Three-dimensional arrays using GlycoPEG tags: Glycan synthesis, purification and immobilisation

Article abstract of DOI:10.1002/chem.201204004

Glycan arrays have become the premier tool for rapidly establishing the binding or substrate specificities of lectins and carbohydrate-processing enzymes. New approaches for accelerating carbohydrate synthesis to address the enormous complexity of natural glycan structures are necessary. Moreover, optimising glycan immobilisation is key for the development of selective, sensitive and reproducible array-based assays. We present a tag-based approach that accelerates the preparation of glycan arrays on all levels by improving the synthesis, the purification and immobilisation of oligosaccharides. Glycan primers were chemically attached to bifunctional polyethyleneglycol (PEG) tags, extended enzymatically with the help of recombinant glycosyltransferases and finally purified by ultrafiltration. When printed directly onto activated glass slides, these glycoPEG tags afforded arrays with exceptionally high sensitivity, low background and excellent spot morphology. Likewise, the conjugation of glycoPEG tags to latex nanoparticles yielded multivalent scaffolds for carbohydrate-binding assays with very low non-specific binding. Got you PEGged: A polyethyleneglycol (PEG)-tag-based approach has been developed that accelerates the preparation of glycan arrays and glyco beads on all levels by improving the chemo-enzymatic synthesis, the purification and immobilisation of oligosaccharides (see figure). Copyright

Full text of DOI:10.1002/chem.201204004

DOI: 10.1002/chem.201204004
Three-Dimensional Arrays Using GlycoPEG Tags: Glycan Synthesis,
Purification and Immobilisation**
Juan Etxebarria,[a] Sonia Serna,[a] Ana Beloqui,[a] Manuel Martin-Lomas,[a, b] and
Niels-Christian Reichardt*[a, b]
Abstract: Glycan arrays have become
the premier tool for rapidly establish-
ing the binding or substrate specificities
of lectins and carbohydrate-processing
enzymes. New approaches for acceler-
ating carbohydrate synthesis to address
the enormous complexity of natural
glycan structures are necessary. More-
over, optimising glycan immobilisation
is key for the development of selective,
sensitive and reproducible array-based
assays. We present a tag-based ap-
proach that accelerates the preparation
of glycan arrays on all levels by im-
proving the synthesis, the purification
and immobilisation of oligosaccharides.
Glycan primers were chemically attach-
ed to bifunctional polyethyleneglycol
(PEG) tags, extended enzymatically
with the help of recombinant glycosyl-
transferases and finally purified by ul-
trafiltration. When printed directly
onto activated glass slides, these glyco-
PEG tags afforded arrays with excep-
tionally high sensitivity, low back-
ground and excellent spot morphology.
Likewise, the conjugation of glycoPEG
tags to latex nanoparticles yielded mul-
tivalent scaffolds for carbohydrate-
binding assays with very low non-spe-
cific binding.
Keywords: carbohydrates
·
en-
zymes · glycan arrays · nanoparti-
cles · oligosaccharides
Introduction
natural sources. (Automated) chemo-enzymatic solid or
tagged solution-phase strategies are particularly promising
for the rapid synthesis of oligosaccharides either directly on
the chip or in solution with the help of solid and soluble
supports.[10–13] Oligosaccharides assembled on a solid support
usually have to be cleaved off the resin, purified and depro-
tected before they can be printed onto microarray slides.[14]
This adds time-consuming steps to any automated solid-
phase oligosaccharide synthesis. Perfluorinated linkers and
protecting groups in combination with solid-phase extraction
have been employed for some time in the tagged solution-
phase synthesis of oligosaccharides and used as linkers for
the noncovalent immobilisation of glycans on Teflon-coated
microarray slides.[15] The low solubility of perfluorinated
compounds (PFCs) and their conjugates in aqueous solu-
tions,[16] however, has limited their synthetic use in organic-
solvent-based chemical synthesis and their generally low bi-
odegradability is now a major environmental concern, there-
by threatening the future commercialisation of fluorous link-
ers.[17] In addition, the non-covalent attachment of glycocon-
jugates is limited to specially prepared perfluorinated surfa-
ces.
Ten years ago, several groups introduced the first glycan
arrays as a new and promising tool for studying the binding
and substrate specificities of carbohydrate-binding proteins
and carbohydrate-processing enzymes with unprecedented
throughput and minimal sample requirements.[1–6] Although
much progress has been made since then, with some arrays
that present over 600 glycan ligands[7,8] having been con-
structed, current formats still cover only a very small frac-
tion of the estimated 10000–30000 biologically relevant
glycan epitopes.[8,9] This is in large part due to the notorious
difficulties in oligosaccharide synthesis or their isolation
from natural sources. To prepare next-generation large
glycan arrays that accommodate far higher numbers of li-
gands, new technologies are required that can accelerate car-
bohydrate synthesis or provide solutions for the high-
throughput isolation, tagging and analysis of glycans from
[a] Dr. J. Etxebarria, Dr. S. Serna, Dr. A. Beloqui,
Prof. Dr. M. Martin-Lomas, Dr. N.-C. Reichardt
Biofunctional Nanomaterials Department
CICbiomaGUNE, Paseo Miramon 182
20009 San Sebastian (Spain)
We were interested in developing an environmentally
friendly, tag-based method for the unified chemo-enzymatic
synthesis, purification and immobilisation of glycans or
other biomolecules on microarrays or other scaffolds. The
chosen tag should be stable under harsh synthetic condi-
tions, allow easy product purification and provide a covalent
attachment point with uniform orientation for all ligands
and low background in protein-binding assays.
E-mail: nreichardt@cicbiomagune.es
[b] Prof. Dr. M. Martin-Lomas, Dr. N.-C. Reichardt
CIBER-BBN, Paseo Miramon 182
20009 San Sebastian (Spain)
[**] GlycoPEG=Glycan primers chemically attached to bifunctional pol-
yethyleneglycol (PEG) tags.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201204004. It includes experi-
mental details and MALDI-MS spectra.
Methylated polyethylene glycol (MPEG) is soluble both
in water and many organic solvents and has been used as a
4776
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 4776 – 4785

FULL PAPER
soluble support for solid-phase synthesis of oligosaccharides,
-nucleotides and -peptides.[10,18,19] Reactions are carried out
under homogeneous solution-phase conditions and inter-
mediates are readily analysed in the presence of the poly-
mer by NMR spectroscopy or mass spectrometry. The addi-
tion of ethanol or ether renders the PEG tag insoluble so re-
agents or other reaction impurities can be removed by
simple filtration. Alternatively, membrane-based ultrafiltra-
tion can purify the tagged compounds in a single step by
size exclusion, an especially important feature for microscale
operations.[20] Furthermore, PEG tags are ideal spacers for
covalently attaching biomolecules to surfaces at a tuned dis-
tance, which helps to improve receptor and enzyme accessi-
bility, and their well-known protein-repelling properties
result in arrays with exceptionally low background.[21]
end products, an important feature for microscale chemistry.
Not surprisingly, then, solid-supported enzymatic oligosac-
charide synthesis has been investigated by a number of
groups over the years with a strong focus on the type of
solid support used for attaching either the enzyme or the
growing sugar chain.[22–33] Generally, good to excellent cou-
pling yields were achieved when soluble polymers[20,26,34]
were used or longer oligoethylene-type linkers attached to
the solid support.[22] Other more rigid polymers were less ac-
cessible for the enzymes and hence provided lower coupling
yields.[24–26] More recent strategies that employed pH-[28] or
thermoresponsive[35] polymers rendered the tagged reaction
products insoluble after a specific stimulus. Enzymes and
co-factors were then removed by simple filtration. In a relat-
ed recent study, ionic liquid tags were applied to monitor
the action of glycosyltransferases on several substrates by
liquid chromatography (LC)-MS.[36] These techniques, how-
ever, are prone to difficulties when employed in microscale
reactions.
Results and Discussion
These clear advantages of polyethylene glycol polymers
prompted us to evaluate bifunctional PEG tags 1 and 2
(Figure 1) in the tag-supported chemo-enzymatic synthesis
of oligosaccharides as well as a protein-repelling spacer for
the direct attachment of glycans onto microarrays.
The use of glycosyltransferases with well-known substrate
specificities avoids the use of protecting groups and compen-
sates the lack of universal chemical methods for stereoselec-
tive formation of glycosidic bonds, the current limitations of
chemical oligosaccharide synthesis. It also justifies the use of
MALDI-TOF MS as the single method for the analysis of
Essential to our strategy, which is schematically depicted
in Figure 1, was the efficient immobilisation of glycans
tagged with a large PEG chain (approximately 5000 Da) by
robotic spotting.
As a proof of concept, PEG-tagged mannose 3 and N-ace-
tylglucosamine (GlcNAc) 4 were prepared (Scheme 1). The
carbobenzyloxy (Cbz)-protected hetero-bifunctional poly-
mer 1 was conjugated with trichloroacetimidates 24 and 25
to furnish conjugates 26 and 27, respectively. Stepwise de-
protection of glycoside 26 by aminolysis of the phthalimido
group, acetylation and Birch reduction followed by rapid pu-
Figure 1. PEG-tag-supported synthesis, purification and immobilisation of glycans for 3D microarrays and functionalised microbeads.
Chem. Eur. J. 2013, 19, 4776 – 4785
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4777

N.-C. Reichardt et al.
Scheme 1. Examples of the chemo-enzymatic synthesis of glycoPEGtags: a) Compound 1, TMSOTf, CH2Cl2; b) 1) NH2EtNH2, nBuOH, MW, 2) Ac2O,
pyridine; 3) Na, NH3,THF, ꢀ788C; c) UDP-GalNAc, C. elegans b-1,4-GalNAcT, BSA, MnCl2 , alkaline phosphatase (EC 3.1.3.1) in cacodylate buffer
80 mm, pH 7.4, 258C, 18 h; d) b-1,4-GalT, UDP-Gal, MnCl2, HEPES 50 mm, pH 7.4, 378C; e) C. elegans a-1,3-fucosyltransferase, GDP-Fuc, MnCl2, MES
80 mm, pH 6.5, RT; f) Na, NH3,THF, ꢀ788C.
rification through precipitation afforded PEG-tagged
GlcNAc 4. PEG-tagged mannose 3 was obtained after
single-step Birch reduction of conjugate 27. The deprotec-
tion and PEG-tag-based purification of both glycoconjugates
proceeded with excellent conversion as confirmed by
MALDI-MS analysis of the final products (3 and 4).
precipitated, and the PEG-tagged N-acetyl lactosamine 5
was purified by simple ultrafiltration by taking advantage of
the large 5 kD PEG tag for the separation from excess
amounts of nucleotide donor and buffer salts[20] (Scheme 1).
The tedious chromatographic purification steps required in
the traditional solution-phase enzymatic oligosaccharide
synthesis were thus avoided. MALDI-TOF MS analysis,
which did not require more than 0.5 mg aliquot of sugar–
PEG conjugate, showed complete reaction to the PEG-
tagged N-acetyl lactosamine 5 by a 162 Da shift of the peak
pattern[37] (see Figure S25 in the Supporting Information).
Subsequent incubation with a recombinant a-1,3-fucosyl-
transferase from Caenorhabditis elegans produced cleanly
the Lewis X trisaccharide polymer conjugate 6, which was
purified by filtration from enzyme and excess amount of nu-
cleotide donor. Treatment of the PEG-tagged GlcNAc
primer 4 with GalNAc transferase from C. elegans and
UDP-GalNAc as cofactor produced the PEG-tagged Gal-
NAcb1!4GlcNAc disaccharide 7 (Scheme 1).
Both constructs were then robotically printed onto an N-
hydroxysuccinimide (NHS)-activated glass slide at concen-
trations between 50 and 500 mm and non-printed areas passi-
vated with a shorter hetero-bifunctional amino-PEG deriva-
tive (average mass 600 Da). Incubation of the slide with flu-
orescently tagged a-mannose binding lectin Concanavalin A
(ConA) and a glucosamine binding lectin from Bandeira
simplicifolia (BS II) clearly demonstrated the successful at-
tachment of the large PEG glycoconjugates to the activated
surface with excellent spot-to-spot reproducibility and ho-
mogeneity (see Figure S32 in the Supporting Information).
Once PEG-tagged sugars had proven to be excellent li-
gands for the robotic printing of glycan arrays, we next in-
vestigated their enzymatic extension by exploiting the large
bifunctional PEG tag (5 kD) for rapid purification. Conju-
gated to the large soluble tag, glycans can be handled on the
low microgram scale with great ease by employing standard
biochemistry laboratory equipment. PEG-tagged GlcNAc 4
(300 mg, 15 mg of sugar) was incubated with a bovine milk b-
1,4-galactosyltransferase and uridine 5’-diphosphogalactose
disodium salt (UDP-galactose) for 20 h. The enzyme was
To demonstrate the potential of PEG-supported micro-
scale enzymatic oligosaccharide synthesis for the prepara-
tion of glycan arrays, we envisaged the formation of a small
library of N-glycans (8–14) with variations in terminal
sugars and core modification starting from a single N-glycan
core structure (Figure 2). Towards this end we coupled bi-
antennary C5-amino-linked N-glycan 21[38] (1 mg; Figure 1)
to the NHS-activated polymer 2 in nearly quantitative yield
4778
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 4776 – 4785

GlycoPEG Tags
FULL PAPER
Figure 2. Enzymatic diversification of PEGylated bi-antennary glycan 8 and MALDI-TOF MS of selected compounds 8, 9 and 10.
and removed the tert-butyloxycarbonyl (Boc) protection to
obtain PEG-tagged sugar 8 as the scaffold for the enzymatic
library synthesis (see Figure S23 and Scheme S6 in the Sup-
porting Information). Aliquots of sugar–polymer (200–
300 mg) were dissolved in buffer and incubated with six re-
combinant glycosyltransferases to produce PEG-tagged gly-
cans 8–14 by enzymatic soluble-polymer-supported synthesis
in only a few days.
Incubation of 8 with GalT and UDP galactose gave rise to
bis-galactosylated conjugate 9, which was further elongated
with a Lewis-type a-1,3-fucosyltransferase from C. elegans
(CeFUT6) to produce conjugate 10. Treatment of 8 with a
GalNAc-transferase from C. elegans and UDP-GalNAc as
the nucleotide donor furnished quantitatively 11 with two
terminal GalNAc residues. The a-1,6- and a-1,3-core-fucosy-
lated conjugates 12 and 13 were accessible by employing
two recombinant fucosyltransferases from Arabidopsis thali-
ana (AtFucTA) and C. elegans (CeFUT8), respectively. Se-
quential incubation of 8 with both enzymes furnished the
bis-fucosylated conjugate 14[39] (Figure 2).
Thus, by means of PEG-tag-supported enzymatic synthe-
sis, we were able to produce seven different N-glycan conju-
gates in a few days from a single compound. The soluble tag
allowed a rapid purification of the glycoconjugate by filtra-
tion of the precipitated enzyme and then removal of low-
molecular-weight impurities on a standard ultrafiltration
device.[40] PEGylated trisaccharide 15 obtained by coupling
the trimannoside 20 with linker 2 (Figure 1) followed by
Boc deprotection was included as a larger mannoside for
the lectin-binding studies into the ligand base.
All PEG-tagged glycans 3–15 and C5-amino-linked gly-
cans 16–23 (see Figure 1) were then arrayed at 200, 100, 50
and 25 mm concentrations onto NHS-activated slides for a
comparison of linker length and printing concentration on
the sensitivity and spot homogeneity. Individual subarrays
that contained all ligands were incubated with the fluores-
cently marked lectins ConA, Ricinus communis agglutinin
(RCA), Aleuria aurantia lectin (AAL), Wisteria floribunda
lectin (WFL), Vicia villosa lectin (VVL), wheat germ agglu-
tinin (WGA), Erythrina cristagalli lectin (ECA), Lens culi-
naris agglutinin (LCA), Galanthus nivalis agglutinin (GNA),
Narcissus pseudonarcissus lectin (NPL), BS II and an anti-
body against horseradish peroxidase (anti-HRP); the fluo-
rescence intensity was quantified and used as a measure for
the lectin–ligand binding strength (Figure 3).
Any differences in the immobilisation between ligands 3–
15 were compensated by the large PEG tag and, in general,
well-defined intense spots with excellent homogeneity were
obtained for all printed samples at all concentrations. C5-
amino-linked printed glycans 16–23 gave smaller and parti-
ally misaligned spots with a concentration- and ligand-de-
pendent variation in size.
In general, all lectins bound with the expected selectivity
but usually with stronger bonding to the PEGylated glycans
(3–15) than to their respective C5-amino-linked homologues
(16–23), which suggests a pronounced effect of the PEG pol-
ymers on the ligand presentation. Monosaccharides and
some smaller glycan fragments were bound in their PEGy-
lated form only, thus highlighting the need for a longer
spacer to access the surface-bound glycans.
More precisely, PEGylated GlcNAc 4 bound to BS II, a
terminal GlcNAc-recognising lectin, with considerable
strength down to 50 mm spotting concentration, whereas the
C5-linked analogue 17 was not bound at all. PEGylated
GlcNAc-terminating N-glycans 8 and 12 bound nearly twice
as strong to BS II than C5-linked congeners 21 and 23, thus
showing that the effect was preserved for more complex
structures as well (Figure 3a). Similar results were obtained
for the terminal galactose binding lectins ECA and RCA.
PEGylated N-acetyl lactosamine
5
and GalNAcb1!
Chem. Eur. J. 2013, 19, 4776 – 4785
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4779

N.-C. Reichardt et al.
Figure 3. Comparison of immobilised glyco-PEG tags (3–15) and C5-linked glycans (16–23) in lectin binding assays. a) Fluorescence scans for arrays incu-
bated with BS II, ECA, RCA and AAL. b) Histogram view of binding data at four different ligand printing concentrations.
4GlcNAc 7 bound to ECA with medium intensity at all spot-
ting concentrations, whereas almost no interaction was de-
tected for disaccharide 18 immobilised through the short C5
linker. RCA bound strongly to both PEGylated galactosylat-
ed bi-antennary glycan 9 and its C5-linked homologue 22,
independent of the linker used. Reduced but still significant
binding was observed in the presence of a-1,3-fucose pres-
ent in glycans 6 and 10; the C5-linked homologue, however,
showed no measurable interaction with the lectin (Figure 3).
As an example of the fucose binding probes tested, AAL
showed the known broad specificity including core a-1,3-
and core a-1,6-fucosylation and Lewis-type fucosylation
present on the array. Significantly, AAL binding to the Le X
epitope was only observed for its PEGylated form 6 with
fluorescence values approaching saturation. Interestingly,
AAL binding to the 1,6-core-fucosylated bi-antennary C5-
linked glycan (19) remained strong, which suggests that
other epitopes apart from the core fucose are recognised by
the lectin.
tions for PEGylated glycans that present terminal GlcNAc
and chitobiose, as in N-glycans 8–14 (see the Supporting In-
formation). Fucosylation and specifically core fucosylation
decreased binding to WGA significantly.[41] Single GlcNAc
was only recognised in its PEGylated form 4 and fluores-
cence values for WGA binding to PEGylated GalNAcb1!
4GlcNAc (LacdiNAc; 7) reached saturation levels, whereas
binding to the C5-linked homologue 18 was moderate. Anti-
HRP is a highly specific probe for core a-1,3-fucose, which
was confirmed by binding exclusively to mono- and bis-fuco-
sylated glycans 13 and 14 with very good signal intensity at
all spotting concentrations. The higher specificity of WFL
for GalNAc residues was reflected in a stronger binding to
the two PEGylated ligands 7 and 11 (see the Supporting In-
formation).
To determine the limit of detection for the immobilised
glycoPEG tags, we printed a dilution series of the Le X–
PEG conjugate 6 (Figure 4a) and LacdiNAc disaccharide 7
(Figure 4b). Both arrays were incubated with fucose binding
lectin from Aleuria aurantia and GalNAc binding Wisteria
floribunda lectin at a nanomolar concentration. The interac-
tion of AAL with ligand 6 could be detected down to 25 at-
tomole of printed glycan at a S/N>5. (Figure 4a). This sen-
Pronounced differences in binding to PEGylated and non-
PEGylated glycans were also observed for the other lectins
tested (see Figures S34 and S35 in the Supporting Informa-
tion). WGA showed strong binding at all printing concentra-
4780
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 4776 – 4785

GlycoPEG Tags
FULL PAPER
mats in which ligands are attached by means of short linkers
directly onto a flat Teflon or metal oxide surface,[17,45] the
immobilisation of glycans through long PEG tags reaches a
sensitivity otherwise only achieved by employing slide surfa-
ces functionalised with a 3D hydrogel matrix.[42]
Although glycan arrays are valuable surface-based assays
for the high-throughput screening of weak carbohydrate–
protein interactions, other applications like suspension
assays, cell labelling, flow cytometry, immuno-separation of
biomolecules or in vivo imaging require spherical 3D scaf-
folds like functionalised nanoparticles or microbeads. We
functionalised aliquots of fluorescent 300 nm-sized latex
beads with equimolar amounts of either glycoPEG tags (3,
5, 6, 7, 12 and 15) or the corresponding C5-linked glycans
(16, 19–21) to study the influence of the PEG spacer on the
binding to a lectin array that comprised ConA, WGA,
RCA, SNA, MAL-1, AAL, PSA, LCA, WFL, VVL and
control buffer (Figure 5). Specific binding of beads to the
lectin-functionalised wells could be visualised by scanning
electron microscopy (SEM) after sputter-coating the slide
with a thin layer of gold. At 200-fold higher resolution than
scanometric fluorescence measurements, single beads can be
visualised and the images obtained can be used for quantifi-
cation of binding interactions.[46] Specific carbohydrate–car-
bohydrate interaction between Le X-functionalised beads
could be the cause for the observed clustering and multilay-
er stacking of beads (Figure 5b).[47]
PEGylated Le X beads bound strongly only to RCA and
fucose binding lectin AAL; beads functionalised with PE-
Gylated mannose 3 selectively bound to ConA, whereas
PEG-tagged trimannoside 15 beads also bound to mannose
binding lectins PSA and LCA. Beads functionalised with
C5-linked glycans, however, did not bind to the lectin array
in a selective manner unless a tenfold-higher glycan concen-
tration was used for their preparation (see the Supporting
Information). Beads functionalised with C5-linked glycans 3
and 15 at tenfold-higher concentration bound to ConA,
PSA and LCA with similar intensity but also showed non-
specific binding to AAL, WGA and RCA.
Figure 4. Assay sensitivity measured for binding of a) AAL to Le X-PEG
tag and b) WFL to LacDiNAc-PEG tag at different spotting concentra-
tions.
Beads functionalised with PEGylated GalNAcb1!
4GlcNAc 7 bound with high affinity and excellent selectivity
to RCA, WGA, ECA and WFL. PEGylated N-acetyl lactos-
amine beads 5 bound exclusively to RCA, whereas the a-
1,6-core-fucosylated bi-antennary glycoPEG tag 12 showed
a strong interaction with WGA, AAL, PSA, LCA and, to a
lower degree, ConA lectins. These examples show that gly-
coPEG tags provide superior sensitivity and selectivity also
in bead-based assays with interesting prospects for array-
based studies of weak carbohydrate–carbohydrate interac-
tions.
sitivity is comparable to reported values for the mannose/
ConA ligand receptor pair studied on three-dimensional hy-
drogel slides, which are frequently used in glycan array re-
search and are considered the gold standard.[42]
The above results clearly show that printed PEG-tagged
glycans outperformed the smaller linkers in spot morpholo-
gy and sensitivity for all lectins studied. The random coil
conformation of large PEG chains upon immobilisation
(“entangled mushroom-like PEG”, 25% polymer exten-
sion)[43] most probably leads to a reduced ligand density for
the PEGylated glycans relative to the C5-tagged glycans
with a smaller footprint.[43] The lower sugar content per
area, however, is more than compensated by the excellent
accessibility of proteins to the binding epitope,[44] most clear-
ly demonstrated by the exclusive binding of PEG-tagged
small glycans 3–7. Relative to other reported microarray for-
Conclusion
We have shown that bifunctional PEG tags provide a versa-
tile handle that can streamline the preparation of glycan
arrays and glycobeads from chemo-enzymatic synthesis and
Chem. Eur. J. 2013, 19, 4776 – 4785
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4781

N.-C. Reichardt et al.
Figure 5. Binding of glyco-PEG-tag-functionalised fluorescent latex nanoparticles on lectin arrays. a) Fluorescence scan of Le X-PEG-tag-functionalised
particles binding to lectin array. b) SEM images of beads binding to lectin-functionalised surface and close up. c) Quantification of binding to lectin
arrays for beads functionalised with glycoPEG-tags 6, 15, 7, 5, 3 and 12.
freshly distilled from Na/benzophenone and dichloromethane was freshly
distilled from CaH2. Thin-layer chromatography was carried out using
purification until the final immobilisation of complex gly-
cans. Like no other tag-based system, PEG tags provide
simple purification by size exclusion or filtration, compati-
bility with enzymatic and chemical synthesis in a homogene-
ous reaction medium and a superior linkage for biomole-
cules to surfaces or beads to result in assays with exception-
ally low background, greatly improved sensitivity and excel-
lent spot morphology. The latter is particularly relevant for
quantitative diagnostic tests that require a high degree of re-
producibility. To increase the synthetic throughput, the de-
velopment of automated PEG-tag-based chemo-enzymatic
synthesis that combines membrane filtration and robotic dis-
pensing technology should be feasible and is currently being
studied in our laboratory.
Merck aluminium sheets silica gel 60 F254 and visualised by UV irradia-
tion (254 nm) or by staining (vanillin (15 g) and H2SO4 (2.5 mL) concen-
trated in EtOH (250 mL)). Microwave irradiation was performed using a
Biotage Initiator monomode oven, Biotage AB, Uppsala, Sweden. Purifi-
cation of compounds was performed by flash chromatography using
Merck 62 ꢁ 230–400 mesh silica gel. Size-exclusion chromatography was
performed using a Biorad P2 gel, Biorad, Hercules, USA. Pooled glycan-
containing fractions were lyophilised using an ALPHA-2-4 LSC freeze-
dryer from Christ, Osterode, Germany. All organic solvents were concen-
trated using rotary evaporation. 1H and 13C spectra were acquired using
Bruker 500 MHz spectrometer and chemical shifts (d) are given in ppm
relative to the residual signal of the solvent used. Coupling constants (J)
are reported in Hz. The mass spectrometric data were obtained using a
Waters LCT Premier XE instrument, Waters, Manchester, UK with a
standard ESI source by direct injection. The instrument was operated
with a capillary voltage of 1.0 kV and a cone voltage of 200 V. Cone and
desolvation gas flow were set to 50 and 600 Lhꢀ1, respectively; source
and desolvation temperatures were 1008C. MALDI-TOF mass analyses
were performed using an Ultraflextreme III time-of-flight mass spectrom-
eter equipped with a pulsed N2/Nd laser (355 nm) and controlled by Flex-
Control 3.3 software (Bruker Daltonics). The acquisitions were carried
out in positive-ion reflectron mode at a laser frequency of 500 Hz. Micro-
arrays were printed on glass slides employing a robotic non-contact spot-
ter S11 from Scienion, Berlin, Germany. Aminosilane-coated glass slides,
Nexterion Slide A+, and NHS-activated glass slides, Nexterion H, were
purchased from Schott AG, Mainz, Germany. Lectins were purchased
Experimental Section
The synthetic procedure and characterisation of key compounds are de-
scribed below. Synthetic details and spectroscopic data for all other com-
pounds can be found in the Supporting Information.
General methods: Chemicals were purchased from Sigma–Aldrich or
Acros Organics and were used without further purification. THF was
4782
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 4776 – 4785

GlycoPEG Tags
FULL PAPER
from Vector Laboratories, Burlingame, USA and labelled with Hilyte
Plus 647 and Hilyte Plus 555 protein-labelling kits from AnaSpec, Free-
mont, USA. Lectin incubations were performed using the Fast Frame in-
cubation chambers from Whatman, Kent, UK. Fluorescence measure-
ments were performed using an Agilent G265BA microarray scanner
system, Agilent Technologies, Santa Clara, USA. Quantification was per-
formed using ProScanArray Express software, Perkin–Elmer, Shelton,
USA. SEM images were recorded using a JSM-6490LV, implemented
with an ion beam sputtering unit.
(97%) by ultra performance liquid chromatography (UPLC) from the re-
action crude by derivatisation with a 9-fluorenylmethoxycarbonyl (Fmoc)
reporter group. The isolated compound was analysed by MALDI-TOF
MS using pre-spotted MALDI targets with a-cyano-4-hydroxy-cinnamic
acid (HCCA) matrix (Anchor Chip, Bruker Daltonics): m/z: 6232.04
[M+Na], most abundant peak.
Core-type a-1,6-fucosylation: A solution (200 mL) of compound 8 (0.3 mg,
46 nmol), GDP-fucose (200 nmol), and C. elegans core-type a-1,6-fucosyl-
transferase (55 mg) in MES buffer (80 mm, pH 6.5), MnCl2 (20 mm) was
incubated at 258C for 24 h. The resulting mixture was heated at 958C for
5 min. The precipitate was removed, the solution was purified and the
buffer exchanged to water using an Amicon spin filter of 10 kDa. The
Chemo-enzymatic synthesis of PEG-tagged Lewis X derivative 6
PEG-tagged monosaccharide 26: Trichloroacetimidate 24 (50 mg,
0.01 mmol) and 1 (18 mg, 0.003 mmol) were dissolved in dry CH2Cl2
(2 mL) in the presence of 4 ꢁ MS. The mixture was stirred for 1 h,
cooled to ꢀ208C and then trimethylsilyl trifluoromethanesulfonate
(TMSOTf; 1 mL, 0.007 mmol) was added. The mixture was stirred for 2 h
from ꢀ20 to 08C, then quenched with a drop of Et3N and diluted with
CH2Cl2 (10 mL). The reaction mixture was evaporated and the product
was precipitated by stirring the crude material in CH2Cl2/Et2O (1:50) at
isolated polymer was freeze-dried to obtain 12 as
a white powder
(0.3 mg, 95% yield). MALDI-TOF MS: m/z: 6377.16 [M+Na].
Core-type a-1,3-fucosylation: A solution (200 mL) of compound 8 (0.3 mg,
46 nmol), GDP-fucose (100 nmol), and A. thaliana core-type a-1,3-fuco-
syltransferase (70 mg) in MES buffer (80 mm, pH 6.5), MnCl2 (20 mm)
was incubated at 258C for 40 h. The resulting mixture was heated at
958C for 5 min. The precipitate was removed, the solution was purified
and the buffer exchanged to water using an Amicon spin filter of 10 kDa.
The isolated polymer was freeze-dried to obtain 13 as a white powder
(0.3 mg, 95% yield). MALDI-TOF MS: m/z: 6377.28 [M+Na].
08C to afford compound 26 as
a white solid (40 mg, 80% yield).
MALDI-TOF: m/z: 5354.94 [M+Na], most abundant peak.
Amino-polyethylene glycolyl 2-acetamido-2-deoxy-b-d-glucopyranoside
(4):
A solution of 26 (0.046 g, 0.0086 mmol) and ethylenediamine
(0.4 mL) in nBuOH (0.6 mL) was irradiated three times with microwaves
at 1208C for 30 min. Solvent was evaporated and the reaction crude was
subjected to acetylation with Ac2O (0.4 mL) in pyridine (0.8 mL) at 08C
for 18 h. The solvent was evaporated and the product was precipitated by
stirring the crude material in CH2Cl2/Et2O (1:50) at 08C to afford the
compound as yellowish solid (0.039 g). This material was dissolved in
THF (1 mL) and added to a suspension of Na in liquid NH3 at ꢀ788C.
The mixture was stirred for 15 min at this temperature and then the reac-
tion was quenched by the addition of NH4Cl. The solvent was then
evaporated. Purification by size-exclusion chromatography (Biogel-P2,
NH4HCO3 as eluent) afforded product 4 as a colourless oil (23 mg, 60%
yield). MALDI-TOF MS: m/z: 4962.38 [M+Na], most abundant peak.
Core-type a-1,6-fucosylation and a-1,3-fucosylation: A solution (200 mL)
of compound 12 (0.15 mg, 23 nmol), GDP-fucose (100 nmol) and A. thali-
ana core-type a-1,3-fucosyltransferase (70 mg) in MES buffer (80 mm,
pH 6.5) was incubated at 258C for 40 h. The resulting mixture was heated
at 958C for 5 min. The precipitate was removed, the solution was purified
and the buffer exchanged to water using an Amicon spin filter of 10 kDa.
The isolated polymer was freeze-dried to obtain 14 as a white powder
(0.1 mg, 67% yield). MALDI-TOF MS: m/z: 6523.20 [M+Na].
3D microarray preparation
Functionalisation of aminosiloxane surfaces: Amino slides (Nexterion
slide-A star) were blocked with succinic anhydride solution (5.0 g) in N-
methyl-2-pyrrolidone (NMP) (250 mL) and borate buffer (30 mL, 0.2m),
pH 9) for 45 min at RT. The slides were washed with phosphate-buffered
saline (PBS) and water and dried with Ar. Activation of the carboxylic
acid slides as succinimide esters was performed by treatment with 0.2m
1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC) and 0.05m NHS
in anhydrous DMF for 2 h at RT. The slides were rinsed with anhydrous
DMF and dried with Ar. After the washing step with dry DMF, the surfa-
ces were ready for the printing of the glycan arrays.
b-1,4-Galactosylation: A solution (200 mL) of 4 (2 mg, 0.39 mmol), UDP-
Gal (uridine 5’-diphosphogalactose disodium salt; 1.6 mg), bovine serum
albumin (BSA; 1 mg), bovine milk b-1,4-galactosyltransferase (50 mU;
EC 2.4.1.22), alkaline phosphatase (2.3 U; EC 3.1.3.1) and MnCl2
(10 mm) in 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES)
buffer (50 mm, pH 7.4), was incubated at 378C for 20 h. The resulting
mixture was heated at 958C for 5 min. The yellowish precipitate was re-
moved and the solution was purified and buffer exchanged to water using
an Amicon spin filter of 10 kDa. The isolated polymer was freeze-dried
to obtain compound 5 as a yellowish powder (1.7 mg, 85% yield).
MALDI-TOF MS: m/z: 5119.143 [M+Na].
Microarray printing: Glycoconjugates 3–15 and the aminopentyl-linked
glycans 16–23 were printed at different concentrations 200, 100, 50 and
25 mm in phosphate buffer (100 mm) that contained 0.002% of Tween-20
at pH 8.4, on NHS-functionalised glass slides with a non-contact spotter.
A number (3) of 270 pL drops of the buffered solutions were printed
with a distance of 300 mm. Five replicates were printed of each sample as
21ꢂ20 matrixes. After printing, the slides were kept in a 75% humidity
chamber (saturated NaCl solution) at 258C for 18 h. Unreacted NHS
groups were quenched by placing the slides in a 50 mm solution of PEG-
600-NH2 in sodium borate buffer 50 mm, pH 9.0, for 2 h. The standard
washing of the slides was performed with PBST (PBS solution that con-
tained 0.5% Tween 20), PBS and water. The slides were dried in a slide
spinner.
a-1,3-Fucosylation: A solution (250 mL) of 5 (0.4 mg, 75 nmol), GDP-Fuc
(0.5 mmol), C. elegans Lewis X type a-1,3-fucosyltransferase (CeFuT6;
69 mg) and MnCl2 (20 mm) in MES buffer ( 80 mm, pH 6.5) was incubated
at 258C for 40 h. The resulting mixture was heated at 958C for 5 min.
The precipitate was removed and the solution was purified and the
buffer exchanged to water using an Amicon spin filter of 3 kDa. The iso-
lated polymer was freeze-dried to obtain compound 6 as a white powder
(0.3 mg, 75% yield). MALDI-TOF MS: m/z: 5266.52 [M+Na].
Chemo-enzymatic synthesis of core-type fucosylated PEG-tagged N-
glycan derivatives 12, 13 and 14
Labelled lectin incubation of the 3D microarray: Lectin stock solutions
(4.0–5.0 mgmLꢀ1) were diluted 1:2 with phosphate buffer 300 mm, pH 8.5
and labelled with 1 mL of Hilyte Plus 647 and 555 (cy5/cy3 analogues)
protein-labelling kit from AnaSpec, Freemont, USA, for each 100 mL of
sample for 2 h at RT. Excess amounts of dye were removed by buffer ex-
change to PBS by employing 3 kDa Amicon filters, Millipore, Carrigtwo-
hill, Co. Cork, Ireland. The labelled lectins were stored as 1–2 mgmLꢀ1
solutions in PBS at 48C (D/P ratio: 1–2). ConA, RCA, AAL, WFL, VVL
lectins and anti-HRP were labelled with the cyanine-3 analogue and
WGA, ECA, LCA, GNA, NPL and BS II were labelled with the cya-
nine-5 analogue.
Bi-antennary N-glycan coupling to the PEG tag: A mixture of glycan 21
(0.23 mg, 0.16 mmol), the N-Boc-protected NHS-activated PEG
2
(1.00 mg, 0.20 mmol) and N,N-diisopropylethylamine (DIPEA; 5 mL,
28.70 mmol) in dry DMF (0.2 mL) was stirred for 2 h at RT and then the
excess amount of activated polymer was removed by addition of Tenta-
gel-NH2 resin (20 mg, 8.6 mmol) overnight at RT. The resin was filtered
off and the solvent removed under vacuum. The crude was re-dissolved
in dH2O (200 mL) and filtered using Amicon spin filters (10 kDa). The
polymeric conjugate was treated with aqueous 10% trifluoroacetic acid
(TFA) solution (0.5 mL) for 2 h at 378C. The solution was then filtered
using Amicon spin filters (10 kDa) to remove the excess amount of TFA.
The polymer aqueous solution was lyophilised to obtain compound 8 as a
white solid (0.78 mg, 76% yield). The coupling conversion was calculated
The microarrays were compartmentalised with a 16-well gasket. Lectin
solutions were prepared in PBST (PBS with 0.05% Tween-20) with
Chem. Eur. J. 2013, 19, 4776 – 4785
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4783

N.-C. Reichardt et al.
CaCl2 (1 mm) and MgCl2 (1 mm). Solutions (100 mL) of the corresponding
labelled lectins were incubated in the dark over each microarray for 1 h
at room temperature. Concentrations of 10 mg lectin per mL were used
for all lectins except for BS II, which was incubated at 50 mgmLꢀ1. The
slides were washed under standard conditions, dried with a slide spinner
and fluorescence emission was analysed with a microarray scanner.
Acknowledgements
Funding from the Ministerio de Ciencia
e Innovaciꢃn (projects
CTQ2008-04444/BQA and CTQ2011-27874), the government of the
Basque Country, an Etortek grant and the European Union (ITN-Euro-
glycoarrays grant) is gratefully acknowledged. We thank Dr. Ilya Revia-
kine and Dr. Ralf Richter, both CIC biomaGUNE, for helpful discus-
sions and Dr. Javier Calvo for the acquisition of high-resolution mass
spectra. We thank Prof. Dr. Iain B. H. Wilson from BOKU, Vienna, for
the P. pastoris clones expressing C. elegans glycosyltransferases and A.
thaliana fucosyltransferase.
Histograms that show the carbohydrate–lectin interactions were prepared
using Microsoft Excel software. Fluorescence measurements were per-
formed using an Agilent G265BA microarray scanner, Agilent Technolo-
gies, Santa Clara, USA. Image quantification was performed with Pro-
ScanArray Express software from Perkin–Elmer, Shelton, USA and in-
teraction profiles with the lectin array are displayed as histograms. The
adaptive circle method with a diameter range of 65–70 mm was employed
for spot quantification. The median value was used for the fluorescence
of each spot and for every carbohydrate an average of five replicate
spots was used to construct histograms that show the lectin-binding pro-
file. Error bars are included that show the standard deviation for each
carbohydrate–lectin interaction.
[1] T. Kristensen, J. D. Mikkelsen, J. P. Knox, Proteomics 2002, 2, 1666–
1671.
[2] F. Fazio, M. C. Bryan, O. Blixt, J. C. Paulson, C. H. Wong, J. Am.
Chem. Soc. 2002, 124, 14397–14402.
[3] B. T. Houseman, M. Mrksich, Chem. Biol. 2002, 9, 443–454.
[4] S. Fukui, T. Feizi, C. Galustian, A. M. Lawson, W. Chai, Nat. Bio-
technol. 2002, 20, 1011–1017.
[5] S. Park, I. Shin, Angew. Chem. 2002, 114, 3312–3314; Angew. Chem.
Int. Ed. 2002, 41, 3180–3182.
[6] D. Wang, S. Liu, B. J. Trummer, C. Deng, A. Wang, Nat. Biotechnol.
2002, 20, 275–281.
[7] O. Blixt, S. Head, T. Mondala, C. Scanlan, M. E. Huflejt, R. Alvarez,
M. C. Bryan, F. Fazio, D. Calarese, J. Stevens, Proc. Natl. Acad. Sci.
USA 2004, 101, 17033–17038.
Lectin array analysis of PEG-tagged glycan-coated fluorescent beads
Glycan immobilisation on fluorescent beads: Fluorescent (cy3) latex
beads 300 nm in diameter (50 mL, 5ꢂ109 beads) functionalised with car-
boxylic acid groups (1.08ꢂ106 groups per bead) were activated as NHS
ester by treatment with solutions of EDC (2 mm, 200 mL) and NHS
(1 mm) in H2O overnight at RT. After washing by centrifugation, the acti-
vated beads were incubated with the corresponding solution of amino-
pentyl glycan (three concentrations: 100 nmol per 200 mL, 10 nmol per
200 mL or 5 nmol per 200 mL) or PEGylated glycan (two concentrations:
10 nmol per 200 mL or 5 nmol per 200 mL) in phosphate buffer (30 mm,
pH 8.4) that contained 0.005% of Tween-20 overnight at RT. The beads
were quenched by the addition of a solution (50 mm, 500 mL) of ethanola-
mine in borate buffer (50 mm, pH 9.0). After centrifugation and washing
with phosphate buffer, the coated beads were stored in a final volume of
50 mL at 48C.
[8] C. D. Rillahan, J. C. Paulson, Annu. Rev. Biochem. 2011, 80, 797–
823.
[9] R. D. Cummings, Mol. Biosyst. 2009, 5, 1087–1104.
[10] H. Patrick, K. D. Janda, Acc. Chem. Res. 2000, 33, 546–554.
[11] T. Matsushita, I. Nagashima, M. Fumoto, T. Ohta, K. Yamada, H.
Shimizu, H. Hinou, K. Naruchi, T. Ito, H. Kondo, J. Am. Chem. Soc.
2010, 132, 16651–16656.
[12] N. C. Reichardt, M. Martin-Lomas, Angew. Chem. 2003, 115, 4822–
4825; Angew. Chem. Int. Ed. 2003, 42, 4674–4677.
[13] M. T. C. Walvoort, H. van den Elst, O. J. Plante, L. Krçck, P. H. See-
berger, H. S. Overkleeft, G. A. van der Marel, J. D. C. Codꢄe,
Angew. Chem. Int. Ed. 2012, 51, 4393–4396.
[14] L. Krçck, D. Esposito, B. Castagner, C. C. Wang, P. Bindschaedler,
P. H. Seeberger, Chem. Sci. 2012, 3, 1617–1622.
[15] K. S. Ko, F. A. Jaipuri, N. L. Pohl, J. Am. Chem. Soc. 2005, 127,
13162–13163.
[16] W. Reindl, K. Deng, J. Gladden, G. Cheng, A. Wong, S. Singer, S.
Singh, J. C. Lee, C. H. Yao, T. Hazen, Energy Environ. Sci. 2011, 4,
2884–2893.
[17] H. D. Edwards, S. K. Nagappayya, N. L. B. Pohl, Chem. Commun.
2012, 48, 510–512.
[18] T. Zhu, G. J. Boons, J. Am. Chem. Soc. 2000, 122, 10222–10223.
[19] R. Ojeda, J. L. de Paz, M. Martꢅn-Lomas, Chem. Commun. 2003,
2486–2487.
[20] N. Brinkmann, M. Malissard, M. Ramuz, U. Rçmer, T. Schumacher,
E. G. Berger, L. Elling, C. Wandrey, A. Liese, Bioorg. Med. Chem.
Lett. 2001, 11, 2503–2506.
[21] K. L. Prime, G. M. Whitesides, Science 1991, 252, 1164–1167.
[22] O. Blixt, T. Norberg, J. Org. Chem. 1998, 63, 2705–2710.
[23] S. Kopper, Carb. Res. 1994, 265, 161–166.
Lectin microarray preparation: Lectin solutions in print buffer
(0.4 mgmLꢀ1; PBS that contained 1% glycerol and 0.001% of Tween-20)
were freshly prepared for every print run to avoid activity loss due to
protein denaturation in a source plate (384-well plate) loaded with ali-
quots (20 mL) of every lectin, which could be stored overnight at 48C
without loss of function. These dilutions were spotted onto NHS-activat-
ed hydrogel glass slides (Nexterion H, Schott AG, Mainz, Germany) by
employing a dispensing volume of 0.8 nL and at 300 mm pitch between
spot centres and a robotic non-contact piezoelectric spotter (S11, Scien-
ion, Berlin).
Lectins ConA, WGA, RCA, ECA, SNA, MAL-I, AAL, PSA, LCA,
WFL and VVL were printed in 15 replicates as 15ꢂ12 matrixes in well
format. Printing buffer was also spotted as a negative control. Humidity
for the slide printing was maintained around 50% before and during
printing. The printed slides were incubated at 75% humidity in a custom-
made humidity chamber filled with saturated NaCl solution at 188C over-
night and stored at ꢀ208C without quenching unless the slides were used
immediately.
Before use of the slides for lectin-array blotting, unreacted NHS-activat-
ed carboxylate groups were quenched by placing the slides in a 50 mm
ethanolamine solution in 50 mm borate buffer for 1 h and washed with
PBS for 5 min. Slides were dried by centrifugation without any additional
washing step.
Interaction of the glyco-conjugated beads with the lectin array: The bead
solutions were diluted 1:100 in PBS buffer that contained 0.005% Tween-
20 and these solutions (100 mL) were loaded into the microarray wells
and incubated at RT for 1 h. The slides were washed under standard con-
ditions, dried with a slide spinner. Fluorescence emission was analysed
with a microarray scanner.
[24] U. Zehavi, M. Herchman, Carb. Res. 1984, 128, 160–164.
[25] U. Zehavi, S. Sadeh, M. Herchman, Carb. Res. 1983, 124, 23–34.
[26] S. Kçpper, U. Zehavi, React. Polym. 1994, 22, 171–180.
[27] O. Seitz, C. H. Wong, J. Am. Chem. Soc. 1997, 119, 8766–8776.
[28] W. Wang, L. Li, C. Jin, Y. Niu, S. Li, J. Ma, L. Li, Y. Liu, L. Cai, W.
Zhao, Bioorg. Med. Chem. Lett. 2011, 21, 5041–5044.
[29] M. Schuster, P. Wang, J. C. Paulson, C. H. Wong, J. Am. Chem. Soc.
1994, 116, 1135–1136.
[30] D. Schmidt, J. Thiem, Chem. Commun. 2000, 1919–1920.
[31] K. Yamada, E. Fujita, S. I. Nishimura, Carb. Res. 1997, 305, 443–
461.
[32] O. Blixt, T. Norberg, Carb. Res. 1999, 319, 80–91.
4784
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 4776 – 4785

GlycoPEG Tags
FULL PAPER
[33] J. F. Tolborg, L. Petersen, K. J. Jensen, C. Mayer, D. L. Jakeman,
R. A. J. Warren, S. G. Withers, J. Org. Chem. 2002, 67, 4143–4149.
[34] K. Yamada, S. I. Nishimura, Tetrahedron Lett. 1995, 36, 9493–9496.
[35] X. Huang, K. L. Witte, D. E. Bergbreiter, C. H. Wong, Adv. Synth.
Catal. 2001, 343, 675–681.
[42] P. H. Liang, S. K. Wang, C. H. Wong, J. Am. Chem. Soc. 2007, 129,
11177–11184.
[43] Z. Yang, J. A. Galloway, H. Yu, Langmuir 1999, 15, 8405–8411.
[44] S. Park, M. Lee, S. J. Pyo, I. Shin, J. Am. Chem. Soc. 2004, 126,
4812–4819.
[36] M. C. Galan, A. T. Tran, K. Bromfield, S. Rabbani, B. Ernst, Org.
Biomol. Chem. 2012, 10, 7091–7097.
[37] H. Ando, S. Manabe, Y. Nakahara, Y. Ito, Angew. Chem. 2001, 113,
4861; Angew. Chem. Int. Ed. 2001, 40, 4725.
[38] S. Serna, J. Etxebarria, N. Ruiz, M. Martin-Lomas, N.-C. Reichardt,
Chem-Eur. J. 2010, 16, 13163–13175.
[39] S. Serna, S. Yan, M. Martin-Lomas, I. Wilson, N. Reichardt, J. Am.
Chem. Soc. 2011, 133, 16495–16502.
[45] S. H. Chang, J. L. Han, S. Y. Tseng, H. Y. Lee, C. W. Lin, Y. C. Lin,
W. Y. Jeng, A. H. J. Wang, C. Y. Wu, C. H. Wong, J. Am. Chem. Soc.
2010.
[46] N. Levit-Binnun, A. B. Lindner, O. Zik, Z. Eshhar, E. Moses, Anal.
Chem. 2003, 75, 1436–1441.
[47] J. de La Fuente, A. Barrientos, T. Rojas, J. Rojo, A. Fernꢆndez, S.
Penadꢄs, Angew. Chem. 2001, 113, 2317; Angew. Chem. Int. Ed.
2001, 40, 2257.
[40] K. C. K. Swamy, N. N. B. Kumar, E. Balaraman, K. V. P. P. Kumar,
Chem. Rev. 2009, 109, 2551–2651.
[41] J. Etxebarria, J. Calvo, M. Martin-Lomas, N. C. Reichardt, ACS
Chem. Biol. 2012, 7, 1729–1737.
Received: November 8, 2012
Revised: January 10, 2013
Published online: February 11, 2013
Chem. Eur. J. 2013, 19, 4776 – 4785
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4785