Detection of intact influenza viruses using biotinylated biantennary S-sialosides

Article abstract of DOI:10.1021/ja800842v

We report the modular synthesis of robust, biotinylated biantennary sialylglycoconjugates and their ability to differentiate between two type A influenza strains. This is the first demonstration of glycoconjugate-based discriminatory capture and detection of two strains of intact influenza virus, in the presence of the innate enzymatic activity of viral neuraminidases. We also demonstrate a carboassay using glycoconjugates as capture and reporter elements, which therefore, does not require antibodies. The capture of intact influenza viruses is of potential benefit for clinical diagnostics. Copyright

Full text of DOI:10.1021/ja800842v

Published on Web 06/05/2008
Detection of Intact Influenza Viruses using Biotinylated Biantennary
S-Sialosides
Ramesh R. Kale,† Harshini Mukundan,‡ Dominique N. Price,§ J. Foster Harris,§ Daniel M. Lewallen,†
Basil I. Swanson,‡ Jurgen G. Schmidt,*,§ and Suri S. Iyer*,†
Chemical and Biosensors group, 805 Crosley, Department of Chemistry, UniVersity of Cincinnati, Cincinnati, Ohio
45221, Physical Chemistry and Applied Spectroscopy (C-PCS), Los Alamos National Laboratory, P.O. Box 1663,
MS J567, Los Alamos, New Mexico 87545, and Bioscience DiVision (B-7 and B-9), Los Alamos National
Laboratory, PO Box 1663, MS E 529, Los Alamos, New Mexico 87545
Received February 7, 2008; E-mail: jschmidt@lanl.gov; suri.iyer@uc.edu
Influenza, a highly pathogenic virus that belongs to the orth-
omyxoviridae family, accounts for over 150 million infections and
200000 hospitalizations every year in the U.S. alone.1 Antivirals
are a welcome tool in the fight against this contagious disease.2
However, they must be administered within 48 h of infection for
optimal efficacy.3 Early and accurate diagnosis of the specific strain
in both a clinical and point-of-care setting would significantly
facilitate a targeted and rapid response by healthcare providers.
Commercial antibody-based rapid diagnostic kits offer results in
15 min., but suffer from sensitivity, selectivity, and variability
issues,4–6 which could potentially be attributed to antibody degrada-
tion upon long-term storage at ambient temperature.7 Stability of
recognition elements for point of care diagnostics is often under-
appreciated. It is a major problem in the case of influenza, as recent
H5N1 outbreaks have originated in remote areas of the world, where
refrigeration for reagents is not always available. Clearly, there is
a pressing need to develop synthetic recognition molecules that
exhibit antibody-like selectivity and sensitivity, are robust, inex-
pensive, amendable to scale up, and require no refrigeration for
application in influenza diagnostics and therapeutics.
Figure 1. Structures of the two glycoconjugates, BG-1 (top) and BG-2
(bottom).
All strains of influenza type A virus bind to terminal sialic acid
derivatives. Within type A viruses, there are several serotypes such
as H1N1 and H3N2, etc. which are distinguished on the basis of
the antibody response elicited in the host. Each serotype is further
subdivided into several strains accounting for the enormous
antigenic diversity. Binding between a particular strain and the host
cell is highly dependent on the interaction of viral glycoproteins,
hemagglutinin (HA) and neuraminidase (NA), with specific struc-
tural features of the glycan.8 Glycan microarrays (developed by
the Consortium for Functional Glycomics), that utilize over sixty
unique O-sialosides, have shown that structure (sulfation, fucosy-
lation, etc.) of the internal sugars plays a critical role in defining
HA binding events.9,10 While these studies have been extremely
valuable in providing insight into HA specificity, NA and intact
virus specificities have not been determined as NA cleaves
O-sialosides rapidly.11 Therefore, detection of the intact virus in
the presence of the innate enzymatic activity requires stable
glycoside analogues. Here, we report the modular synthesis of viral
NA resistant S-sialosides and demonstrate their ability to differenti-
ate between two selected influenza strains belonging to H1N1 and
H3N2 serotypes. These results show glycoconjugate-based viral
capture and a potential path for the discriminatory detection of intact
influenza viruses in the presence of the innate enzymatic activity
of viral NAs.
Since the position of sialic acid, orientation, density, tether length,
and ancillary groups influence binding efficiency and selectivity,
we designed a modular, flexible synthetic strategy that accom-
modates these variables to yield stable glycoconjugates. (Figure 1)
The three important components of the designer ligands are (i) A
biotinylated divalent scaffold. Biotin was used because it can attach
to any streptavidin matrix. The biotin further affords additional
multivalency as avidin binds up to four biotinylated molecules.
Indeed, the reporter element was prepared by incubation of the
biotinylated ligand with a fluorescently labeled streptavidin molecule
to yield a streptavidin-(biotin-saccharide)4 conjugate. (ii) A robust
recognition element. We synthesized S-sialosides as these have been
shown to be more resistant than O-glycosides.11,12 (iii) A critical
variable spacer that connects and separates the biotin from the
recognition motifs. Recent reports have suggested that the topology
of the sugars is very critical for differentiation between influenza
strains.13 We chose oligoethylene glycols to reduce nonspecific
binding, to impart a degree of flexibility, and to facilitate develop-
ment of glycoconjugates with distance optimized between adjacent
recognition sites.
We used this modular strategy to synthesize biotinylated bian-
tennary glycoconjugates, BG-1 and BG-2 that differ in the
presentation of the sialic acid. BG-1 was designed to exhibit no
discrimination, while we expected the sialic acid attached to the 6′
† University of Cincinnati.
‡ Physical Chemistry and Applied Spectroscopy (C-PCS), Los Alamos National
Laboratory.
§ Bioscience Division (B-7 and B-9), Los Alamos National Laboratory.
9
10.1021/ja800842v CCC: $40.75
2008 American Chemical Society
J. AM. CHEM. SOC. 2008, 130, 8169–8171 8169

C O M M U N I C A T I O N S
Scheme 1. Synthesis of the Biantennary R-2,6-S-Sialosidea
a Reagents and conditions: (a) H(OCH2CH2)4N3, TMSOTf, DCM, 0 °C,
1.5 h, 74%; (b) NaOMe, MeOH, 12 h; (c) PhCH(OMe)2, p-TSA, CH3CN.
87% over two steps; (d) Ac2O, DMAP, pyridine, 16 h, 88%; (e) TFA/H2O
(3:2), DCM, 60%; (f) Tf2O, pyridine, DCM, -25 °C, 1 h; (g) Et2NH, DMF,
-25 °C, 2 h. 65% over two steps; (h) CuSO4, sodium ascorbate, THF:
H2O, 36 h, 67%; (i) NaOMe, MeOH, 12 h; (j) NaOH, H2O, 10 h, 91%
over two steps.
Figure 2. Schematic of the sandwich “carboassay” on a planar optical
waveguide biosensor. A biotinylated carbohydrate ligand conjugate to
fluorescent streptavidin is shown as the detector. Fluorescent antibodies as
reporter were also used.
the efficient capture and provide a baseline for the follow up study,
in which glycans were used both for capture and detection. Briefly,
BG-1 (or BG-2) was anchored onto the biosensor surface via
streptavidin coupling. After washing, inactivated virus [A/Beijing/
262/95 (H1N1) or A/Sydney/26/95 (H3N2)] from cell extracts was
injected and incubated for 30 min. Following washing, injection
of the fluorescently labeled reporter, Alexafluor-647 antibody or
Alexafluor-647 streptavidin-(biotin-glycoconjugate)4, gave the ob-
served signal. The exact sequence of steps, without the virus, was
performed to measure nonspecific binding (see Supporting Informa-
tion).
position of lactose in BG-2 to allow differential binding to viral
strains. The synthesis of the R-2,6 analogue BG-2 is shown in
Scheme 1. Lactose septaacetate ꢀ-imidate 1 was reacted with a
tetraethylene glycol spacer, 1-azido-(2-(2-ethoxy)ethoxy)ethanol to
yield the beta isomer 2 in 74% yield. Saponification using NaOMe
in MeOH was followed by benzylidene protection of the 4,6
hydroxyl groups to yield 3 in 87% over two steps. Reprotection of
free hydroxyls with acetate groups followed by removal of the acetal
furnished 4. Next, the primary alcohol in 4 was selectively activated
to the triflate and reacted with the known thio-N-acetylneuraminic
acid, 5, in the presence of diethyl amine to yield trisaccharide, 6.
The azide bearing trisaccharide, 6, was reacted with a biotinylated
scaffold carrying two alkyne functionalities 7, in the presence of
CuSO4 and sodium ascorbate to give 8 in reasonable yield. A two-
step procedure, saponification followed by deesterification resulted
in the desired product, BG-2. This final product was purified using
a Biogel P-2 column. Similarly, BG-1 was synthesized and purified
(see Supporting Information).
Next, we developed the carbohydrate analogue of a sandwich
immunoassay (the virus is sandwiched between capture and
fluorescent reporter molecules) on the surface of a waveguide-based
optical biosensor that facilitates rapid and sensitive detection of
biological agents.14 Briefly, the technology involves the use of
planar optical waveguides comprising thin (∼120 nm) high refrac-
tive index (RI) dielectric materials deposited on a substrate with a
lower RI. Etched diffraction gratings provide facile means of
coupling laser light into the thin waveguide film (see Figure 2).
Although a majority of the incident light is present in the guided
mode, the evanescent field extends out into the medium and excites
the fluorescent reporter. Since the evanescent field fades with
increasing distance from the surface, molecules >200 nm from the
surface are not excited eliminating background fluorescence in
complex matrices. We have successfully demonstrated the applica-
tion of the waveguide biosensor to the rapid and sensitive detection
of nanomolar quantities of B. anthracis and breast cancer proteins
in complex samples.14 Here, we demonstrate the capability of this
assay platform to detect larger particles such as an intact Virus.
The first set of experiments established the ability of synthetic
glycans to capture viral particles, while a fluorescent monoclonal
anti-HA antibody was used as reporter. These experiments establish
The results of the A/Beijing/262/95 (H1N1) binding with BG-1
are shown in Figure 3a and 3b. The top curve of Figure 3a shows
the spectral response in relative fluorescence units (RFU) of the
excited waveguide following exposure of the viral particles (108
particles/ml) to BG-1. The bottom curve of Figure 3a is the
background signal. A response of 2500 RFU was obtained when
virus was injected. BG-1 conjugated to fluorescent streptavidin can
also be used as a reporter because of the presence of over 100 copies
of multimeric HA and NA on the virus providing multivalent
binding enhancement. In this case, a discernible signal of >250
RFU with a signal-to-noise ratio of 3.8 was recorded. (Figure 3b)
While signal intensities are considerably lower than an antibody
reporter, these results demonstrate for the first time, an influenza
capture and reporter pair entirely based on glycans. We anticipate
that sensitivities can be improved by adjusting the multivalency
and distances of sialic acids on the scaffold. Testing the ability of
BG-1 to bind to A/Sydney/262/95 (H3N2) yielded similar results,
supporting our initial expectation that this presentation of sialic acid
on the scaffold does not provide strain selective binding. We also
tested the binding of the two strains to BG-2. While A/Sydney/
26/95 (H3N2) bound well to BG-2, we obserVed no detectable
signal with A/Beijing/262/95 (H1N1) which demonstrates an
absolute discrimination of this sialoside between these two specific
strains. (Figure 3c) There are several possible explanations why
A/Beijing/262/95 (H1N1) did not bind to BG-2. (a) Previous studies
have been carried out with glycoconjugates that use 6′sialyl-N-
acetyllactosamine, whereas we have used 6′sialyllactose. We have
observed a similar behavior with Shiga toxins where the N-acetyl
group at the 2 position plays a significant role in the discriminatory
binding event.15 Indeed, modeling of the 6′-sialyl-N-acetyllac-
tosamine with HAs of a different H1N1 strain, A/SouthCarolina/
1/18, indicates hydrogen bonds from the NHAc to Asp 190 of HA,9
which should be absent in 6′sialyllactose. (b) Differences between
9
8170 J. AM. CHEM. SOC. VOL. 130, NO. 26, 2008

C O M M U N I C A T I O N S
Figure 3. Influenza detection using (A) BG-1 as capture and Alexafluor 647-antibody as reporter with A/Beijing/262/95 (H1N1). (B) BG-1 as capture and
Alexafluor-647-streptavidin-(BG-1)4 as reporter with A/Beijing/262/95 (H1N1). (C) Differential binding of A/Beijing/262/95 (H1N1) and A/Sydney/26/95
(H3N2) using BG-2. (D) Limit of detection using BG-2 as capture and Alexafluor-647-antibody as reporter with H3N2A/Sydney/26 /95 (H3N2).
the NAs (N1 versus N2) could also contribute in the binding event.
(c) The topology of the glycoconjugate adopts a conformation that
the viral glycoproteins cannot access, thereby precluding A/Beijing/
262/95 (H1N1) from binding. We also determined the detection
limit. This nascent technology allows us to detect 106 viral particles/
ml using the reporter antibody. (Figure 3d)
Acknowledgment. We thank the Center for Sensors and
Biosensors, Department of Chemistry, University of Cincinnati
(SSI) and the Los Alamos LDRD Research program (B.I.S. and
J.G.S.) for funding. We thank Drs. Steven Macha and Karen Grace
for ESI-MS analysis and waveguide fabrication, respectively, and
Dr. Hong Cai, Los Alamos National laboratory, for the generous
use of facilities for viral culture and characterization.
In summary, we have developed a modular strategy that allows
for the development of synthetic S-sialosides. The molecules have
been used as capture and reporter recognition elements for rapid
detection of two strains of the influenza type A virus. The
glycoconjugates are highly robust over extended time periods (>6
months) at ambient temperatures and require no refrigeration. One
of the glycoconjugates, BG-2 differentiates between A/Beijing/262/
95 (H1N1) and A/Sydney/26/95 (H3N2) using a waveguide-based
biosensor platform. While these initial studies are promising, it is
important to note that one cannot extrapolate these results to suggest
that BG-2 will not bind to other sialic acid recognizing pathogens
present in the respiratory tract or other strains. Several serotypes
and strains within each serotype must be tested to explore the
discriminatory potential of these thio-glycoconjugates. The binding
of multiple and diverse classes of pathogens to sialic acid on host
cell surfaces clearly poses challenges to carbohydrate-based dif-
ferential detection. Subtle differences in binding affinities can be
explored to derive selective binding as demonstrated in this limited
preliminary study of two influenza type A strains belonging to two
different serotypes. Currently, we are developing a focused glycan
library comprising stable S-sialosides to evaluate the binding
affinities of different strains of the Influenza virus and other
pathogens. On the instrumentation side, we have extended our
biosensor platform to demonstrate the capture of intact viral particles
and are developing a multichannel adaptation of the optical
biosensor. Advances, both in the availability of ligands and sensor
refinement, will facilitate future exploration of selectivity, affinity,
and discriminatory ability of our approach for respiratory infectious
agents that share sialic acid binding to infect the host.
Supporting Information Available: Details of the synthesis and
the sandwich assay. This material is available free of charge via the
Internet at http://pubs.acs.org.
References
(1) Smith, N. M.; Bresee, J. S.; Shay, D. K.; Uyeki, T. M.; Cox, N. J.; Strikas,
R. A. MMWR Recomm. Rep. 2006, 55, 1–42.
(2) von Itzstein, M. Nat. ReV. Drug DiscoVery 2007, 6, 967–74.
(3) Englund, J. A. Semin. Pediatr. Infect. Dis. 2002, 13, 120–8.
(4) Smit, M.; Beynon, K. A.; Murdoch, D. R.; Jennings, L. C. Diagn. Microbiol.
Infect. Dis. 2007, 57, 67–70.
(5) Rahman, M.; Kieke, B. A.; Vandermause, M. F.; Mitchell, P. D.; Greenlee,
R. T.; Belongia, E. A. Diagn. Microbiol. Infect. Dis. 2007, 58, 413–8.
(6) Suarez, D. L.; Das, A.; Ellis, E. AVian Dis. 2007, 51, 201–8.
(7) Ngundi, M. M.; Kulagina, N. V.; Anderson, G. P.; Taitt, C. R. Expert ReV.
Proteomics 2006, 3, 511–24.
(8) Stevens, J.; Blixt, O.; Tumpey, T. M.; Taubenberger, J. K.; Paulson, J. C.;
Wilson, I. A. Science 2006, 312, 404–10.
(9) Stevens, J.; Blixt, O.; Paulson, J. C.; Wilson, I. A. Nat. ReV. Microbiol.
2006, 4, 857–64.
(10) Paulson, J. C.; Blixt, O.; Collins, B. E. Nat. Chem. Biol. 2006, 2, 238–48.
(11) Wilson, J. C.; Kiefel, M. J.; Angus, D. I.; von Itzstein, M. Org. Lett. 1999,
1, 443–6.
(12) Rich, J. R.; Wakarchuk, W. W.; Bundle, D. R. Chem.sEur. J. 2006, 12,
845–58.
(13) Chandrasekaran, A.; Srinivasan, A.; Raman, R.; Viswanathan, K.; Raguram,
S.; Tumpey, T. M.; Sasisekharan, V.; Sasisekharan, R. Nat. Biotechnol.
2008, 26, 107–13.
(14) Martinez, J. S.; Grace, K. W.; Grace, K. M.; Hartman, N.; Swanson, B. I.
J. Mater. Chem. 2005, 15, 4639–4647.
(15) Kale, R. R.; McGannon, C. M.; Fuller-Schaefer, C.; Hatch, D. M.; Flagler,
M. J.; Gamage, S. D.; Weiss, A. A.; Iyer, S. S. Angew. Chem., Int. Ed.
2008, 47, 1265–1268.
JA800842V
9
J. AM. CHEM. SOC. VOL. 130, NO. 26, 2008 8171