Photo-click immobilization on quartz crystal microbalance sensors for selective carbohydrate-protein interaction analyses

Article abstract of DOI:10.1021/ac102781u

A photoclick method based on azide photoligation and Cu-catalyzed azide-alkyne cycloaddition has been evaluated for the immobilization of carbohydrates to polymeric materials. The biomolecular recognition properties of the materials have been investigated with regard to applicable polymeric substrates and selectivity of protein binding. The method was used to functionalize a range of polymeric surfaces (polystyrene, polyacrylamide, polyethylene glycol), poly(2-ethyl-2-oxazoline), and polypropene) with various carbohydrate structures (based on α-D-mannose, β-D-galactose, and N-acetyl-β-D-glucosamine). The functionalized surfaces were evaluated in real-time studies of protein-carbohydrate interactions using a quartz crystal microbalance flow-through system with a series of different carbohydrate-binding proteins (lectins). The method proved to be robust and versatile, resulting in a range of efficient sensors showing high and predictable protein selectivities.

Full text of DOI:10.1021/ac102781u

ARTICLE
pubs.acs.org/ac
Photo-Click Immobilization on Quartz Crystal Microbalance Sensors
for Selective Carbohydrate-Protein Interaction Analyses
Oscar Norberg,†,‡ Lingquan Deng,† Teodor Aastrup,‡ Mingdi Yan,*,§ and Olof Ramstr€om*,†
†Department of Chemistry, KTH - Royal Institute of Technology, Teknikringen 30, S-10044, Stockholm, Sweden
‡Attana AB, Bj€ornn€asv€agen 21, S-11347, Stockholm, Sweden
§Department of Chemistry, Portland State University, P.O. Box 751, Portland, Oregon 97207-0751, United States
ABSTRACT: A photoclick method based on azide photoligation
and Cu-catalyzed azide-alkyne cycloaddition has been evaluated
for the immobilization of carbohydrates to polymeric materials. The
biomolecular recognition properties of the materials have been
investigated with regard to applicable polymeric substrates and
selectivity of protein binding. The method was used to functionalize
a range of polymeric surfaces (polystyrene, polyacrylamide, poly-
(ethylene glycol), poly(2-ethyl-2-oxazoline), and polypropene)
with various carbohydrate structures (based on R-D-mannose, β-D-galactose, and N-acetyl-β-D-glucosamine). The functionalized surfaces
were evaluated in real-time studies of protein-carbohydrate interactions using a quartz crystal microbalance flow-through system with a series
of different carbohydrate-binding proteins (lectins). The method proved to be robust and versatile, resulting in a range of efficient sensors
showing high and predictable protein selectivities.
lycobiology is a rapidly growing research field encompassing
the effects of carbohydrates and glycoconjugates (glycans) in
compounds and are extensively used as precursors to obtain
amine-functionalized molecules, a feature which has resulted in
large libraries of available carbohydrate azides.27 The use of CuAAC
to immobilize carbohydrate azides in array formats have also been
applied to alkyne-functionalized microtiter plates,28 alkyne-linkers on
glass slides,29 and alkyne-functionalized gold surfaces.30
New methodologies for glycan arrays require rigorous valida-
tion procedures to ensure that the measured response indeed
correlates to the biological response. Most methodologies are
evaluated using glycans and receptors with known and corre-
sponding specificity for the chosen glycans. Lectins (unmodified
and FITC-labeled)10-12,23,28-35 and antibodies (mono- and
polyclonal)14,18,35-37 are the most commonly used validation
targets for glycan array methodologies. Lectins are nonenzymatic
proteins that bind oligosaccharides with high specificity, often in
di- or polyvalent complexes. Many lectins are selective for mono-
saccharides (e.g., R-D-mannose, β-D-galactose, or N-acetyl-β-D-
glucosamine) but they show up to a 1000-fold increase in affinity
for the corresponding di-, tri-, and tetrasaccharides.38
Glycan array methodologies are primarily designed for use in
high-throughput microarray printing and analysis with fluores-
cently tagged targets, or targets with fluorescently tagged anti-
bodies known to bind specifically to the target. To avoid the need
for fluorescently tagged targets, which are both expensive and
demanding to produce, many glycan array methodologies are
evaluated using techniques such as surface plasmon resonance
(SPR)30,39-43 and quartz crystal microbalance (QCM)23,32,44,45
G
biological systems and their interactions with proteins, cells, and other
biomolecules. The complex functions of glycans have so far been
difficult to explore due to several challenges, most notably the syn-
thesis and/or purification of glycans and the following analysis of
the biomolecular interactions.1-4 In order to robust and efficient
high-throughput analysis methods are required. Several potent glycan
array formats have for this reason been developed, where more or less
specific chemical ligation techniques are used to selectively immobi-
lizeglycansinanarrayformat.5-11 These techniques have accelerated
the development of the field, and glycan arrays have for example been
used to discover multivalency effects,12 evaluate blood serum glycan
binding13 and antibodies toward cancer,14 determine detailed binding
specificity for glycan-binding proteins15 and macrophage lectins,16
investigate the receptor binding properties of the influenza virus
H2N2,17 and identify cellular markers,18 enzymes involved in wound
healing,19 proteins involved in cancer metastasis,20 glycans modulat-
ing galectin-1 T cell death,21 and the glycome of HIV-1 virions.22
Herein we present the exploration of the photoclick immobi-
lization technique previously developed in our laboratory as a
general polymeric platform for protein-carbohydrate interac-
tions.23 Polymeric materials are both cheap and easily moldable,
qualities which are highly desirable in life sciences. Methodo-
logies based on polymeric materials are thus potentially useful for
large scale production. In addition, the physicochemical proper-
ties of polymers can be easily tuned. Copper-catalyzed azide-
alkyne cycloaddition (CuAAC) is a highly selective reaction
between azides and alkynes, chemical moieties that are both rare
and generally inert in biological systems, which forms a strong non-
hydrolyzable triazole linkage between the two molecules.24-26
Additionally, azides are easy to incorporate synthetically in organic
Received: October 22, 2010
Accepted: December 1, 2010
Published: December 16, 2010
r
2010 American Chemical Society
1000
dx.doi.org/10.1021/ac102781u Anal. Chem. 2011, 83, 1000–1007
|

Analytical Chemistry
ARTICLE
Scheme 1. Synthesis of Azide-Functionalized Ligands for Photoclick Immobilizationa
a (a) Ac2O, pyridine, r.t., 46 h (quant.); (b) 2-azidoethanol, BF3 Et2O, DCM, 0 °C, 18 h (83%); (c) NaOMe, MeOH, rt, 2.5 h (95%);
3
(d) 2-azidoethanol, BF3 Et2O, DCM, -40 °C, 24 h (69%); (e) (i) TMSOTf, 50 °C, 30 min, (ii) TEA, r.t., 10 min (98%); (f) 2-azidoethanol,
3
H2SO4, r.t., 19 h (23%); (g) NaN3, TBABr, 110 °C, 18 h (99%).
which instead rely on the physical properties of surface plasmon
absorption and mass change to detect target binding.
In this publication we explore the versatility of the platform
and protein selectivity of the photoclick immobilization techni-
que. The methodology was evaluated by protein binding analyses
on a range of polymeric materials functionalized with a series of
monosaccharides, using a QCM flow-though instrumentation.
with dichloromethane (DCM). The organic phase was washed with
saturated NaHCO3 (aq) and brine and dried over MgSO4. The
solvent was evaporated under reduced pressure yielding D-mannose
pentaacetate (2) as a pale yellow oil (10.3 g, quant.), with an R/β
ratio of 1.37:1. A solution of 2-azidoethanol (13, 424 mg, 4.88 mmol)
in DCM (20 mL) was subsequently added to compound 2 (1.59 g,
4.06 mmol). The mixture was stirred at 0 °C, and boron trifluoride
diethyl etherate (BF3 Et2O, 2.04 mL, 16.2 mmol) was added slowly.
3
After 1 h, the mixture was allowed to reach r.t., and after 18 h added to
ice-water (20 mL). The resulting mixture was extracted with DCM
(2 ꢀ 20 mL), and the combined organic phase was washed with
ice-water, saturated NaHCO3 (aq), and ice-water. The organic
phase was dried over Na2SO4 and the solvent evaporated under
reduced pressure yielding product 3 (1.41 g, 83%). 1H NMR (500
MHz, CDCl3): δ 5.34 (dd, 1 H, J = 3.5 and 9.8 Hz, H-3), 5.28 (d, 1
H, J = 10.1 Hz, H-4), 5.25 (dd, 1 H, J = 1.9 and 3.5 Hz, H-2), 4.85 (d,
1 H, J = 1.9 Hz, H-1), 4.26 (dd, 1 H, J = 5.4 and 12.3 Hz, H-6), 4.10
(dd, 1 H, J = 2.2 and 12.3 Hz, H-60), 4.02 (ddd, 1 H, J = 2.5, 5.4, and
9.8 Hz, H-5), 3.85 (ddd, 1 H, J = 3.8, 6.9, and 10.7 Hz, CH2CH2N3),
3.65 (ddd, 1 H, J = 3.9, 6.2, and 10.3 Hz, CH2CH2N3), 3.47 (ddd, 1
H, J = 3.5, 6.9, and 13.2 Hz, CH2N3), 3.42 (ddd, 1 H, J = 3.8, 6.0, and
13.2 Hz, CH2N3), 2.14 (s, 3 H, Ac), 2.08 (s, 3 H, Ac), 2.03 (s, 3 H,
Ac), 1.97 (s, 3 H, Ac). 13C NMR (125 MHz, CDCl3): δ 170.74,
170.13, 169.93, 169.88, 97.88, 69.53, 69.00, 68.98, 67.18, 66.14, 62.59,
50.49, 21.00, 20.87, 20.83, 20.78.
’ MATERIALS AND METHODS
General. All commercially available starting materials were of
reagent grade and used as received. Lectins were purchased from
Sigma-Aldrich and Vector Laboratories. QCM crystals were
obtained from Attana AB. 1H and 13C NMR data were recorded
on a Bruker Avance 400 instrument at 400 MHz (1H) or a Bruker
DMX 500 instrument at 500 MHz (1H) or 125 MHz (13C).
Chemical shifts are reported as δ values (ppm) with either CDCl3
(1H: δ = 7.26, 13C = 77.16) or D2O (1H: δ = 4.79) as internal
1
standard. J values are given in Hertz. H peak assignments were
made by first-order analysis of the spectra supported by standard
1H-1H correlation spectroscopy (COSY). Thin layer chromatog-
raphy (TLC) was performed on precoated Cromatofolios AL Silica
gel 60 F254 plates (Merck). Flash column chromatography was
performed on silica gel 60, 0.040-0.063 mm (SDS).
Synthesis. All compounds were synthesized in a few steps
with high yields according to previously published procedures
(Scheme 1). Compounds 4 and 7 were synthesized from the
corresponding pentaacetates (2 and 5) using a procedure by
Wong et al.28 Compound 11 was prepared from the peracety-
lated N-acetyl-β-D-glucosamine (8) through the oxazoline inter-
mediate according to Nakabayashi et al.46 and Åberg et al.47
2-Azidoethanol (13) was synthesized from 2-chloroethanol
according to Pfaendler and Weimar.48 Compounds 14, 15, and
16 were synthesized as previously reported by our group.23,32
1-(2-Azidoethyl)-2,3,4,6-tetra-O-acetyl-r-D-mannopyrano-
side (3). D-Mannose (1, 4.64 g, 25.7 mmol) and pyridine (1.0 mL,
12.9 mmol) were mixed in acetic anhydride (Ac2O, 150 mL) and
stirred at r.t. for 46 h, after which the solution turned clear. The
mixture was then poured on ice, and the water phase was extracted
1-(2-Azidoethyl)-r-D-mannopyranoside (4). 1-(2-Azidoethyl)-
2,3,4,6-tetra-O-acetyl-R-D-mannopyranoside (3, 66 mg, 0.16 mmol)
was dissolved in MeOH (30 mL). Then sodium methoxide (NaOMe,
25 mg, 0.47 mmol) was added, and the reaction mixture was
stirred at r.t. for 2.5 h until TLC indicated full conversion. Then
Amberlyst 15 was added until pH reached ∼7. The Amberlyst
was then filtered off, and the solvent was evaporated under
reduced pressure giving the pure product 4 (37 mg, 95%).
1H NMR (500 MHz, D2O): δ 4.93 (d, 1 H J = 1.6 Hz, H-1),
3.99 (dd, 1 H, J = 1.6 and 3.5 Hz, H-2), 3.93 (ddd, 1 H, J = 3.4, 7.0,
and 10.8 Hz, CH2CH2N3), 3.91 (dd, 1 H, J = 1.3 and 12.3 Hz,
H-6), 3.85 (dd, 1 H, J = 3.5 and 9.1 Hz, H-3), 3.77 (dd, 1 H, J =
5.7 and 12.0 Hz, H-60), 3.73 (ddd, 1 H, J = 3.2, 6.3, and 11.0 Hz,
CH2CH2N3), 3.70-3.65 (m, 2 H, H-4 and H-5), 3.56 (ddd, 1 H,
1001
dx.doi.org/10.1021/ac102781u |Anal. Chem. 2011, 83, 1000–1007

Analytical Chemistry
ARTICLE
J = 3.1, 6.9 and 13.6, CH2N3), 3.50 (ddd, 1 H, J = 3.2, 6.4
and 13.6, CH2N3). 13C NMR (125 MHz, D2O): δ 99.80, 72.88,
70.37, 69.92, 66.67, 66.29, 60.90, 50.19.
(5 mL) containing molecular sieves (4 Å). The mixture was stirred
for 30 min at r.t. after which concentrated sulfuric acid (three drops)
was added. The mixture was then stirred at r.t. for 19 h after which it
was diluted with dichloromethane and filtered through Celite. The
organic phase was washed with saturated NaHCO3 (aq) and ice-
water. The organic phase was dried (MgSO4) and the solvent
evaporated under reduced pressure. The crude product was sepa-
rated by flash column chromatography with solvent system Hex/
1-(2-Azidoethyl)-2,3,4,6-tetra-O-acetyl-β-D-galactopyrano-
side (6). β-D-Galactose pentaacetate (5, 652 mg, 1.67 mmol)
was dissolved in DCM (10 mL), after which 2-azidoethanol (13, 250
μL, 3.34 mmol) was added. The mixture was then stirred at -40 °C
and BF3 Et2O (1.07 mL, 8.52 mmol) was added slowly. After 24 h,
3
1
the temperature had reached 10 °C. The mixture was then added to
ice-water (20 mL) and extracted with DCM (2 ꢀ 20 mL). The
combined organic phases were washed with ice-water (20 mL),
saturated NaHCO3 (aq), and ice-water. The organic phase was
dried (MgSO4) and the solvent evaporated under reduced pressure.
The crude product was separated by flash column chromatography
with solvent system hexanes/ethyl acetate (Hex/EtOAc) 1:1 giving
pure 6 (482 mg, 69%). 1H NMR (500 MHz, CDCl3): δ 5.40 (d, J =
3.4 Hz, 1H, H-4), 5.24 (dd, J = 8.0 and 10.4 Hz, 1H, H-2), 5.03 (dd,
J= 3.4 and 10.5 Hz, 1H, H-3), 4.56 (d, J= 8.0 Hz, 1H, H-1), 4.19 (dd,
J = 6.6 and 11.3 Hz, 1H, H-6), 4.13 (dd, J = 6.8 and 11.2 Hz, 1H,
H-60), 4.04 (dt, J= 4.2 and 10.6 Hz, 1H, CH2), 3.93(t,J= 6.7 Hz, 1H,
H-5), 3.72-3.67 (m, 1H, CH2), 3.53-3.48 (m, 1H, CH2), 3.31 (dt,
J= 4.0 and 13.4 Hz, 1H, CH2),2.16(s,1H, Ac),2.07(s, 1H,Ac),2.05
(s, 1H, Ac), 1.99 (s, 1H, Ac). 13C NMR (125 MHz, CDCl3): δ
170.50, 170.35, 170.27, 169.60, 101.26, 71.02, 70.95, 68.66, 68.50,
67.14, 61.38, 50.69, 20.88, 20.78, 20.77, 20.68.
EtOAc 1:8 giving pure 10 (26 mg, 23%). H NMR (500 MHz,
CDCl3): δ 5.63 (d, J = 8.6 Hz, 1H, NH), 5.36 (t, J = 9.9 Hz, 1H,
H-3), 5.07 (t, J= 9.6 Hz, 1H, H-4), 4.83 (d, J= 8.3 Hz, 1H, H-1), 4.25
(dd, J= 4.8 and 12.3 Hz, 1H, H-6), 4.15 (dd, J = 2.4 and 12.3 Hz, 1H,
H-60), 4.04 (app. dt, J = 3.9 and 10.8 Hz, 1H, CH2), 3.81 (app. q,
J = 9.3 Hz, 1H, H-2), 3.74-3.68 (m, 2H, H-5 þ CH2), 3.53-3.48
(m, 1H, CH2), 3.26 (app. dt, J = 3.9 and 13.2 Hz, 1H, CH2), 2.08
(s, 3H, Ac), 2.02 (s, 3H, Ac), 2.02 (s, 3H, Ac), 1.95 (s, 3H, Ac). 13C
NMR (125 MHz, CDCl3): δ170.93, 170.84, 170.61, 169.57, 100.64,
72.22, 72.14, 68.72, 68.57, 62.15, 55.09, 50.77, 23.53, 20.90, 20.83,
20.79.
1-(2-Azidoethyl)-2-acetamido-β-D-glucopyranoside (11).
1-(2-Azidoethyl)-2-acetamido-3,4,6-tri-O-acetyl-β-D-glucopyrano-
side (10, 101 mg, 0.241 mmol) was dissolved in MeOH (15 mL).
Then sodium methoxide (14 mg, 0.27 mmol) was added, and the
reaction mixture was stirred at r.t. for 4 h until TLC indicated full
conversion. Then Amberlyst 15 was added until the pH reached
∼7. The Amberlyst was then filtered off, and the solvent was
evaporated under reduced pressure giving the pure product 11
(67 mg, 95%). 1H NMR (500 MHz, D2O): δ 4.58 (d, 1H, J =
8.6 Hz, H-1), 4.05 (ddd, 1H, J = 2.8, 5.3, and 11.2 Hz,
CH2CH2N3), 3.92 (app. dd, 1H, J = 1.6 and 12.5 Hz, H-6),
3.87-3.70 (m, 3H, H-2, H-5 and CH2CH2N3), 3.54 (app. t, 1H,
J = 9.2 Hz, H-3), 3.50-3.39 (m, 4H, H-4, H-60 and 2 ꢀ CH2N3),
2.04 (s, 3H, NAc). 13C NMR (125 MHz, D2O): δ 169.63, 95.96,
70.83, 68.79, 64.81, 63.65, 55.62, 45.28, 25.14, 17.15.
2-Azidoethanol (12). 2-Chloroethanol (2.09 g, 26.0 mmol),
sodium azide (5.1 g, 78 mmol), and tetrabutylammonium bromide
(TBABr, 836 mg, 2.60 mmol) were mixed in a round-bottom
flask equipped with a reflux condenser and stirred at 110 °C for
18 h (using a safety shield). The mixture was then diluted with
diethyl ether, and the solid byproducts were filtered off. The
solvent was evaporated under reduced pressure (no heating),
giving a yellow oil. The crude product was purified by distillation
at 12 mbar, yielding 13 as a colorless oil (2.23 g, 99%). 1H NMR
(500 MHz, CDCl3): δ 3.78 (app. q, 2H, J = 5.2 Hz, HOCH2-
CH2N3), 3.45 (t, 2H, J = 5.2 Hz, HOCH2CH2N3), 1.77 (t, 1H,
J = 5.7 Hz, HOCH2CH2N3). 13C NMR (125 MHz, CDCl3):
δ 61.71, 53.71.
General Surface Modification. All QCM experiments were
performed on gold-plated 10 MHz quartz crystals (Attana) coated
with polymeric layers. Polystyrene-coated crystals were purchased
from Attana. All other surfaces were spin-coated using a Cookson
Electronics Specialty Coating Systems Spincoater model P6708D
and solutions of the selected polymers. The photoreaction step
was performed at 240-400 nm at a measured intensity of 13.3-
13.5 mW/cm2 with a LC8 equipped Hg-Xe UV-lamp from
Hamamatsu Photonics. The fabricated crystals were mounted in
a flow-through QCM system (Attana A100 C-Fast).
1-(2-Azidoethyl)-β-D-galactopyranoside (7). 1-(2-Azidoethyl)-
2,3,4,6-tetra-O-acetyl-β-D-galactopyranoside (6, 480 mg, 1.15 mmol)
was dissolved in MeOH (10 mL) under nitrogen atmosphere.
Then sodium methoxide (74 mg, 1.4 mmol) in methanol was
added with syringe, and the reaction mixture was stirred at r.t. for
4 h after which TLC indicated full conversion. Then Amberlyst
15 was added under gentle stirring until the pH reached ∼7. The
Amberlyst was then filtered off, and the solvent was evaporated
under reduced pressure to give pure 7 (286 mg, quant.). 1H NMR
(500 MHz, D2O): δ 4.43 (d, J = 7.9 Hz, 1H, H-1), 4.06 (dt, J = 4.9
and 11.3, 1H, CH2), 3.92 (d, J = 3.4 Hz, 1H, H-4), 3.86-3-73
(m, 3H, CH2, H-6 and H-60), 3.69(dd, J= 4.4 and 7.8 Hz, 1H, H-5),
3.65 (dd, J = 3.5 and 10.0 Hz, 1H, H-3), 3.56-3.51 (m, 3H, 2 ꢀ
CH2 and H-2). 13C NMR (125 MHz, D2O): δ 152.92, 102.88,
99.77, 75.17, 72.69, 70.67, 68.61, 68.37, 60.93, 56.11, 50.54, 20.27.
2,3-Dihydrooxazole-3,4,6-tri-O-acetyl-β-D-glucopyrano-
side (9). 2-Acetamido-2-deoxy-1,3,4,6-tetra-O-acetyl-β-D-gluco-
pyranose (8, 210 mg, 0.539 mmol) was dissolved in dichloroethane
(15 mL). Then trimethylsilyl trifluoromethanesulfonate (TMSOTf,
97 μL, 0.59 mmol) was added, and the mixture was stirred at 50 °C
for 30 min after which TLC indicated full conversion. The mixture
was then removed from the heat, and triethylamine (TEA, 225 μL,
1.62 mmol) was added. The mixture was then stirred at r.t. for
10 min after which it was passed through a short plug of silica and
washed carefully with DCM and EtOAc. The solvent was then
evaporated under reduced pressure, and the crude product was
separated by flash column chromatography with EtOAc as solvent
1
giving 9 as a colorless oil (174 mg, 98%). H NMR (500 MHz,
CDCl3): δ5.86 (d, J= 7.4 Hz, 1H, H-1), 5.14 (app. t, J=2.1 Hz, 1H,
H-5), 4.80 (d, J = 9.3 Hz, 1H, H-4), 4.07-3.99 (m, 3H, H-2, H-6
and H-60), 3.51-3.57 (m, 1H, H-3), 2.00 (s, 3H, Ac), 1.98 (s, 6H,
2 ꢀ Ac), 1.97 (s, 3H, Ac). 13C NMR (125 MHz, CDCl3): δ 170.41,
169.39, 169.08, 166.51, 99.27, 70.24, 68.27, 67.38, 64.81, 63.22,
20.76, 20.69, 20.59, 13.80.
1-(2-Azidoethyl)-2-acetamido-3,4,6-tri-O-acetyl-β-D-gluco-
pyranoside (10). 2,3-Dihydrooxazole-3,4,6-tri-O-acetyl-β-D-gluco-
pyranoside (9, 90 mg, 0.27 mmol) was dissolved in dichloromethane
Surface Functionalization (cf. Figure 1). All polymeric
surfaces except for commercial polystyrene-coated surfaces were
produced according to Figure 1 by a modified procedure pre-
viously developed in our laboratory.32 The gold-plated quartz
crystals were soaked in a solution of PFPA-disulfide (14, 14 mM)
1002
dx.doi.org/10.1021/ac102781u |Anal. Chem. 2011, 83, 1000–1007

Analytical Chemistry
ARTICLE
Figure 1. General surface functionalization procedure. Gold surfaces were soaked in a solution of compound 14 overnight, after which they were spin-
coated with polymer solutions, irradiated with UV light for 5 min and subsequently washed. The polymeric surfaces were then spin-coated with a solution
of compound 15 and irradiated with UV light for 5 min. The NHS-activated ester surfaces were linked through an amidation process with amine 16 to
produce the alkyne functionalized surfaces. The alkyne surfaces were then differentiated with several azide functionalized molecules through CuAAC
chemistry.
Table 1. Polymer Spin-Coating Parameters
polymer
Mw
solvent
concn, mg/mL
drop size, μL
wash solvent
polyacrylamide (PAAm)
poly(ethylene glycol) (PEG)
poly(2-ethyl-2-oxazoline) (PEOX)
polypropylene (PP)
5-6 ꢀ 106
20 ꢀ 103
ethylene glycol
dimethoxyethane
10
10
10
10
10
15
15
10
EtOH
EtOH
CHCl3
R-130
200 ꢀ 103
250 ꢀ 103
chloroform
1,1,2,2-tetrachloroethane (R-130)
in dichloromethane overnight (protected from light). The next
day, they were removed from the solution and dried under a
gentle stream of nitrogen, spin-coated with a solution of the
selected polymer (see Table 1) at 2000 rpm for 120 s, and
irradiated for 5 min with UV light. The crystals were then washed
(see Table 1) and dried under a gentle stream of nitrogen.
Subsequently, the crystals (including polystyrene) were spin-
coated with a solution of PFPA-NHS (15, 10 mM, 10 μL) in
ethanol at 1500 rpm for 180 s followed by immersion in a
solution of linker 16 (53 mM) in acetonitrile for 8 h. The crystals
were then washed with acetonitrile before the CuAAC-step. The
crystals were soaked in a solution of azide 4, 7, 11, or 13
(32 mM), CuI (1 mM), and diisopropylethylamine (DIPEA,
13 mM) in water:acetonitrile (1:1) for 18 h after which they were
thoroughly rinsed with water:acetonitrile (1:1) and methanol,
dried under a gentle stream of nitrogen, and placed in the QCM
flow-through system.
lectin II (BS-II) were prepared in the same buffer. The crys-
tals were washed/equilibrated with buffer solution prior to
manipulations/measurements. After the crystal was equilibrated
in the flow-through system, it was subjected to five injections of
Bovine Serum Albumin (BSA, 0.5 mg/mL), three injections of
low pH buffer (pH 1.5), and finally two additional injections of
Bovine Serum Albumin (BSA, 0.5 mg/mL) to fully block any
nonfunctionalized surface. Binding to the surface was monitored
by frequency logging with Attester 1.1 (Attana), and adsorption/
desorption to the surface recorded as the resulting frequency
shifts. Solutions (0.2 mg/mL) of each lectin were then injected
into the system, where the resulting shift in frequency corresponds
to the binding to the surface. Bound lectin was released from the
surface between measurements by eight successive injections of low
pH buffer (PBS 10 mM, pH 1.5). The procedure was then repeated
two times for each lectin to give an average value and determine
the surface stability over time.
General Surface Analysis. A continuous flow of running
buffer (PBS 10 mM, pH 7.4, 25 μL/min) was used throughout
the experiments, and samples of concanavalin A (Con A), wheat
germ agglutinin (WGA), Pisum sativum agglutinin (PSA), Ricinus
communis agglutinin I (RCA-I), and Bandeiraea simplicifolia
’ RESULTS AND DISCUSSION
The photoclick immobilization procedure presented in
Figure 1 is a convergent approach that consists of three steps;
1003
dx.doi.org/10.1021/ac102781u |Anal. Chem. 2011, 83, 1000–1007

Analytical Chemistry
ARTICLE
Table 2. Properties of Lectins Used in the Binding Study
name
Mw, kDa
binding valency
carbohydrate specificity
family
reported pI
measured pI
refs
52-55
56
Con A
WGA
104
36
homotetramer
homodimer
Man, Glc
GlcNAc
Man, Glc
Gal
legume
cereal
8.3552
8.756
5.9 and 7.057
6.5-8
n.d.a
57,58
59,60
61
PSA (seeds)
RCA-I (seeds)
BS-II (seeds)
48
heterotetramer
heterotetramer
homotetramer
legume
4.5-7.3
6.6-9.3
5-7.3
120
120
Euphorbiaceae
legume
-
-
GlcNAc
a n.d. = not detectable.
photochemical insertion of perfluorophenyl azides into the
polymeric material, amide coupling to produce a triethylene
glycol-linked alkyne surface, and, the last step, copper-catalyzed
azide-alkyne cycloaddition, which is the diversifying step of the
procedure. Since only the last step is dissimilar for production of
different carbohydrate-functionalized surfaces, the methodology
allows for large scale production of the intermediate alkyne
surfaces, which, after production, can be stored for a long time
before use (no detectable reduction in activity after weeks of
storage at þ4 °C in water). The photochemical step being very
fast, the overall surface preparation time is governed primarily by
the amidation and the CuAAC reaction steps. In the present
setup, the crystals were submerged during these steps in a
solution of the reactants (g300 μL) to ensure full coverage of
the crystal surface, and the reaction times can likely be further
reduced with smaller crystal formats and higher concentrations.
One of the advantages of the functionalization method is that
relatively simple organic molecules, which require short and
facile synthesis, can be employed. The NHS-activated ester (15),
the alkyne linker (16), and the azide derivatives (4, 7, 11 and 13)
were all synthesized in few steps with good overall yields. Both
the NHS-activated ester (15) and the alkyne linker (16) can be
stored in solution at low temperature for extended times,
enabling repeated use of one batch of compounds. Since the
functionalization is only diversified in the final step, expansion of
the format can be efficiently achieved using arrays of carbo-
hydrate azides.
For instrumentation reasons, gold-plated quartz-crystals were
used as platforms for the functionalization methodology, thus
requiring the production of polymeric layers on the gold surfaces.
The polymeric surfaces used in the study were therefore pro-
duced by the spin-coating of dissolved polymeric materials on
PFPA-derivatized polished gold surfaces and subsequent photo-
irradiation to covalently attach the polymers to the gold surface
(Figure 1).32 The polymers were chosen to cover a range of
polymeric materials to illustrate the versatility of the method.
Both hydrophobic (PS and PP) and hydrophilic (PAAm, PEG,
and PEOX) polymers with a wide range of molecular weights were
used (Table 1). The solvents used to dissolve the polymers for spin-
coating were chosen based on previous experience in spin-coating
(PEG and PEOX49) and solubility properties of the polymers. The
solvation of the hydrophobic polymer polypropene proved in this
context especially difficult. Only with a combination of 1,1,2,2-
tetrachloroethane and heating could the polymer be dissolved. After
the spin-coating, the crystals were irradiated with UV light for 5 min
before washing with a suitable solvent. This time, the solvent was
chosen so that it would be possible to wash off any unbound
polymer as well as solvent residues from the spin-coating (ethylene
glycol and dimethoxyethane) and at the same time dry the surface
on a reasonable time scale.
The surfaces produced were then functionalized with alkyne
moieties in two steps. First, the polymeric surfaces were spin-
coated with an ethanol solution of PFPA-NHS 15 followed by
irradiation with UV light. Second, the QCM-crystals were
immersed in an acetonitrile solution of alkyne linker 16, where
the amine was coupled to the activated ester, forming a stable
amide bond. The two-step method was chosen over the faster
direct alkyne functionalization method (previously reported by
our group23) since the two-step methodology produced more
consistent surfaces with shorter response times in the QCM
instrumentation. The alkyne surfaces were then immersed in a
water/acetonitrile (1:1) solution of the carbohydrate azides in
the presence of a catalytic amount of CuI and diisopropylethyl-
amine. CuI and diisopropylethylamine were chosen over the
more traditional catalyst, in situ reduction of CuSO4 by sodium
ascorbate, in order to avoid the use of ascorbate in the metho-
dology. Since the reduction of CuII by ascorbate is a radical two-
step oxidation process of ascorbate, possible side reactions of the
radical intermediate to the surface are highly undesirable. The
functionalized surfaces were subsequently washed with water/
acetonitrile and methanol to remove unreacted reagents and dry
the surface before analysis. The surfaces were thus functionalized
with three different monosaccharide moieties (R-D-mannose,
β-D-galactose, or N-acetyl-β-D-glucosamine) which encom-
passed a short spacer between the carbohydrate and the azide
in order to reduce potential interference in the subsequent
protein binding event, due to a too closely attached triazole
moiety. The polymeric surfaces were also functionalized with
2-azidoethanol to produce control surfaces.
The method was evaluated using a QCM flow-through
instrumentation with repeated injections of lectins specific for
the monosaccharides used in the study. The lectins and carbo-
hydrates were chosen based on binding specificity, molecular
weight, and previous use of lectins in surface analysis methodol-
ogy (Table 2). Con A and PSA were chosen because of their
specificity toward R-D-mannose-residues, whereas WGA and BS-
II bind selectively to N-acetyl-β-D-glucosamine and RCA-I to
β-D-galactose. To minimize nonspecific binding, BSA was used to
block all nonderivatized surfaces by several injections in the
beginning of the experiment. Previous experience with the
instrumentation shows that blocking with BSA in the initial
phase of the experiment lasts throughout the whole experiment,
thus obviating the need for BSA in the running buffer.23,50,51
The protein-carbohydrate interactions were monitored in
real time, and both association and dissociation of the lectins to
the surfaces could be observed (Figure 2). Small levels of potential
nonspecific binding of several of the lectins were noticed, hypo-
thetically owing to, for example, surface structure and composi-
tion, blocking coverage by BSA, pI values of the lectins, etc.
A strategy to correct for this nonspecific binding was therefore
1004
dx.doi.org/10.1021/ac102781u |Anal. Chem. 2011, 83, 1000–1007

Analytical Chemistry
ARTICLE
adopted, based on the binding to surfaces produced using
2-azidoethanol, resulting in good specific affinities for all carbo-
hydrate surfaces tested. An average binding value was thus
obtained for each lectin on each surface with good reproduci-
bility over time (Figure 3). High binding selectivities were found
for all lectins evaluated and followed the predicted trends
indicated in Table 2. Thus, Con A and PSA displayed binding
to the mannose surfaces, WGA and BS-II to the N-acetylgluco-
samine surfaces, and RCA-I to the galactose surfaces (Figure 3).
The difference in quantitative binding between the different
lectins is primarily due to parameters such as the lectin molec-
ular weight and the affinity for the monosaccharides used in the
study.
The lectin binding profiles to all carbohydrate-functionalized
sensor surfaces are represented in Figure 4. From the binding
analysis it can be noted that Con A proved to be the most
consistent lectin in the study with more than 40 Hz corrected
binding to all mannose-functionalized surfaces and with insignifi-
cant nonspecific binding. PSA followed the pattern of Con A
although the selective binding was lower than for Con A. The
other lectins showed a variance in binding between the polymeric
materials; e.g., low binding of WGA and BS-II to the GlcNAc/
PAAm surface and high binding to GlcNAc/PEG. RCA-I also
showed some variance in binding to the galactose-functionalized
surfaces although less than for WGA/BS-II to the GlcNAc-
surfaces. The binding analysis of RCA-I resulted in a low residual
negative binding to some of the mismatching carbohydrate
surfaces (Man/PS, GlcNAc/PS, and GlcNAc/PAAm), within
the limits of sensitivity.
Figure 2. General lectin binding behavior to functionalized surfaces, in
this case to GlcNAc-functionalized polystyrene.
Comparison of the differences in sensitivity between the
different polymeric surfaces shows that the polystyrene surfaces
in general had a resulting negative binding to the mismatched
carbohydrate-functionalized surfaces (e.g., RCA-I to Man and
PSA to GlcNAc) whereas the polypropylene surfaces had a
resulting positive binding (e.g., PSA to GlcNAc and Con A to Gal).
The water-soluble polymeric surfaces showed a more coherent
binding than the hydrophobic polymeric surfaces, shown clearly by
the more coherent corrected lower binding values to the corrected
mismatched surfaces (except for WGA to Man/PEOX and RCA-I to
GlcNAc/PAAm).
Figure 3. Corrected lectin binding to PEG-functionalized surfaces. Each
value represents an averaged binding of two separate injections of protein,
subtracted with the averaged binding to the 2-azidoethanol surface.
Figure 4. Corrected binding values of lectins to the functionalized surfaces: (a) poly(ethylene glycol) (PEG), (b) poly(2-ethyl-2-oxazoline) (PEOX),
(c) polyacrylamide (PAAm), (d) polystyrene (PS), and (e) polypropylene (PP).
1005
dx.doi.org/10.1021/ac102781u |Anal. Chem. 2011, 83, 1000–1007

Analytical Chemistry
ARTICLE
A hypothetical reason for nonspecific binding is ionic binding
of the lectins to the blocking agent on the surface (immobilized
BSA in the primary blocking step). Since BSA has a relatively low
pI value (4.7) it would essentially be positively charged at the
buffer pH (7.4), thereby creating a partial positively charged
surface. The lectins which showed the most nonspecific binding
(WGA, Con A, and RCA-I) all have pI values above 7.4, resulting
in a net negative charge at the buffer pH. The lectins with a pI
value below the buffer pH (PSA and BS-II) both show a lower
nonspecific binding which is expected since there would be no
ionic attraction between the lectin and the surfaces. Further, both
PSA and BS-II showed a general smaller binding even to the
matched carbohydrate surfaces, which fits well with the expected
behavior if a repulsive ionic interaction exists, in coherence with
recently published material regarding antibody immobilization to
functionalized surfaces.62 Any potential ionic attraction by lectins
with a pI value above the buffer pH is however not a problem per
se, since it can be easily corrected for. Ionic repulsion, on the
other hand, may present a problem if it results in reduced binding
below the detection limit. To circumvent this possible problem,
the use of other blocking agents such as Tween-20 is currently
being investigated.
’ REFERENCES
(1) Gabius, H.-J., Ed. The Sugar Code: Fundamentals of Glycosciences;
Wiley-VCH: Weinheim, 2009.
(2) Moran, A. P.; Holst, O.; Brennan, P. J.; von Itzstein, M., Eds.
Microbial Glycobiology; Elsevier: London, 2009.
(3) Seeberger, P. H. Nat. Chem. Biol. 2009, 5, 368–372.
(4) Varki, A.; Cummings, R. D.; Esko, J. D.; Freeze, H. H.; Stanley,
P.; Bertozzi, C. R.; Hart, G. W.; Etzler, M. E., Eds. Essentials of
glycobiology; Cold Spring Harbor Laboratory Press: New York, 2009.
(5) Horlacher, T.; Seeberger, P. H. Chem. Soc. Rev. 2008, 37, 1414–
1422.
(6) Park, S.; Lee, M.-R.; Shin, I. Chem. Soc. Rev. 2008, 37, 1579–
1591.
(7) Krishnamoorthy, L.; Mahal, L. K. ACS Chem. Biol. 2009, 4, 715–
732.
(8) Oyelaran, O.; Gildersleeve, J. C. Curr. Opin. Chem. Biol. 2009, 13,
406–413.
(9) Wu, C.-Y.; Liang, P.-H.; Wong, C.-H. Org. Biomol. Chem. 2009,
7, 2247–2254.
(10) Lee, M.-r.; Shin, I. Angew. Chem., Int. Ed. 2005, 44, 2881–2884.
(11) Park, S.; Shin, I. Org. Lett. 2007, 9, 1675–1678.
(12) Branderhorst, H. M.; Ruijtenbeek, R.; Liskamp, R. M. J.; Pieters,
R. J. ChemBioChem 2008, 9, 1836–1844.
(13) Huflejt, M. E.; Vuskovic, M.; Vasiliu, D.; Xu, H.; Obukhova, P.;
Shilova, N.; Tuzikov, A.; Galanina, O.; Arun, B.; Lu, K.; Bovin, N. Mol.
Immunol. 2009, 46, 3037–3049.
(14) Huang, C.-Y.; Thayer, D. A.; Chang, A. Y.; Best, M. D.;
Hoffmann, J.; Head, S.; Wong, C.-H. Proc. Natl. Acad. Sci. U.S.A.
2006, 103, 15–20.
’ CONCLUSIONS
It has been shown that the photoclick surface functionalization
method is highly versatile, resulting in robust carbohydrate sensor
surfaces with high protein selectivities. Sensors were fabricated
using a panel of carbohydrate structures and evaluated using a
QCM setup for lectin binding. The predicted protein binding
could thus be efficiently demonstrated, validating the technique
with other formats. A variety of polymeric substrate materials
were further evaluated, generally showing good compatibility
with the method. As expected, poly(ethylene glycol)-based
sensors proved very good, but other formats also showed high
potential. The polymeric substrates enable control of the non-
specific binding, and only low levels were found in the present
study. Further optimization with respect to polymer structure,
composition, and size is ongoing. The method is inherently
robust, requiring generally low-cost starting materials, and can be
performed under ambient conditions. The photochemical step
also allows for the use of photolithography to easily produce
alkyne-functionalized arrays. In addition, the convergent fabrica-
tion approach constitutes a versatile platform for sensing and
glycoarray development.
(15) Porter, A.; Yue, T.; Heeringa, L.; Day, S.; Suh, E.; Haab, B. B.
Glycobiology 2010, 20, 369–380.
(16) Singh, S. K.; Streng-Ouwehand, I.; Litjens, M.; Weelij, D. R.;
García-Vallejo, J. J.; van Vliet, S. J.; Saeland, E.; van Kooyk, Y. Mol.
Immunol. 2009, 46, 1240–1249.
(17) Xu, R.; McBride, R.; Paulson, J. C.; Basler, C. F.; Wilson, I. A.
J. Virol. 2010, 84, 1715–1721.
(18) Wang, D.; Liu, S.; Trummer, B. J.; Deng, C.; Wang, A. Nat.
Biotechnol. 2002, 20, 275–281.
(19) Saravanan, C.; Cao, Z.; Head, S. R.; Panjwani, N. Glycobiology
2010, 20, 13–23.
(20) Hatakeyama, S.; Sugihara, K.; Nakayama, J.; Akama, T. O.;
Wong, S.-M. A.; Kawashima, H.; Zhang, J.; Smith, D. F.; Ohyama, C.;
Fukuda, M.; Fukuda, M. N. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 3095–
3100.
(21) Earl, L. A.; Bi, S.; Baum, L. G. J. Biol. Chem. 2010, 285, 2232–2244.
(22) Krishnamoorthy, L.; Bess, J. W.; Preston, A. B.; Nagashima, K.;
Mahal, L. K. Nat. Chem. Biol. 2009, 5, 244–250.
(23) Norberg, O.; Deng, L.; Yan, M.; Ramstr€om, O. Bioconjugate
Chem. 2009, 20, 2364–2370.
(24) Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B.
Angew. Chem., Int. Ed. 2002, 41, 2596–2599.
(25) Tornøe, C. W.; Christensen, C.; Meldal, M. J. Org. Chem. 2002,
67, 3057–3064.
’ AUTHOR INFORMATION
Corresponding Author
*O.R.: E-mail: ramstrom@kth.se. Fax: þ46 8 7912333. M.Y.:
(26) Wang, Q.; Chan, T. R.; Hilgraf, R.; Fokin, V. V.; Sharpless,
K. B.; Finn, M. G. J. Am. Chem. Soc. 2003, 125, 3192–3193.
(27) Consortium for Functional Glycomics (CFG), http://www.
functionalglycomics.org/.
E-mail: yanm@pdx.edu. Fax: þ1 503 725-9525.
(28) Fazio, F.; Bryan, M. C.; Blixt, O.; Paulson, J. C.; Wong, C.-H.
J. Am. Chem. Soc. 2002, 124, 14397–14402.
(29) Sun, X.-L.; Stabler, C. L.; Cazalis, C. S.; Chaikof, E. L. Bio-
conjugate Chem. 2006, 17, 52–57.
(30) Zhang, Y.; Luo, S.; Tang, Y.; Yu, L.; Hou, K.-Y.; Cheng, J.-P.;
Zeng, X.; Wang, P. G. Anal. Chem. 2006, 78, 2001–2008.
(31) Love, K. R.; Seeberger, P. H. Angew. Chem., Int. Ed. 2002, 41,
3583–3586.
(32) Pei, Y.; Yu, H.; Pei, Z.; Theurer, M.; Ammer, C.; Andre, S.;
Gabius, H.-J.; Yan, M.; Ramstr€om, O. Anal. Chem. 2007, 79, 6897–6902.
’ ACKNOWLEDGMENT
This study was supported in part by the Swedish Research
Council, the Royal Institute of Technology, the National In-
stitutes of General Medical Science (NIGMS) under NIH Award
Numbers R01GM080295 and 2R15GM066279, ARL-ONAMI
Center for Nanoarchitectures for Enhanced Performance,
and the European Commission (MRTN-CT-19561). L.D.
thanks the China Scholarship Council for a special scholarship
award.
1006
dx.doi.org/10.1021/ac102781u |Anal. Chem. 2011, 83, 1000–1007

Analytical Chemistry
ARTICLE
(33) Pei, Z.; Yu, H.; Theurer, M.; Walden, A.; Nilsson, P.; Yan, M.;
Ramstr€om, O. ChemBioChem 2007, 8, 166–168.
(34) Michel, O.; Ravoo, B. J. Langmuir 2008, 24, 12116–12118.
(35) Park, S.; Lee, M.-R.; Shin, I. Bioconjugate Chem. 2009, 20,
155–162.
(36) Fukui, S.; Feizi, T.; Galustian, C.; Lawson, A. M.; Chai, W. Nat.
Biotechnol. 2002, 20, 1011–1017.
(37) Manimala, J. C.; Roach, T. A.; Li, Z.; Gildersleeve, J. C.
Glycobiology 2007, 17, 17C–23.
(38) Lis, H.; Sharon, N. Chem. Rev. 1998, 98, 637–674.
ꢀ
(39) Mu~noz, F. J.; Pꢀerez, J.; Rumbero, A.; Santos, J. I.; Ca~nada, F. J.;
Andrꢀe, S.; Gabius, H.-J.; Jimꢀenez-Barbero, J.; Sinisterra, J.; Hernꢀaiz, M. J.
Bioconjugate Chem. 2009, 20, 673–682.
(40) Dhayal, M.; Ratner, D. M. Langmuir 2009, 25, 2181–2187.
(41) Uzawa, H.; Ohga, K.; Shinozaki, Y.; Ohsawa, I.; Nagatsuka, T.;
Seto, Y.; Nishida, Y. Biosens. Bioelectron. 2008, 24, 923–927.
(42) Linman, M. J.; Taylor, J. D.; Yu, H.; Chen, X.; Cheng, Q. Anal.
Chem. 2008, 80, 4007–4013.
(43) de Boer, A.; Hokke, C.; Deelder, A.; Wuhrer, M. Glycoconjugate
J. 2008, 25, 75–84.
(44) Mahon, E.; Aastrup, T.; Barboiu, M. Chem. Commun. 2010, 46,
2441–2443.
(45) Lyu, Y.-K.; Lim, K.-R.; Lee, B. Y.; Kim, K. S.; Lee, W.-Y. Chem.
Commun. 2008, 4771–4773.
(46) Nakabayashi, S.; Warren, C. D.; Jeanloz, R. W. Carbohydr. Res.
1986, 150, c7–c10.
(47) Eklind, K.; Gustafsson, R.; Tidꢀen, A. K.; Norberg, T.; Åberg,
P. M. J. Carbohydr. Chem. 1996, 15, 1161–1178.
(48) Pfaendler, H. R.; Weimar, V. Synthesis 1996, 1996, 1345–1349.
(49) Yan, M.; Ren, J. Chem. Mater. 2004, 16, 1627–1632.
(50) Pei, Z.; Larsson, R.; Aastrup, T.; Anderson, H.; Lehn, J.-M.;
Ramstr€om, O. Biosens. Bioelectron. 2006, 22, 42–48.
(51) Pei, Z.; Anderson, H.; Aastrup, T.; Ramstr€om, O. Biosens.
Bioelectron. 2005, 21, 60–66.
(52) Bhattacharyya, L.; Brewer, C. F. J. Chromatogr. A 1990, 502,
131–142.
(53) Kalb, A. J.; Lustig, A. Biochim. Biophys. Acta 1968, 168,
366–367.
(54) Wang, J. L.; Cunningham, B. A.; Edelman, G. M. Proc. Natl.
Acad. Sci. U.S.A. 1971, 68, 1130–1134.
(55) Wang, J. L.; Edelman, G. M. J. Biol. Chem. 1978, 253, 3000–
3007.
(56) Rice, R. H.; Etzler, M. E. Biochemistry 1975, 14, 4093–4099.
(57) Entlicher, G.; Kocourek, J. Biochim. Biophys. Acta 1975, 393,
165–169.
(58) Díaz, C. L.; Hosselet, M.; Logman, G. J. J.; Driessche, E.;
Lugtenberg, B. J. J.; Kijne, J. W. Planta 1990, 181, 451–461.
(59) Ready, M.; Wilson, K.; Piatak, M.; Robertus, J. D. J. Biol. Chem.
1984, 259, 15252–15256.
(60) Sweeney, E. C.; Tonevitsky, A. G.; Temiakov, D. E.; Agapov,
I. I.; Saward, S.; Palmer, R. A. Proteins: Struct., Funct., Genet. 1997, 28,
586–589.
(61) Iyer, P. N. S.; Wilkinson, K. D.; Goldstein, I. J. Arch. Biochem.
Biophys. 1976, 177, 330–333.
(62) Pei, Z.; Anderson, H.; Myrskog, A.; Dunꢀer, G.; Ingemarsson, B.;
Aastrup, T. Anal. Biochem. 2010, 398, 161–168.
’ NOTE ADDED AFTER ASAP PUBLICATION
This manuscript was published on the Web on December 16,
2010 without final corrections. The corrected version was
reposted on December 21, 2010.
1007
dx.doi.org/10.1021/ac102781u |Anal. Chem. 2011, 83, 1000–1007