir, J. W.; Fairbanks, A. J. ChemBioChem
2006, 7, 1177–1180. (k) Heidecke, C. D.; Ling, Z.; Bruce, N. C.; Moir,
J. W.; Parsons, T. B.; Fairbanks, A. J. ChemBioChem 2008, 9, 2045–2051.
(l) Rising, T. W.; Heidecke, C. D.; Moir, J. W.; Ling, Z.; Fairbanks, A. J.
Chem.—Eur. J. 2008, 14, 6444–6464. (m) Parsons, T. B.; Patel, M. K.;
Boraston, A. B.; Vocadlo, D. J.; Fairbanks, A. J. Org. Biomol. Chem. 2010,
8, 1861–1869.
’ ASSOCIATED CONTENT
S
Supporting Information. HPLC profile of enzymatic
b
transglycosylation (Figure S1); HPAEC-PAD monosaccharide
composition analysis of glycoprotein 2 (Figure S2); reverse-phase
HPLC profiles of folded and denatured RNase B (Figure S3);
hydrogen-bonding network in the structure of CRT in complex
with Glc1Man3 tetrasaccharide (Figure S4); and experimental
(9) Li, B.; Song, H.; Hauser, S.; Wang, L. X. Org. Lett. 2006, 8, 3081–
3084.
(10) Ochiai, H.; Huang, W.; Wang, L. X. J. Am. Chem. Soc. 2008,
130, 13790–13803.
14416
dx.doi.org/10.1021/ja204831z |J. Am. Chem. Soc. 2011, 133, 14404–14417

Journal of the American Chemical Society
ARTICLE
(11) Huang, W.; Li, C.; Li, B.; Umekawa, M.; Yamamoto, K.; Zhang,
X.; Wang, L. X. J. Am. Chem. Soc. 2009, 131, 2214–2223.
(12) Huang, W.; Yang, Q.; Umekawa, M.; Yamamoto, K.; Wang,
L. X. ChemBioChem 2010, 11, 1350–1355.
(13) Umekawa, M.; Li, C.; Higashiyama, T.; Huang, W.; Ashida, H.;
Yamamoto, K.; Wang, L. X. J. Biol. Chem. 2010, 285, 511–521.
(14) (a) Umekawa, M.; Huang, W.; Li, B.; Fujita, K.; Ashida, H.;
Wang, L. X.; Yamamoto, K. J. Biol. Chem. 2008, 283, 4469–4479. (b)
Wang, L. X.; Huang, W. Curr. Opin. Chem. Biol. 2009, 13, 592–600.
(15) Heng, L.; Ning, J.; Kong, F. J. Carbohydr. Chem. 2001,
20, 283–296.
(16) Zhu, Y.; Chen, L.; Kong, F. Carbohydr. Res. 2002, 337, 207–215.
(17) Matsuo, I.; Isomura, M.; Miyazaki, T.; Sakakibara, T.; Ajisaka,
K. Carbohydr. Res. 1998, 305, 401–413.
(18) Suzuki, K.; Daikoku, S.; Ako, T.; Shioiri, Y.; Kurimoto, A.;
Ohtake, A.; Sarkar, S. K.; Kanie, O. Anal. Chem. 2007, 79, 9022–9029.
(19) Matsuo, I.; Isomura, M.; Miyazaki, T.; Sakakibara, T.; Ajisaka,
K. Carbohydr. Res. 1997, 305, 401–413.
(20) Allen, J. R.; Allen, J. G.; Zhang, X. F.; Williams, L. J.; Zatorski,
A.; Ragupathi, G.; Livingston, P. O.; Danishefsky, S. J. Chem.—Eur. J.
2000, 6, 1366–1375.
(21) (a) Zhang, Y. M.; Mallet, J. M.; Sinay, P. Carbohydr. Res. 1992,
236, 73–88. (b) Geng, X.; Dudkin, V. Y.; Mandal, M.; Danishefsky, S. J.
Angew. Chem., Int. Ed. 2004, 43, 2562–2565.
(22) Noguchi, M.; Tanaka, T.; Gyakushi, H.; Kobayashi, A.; Shoda,
S. I. J. Org. Chem. 2009, 74, 2210–2212.
(23) (a) Lees, J. G.; Miles, A. J.; Wien, F.; Wallace, B. A. Bioinfor-
matics 2006, 22, 1955–1962. (b) Whitmore, L.; Wallace, B. A. Nucleic
Acids Res. 2004, 32, W668–673. (c) Lobley, A.; Whitmore, L.; Wallace,
B. A. Bioinformatics 2002, 18, 211–212.
(24) Toumi, M. L.; Go, E. P.; Desaire, H. J. Pharm. Sci. 2009,
98, 2581–2591.
(25) Kunitz, M. J. Biol. Chem. 1946, 164, 563–568.
(26) Rudd, P. M.; Joao, H. C.; Coghill, E.; Fiten, P.; Saunders, M. R.;
Opdenakker, G.; Dwek, R. A. Biochemistry 1994, 33, 17–22.
(27) (a) Kumar, T. K.; Gopalakrishna, K.; Ramakrishna, T.; Pandit,
M. W. Int. J. Biol. Macromol. 1994, 16, 171–176. (b) Ristow, S. S.;
Wetlaufer, D. B. Biochem. Biophys. Res. Commun. 1973, 50, 544–550.
(28) (a) Rodan, A. R.; Simons, J. F.; Trombetta, E. S.; Helenius, A.
EMBO J. 1996, 15, 6921–6930. (b) Cannon, K. S.; Helenius, A. J. Biol.
Chem. 1999, 274, 7537–7544.
(29) (a) Saito, Y.; Ihara, Y.; Leach, M. R.; Cohen-Doyle, M. F.;
Williams, D. B. EMBO J. 1999, 18, 6718–6729. (b) Williams, D. B. J. Cell
Sci. 2006, 119, 615–623. (c) Michalak, M.; Groenendyk, J.; Szabo, E.;
Gold, L. I.; Opas, M. Biochem. J. 2009, 417, 651–666.
(30) Ito, Y.; Hagihara, S.; Arai, M. A.; Matsuo, I.; Takatani, M.
Glycoconjugate J. 2004, 21, 257–266.
(31) Kozlov, G.; Pocanschi, C. L.; Rosenauer, A.; Bastos-Aristizabal,
S.; Gorelik, A.; Williams, D. B.; Gehring, K. J. Biol. Chem. 2010,
285, 38612–38620.
14417
dx.doi.org/10.1021/ja204831z |J. Am. Chem. Soc. 2011, 133, 14404–14417