A general strategy for the chemoenzymatic synthesis of asymmetrically branched N-glycans

Article abstract of DOI:10.1126/science.1236231

A systematic, efficient means of producing diverse libraries of asymmetrically branched N-glycans is needed to investigate the specificities and biology of glycan-binding proteins. To that end, we describe a core pentasaccharide that at potential branching positions is modified by orthogonal protecting groups to allow selective attachment of specific saccharide moieties by chemical glycosylation. The appendages were selected so that the antenna of the resulting deprotected compounds could be selectively extended by glycosyltransferases to give libraries of asymmetrical multi-antennary glycans. The power of the methodology was demonstrated by the preparation of a series of complex oligosaccharides that were printed as microarrays and screened for binding to lectins and influenza-virus hemagglutinins, which showed that recognition is modulated by presentation of minimal epitopes in the context of complex N-glycans.

Full text of DOI:10.1126/science.1236231

A General Strategy for the Chemoenzymatic Synthesis of
Asymmetrically Branched N-Glycans
Zhen Wang et al.
Science 341, 379 (2013);
DOI: 10.1126/science.1236231
This copy is for your personal, non-commercial use only.
If you wish to distribute this article to others, you can order high-quality copies for your
colleagues, clients, or customers by clicking here.
Permission to republish or repurpose articles or portions of articles can be obtained by
following the guidelines here.
The following resources related to this article are available online at
www.sciencemag.org (this information is current as of April 2, 2014 ):
Updated information and services, including high-resolution figures, can be found in the online
version of this article at:
http://www.sciencemag.org/content/341/6144/379.full.html
Supporting Online Material can be found at:
http://www.sciencemag.org/content/suppl/2013/07/24/341.6144.379.DC1.html
A list of selected additional articles on the Science Web sites related to this article can be
found at:
http://www.sciencemag.org/content/341/6144/379.full.html#related
This article cites 46 articles, 9 of which can be accessed free:
http://www.sciencemag.org/content/341/6144/379.full.html#ref-list-1
This article has been cited by 4 articles hosted by HighWire Press; see:
http://www.sciencemag.org/content/341/6144/379.full.html#related-urls
This article appears in the following subject collections:
Chemistry
http://www.sciencemag.org/cgi/collection/chemistry
Science (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by the
American Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005. Copyright
2013 by the American Association for the Advancement of Science; all rights reserved. The title Science is a
registered trademark of AAAS.

REPORTS
14. A. R. Yavari, Nat. Mater. 4, 2–3 (2005).
33. G. Tarjus, S. Kivelson, Z. Nussinov, P. Viot, J. Phys.
solving the puzzles of glass transition (34) and
may provide a method to explore the atomic struc-
ture and atomic-scale properties of disordered
materials.
Condens. Matter 17, R1143–R1182 (2005).
34. J. S. Langer, Phys. Today 60, 8–9 (2007).
15. J. F. Sadoc, R. Mosseri, Geometrical Frustration
(Cambridge Univ. Press, Cambridge, 1999).
16. D. R. Nelson, Defects and Geometry in Condensed Matter
(Cambridge Univ. Press, Cambridge, 2002).
17. D. R. Nelson, Phys. Rev. Lett. 50, 982–985 (1983).
18. A. Di Cicco, A. Trapananti, S. Faggioni, A. Filipponi,
Phys. Rev. Lett. 91, 135505 (2003).
Acknowledgments: We thank Y. Hirotsu and T. Egami for
constructive discussion and H. Sheng for providing Zr-Pt
embedded atom model potentials and J. Saida for Zr80Pt20
ribbons. This work was sponsored by “WPI Research Center
Initiative for Atoms, Molecules and Materials,” Ministry of
Education, Culture, Sports, Science and Technology of Japan;
Grants-in-Aid for Scientific Research (24360260 and 24656400),
Japan Society for the Promotion of Science; “A mathematical
challenge to a new phase of materials science,” Japan
Science and Technology Agency–Core Research for Evolutional
Science and Technology; and the National Natural Science
Foundation of China (grant 51271113).
References and Notes
1. F. C. Frank, Proc. R. Soc. London Ser. A 215, 43–46
(1952).
19. D. R. Nelson, Phys. Rev. B 28, 5515–5535 (1983).
20. T. Schenk, D. Holland-Moritz, V. Simonet, R. Bellissent,
D. M. Herlach, Phys. Rev. Lett. 89, 075507 (2002).
21. K. Saksl et al., Appl. Phys. Lett. 83, 3924 (2003).
22. K. F. Kelton et al., Phys. Rev. Lett. 90, 195504 (2003).
23. A. Hirata et al., Nat. Mater. 10, 28–33 (2011).
24. N. A. Mauro et al., Phys. Rev. B 83, 184109 (2011).
25. J. Saida et al., J. Phys. Condens. Matter 21, 375104 (2009).
26. P. J. Steinhardt, D. R. Nelson, M. Ronchetti, Phys. Rev. B
28, 784–805 (1983).
2. W. Hume-Rothery, E. Anderson, Philos. Mag. 5, 383–405
(1960).
3. J. D. Bernal, Nature 185, 68–70 (1960).
4. G. D. Scott, Nature 194, 956–957 (1962).
5. J. L. Finney, Nature 266, 309–314 (1977).
6. P. J. Steinhardt, D. R. Nelson, M. Ronchetti, Phys. Rev.
Lett. 47, 1297–1300 (1981).
7. T. Tomida, T. Egami, Phys. Rev. B 52, 3290–3308 (1995).
8. M. W. Chen, T. Zhang, A. Inoue, A. Sakai, T. Sakurai,
Appl. Phys. Lett. 75, 1697 (1999).
Supplementary Materials
www.sciencemag.org/cgi/content/full/science.1232450/DC1
Materials and Methods
Supplementary Text
Figs. S1 to S12
27. T. Egami, S. Aur, J. Non-Cryst. Solids 89, 60–74 (1987).
28. Y. Q. Cheng, E. Ma, Prog. Mater. Sci. 56, 379–473
(2011).
9. D. B. Miracle, Nat. Mater. 3, 697–702 (2004).
10. H. W. Sheng, W. K. Luo, F. M. Alamgir, J. M. Bai, E. Ma,
Nature 439, 419–425 (2006).
29. T. Kaczynski, K. Mischaikow, M. Mrozek, Computational
Homology (Springer, New York, 2003).
11. P. F. Guan, T. Fujita, A. Hirata, Y. H. Liu, M. W. Chen,
Phys. Rev. Lett. 108, 175501 (2012).
References (35–40)
30. M. D. Ediger, P. J. Harrowell, J. Chem. Phys. 137,
080901 (2012).
12. D. Ma, A. D. Stoica, X.-L. Wang, Nat. Mater. 8, 30–34
(2009).
6 November 2012; accepted 29 April 2013
Published online 11 July 2013;
10.1126/science.1232450
31. L. Berthier, G. Biroli, Rev. Mod. Phys. 83, 587–645 (2011).
13. T. Takagi et al., Appl. Phys. Lett. 79, 485 (2001).
32. H. Shintani, H. Tanaka, Nat. Phys. 2, 200–206 (2006).
The need for more efficient approaches has stim-
ulated the development of chemo-enzymatic meth-
ods in which a synthetic oligosaccharide precursor
is modified by a range of glycosyltransferases to
give more complex derivatives (10, 11). Such an
approach can, however, only provide symmetri-
cally branched oligosaccharides.
Naturally occurring branched oligosaccha-
rides often bear distinctive appendages at each
branching point (12). In this respect, the bio-
synthesis of N-linked oligosaccharides is initiated
A General Strategy for the
Chemoenzymatic Synthesis of
Asymmetrically Branched N-Glycans
Zhen Wang,1* Zoeisha S. Chinoy,1,2* Shailesh G. Ambre,1,2 Wenjie Peng,3 Ryan McBride,3
Robert P. de Vries,3 John Glushka,1 James C. Paulson,3 Geert-Jan Boons1,2
†
A systematic, efficient means of producing diverse libraries of asymmetrically branched N-glycans is needed in the endoplasmic reticulum where a dolichol-
to investigate the specificities and biology of glycan-binding proteins. To that end, we describe a core
linked Glc3Man9GlcNAc2 oligosaccharide pre-
pentasaccharide that at potential branching positions is modified by orthogonal protecting groups to allow cursor is transferred en bloc to an Asn-X-Ser/Thr
selective attachment of specific saccharide moieties by chemical glycosylation. The appendages were
sequon, where X is any amino acid, on newly
selected so that the antenna of the resulting deprotected compounds could be selectively extended by
synthesized polypeptides. Subsequent trimming
glycosyltransferases to give libraries of asymmetrical multi-antennary glycans. The power of the methodology and processing of the transferred oligosaccharide
was demonstrated by the preparation of a series of complex oligosaccharides that were printed as
microarrays and screened for binding to lectins and influenza-virus hemagglutinins, which showed
results in a GlcNAcMan3GlcNAc2 core structure,
which is transported to the Golgi where addi-
that recognition is modulated by presentation of minimal epitopes in the context of complex N-glycans. tional N-acetylglucosamine moieties (O-GlcNAc)
can be added. Subsequent conversion of the
ost cell surface and secreted proteins molecular mechanisms by which these compounds O-GlcNAc stubs into N-acetyllactosamine [bGal(1,4)
are modified by covalently linked glycans, exert their functions have been difficult to de- GlcNAc, LacNAc], provide precursors that can
M
which are essential mediators of bio- fine. The latter is due to a lack of comprehensive be elaborated by various glycosyltransferases to
logical processes such as protein folding, cell libraries of well-defined complex oligosaccharides give rise to enormous structural diversity.
signaling, fertilization, embryogenesis, and the that are needed as standards to determine exact
The biosynthesis of complex branched oligo-
proliferation of cells and their organization into structures of glycans in complex mixtures (4, 5) saccharides generally leads to positional isomers,
specific tissues (1). Overwhelming data support the and to examine specificities and biology of glycan- which are structurally difficult to assign by mass
relevance of glycosylation in pathogen recogni- binding proteins that occur in nature (6–8).
spectrometry (4, 5). Furthermore, glycan micro-
tion, inflammation, innate immune responses, and Naturally occurring glycans are typically iso- array technology has shown that terminal oligo-
the development of autoimmune diseases and can- lated in small quantities as mixtures of closely saccharide motifs of complex glycans mediate
cer (2, 3). Although the functional importance related structures that are difficult to separate, biological recognition (13). However, recent studies
of glycoprotein glycosylation is well established, and therefore do not provide a reliable source of indicate a more complex picture in which the core
well-defined oligosaccharides. Thus, it is widely structure can influence terminal glycan recognition
1Complex Carbohydrate Research Center, University of Georgia,
315 Riverbend Road, Athens, GA 30602, USA. 2Chemistry De-
partment, University of Georgia, Athens, GA 30602, USA. 3De-
partments of Cell and Molecular Biology, and Chemical
Physiology, The Scripps Research Institute, 10550 North Torrey
Pines Road, La Jolla, CA 92037, USA.
accepted that chemical- or enzymatic approaches (14). A synthetic technology that can give libraries
must be used for the preparation of diverse gly- of asymmetrically substituted glycans will make it
can libraries needed for biological and struc- possible to fabricate the next generation of glycan
tural studies (7–11). Despite ongoing progress, microarray to examine in detail glycan-protein
the chemical synthesis of complex oligosaccha- recognition, to develop algorithms for the assign-
rides remains very time consuming, especially ment of mass spectra, and to design probes for
when highly complex structures are targeted (7). elucidating pathways of glycoconjugate biosyn-
*These authors contributed equally to this work.
†Corresponding author. E-mail: gjboons@ccrc.uga.edu
www.sciencemag.org SCIENCE VOL 341 26 JULY 2013
379

REPORTS
thesis. Despite the urgent need for libraries of tion of libraries of asymmetrically branched fore it was possible to generate libraries of complex
asymmetrically branched N-glycans (15, 16), none N-glycans (Fig. 1). This pentasaccharide resem- branched bi-, tri-, and tetra-antennary structures by
of the currently available methods can produce bles the core structure common to all eukaryotic sequential removal of the protecting groups fol-
collections of such compounds, and previous syn- N-linked glycans (12) and is modified at po- lowed by chemical glycosylations using a diverse
thetic efforts have almost exclusively focused on sitions where branching points can occur with the set of glycosyl donors. Furthermore, we anticipated
the preparation of symmetrically branched com- protecting groups levulinoyl (Lev), fluorenylmethyl- that the use of LacNAc and GlcNAc donors 2 to 5,
pounds (17–23).
oxycarbonate (Fmoc), allyloxycarbonate (Alloc), followed by removal of all protecting groups ex-
We envisaged that oligosaccharide 1 would and 2-naphthylmethyl (Nap). We show here that cept the acetyl esters, would give precursor glycans
be an attractive starting material for the prepara- these protecting groups are orthogonal, and there- that at each antenna could be selectively extended
by a panel of glycosyltransferases to rapidly give
large numbers of highly complex asymmetrically
Fig. 1. Orthogonally pro-
tected core pentasaccharide
1 and glycosyl donors 2
to 5. Coupling of 1 with
these reagents in a parallel
combinatorial manner gives
oligosaccharide precursors to
enzyme substrates.
LevO
substituted N-glycans. Selective extension was ex-
pected to be feasible because many relevant glyco-
syltransferases recognize LacNAc but not GlcNAc
as a substrate (18). The latter moiety can, however,
be converted into LacNAc by enzymatic galacto-
sylation, and the resulting derivative can then be
elaborated by other glycosyltransferases. Fur-
thermore, acetylation should render LacNAc and
GlcNAc moieties inactive for enzymatic modi-
fication; however, the removal of these esters would
give an appropriate substrate for extension by
glycosyltransferases.
Some applications, such as the use of syn-
thetic glycans as standards for mass spectrometry,
require compounds having an unmodified reduc-
ing end. Other uses, such as the development of
glycan microarrays, need compounds modified
with a reactive anomeric linker. To ensure that
the glycans prepared by the chemo-enzymatic
approach can be employed for multiple pur-
OAlloc
O
BnO
BnO
OBn
OBn
OBn
O
O
BnO
O
O
O
O
O
OBn
BnO
TrocHN
BnO
NHTroc
BnO
1
NapO
O
BnO
OFmoc
OBn
OBn
BnO
BnO
OBn
O
O
BnO
AcO
O
O
BnO
BnO
TrocHN
BnO
TrocHN
OC(NPh)CF3
OC(NPh)CF3
2
4
OAc
OAc
O
AcO
AcO
OBn
O
O
O
AcO
BnO
TrocHN
AcO
TrocHN
OC(NPh)CF3
OC(NPh)CF3
5
3
R4O
A
OR3
O
BnO
BnO
1. R1 = Fmoc, R2 = Nap, R3
6. R1 = H, R2 = Nap, R3 Alloc, R4 = Lev (94%)
7. R1 = Fmoc, R2 = Nap, R3 Alloc, R4 = H (85%)
=
Alloc, R4 = Lev
OBn
OBn
OBn
O
=
O
BnO
O
O
=
O
O
OBn
NHTroc
BnO
TrocHN
8. R1 = Fmoc, R2 = Nap, R3 = H, R4 = Lev (96%)
O
BnO
9. R1 = Fmoc, R2 = H, R3
=
Alloc, R4 = Lev (86%)
BnO
R2O
BnO
O
R1O
LevO
RO
OAlloc
OAlloc
B
O
O
BnO
BnO
BnO
BnO
OBn
OBn
OBn
O
OBn
OBn
OBn
O
O
O
2, TfOH, DCM
3, TfOH, DCM
-60oC (94%)
BnO
O
O
BnO
O
O
O
O
O
O
OBn
NHTroc
-60oC (97%)
O
BnO
OAc
O
O
OBn
NHTroc
BnO
TrocHN
AcO
AcO
BnO
TrocHN
BnO
6
OBn
O
BnO
O
BnO
RO
BnO
OBn
O
O
O
BnO
BnO
OAc
BnO
TrocHN
10. R = Nap
11. R = H
12. R = Lev H2NNH2, AcOH
DDQ, DCM,
H2O (76%)
OBn
O
BnO
BnO
13. R = H
OBn
O
DCM (85%)
OBn
O
O
O
O
O
BnO
BnO
O
OBn
BnO
TrocHN
OBn
TrocHN
HO
BnO
BnO
BnO
O
i) Pd(PPh3)4,
ii) Zn, AcOH
iii) Ac2O, MeOH
iv) Pd(OH)2, H2
O
HO
O
OH
O
OAlloc
O
HO
O
HO
BnO
TrocHN
AcHN
4, TfOH, DCM
-60oC (78%)
HO
BnO
OH
OBn
OH
OH
O
OBn
OBn
O
O
O
(46%, 4 steps)
HO
O
BnO
O
O
O
O
O
O
O
OAc
OAc
O
O
OBn
NHTroc
AcO
AcO
HO
AcO
AcO
BnO
OH
HO
OBn
BnO
AcHN
TrocHN
AcHN
HO
BnO
O
O
OH
O
O
O
O
O
O
O
O
HO
BnO
14
15
HO
BnO
OAc
HO
OAc
BnO
AcHN
TrocHN
OBn
OH
OH
OBn
O
O
O
O
O
O
HO
O
BnO
O
HO
BnO
OBn
OH
AcHN
TrocHN
Fig. 2. Chemical synthesis of decasaccharide 15 for branch-specific enzymatic extensions. (A) Selective removal of temporary protecting groups. (B)
Preparation of glycan precursor for enzymatic extension.
380
26 JULY 2013 VOL 341 SCIENCE www.sciencemag.org

REPORTS
poses, the anomeric center of compound 1 was Alloc protecting group providing the correspond- with 4 to give fully protected decasaccharide 14.
protected as a benzyl glycoside. This protecting ing hydroxyl 8, and oxidation with 2,3-dichloro- Partial deprotection of 14 to give target com-
group will be removed during the deprotection 5,6-dicyano-1,4-benzoquinone (DDQ) resulted in pound 15 was accomplished by cleavage of the
stage to give glycans having an unmodified re- the removal of the Nap ether to give 9 in high yield. Alloc carbonate with Pd(PPh3)4 followed by re-
ducing end. The latter type of compound can, how-
Having demonstrated the orthogonality of the moval of the 2,2,2-trichloroethyoxycarbamate
ever, easily be derivatized by a reactive anomeric temporary protecting groups, we focused on the (Troc) groups with Zn in acetic acid, acetylation
linker by reaction with an appropriate reagent preparation of tri-antennary oligosaccharide 15, of the resulting free amines with acetic anhydride,
such as 2-[(methylamino)oxy]ethanamine (24).
which was expected to be an appropriate precur- and catalytic hydrogenolysis of the benzyl ethers.
Pentasaccharide 1 was readily assembled from sor for branch-specific enzymatic modification Detailed nuclear magnetic resonance (NMR)
appropriately protected monosaccharide building (Fig. 2). Glycosyl acceptor 6 was coupled with 2 analysis of 15 showed that the acetyl esters were
blocks (fig. S2). The Fmoc group of 1 could be by using trifluoromethanesulfonic acid (TfOH) still intact, and thus a compound was obtained that
selectively removed by the non-nucleophilic base (25, 26) as the promoter to give heptasaccharide has characteristic saccharide appendages at each
triethylamine to give 6, whereas treatment with the 10. The Nap ether of 10 was removed by oxida- antenna, allowing selective modification by a
nucleophilic base hydrazine acetate led to cleavage tion with DDQ, and the resulting acceptor 11 was panel of glycosyltransferases.
of the Lev ester to provide 7 without affecting the glycosylated with 3 to provide nonasaccharide
In addition to compound 15, pentasaccharide
other base-sensitive protecting groups (Fig. 2A). 12. Next, the Lev ester of 12 was cleaved with 1 is an appropriate starting material for the
Treatment of 1 with Pd(PPh3)4 affected only the hydrazine acetate to give 13, which was coupled chemical synthesis of other bi-, tri-, and tetra-
Fig. 3. Chemoenzymatic synthesis of complex oligosaccharides from (see figs. S14 and S15 for synthetic intermediates). N-Acetyl neuraminic acid
15. (A) Synthetic route to asymmetrically substituted multi-antennary glycan (Neu5Ac, diamonds); D-galactose (Gal, circles); N-acetyl-D-glucosamine (GlcNAc,
22. (B) Structures of compounds 23 to 26 prepared by an analogous approach squares); D-mannose (Man, circles); L-fucose (Fuc, triangles).
www.sciencemag.org SCIENCE VOL 341 26 JULY 2013
381

REPORTS
antennary precursor oligosaccharides by chang- clusively compound 16. Next, the acetyl esters biological or biophysical studies. The precursor
ing the number and sites of attachment of the of 16 were removed by treatment with aqueous oligosaccharide 15 is an attractive starting material
appendages (2 to 5). For example, a positional ammonia to give compound 17, which then for the preparation of many other highly complex
isomer of 15 was readily prepared by the sequential had an unmasked LacNAc moiety at the C-4 of glycans. To illustrate this feature, we prepared
removal of the Fmoc, Alloc, and Lev groups of the Man-a3 arm that was expected to be avail- compounds 23 to 27 (figs. S14 and S15), which
1 and glycosylations with glycosyl donors 2, 3, able for enzymatic transformations. Indeed, are asymmetrical and have varying numbers of
and 4, respectively (fig. S3).
fucosylation of 17 with a1,3-fucosyltransferase 2,3- or 2,6-linked sialic acids at the various anten-
The precursor oligosaccharide 15 was further (a3FucT) (30) resulted in the modification of nae (27). Thus, subsequent deacetylation and
extended by glycosyltransferases to demonstrate the LacNAc and sialyl-LacNAc moieties to give bis-fucosylation of 15 to give Lex moieties at the
the possibility of selective modification of each bis-fucosylated derivative 18. The GlcNAc moiety b2 and b4 arm were followed by galactosylation
antenna to form highly complex asymmetrically at the C-6 antenna of 18 was converted into a to form a LacNAc moiety at the b6 arm that was
branched N-glycans (Fig. 3). Many human N-glycans LacNAc moiety by using b1,4-galactosyltransferase capped with 2,6-Neu5Ac to form 23 or further
contain terminal sialic acids either exclusively (GalT-1), uridine 5′-diphosphogalactose (UDP-Gal), extended with 2,6-Neu5Ac-LacNAc to provide
a(2,3)- or a(2,6)-linked to N-acetyllactosamine and CIAP to give 19. Treatment of 19 with b1,3-N- 24. Similarly, compounds 25 to 27 were synthe-
or a combination of these two linkages (27). Fur- acetylglucosaminyltransferase (b1,3GlcNAcT) sized by either bis-a(2-3) (to give 27) or bis-a(2-6)-
thermore, Lewis antigens such as Lewisy (Ley), (31), UDP-GlcNAc, and CIAP resulted in a selec- sialylation, followed by extension of the b6 arm to
Lex, and sialyl Lewisx (SLex) are found on many tive addition of a b(1,3)-linked GlcNAc moiety provide 25 and 26 (fig. S15).
biologically important glycans. Therefore, we to the LacNAc moiety of the b1-6 branch to give
It was anticipated that compounds 22 to 27
focused on the preparation of heptadecasaccha- 20. The Lex moiety of 19 was unaffected, high- would be useful for examining the activity of the
ride 22, which has SLex and Lex appendages lighting the feasibility of exploiting inherent sub- various biologically relevant glycan epitopes in
at the C-2 and C-4 arm, respectively, and a strate specificities of glycosyltransferases for the the context of their presence on multiantennary
di-LacNAc moiety extended by a(2,6)-linked selective modification of multi-antennary gly- asymmetric structures. Thus, a glycan microarray
sialoside at the C-6 arm. A key aspect of this cans. The b1,6-branch was further extended by was constructed composed of the asymmetrical
strategy is that relatively few glycosyltransferases GalT-1 and a2,6-sialyltransferase (ST6Gal-1) to tri-antennary glycans (22 to 27) and previously
are needed to elaborate these terminal glycan provide target compound 22, which has distinc- prepared linear and bi-antennary glycans having
sequences, and enzyme expression systems that tive oligosaccharide appendages at each of the a terminal b(1-4)Gal (A to D), a(1-3)-Fuc (E and
produce these and many other mammalian and three antennae.
bacterial glycosyltransferases useful in chemo- After each step, the product was purified by (M to Q) moiety (table S14). Compounds 22 to 27
enzymatic synthesis have already been described size exclusion chromatography and the resulting were modified with an amino-containing linker by
(28, 29). compound fully characterized by NMR and mass treatment with 2-[(methylamino)oxy]ethanamine
F), a(2-6)-Neu5Ac (G to L), or a(2-3)-Neu5Ac
The LacNAc moiety of decasaccharide 15 spectrometry of the permethylated derivative. If (24), and the resulting derivatives were printed on
was sialylated by a2,3-sialyltransferase (ST3Gal- any starting material was observed, the compound N-hydroxysuccinimide (NHS)–activated glass
IV), cytidine-5′-monophospho-N-acetylneuraminic was resubjected to the enzyme until a homoge- slides with the reference compounds (32).
acid (CMP-Neu5Ac), and calf intestine alkaline neous product was obtained. In addition to target
Probing the array with the Erythrina crystagalli
phosphatase (CIAP), and as expected, only one compound 22, each intermediate of the enzymatic agglutinin (ECA) specific for terminal LacNAc
of the three antennae was modified to give ex- extension (17 to 21) can in principle be used for sequences detected the corresponding reference
Fig. 4. Glycan microarray binding analyses. Fluorescently labeled lectins Shown is the mean signal and standard error calculated for six indepen-
(ECA, AAL, and SNA), and recombinant avian (VN/04) and human influ- dent replicates on the array. Structures of each of the lettered glycans are
enza A (KY/07 and CA/05) HA were assessed for binding to the array. found in table S14.
382
26 JULY 2013 VOL 341 SCIENCE www.sciencemag.org

REPORTS
18. S. Hanashima, S. Manabe, Y. Ito, Angew. Chem. Int.
compounds A to D and compounds I and J; two this motif is also present in triantennary glycans
biantennary compounds that have one branch 22 and 24, which are not recognized by this HA.
modified with a LacNAc structure (Fig. 4). Of the Compounds L and 26 also have in common at
synthetic triantennary compounds, ECA lectin least two Neu5Aca(2–6) epitopes on different
bound strongly to 25 and weakly to 22 to 24. The antennae, but so do compounds K and 25, which
latter compounds contain LacNAc substituted have a single LacNAc extension and are not recog-
with a fucoside, which is known to reduce the af- nized. These results reflect differences in the spec-
finity of ECA (33). By contrast, the fucose-specific ificity of these HAs, and not simple differences
Aleuria aurantia lectin (AAL) robustly recognized in avidity, because similar array results were ob-
the fucoside containing glycans 22 to 24 as well tained when the concentration of the HA applied
as the three reference compounds containing a to the array was titrated down in twofold dilu-
Lex epitope (E, F, and M). Sambuccus nigra ag- tions from 100 to 6 mg/ml (fig. S24).
Ed. Engl. 44, 4218–4224 (2005).
19. S. Jonke, K. G. Liu, R. R. Schmidt, Chemistry 12, 1274–1290
(2006).
20. B. Sun, B. Srinivasan, X. F. Huang, Chemistry 14, 7072–7081
(2008).
21. C. Unverzagt et al., Chemistry 15, 12292–12302 (2009).
22. S. Serna, J. Etxebarria, N. Ruiz, M. Martin-Lomas,
N. C. Reichardt, Chemistry 16, 13163–13175 (2010).
23. M. A. Walczak, S. J. Danishefsky, J. Am. Chem. Soc. 134,
16430–16433 (2012).
24. O. Bohorov, H. Andersson-Sand, J. Hoffmann, O. Blixt,
Glycobiology 16, 21C–27C (2006).
25. R. R. Schmidt, W. Kinzy, Adv. Carbohydr. Chem. Biochem.
50, 21–123 (1994).
glutinin (SNA) specific for terminal a(2-6)Neu5Ac
recognized all structures containing this epitope topes presented on asymmetrically branched
(G to L and 22 to 26). N-linked glycans can be distinguished from the
These results demonstrate that glycan epi-
26. B. Yu, H. C. Tao, Tetrahedron Lett. 42, 2405–2407
(2001).
27. G. Spik, V. Debruyne, J. Montreuil, H. van Halbeek,
J. F. Vliegenthart, FEBS Lett. 183, 65–69 (1985).
28. More then 800 expression constructs of human
glycosylation enzymes in multiple expression vectors and
protocols to produce the enzymes for biochemical,
enzymatic synthesis, and structural studies are available
at http://glycoenzymes.ccrc.uga.edu.
Influenza viruses recognize sialic acids as same epitopes on linear or symmetrically branched
receptors, and it is well documented human and glycans. Such context-dependent recognition can
avian viruses exhibit differential specificity for be due to extended binding sites, unfavorable in-
glycans with Neu5Aca(2-6)Gal and Neu5Aca teractions by neighboring antennae, and multiva-
(2-3)Gal linkages, respectively. This difference in lency by proper spacing of minimal epitopes at
specificity represents a major barrier for trans- two or more antennae. As illustrated by the se-
mission of avian viruses to humans (34, 35), and lected influenza HAs, these differences are rele-
increasing attention is placed on glycan micro- vant to the recognition of receptors by human
array analysis to understand the receptor require- pathogens. A complete understanding of influ-
ments of avian and human virus hemagglutinins enza receptor specificity and its relevance to
(HAs) required for species tropism (36–38). To adaptation of animal viruses to human hosts will
assess the potential for influenza HA to distin- require an extensive panel of asymmetric and sym-
guish between symmetric and asymmetric gly- metric glycan structures representative of those
cans, we evaluated the specificity of an HA from found on human and animal airway epithelia (38).
an exemplary H5N1 avian virus (VN/04), a human Such libraries of glycans, which can be produced
seasonal H1N1 virus (KY/07), and an H1N1 virus by the methodology presented here, will begin to
29. O. Blixt, N. Razi, Methods Enzymol. 415, 137–153
(2006).
30. W. Wang et al., Proc. Natl. Acad. Sci. U.S.A. 106,
16096–16101 (2009).
31. B. Sauerzapfe et al., Glycoconj. J. 26, 141–159 (2009).
32. O. Blixt et al., Proc. Natl. Acad. Sci. U.S.A. 101,
17033–17038 (2004).
33. Y. Itakura et al., J. Biochem. 142, 459–469 (2007).
34. A. Chandrasekaran et al., Nat. Biotechnol. 26, 107–113
(2008).
35. M. Imai, Y. Kawaoka, Curr. Opin. Virol. 2, 160–167 (2012).
36. L. M. Chen et al., Virology 422, 105–113 (2012).
37. M. B. Pearce et al., Proc. Natl. Acad. Sci. U.S.A. 109,
3944–3949 (2012).
38. T. Walther et al., PLoS Pathog. 9, e1003223 (2013).
39. J. Stevens et al., Science 312, 404–410 (2006).
40. Y. van Kooyk, G. A. Rabinovich, Nat. Immunol. 9,
593–601 (2008).
from the 2009 influenza pandemic (CA/05).
define the human glycome and provide tools to
The H5 HA from VN/04 recognized com- understand the biology mediated by both micro-
pounds N to Q and 27, which contain the bial and mammalian glycan-binding proteins that
Neu5Aca(2-3)Gal, consistent with the consensus mediate host pathogen interactions and innate and
receptor specificity of avian viruses (34, 39). adaptive immune responses (13, 40).
Notably, this cloned HA did not recognize the
Acknowledgments: This research was supported by NIH grant
P41RR005351 from the National Center for Research
Resources (G.J.-B. and J.G.), National Institute of General
Medical Sciences grants P41GM103390 (G.J.B. and J.G.) and
R01GM090269 (G.J.-B.), Institute of Allergy and Infectious
Disease grant AI058113 (J.C.P.), and a contract from the
Centers for Disease Control (J.C.P.). R.P.d.V. is a recipient of
a Rubicon grant from the Netherlands Organization for
Scientific Research (NWO). We thank A. Crie and M. Wolfert for
assistance in preparation of the manuscript, K. Moremen
(University of Georgia) for providing ST6Gal-1 and ST3Gal-IV,
and P. Wu (Albert Einstein College of Medicine) and
Neu5Aca(2-3)Gal in the fucosylated sequence
References and Notes
SLex in compound 22 or the reference compound
1. G. W. Hart, R. J. Copeland, Cell 143, 672–676 (2010).
2. H. H. Freeze, Nat. Rev. Genet. 7, 537–551 (2006).
3. K. Ohtsubo, J. D. Marth, Cell 126, 855–867 (2006).
4. S. J. North, P. G. Hitchen, S. M. Haslam, A. Dell,
Curr. Opin. Struct. Biol. 19, 498–506 (2009).
5. K. Mariño, J. Bones, J. J. Kattla, P. M. Rudd, Nat. Chem. Biol.
6, 713–723 (2010).
M. By contrast, the HA from the two human in-
fluenza viruses exhibited binding only to gly-
cans containing the Neu5Aca(2-6)Gal epitope
(Fig. 4), but otherwise exhibited different fine
specificities. The HA from the H1N1 seasonal
strain A/Kentucky/07 (KY/07) recognized all the
reference compounds (G to L) and all the tri-
antennary compounds (22 to 26) that contained
this linkage. However, relative to the linear ref-
erence compounds (G and H), the compounds
that have a Neu5Aca(2–6)Gal moiety on only
one branch of a biantennary glycan were bound
weakly (I and J), whereas those that had the
Neu5Aca (2–6)Gal sequence on only one branch
of the triantennary glycans (23, 24) were rec-
ognized equally well. Thus, this HA distinguishes
structures with a single sialic acid in the context
of linear or biantennary and triantennary chain
N-linked glycan chains. More pronounced differ-
ences are seen when comparing the seasonal
H1 and the pandemic HA H1 from A/California/
05/09 (CA/05). The CA/05 HA recognized only
reference compounds H and L and a single tri-
antennary glycan, namely 26. These compounds
have in common the Neu5Aca(2–6) epitope linked
to an extended dimeric-LacNAc moiety. However,
6. N. Laurent, J. Voglmeir, S. L. Flitsch, Chem. Commun.
(Camb.) (37): 4400–4412 (2008).
W. Wakarchuk (Ryerson University) for providing the plasmids
for a1,3FucT and b1,3GlcNAcT, respectively. G.J.-B. conceived
the idea, Z.W. performed the chemical synthesis, S.G.A. assisted
with chemical synthesis, Z.S.C. performed the enzymatic
7. T. J. Boltje, T. Buskas, G. J. Boons, Nat. Chem. 1,
611–622 (2009).
8. B. Lepenies, J. Yin, P. H. Seeberger, Curr. Opin. Chem. Biol.
transformations, and W.P. and Z.W. assisted with the enzymatic
transformation. Z.S.C. and J.G. performed the analysis of the
complex glycans. W.P. performed the attachment of the reactive
linker to the glycans, R.M. performed the microarray screening,
R.P.d.V. prepared the influenza HA, and J.C.P. supervised and
analyzed the microarray studies. G.J.-B. and J.C.P. wrote the paper.
The data for this report are archived as supplementary materials
on Science Online. A patent application related to the described
chemoenzymatic approach has been filed by the University of
Georgia Research Foundation and lists G.J.-B and Z.W. as
inventors. The authors declare no competing financial interests.
14, 404–411 (2010).
9. X. M. Zhu, R. R. Schmidt, Angew. Chem. Int. Ed. Engl. 48,
1900–1934 (2009).
10. M. M. Palcic, Curr. Opin. Chem. Biol. 15, 226–233 (2011).
11. R. M. Schmaltz, S. R. Hanson, C. H. Wong, Chem. Rev.
111, 4259–4307 (2011).
12. K. W. Moremen, M. Tiemeyer, A. V. Nairn, Nat. Rev. Mol.
Cell Biol. 13, 448–462 (2012).
13. C. D. Rillahan, J. C. Paulson, Annu. Rev. Biochem. 80,
797–823 (2011).
14. H. J. Gabius, S. André, J. Jiménez-Barbero, A. Romero,
D. Solís, Trends Biochem. Sci. 36, 298–313 (2011).
15. R. Ranzinger, M. Frank, C. W. von der Lieth, S. Herget,
Glycobiology 19, 1563–1567 (2009).
Supplementary Materials
www.sciencemag.org/cgi/content/full/341/6144/379/DC1
Materials and Methods
Figs. S1 to S24
Tables S1 to S14
16. Analysis of the glycome-DB database shows that 85% of
known glycan structures are asymmetrical. Almost all of
the known N-linked glycans have fewer than 25
monosaccharides, and such compounds should be
accessible by the strategy presented here.
17. C. Unverzagt, Angew. Chem. Int. Ed. Engl. 35, 2350–2353
(1996).
References (41–48)
Copies of NMR Spectra
7 February 2013; accepted 14 June 2013
10.1126/science.1236231
www.sciencemag.org SCIENCE VOL 341 26 JULY 2013
383