Facile Chemoenzymatic Synthesis of O-Mannosyl Glycans

Article abstract of DOI:10.1002/anie.201803536

O Mannosylation is a vital protein modification involved in brain and muscle development whereas the biological relevance of O-mannosyl glycans has remained largely unknown owing to the lack of structurally defined glycoforms. An efficient scaffold synthe

Full text of DOI:10.1002/anie.201803536

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Facile chemoenzymatic synthesis of O-mannosyl glycans
Authors: Lei Li, Shuaishuai Wang, Qing Zhang, Congcong Chen, Yuxi
Guo, Madhusudhan Reddy Gadi, Jin Yu, Ulrika Westerlind,
Yunpeng Liu, Xuefeng Cao, and Peng Wang
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.201803536
Angew. Chem. 10.1002/ange.201803536
Link to VoR: http://dx.doi.org/10.1002/anie.201803536
http://dx.doi.org/10.1002/ange.201803536

10.1002/anie.201803536
Angewandte Chemie International Edition
COMMUNICATION
Facile chhemoenzymatic synthesis of O-mannnosyl glycans
Shuaishuai WWang,[a]§ Qing Zhang,[a]§ CongCong Chen,[b] Yuxi Guo,[a] Madhusudhan Reddy Gadi,[a] Jin
Yu,[c] Ulrika Westerlind,[c,d] Yunpeng Liu,[a] Xuefeng Cao,[a] Peng G. Wang,*[a] and Lei Li*[a]
Abstract: O-Mannnosylation is a vital protein modification involved in
brain and musclee development, whereas the biological relevance of
O-mannosyl glycans remains largely unknown, due to the lack of
structurally definned glycoforms. An efficient scaffold synthesis/
enzymatic extenssion (SSEE) strategy was thus developed to prepare
such structures, by combining gram-scale convergent chemical
synthesis of 3 sccaffolds and strictly controlled sequential enzymatic
extension catalyzzed by glycosyltransferases. Totally, 45 O-mannosyl
glycans were obbtained, covering majority of identified mammalian
structures. Subsequent glycan microarray analysis revealed fine
specificities of glyycan-binding proteins and specific anti-sera.
Gal-1,4-GlcNAc, 3SLN), Lewis X [Gal-1,4-(Fuc-1,3-)GlcNAc,
LeX], sialyl-LeX (sLeX), and the human natural killer-1 (3-sulfo-
GlcA-1,3-Gal-1,4-Glc, HNK-1) epitopes (Fig 1B).[1b] These
terminal epitopes largely diversify core M1 and M2 based
glycans and in fact, the majority of identified structures belong to
these two core types.[1b] Additionally, it was found that the
structures centeredd on core M1 and M2 account for 15% and 5%
of all brain O-linked glycans, respectively.[4] Despite the
abundance of core M1 and M2 structures, very little is known
about their functional relevance.[5] Therefore, structurally well-
defined O-mannos
y
y
l glycans not only serve as ideal standards to
identify and charaacterize such glycoforms, but also provide
unique probes to uncover their biological roles.
O-Mannosylationn is a vital protein post-translational modification
that plays essenntial roles in brain and muscle development and
normal tissue fuunction.[1] Alpha-dystroglycan (-DG) is the most
studied O-mannnosyl protein that is widely distributed in muscle
and brain tissuees. Abnormal O-mannosylation on -DG disrupts
the receptor function of dystroglycan and leads to congenital
muscular dystrophy.[2] In addition, O-mannosyl glycans have
been reported too be involved in other human diseases including
arenaviral infection, cancer and metastasis.[3]
In mammaliaan brain tissue, O-mannosyl glycans account for
up to 30% off
all O-linked glycans.[4] Regardless of the
abundance, studdies on mammalian O-mannosylation have been
mainly focused on -DG. Recent advances in glycoproteomics
enabled discovery of an increased number of O-mannosylated
Figure 1. The 4 cores of mammalian O-mannosyl glycans (A), and identified
extensions on core M1 and M2 (B). Gal, galactose; Man, mannose; GlcNAc,
N-acetylglucosamine;
GalNAc,
N-acetylgalactosamine;
Neu5Ac,
N-
5]
proteins and vaarious O-mannosyl glycans.[1b, To date, more
acetylneuraminic acid; Neu5Gc, N-glycolylneuraminic acid; Fuc,
L-fucose;
GlcA, glucuronic acid.
than 20 O-mannosyl glycans have been identified, which are
classified into four core types.[5] As illustrated in Figure 1A, core
M0, identified in cadherin family proteins,[6] represents a single
structure of one mannose (Man) residue -linked to serine (Ser)
To date, only a few O-mannosyl glycans were prepared.[7]
A
core M1 tetrasacchharide (Fig 2, M102) was firstly synthesized in
or threonine (Thr). Core M1 contains a linear 1,2-linked
N-
1999,[7c, 7d] and later two lower homologs (M100, M101). Most
acetylglucosamine (GlcNAc), whereas the branched core M2
contains both 11,2- and 1,6-linked GlcNAc-residue. Core M3 is
a phosphorylated trisaccharide, to which glycosaminoglycan-like
polysaccharidess usually are attached.[5]
recently, Cao and co-workers prepared six core M1 structures
(
M100, M101, M102, M104, M105, and M102G) employing a
chemoenzymatic approach.[7a] Glycopeptides harboring basic O-
mannosyl structures (e.g., M101,[8] M102,[9] core M0,[7e] M000,[10]
The core M1 and M2 are typically extended and have been
or core M3[11]) we
of O-mannosyl gly
r
r
e also generated. Nevertheless, the majority
identified with structures including
N-acetyllactosamine (Gal-
ccans (especially core M2 branched structures)
1,4-GlcNAc, LN), 3-sialyl-N-acetyllactosamine (Neu5Ac-2,3-
are still not accessible, which hampered in-depth understanding
of O-mannosylation. Here an efficient strategy was developed
[a]
S, Wang; Q, Zhang; Y, Guo; Dr. M, Gadi; Dr. Y, Liu; Dr. X, Cao;
Prof. Dr. P.GG. Wang, and Dr. L. Li
Department of Chemistry & Center for Diagnostics & Therapeutics,
Georgia Statte University, Atlanta, GA 30303 (USA)
E-mail: lli22@@gsu.edu (L. Li) or pwang11@gsu.edu (P.G. Wang)
C. Chen
National Glyycoengineering Research Center, School of
Pharmaceuttical Science, Shandong University, Jinan 250012, China
Dr. J Yu and Prof. Dr. U. Westerlind
Leibniz-Instittut für Analytische Wissenschaften - ISAS - e.V.
44227, Dortmmund, Germany
Prof. Dr. U. Westerlind
Department of Chemistry, Umeå University,
901 87 Umeeå, Sweden
S. Wang andd Q. Zhang contributed equally to this work, list in
alphabetical order by last name.
employing scaffoldd synthesis/enzymatic extension (SSEE) to
prepare 45 structurally well-defined core M1 and M2 based O-
mannosyl glycans (Fig 2), covering all identified glycoforms with
the exception of the HNK-1 epitope.[1b]
Considering the structural signatures of O-mannosyl glycans,
it is concluded that all core M1 and M2 structures can be
enzymatically elaborated from the following three scaffolds,
[b]
[c]
[d]
§
M100
, M201 and M301 (Fig 2). We envisioned that the
protected Man derivative 1 (Fig 3A) would serve as a versatile
precursor for the synthesis of these scaffolds. It has a free
hydroxyl group at the C-2 position and the 4,6-hydroxyl groups
are protected as the benzylidene acetal. This arrangement
allows performing chemical glycosylation at the C-2 position
followed by the C-
protecting group.
66 position with the deprotection of the acetal
Supporting information for this article is given via a link at the end of
the documennt.
This article is protected by copyright. All rights reserved.

10.1002/anie.201803536
Angewandte Chemie International Edition
COMMUNICATION
Figure 2. The 45 core M1 and M2 O-mannosyl glycans prepared in this study. The three scaffold structures were squared. TF, Fmoc protected threonine.
As shown in Figure 3, the three scaffolds required for the
enzymatic extension were synthesized in a convergent approach
C-2 position of acceptor
position Then coupling of the obtained oligosaccharide
thiosglycoside 18 with amino acid gave the desired
glycosylated-amino acid 19 in a total yield of 29% over 11 steps.
1 followed by coupling of 2 to the C-6
.,
and were assembled by three simple building blocks,
(please refer Supporting Information I for synthetic details) and
the Fmoc-Thr(OH)-O Bu . To prepare the glyco-amino acid
M100, donor 2 was chemically glycosylated to the glycosyl
acceptor 1 at the C-2 position to obtain the disaccharide in
85% yield. An -phenyltrifluoroacetimidate donor was used
1
,
2
and
3
4
t
4
OAc
O
AcO
HO
OAc
O
OAc
O
OH
O
Ph
O
5
AcO
AcO
O
O
AcO
OC(NPh)CF3
OC(NPh)CF3
FmocHN
CO2tBu
A
B
AcO
PMBO
AcO
T
rocHN
TrocHN
N
SPh
1
2
3
4
R1
instead of the typical trichloroacetimidate, as the latter can
generate over 50% of a glucosamine thioether by-product
through thioether migration. Typical deprotection and protection
strategies were employed to convert the benzylidene acetal, 4-
methoxybenzyl ether (PMB), and trichloroethoxycarbonyl (Troc)
O
O
AcHN
R1
R1
OAc
R1
R1
OAc
O
O
R1
AcO
AcO
O
O
O
O
O
AcO
AcO
OAc
AcO
AcO
R1
O
O
R1
R1
R4
R1
R1
AcO
O
AcO
O
O
TrocHN
AcHN
+4
+3
a
O
HO
HO
AcO
O
R2
AcO
AcO
O
c
O
R4
SPh
SPh
9
FmocHN
: R1 = OAc, R5 = CO2tBu
R5
: R2 = OH, R4 = NHTroc
: R2 = OAc, R4 = NHAc
10
11
12
13
protecting groups on
three steps, with a total yield of 80%. The obtained thiophenyl
glycosyl donor 6 was coupled to the protected threonine amino
acid 4 at 0 °C in the presence of -iodosuccinimide (NIS) and
AgOTf to obtain the protected glyco-amino acid derivative in
excellent yield. The -butyl ( Bu) and Ac groups were then
removed successively, followed by reintroduction of the partially
cleaved Fmoc to provide the final product M100) in gram-
5
to acetyl (Ac) groups to afford
6
over
f
d
(M301)
H
: R1 = OH, R5 = CO2
e
R1
O
O
Ac
OAc
R1
R1
O
O
AcHN
R1
R1
6
O
O
AcO
AcO
AcO
AcO
O
O
Tro
cc
HN
AcHN
N
OH
O
AcO
AcO
AcO
O
Ph
Ph
O
+4
c
O
+2
a
O
O
O
O
R1
7
PMBO
PMB
O
b
O
SPh
SPh
SPh
5
6
1
t
t
FmocHN
R5
: R1 = OAc, R5 = CO2tBu
7
8
d
a
: R1 = OH, R5 = CO2H (M100)
+3
8
(
R1
R1
R1
amounts with a total yield of 90% over three steps.
OAc
O
AcO
O
OAc
OAc
O
O
AcO
AcO
R1
OAc
O
O
R1
O
O
O
R1
To synthesize M301, the PMB group on the disaccharide
5
AcO
O
AcHN
O
AcO
R1
R1
AcO
O
AcO
O
R4
AcO
+4
c
+2
AcO
AcO
AcO
O
R1
R1
TrocHN
R1
was initially converted to Ac, followed by deprotection of the
benzylidene acetal to achieve the diol . Glycosylation of with
the glycosyl donor yielded the protected target tetrasaccharide
O
O
AcHN
R1
R2
R3
O
R2
AcO
a
O
O
R4
9
9
O
SPh
SPh
3
FmocHN
: R1 = OAc, R5 = CO2tBu
R5
: R1, R2 = Benzylidene, R3
: R1, R2 = OH, R3 = OAc
=
O
O
PMB
: R2 = OH, R4 = NHTroc
: R2 = OAc, R4 = NHAc
18
19
14
15
16
17
d
e
f
10 in 79% yield. Successively, the Troc protecting groups were
converted to Ac groups and provided compound 11, which was
: R1 = OH, R5 = CO2H (M201)
used as glycosyl donor for coupling with the amino acid
obtain the tetrasaccharide 12 in 89% yield. Complete
deprotections of Bu and Ac groups were performed under
standard reaction conditions and reprotection of the partially
Fmoc deprotected amine afforded gram-amounts of compound
13 (M301), in a total yield of 85% over three steps. Similarly, 19
4 to
Figure 3. The building blocks (A) for the assembly of the 3 O-mannosyl
scaffold structures, and the synthetic scheme (B). Reagents and conditions:
(a) TMSOTf, DCM, -60 °C, 5: 85%; 10: 79%; 14: 75%; 16: 82%; (b) (i) TFA,
DCM; (ii) Zn, AcOH; (iii) Ac2O, Py, 80% over three steps; (c) NIS, AgOTf,
t
DCM/Et2O=1:1, 0 °C, 7: 93%; 12: 89%; 18: 83%; (d) (i) TFA, DCM; (ii) NaOMe,
MeOH; (iii) FmocOSu, NaHCO3, Acetone/H2O=3:1, 8: 90%; 13: 85%; 19: 87%;
(e) (i) DDQ, DCM/PBS Buffer=9:1; (ii) Ac2O, Py; (iii) EtSH, TsOH, DCM, 9:
77%; 15: 81%; (f) (i) Zn, AcOH; (ii) Ac2O, Py, 11: 87%; 17: 82%.
(M201) was prepared, by coupling of disaccharide donor
3 to the
This article is protected by copyright. All rights reserved.

10.1002/anie.201803536
Angewandte Chemie International Edition
COMMUNICATION
Figure 4. Enzymatic extension of core M1 (A) and core M2 (B) structures. (a) NmLgtB, UDP-Gal, Mg2+; (b) PmST1-M144D, NmCSS, Neu5Ac, CTP, Mg2+; (c)
Pd26ST, NmCSS, Neu5Ac, CTP, Mg2+; (d) Hp3FT, GDP-Fuc, Mg2+; (e) PmST1-M144D, NmCSS, Neu5Gc, CTP, Mg2+; (f) Pd26ST, NmCSS, Neu5Gc, CTP, Mg2+
;
(g) GlcAT-P, UDP-GlcA, Mg2+; (h) PmST1m, NmCSS, Neu5Ac, CTP, Mg2+; (i)
GalD.
To note the Fmoc was reintroduced to facilitate product
detection and purification, as Fmoc has both UV (260 nm) and
fluorescence (Ex. 260 nm, Em. 310 nm) absorbance.[12] The
bulky hydrophobic group also enhanced retention and
separation of hydrophilic O-mannosyl glycans on reversed
phase chromatography. Herein, HPLC was used to detect, purify
and quantify enzymatically extended O-mannosyl glycans
a new peak (TR = 21.71 min) was observed, of which the area
underneath increased while the peak corresponding to M100 (TR
= 23.32 min) became smaller over time. After over 90%
conversion (monitored by HPLC with a 4.6 × 250 mm Inertsil
ODS-4 column), M101 was isolated by HPLC using a semi-
preparative column (Inertsil ODS-4, 20 × 250 mm). The purified
M101 was then lyophilized, and further extended to afford M102
(Supporting Information II).
and M104
,
catalyzed by PmST1-M144D and Hp3FT,
Four robust bacterial GTs and
glucuronyltransferase (GlcAT-P) were used to extend the
scaffold M100 to eight core M1 O-mannosyl glycans (Fig 4A).
a
human

1,3-
respectively (Fig 4). To make glycosylation reactions more
efficient and less costly, the one-pot two-enzyme (OPTE)[18]
approach was adopted when adding sialic acid residues. For
example, in the PmST1-M144D-catalyzed reaction to generate
M102, Neu5Ac, cytidine 5'-triphosphate (CTP), MgCl2, and N.
meningitidis CMP-sialic acid synthetase (NmCSS)[19] were
added to achieve in situ generation of the sugar donor CMP-
Neu5Ac. Finally, M105 was prepared from M104 via the same
OPTE approach. Employing a bacterial 2,6-sialyltransferase
(Pd26ST), we were also able to generate a not yet identified
The bacterial GTs include,

1,4-galactosyltransferase from
2,3-
Neisseria meningitidis (NmLgtB),[13] mutant M144D of

sialyltransferase 1 from Pasteurella multocida (PmST1-M144D)
with decreased donor hydrolysis plus reduced sialidase
activity,[14]

2,6-sialyltransferase from Photobacterium damselae
(Pd26ST),[15] and C-terminal 66 amino acids truncated

1,3-
fucosyltransferase from Helicobacter pylori (Hp3FT) with
increased solubility.[16] These GTs are highly active and
recognize minimal motifs (e.g., GlcNAc for NmLgtB, LN for
Pd26ST and Hp3FT, LN or LeX for PmST1-M144D) on
oligosaccharide acceptors, and have been previously applied for
preparing complex glycans.[17]
structure, M103
, harboring the 6-sialyl-N-acetyllactosamine
(Neu5Ac- 2,6-Gal-1,4-GlcNAc, 6SLN) epitope (Fig 4), which

may serve as an ideal standard for mining such structures.
Core M1 glycans terminated with N-glycolylneuraminic acid
(Neu5Gc) (e.g. 3SGcLN, Fig 1) were also observed on animal -
As illustrated in Figure 4
,
M101 was prepared by the
DG.[4, 20] Structures with a Neu5Gc residue (M102G and M103G
)
NmLgtB-catalyzed reaction, containing M100 (10 mM), uridine
5'-diphosphogalactose (UDP-Gal) (15 mM), MgCl2 (10 mM), and
an appropriate amount of NmLgtB. After overnight incubation, an
were thus prepared, using PmST1-M144D- and Pd26ST-
catalyzed OPTE system similar to the synthesis of M102 and
M103. The HNK-1 epitope is a unique sulfated trisaccharide that
highly expressed in the nervous system and plays critical roles
in neuronal plasticity and diseases.[21] We optimized the codon
m
/
z
peak of 867.2972 was observed on the ESI mass spectrum,
corresponding to M101 [M-H]-. Meanwhile, on the HPLC profile,
This article is protected by copyright. All rights reserved.

10.1002/anie.201803536
Angewandte Chemie International Edition
COMMUNICATION
and cloned the -glucuronosyltransferase gene (GlcAT-P) for
heterogeneous expression in E. coli. Even though pure soluble
proteins were not obtained, we were able to synthesize a
in a strictly controlled sequential manner. The synthesis of
M2XX is illustrated in Figure 4B. Such synthetic routes were
designed according to substrate specificities of corresponding
GTs to avoid undesirable glycosylation. With M234 as example,
the scaffold M201 was first fucosylated to form M204, followed
by galactosylation on the 1,6-branch to provide M214. Such a
synthetic sequence eliminates the addition of Fuc onto the 1,6-
branch, as Hp3FT requires an LN disaccharide motif for its
activity.[16] Lastly, M234 was formed by the Pd26ST-catalyzed
reaction, which attaches a Neu5Ac residue selectively onto the
terminal Gal on the 1,6-branch, as 1,3-fucosylation prevents
Pd26ST-catalyzed 2,6-sialylation[22] on the 1,2-branch. The
synthetic scheme of M3XX is shown in Figure S1. It should be
noted that M224 and M324 were synthesized from M214 and
precursor structure of the HNK-1 epitope on core M1 (Fig 4
,
M106 using GlcAT-P-containing cell lysate Supporting
)
(
Information II). Research in obtaining a 3-sulfotransferase for
generating HNK-1 epitope is undergoing.
Core M2 O-mannosyl glycans can be classified into
symmetric (with same motifs on both 1,2- and 1,6-branches,
e.g., M0X0) and asymmetric (with different motifs on the
branches, e.g., M2XX and M3XX) structures. To generate
symmetric structures, chemically prepared M201 was firstly
galactosylated by NmLgtB to afford M010. Similar as described
for the synthesis of core M1 glycans, M020, M030 and M040
were then prepared starting from M010 via reactions catalyzed
by PmST1-M144D, Pd26ST, and Hp3FT, respectively. On the
other hand, M050 was prepared starting from M040 via the
PmST1-M144D-catalyzed reaction (Fig 4B). Specifically, M000
was generated by treating M201 with a -galactosidase from
Streptococcus pneumoniae (GalD).[20]
The synthesis of asymmetric core M2 O-mannosyl glycans
was performed starting with M201 or M301, by enzymatic
extension of the Gal-containing branch first and then the other,
M314 (Fig 4B, Fig S1) by using mutant E271F/R313Y of
PmST1 (PmST1m) instead of PmST1-M144D, which greatly
prefers the LN disaccharide over the LeX trisaccharide,[23] to
avoid undesired sialylation. Collectively, all 36 possible
combinations of core M2 glycans that harboring LN, 3SLN,
6SLN, LeX or sLeX motifs were prepared. These glycans were
purified by HPLC, and characterized by ESI/MALDI-MS & NMR
to confirm the structures (Supporting Information IV
& V).
Figure 5. Microarray analysis and evaluation of binding to spotted O-mannosyl glycans against Ricinus communis lectin I and Erythrina crystagalli lectin (A), anti-
CD15s antibody (B), and core M2 polyclonal rabbit sera 26560 (C). The x-axis shows the glycans and the y-axis shows the relative fluorescence, readout by (A)
Cy3-streptavidin, (B) goat anti-mouse IgG-Alexa Fluor 647 conjugate, and (D) goat anti-rabbit IgG-Alexa Fluor 647 conjugate. NC, printing buffer; PC1, Biotin-
PEG2-Amine; PC2, Rabbit IgG; M, human IgG-Cy3 conjugate and human IgG Alexa647 conjugate.
This article is protected by copyright. All rights reserved.

10.1002/anie.201803536
Angewandte Chemie International Edition
COMMUNICATION
The synthesized O-mannosyl glycans are well-defined and
closely related glycoforms (Fig 2), providing unique probes for
mining fine specificities of glycan-binding proteins (GBPs). As
Experimental Section
Detailed synthetic procedures, enzymatic reactions, microarray analysis,
nuclear magnetic resonance (NMR) spectroscopy and mass
spectrometry data, and NMR spectra for products are available in the
shown in Figure 5A
,
microarray analysis
(Supporting
Information III) showed that both Ricinus communis lectin I
(RCA-I) and Erythrina cristagalli lectin (ECA) strongly bound to
M010, consistent with its primary specificity towards terminal LN
supporting informationn.
epitope.[24] Moreover, ECA exhibited
a
broader specificity
towards all O-mannosyl glycans harboring a free terminal LN
M101 M201 M21X M301, and M31X), whereas RCA-I seems
to prefer terminal LN on the 1,6-branch (M301 and M21X) over
the 1,2-branch (M101 M201 M314 and M215) (Fig 5A). Such
Acknowledgements
(
,
,
,
The work was supported by National Institutes of Health
(U01GM116263 to P. G. Wang and L. Li). We are grateful to Z
Biotech LLC (Aurora, CO) for printing glycan microarrays
(supported by National Institute of Health under R43GM123820).
Mab(IIH6) was a kind gift from Dr. Kevin Campbell (HHMI,
University of Iowa). We thank Dr. Xi Chen from University of
California, Davis, for providing sialyltransferases. U. Westerlind
is grateful to the Ministerium für Kultur und Wissenschaft des
Landes Nordrhein-Westfalen and the Bundesministerium für
Bildung und Forschung.
,
,
a branch preference was also found for the anti-CD15s antibody
(specific to sLeX epitope), which bound to glycans that contain
sLeX on the 1,6-branch (M050
only on the 1,2-branch (M105
,
M3X5) but not to that with sLeX
, M2X5) (Fig 5B). On the other
hand, Aleuria aurantia lectin (AAL, specific to -Fuc) exhibited a
preference towards the 1,2-branch, as well as other fine
specificities (Supporting Information III, Fig S2).
ConA, an -Man specific lectin commonly used for enriching
tryptic O-mannosyl peptides, strongly bound to M100 but not to
other natural core M1 or any core M2 structures (Fig S3). It thus
can be speculated that detection or enrichment of O-mannosyl
glycans using ConA may miss a substantial amount of complex
structures. Interestingly, weak bindings of ConA to unnatural
Keywords: carbohydrates • chemoenzymatic synthesis •
glycosylation • microarrays • O-mannosyl glycans
[1]
[2]
a) T. Endo, J. Biochem. 2015, 157, 1; b) J. L. Praissman, L. Wells,
Biochemistry 2014, 53, 3066.
2,6-sialylated core M1 glycans (M103
, M103G) was observed
(Fig S3), suggesting that such modification may result in
T. Yoshida-Moriguchi, T. Willer, M. E. Anderson, D. Venzke, T. Whyte,
F. Muntoni, H. Lee, S. F. Nelson, L. Yu, K. P. Campbell, Science 2013,
341, 896.
conformational changes that facilitated ConA binding.
Anti-glycan antibodies, on the other hand, may present as
better detection tools. For example, anti-serum from two rabbits
[3]
[4]
C. M. Dobson, S. J. Hempel, S. H. Stalnaker, R. Stuart, L. Wells, Cell.
Mol. Life. Sci. 2013, 70, 2849.
(26559, 26560) that immunized with a core M2 glycan (M000
)
S. H. Stalnaker, K. Aoki, J. M. Lim, M. Porterfield, M. Liu, J. S. Satz, S.
Buskirk, Y. Xiong, P. Zhang, K. P. Campbell, H. Hu, D. Live, M.
Tiemeyer, L. Wells, J. Biol. Chem. 2011, 286, 21180.
conjugate bound to broad range of core M2-containing
a
glycopeptides regardless of varied peptide sequences.[10] We
further evaluated the anti-sera toward synthesized O-mannosyl
glycans. Our results showed that both sera exhibit binding
specificities towards core M2 O-mannosyl glycans that
[5]
[6]
M. O. Sheikh, S. M. Halmo, L. Wells, Glycobiology 2017, 27, 806.
M. B. Vester-Christensen, A. Halim, H. J. Joshi, C. Steentoft, E. P.
Bennett, S. B. Levery, S. Y. Vakhrushev, H. Clausen, Proc. Natl. Acad.
Sci. U.S.A. 2013, 110, 21018.
containing at least one free terminal GlcNAc residue (M000
M20X M30X) as well as the core M1 disaccharide M100 Fig
6C Fig S4). In addition, anti-sera from rabbit 26560 exhibited
,
[7]
a) Y. Zhang, C. Meng, L. Jin, X. Chen, F. Wang, H. Cao, Chem.
Commun. 2015, 51, 11654; b) J. Seifert, T. Ogawa, S. Kurono, Y. Ito,
Glycoconj. J. 2000, 17, 407; c) J. Seifert, T. Ogawa, Y. Ito, Tetrahedron
Lett. 1999, 40, 6803; d) I. Matsuo, M. Isomura, K. Ajisaka, Tetrahedron
Lett. 1999, 40, 5047; e) D. Live, L. Wells, G. J. Boons, Chembiochem
2013, 14, 2392.
,
(
,
comparable bindings to all those glycans, whereas anti-sera
from rabbit 26559 showed stronger binding to glycans contain a
free GlcNAc on the 1,6-branch (M20X) compared with binding
to the 1,2-branch (M100, M30X) (Fig S4). The results imply
[8]
[9]
J. Yu, U. Westerlind, Chembiochem 2014, 15, 939.
that antibodies generated from different hosts may posses
certain individual heterogeneity (or individual difference).
Nevertheless, our results revealed fine specificities of the anti-
sera towards O-mannosyl glycans. Comparing with ConA, these
anti-sera are advantageous in the detection of branched O-
mannosyl glycans.
In summary, by combining convergent chemical synthesis
with strictly programmed enzymatic synthesis in a stepwise
manner, an efficient scaffold synthesis/enzymatic extension
(SSEE) strategy was developed to access 45 mammalian O-
mannosyl glycans. Such unique glycoforms provide not only
standards for identifying O-mannosyl glycans and revealing their
biological roles, but also ideal probes for mining fine details of
protein-glycan interactions.
R. Sardzik, A. P. Green, N. Laurent, P. Both, C. Fontana, J. Voglmeir,
M. J. Weissenborn, R. Haddoub, P. Grassi, S. M. Haslam, G. Widmalm,
S. L. Flitsch, J. Am. Chem. Soc. 2012, 134, 4521.
[10] J. Yu, O. C. Grant, C. Pett, S. Stahl, R. J. Woods, U. Westerlind, Chem.
Eur. J. 2017, 23, 3466.
[11] S. M. Halmo, D. Singh, S. Patel, S. Wang, M. Edlin, G. J. Boons, K. W.
Moremen, D. Live, L. Wells, J. Biol. Chem. 2017, 292, 2101.
[12] S. Einarsson, B. Josefsson, S. Lagerkvist, J. Chromatogr. A 1983, 282,
609.
[13] K. Lau, V. Thon, H. Yu, L. Ding, Y. Chen, M. M. Muthana, D. Wong, R.
Huang, X. Chen, Chem. Commun. 2010, 46, 6066.
[14] G. Sugiarto, K. Lau, J. Qu, Y. Li, S. Lim, S. Mu, J. B. Ames, A. J. Fisher,
X. Chen, ACS Chem. Biol. 2012, 7, 1232.
[15] H. Yu, S. Huang, H. Chokhawala, M. Sun, H. Zheng, X. Chen, Angew.
Chem. Int. Ed. 2006, 45, 3938.
[16] S. W. Lin, T. M. Yuan, J. R. Li, C. H. Lin, Biochemistry 2006, 45, 8108.
This article is protected by copyright. All rights reserved.

10.1002/anie.201803536
Angewandte Chemie International Edition
COMMUNICATION
[17] a) L. Li, Y. Liu, C. Ma, J. Qu, A. D. Calderon, B. Wu, N. Wei, X. Wang,
Y. Guo, Z. Xiao, J. Song, G. Sugiarto, Y. Li, H. Yu, X. Chen, P. G.
[21] J. Morise, H. Takematsu, S. Oka, Biochim. Biophys. Acta 2017, 1861,
2455.
Wang, Chemm. Sci. 2015
,
6
, 5652; b) Z. Wu, Y. Liu, C. Ma, L. Li, J. Bai,
[22] W. W. Unger, C. T. Mayer, S. Engels, C. Hesse, M. Perdicchio, F.
Puttur, I. Streng-Ouwehand, M. Litjens, H. Kalay, L. Berod, T.
L. Byrd-Leotis, Y. Lasanajak, Y. Guo, L. Wen, H. Zhu, J. Song, Y. Li, D.
A. Steinhauer, D. F. Smith, B. Zhao, X. Chen, W. Guan, P. G. Wang,
Sparwasser, Y. van Kooyk, Oncoimmunology 2015
[23] G. Sugiarto, K. Lau, Y. Li, Z. Khedri, H. Yu, D. T. Le, X. Chen, Mol.
Biosyst. 2011 , 3021.
[24] Y. Itakura, S. Nakamura-Tsuruta, J. Kominami, N. Sharon, K. Kasai, J.
Hirabayashi, J. Biochem. 2007 142, 459.
, 4, e970462.
Org. Biomol. Chem. 2016
[18] H. Yu, X. Chen, Org. Biomol. Chem. 2016
[19] H. Yu, H. Yu, R. Karpel, X. Chen, Bioorg. Med. Chem. 2004
, 14, 11106.
,
14, 2809.
, 7
,
12, 6427.
[20] S. H. Stalnaker, S. Hashmi, J. M. Lim, K. Aoki, M. Porterfield, G.
Gutierrez-Sanchez, J. Wheeler, J. M. Ervasti, C. Bergmann, M.
,
Tiemeyer, L. Wells, J. Biol. Chem. 2010, 285, 24882.
This article is protected by copyright. All rights reserved.

10.1002/anie.201803536
Angewandte Chemie International Edition
COMMUNICATION
Entry for thhe Table of Contents (Please choose one layout)
Layout 1:
COMMUNNICATION
An efficient scaffold
Shuaishuai Wang, Qing Zhang,
synthesis/enzymatic extension
(SSEE) strategy was developed to
prepare 45 structurally well-defined O-
mannosyl glycans, combining
convergent chemistry and strictly
programmed enzymatic extension
catalyzed by glycosyltransferases.
Glycan microarray analysis was also
performed to mine fine specificities of
glycan-binding proteins using these
glycoforms.
CongCong Chen, Yuxi Guo,
Madhusudhan Reddy Gadi, Jin Yu,
Ulrika Westerlind, Yunpeng Liu, Xuefeng
Cao, Peng G. Wang,* and Lei Li*
Page No. –– Page No.
Facile chemoenzymatic synthesis of
O-mannosyl glycans
This article is protected by copyright. All rights reserved.