Content uploaded by Jakub Dostalek
Author content
All content in this area was uploaded by Jakub Dostalek on Dec 15, 2013
Content may be subject to copyright.
Bacterial Pathogen Surface Plasmon Resonance Biosensor Advanced
by Long Range Surface Plasmons and Magnetic Nanoparticle Assays
Yi Wang,
†,‡
Wolfgang Knoll,
†,‡
and Jakub Dostalek*
,†
†
Austrian Institute of Technology, Muthgasse 11/2, 1190 Vienna, Austria
‡
Nanyang Technological University, Centre for Biomimetic Sensor Science, Singapore 637553
*
SSupporting Information
ABSTRACT: A new approach to surface plasmon resonance (SPR) biosensors for rapid and
highly sensitive detection of bacterial pathogens is reported. It is based on the spectroscopy of
grating-coupled long-range surface plasmons (LRSPs) combined with magnetic nanoparticle
(MNP) assay. The interrogation of LRSPs allows increasing the biosensor figure of merit (FOM),
and the employment of MNPs further enhances the sensor response by a fast delivery of the
analyte to the sensor surface and through the amplified refractive index changes associated with the
capture of target analyte. This amplification strategy is particularly attractive for detection of large
analytes that diffuse slowly from the analyzed sample to the sensor surface. The potential of the
presented approach is demonstrated in a model experiment in which Escherichia coli O157:H7 was
detected at concentrations as low as 50 cfu mL−1, 4 orders of magnitude better than the limit of
detection achieved by regular grating-coupled SPR with direct detection format.
Surface plasmon resonance (SPR) biosensors represent
rapidly advancing technology for fast and sensitive
detection of chemical and biological analytes in important
applications areas of medical diagnostics, food control, and
environmental monitoring.
1
SPR offers the advantage of direct
label-free detection method that relies on the measurement of
refractive index changes accompanied with the binding of target
analyte. The specific capture of target analyte on metallic sensor
surface with attached biomolecular recognition elements is
probed by resonantly excited surface plasmons. These modes
originate from coupled oscillations of charge density and the
associated electromagnetic field occurring at a distance up to
approximately hundred nanometers from the metal. Over the
last years, we witnessed extensive research efforts aimed at the
implementation of SPR biosensors for rapid detection of
bacterial pathogens.
2−5
However, they were typically shown to
allow for the analysis of bacterial pathogens at concentrations
above 103colony forming units (cfu) per mL which is not
sufficient for the majority of harmful pathogens. The key
limitations impeding the sensitivity of bacterial pathogen SPR
biosensors are related to small refractive index contrast of the
analyte bound to the surface, slow diffusion-driven mass
transfer from a sample to the sensor surface, and insufficient
depth probed by surface plasmons which is much smaller than
the micrometer characteristic size of bacteria.
Numerous amplification strategies were pursued for advanc-
ing SPR biosensors based enzymatic reactions,
6
metallic
7,8
and
magnetic
9−14
nanoparticle assays, as well as by combining SPR
with fluorescence spectroscopy.
15
Parallel to amplification
strategies, improving the sensitivity through increasing the
accuracy of refractive index measurements was subject to
research in SPR sensors utilizing propagating and localized
surface plasmons on nanostructured metallic surfaces.
16,17
Among these, the employing of long-range surface plasmons
(LRSPs), which propagate along thin metal films embedded in
a refractive index symmetrical layer architecture,
18
were
investigated for high-resolution SPR biosensors.
19
LRSP
modes exhibit lower losses compared to regular surface
plasmons, which translates into narrower resonance. Therefore,
they allow for improving the figure of merit (FOM)
19,20
and
more accurate measurement of refractive index variations. In
addition, the profile of LRSP field can be tuned to probe an
order of magnitude higher distances from the metal surface
compared to regular surface plasmons,
20
which makes them
excellently suited for the analysis of large analytes such as
bacterial pathogens.
21,22
This paper describes a new approach to SPR biosensors for
the detection of bacterial pathogens that utilizes nanostructured
surface architecture comprising low refractive index fluoropol-
ymer and metallic layers for the diffraction grating-based
excitation and interrogation of LRSPs. The developed sensor
chip enables straightforward implementation of amplification
strategies relying on magnetic nanoparticle (MNP) assays. This
approach is employed for the rapid delivery of target analyte
from a sample to the sensor surface by a magnetic field gradient
that is applied through the sensor chip. Let us note that this
feature is particularly advantageous for detection of large
analytes and its implementation to more commonly used
prism-based SPR biosensors is not possible. The performance
characteristics of the developed sensor platform are shown by
Received: July 9, 2012
Accepted: August 29, 2012
Published: August 29, 2012
Article
pubs.acs.org/ac
© 2012 American Chemical Society 8345 dx.doi.org/10.1021/ac301904x |Anal. Chem. 2012, 84, 8345−8350
using an immunoassay experiment for the detection of
Escherichia coli O157:H7, and the achieved results are
compared to those that were determined in an identical
experiment by regular diffraction grating-coupled SPR.
■MATERIALS AND METHODS
Materials. Magnetic iron oxide nanoparticles modified with
a polysaccharide layer (fluidMAG-ARA with diameter of ∼200
nm, magnetic core diameter of ∼175 nm) were purchased from
Chemicell (Berlin, Germany). E. coli O157:H7 standard was
obtained from KPL Inc. (Gaithersburg, MD), and E. coli strain
K12 was cultivated in our laboratory. The average diameters of
heat-killed E. coli O157:H7 and K12 were determined by
dynamic light scattering (DLS) with Zetasizer from Malvern
Instruments (Worcestershire, U.K.) as 1177 and 869 nm,
respectively. Capture antibody against E. coli O157:H7 (cAb,
no. ab75244) was from Abcam (Cambridge, U.K.). Affinity-
purified detection antibody against E. coli O157:H7 (dAb, no.
01-95-90) was obtained from KPL (Gaithersburg, MD). 1-
Ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC) and N-
hydroxysuccinimide (NHS) were from Pierce (Rockford).
Dithiolaromatic PEG6-carboxylate (thiol-COOH) and dithio-
laromatic PEG3 (thiol-PEG) were from SensoPath Technolo-
gies (Bozeman, MT). 2-(N-Morpholino)ethanesulfonic acid
(MES), phosphate buffered saline (PBS) tablets, and Tween-20
were acquired from Sigma-Aldrich (Austria). All detection
experiments were carried out in PBS buffer with 0.05% Tween
20 (PBST). Low-refractive index polymer, Cytop CTL-809 M,
and UV-curable polymer NOA 72 were obtained from Asahi
Inc. (Japan) and Norland (Cranbury, NJ), respectively.
Polydimethylsiloxane (PDMS) and its curing agent SYLGARD
184 were from Dow Corning.
Preparation of Sensor Chip. Nanostructured layer
architectures for the diffraction grating-based excitation of
long-range surface plasmons and regular surface plasmons were
prepared by a variant of nanoimprint lithography (NIL). First, a
silicon master with sinusoidal relief modulation (period of Λ=
510 nm and modulation depth of d= 27 nm) was fabricated by
interference lithography and reactive ion beam etching.
Afterward, the corrugated surface of the silicon master was
casted to a PDMS stamp which was used to transfer the
corrugation onto used sensor chips, see Figure 1a. For the
grating-coupled long-range surface plasmon (GC-LRSP) sensor
chips, a 100 nm gold film was first deposited on a flat glass
substrate by sputtering (UNIVEX 450C from Leybold Systems,
Hanau, Germany) followed by the spin-coating of a low-
refractive index fluoropolymer film. The Cytop fluoropolymer
was used as it exhibits a refractive index of nb= 1.338 (at the
wavelength λ= 633 nm) that is close to that of water ns= 1.332
and its structuring by using nanoimprint lithography was
reported by other groups previously.
23
The Cytop layer was
deposited with the thickness of db= 630 nm followed by the
attachment of the PDMS stamp to its top after a 5-min
predrying. The surface contacted with PDMS stamp was
completely dried overnight in vacuum oven at 50 °C, the
PDMS stamp was detached, and a gold layer with the thickness
of dm= 25 nm was sputtered on the corrugated Cytop surface.
The gold grating on GC-LRSP chip exhibited the same period
Λ= 510 nm as the master and a decreased modulation depth of
12 nm as was determined by atomic force microscopy (AFM).
For the reference sensor chip with regular grating-coupled
surface plasmon (GC-SP) resonance, the silicon master
corrugation was transferred to a UV-curable polymer NOA72
that was subsequently coated by a 60 nm thick gold layer as
described in our previous work.
13
The GC-SP sensor chip
carried a gold grating with the period Λ= 510 nm and a deeper
modulation depth d= 27 nm.
Thiol self-assembled monolayer (SAM) was formed on the
gold surface of GC-SP and GC-LRSP chips by the overnight
incubating in a mixture of thiol-COOH and thiol-PEG
dissolved at a molar ratio of 1:9 in absolute ethanol (total
concentration of 1 mM). On the mixed thiol SAM, cAb was
covalently bound via the amine groups after the activation of
SAM carboxyl groups by a mixture of EDC (37.5 mg mL−1)
and NHS (10.5 mg mL−1).
Optical Setup. An optical setup for angular spectroscopy of
surface plasmon modes shown in Figure 2a was used. This
setup was derived from that previously employed in our
laboratory.
13
Briefly, a light beam with the wavelength of λ=
632.8 nm from a He−Ne laser (Uniphase, CA) was made
incident at a surface of a GC-LRSP or regular GC-SP sensor
chip. The sensor chip was mounted on a rotation stage (Huber
AG, Germany) in order to control the angle of incidence θ. The
intensity of the laser beam that was reflected from the sensor
chip surface was measured by using a photodiode (PD)
connected to a lock-in amplifier (Princeton Applied Research,
TN). A flow-cell with the volume of about 10 μL was clamped
to the surface of the sensor chip. Samples were flowed over the
sensor surface at a flow-rate of 503 μL min−1. This setup
allowed for the measurements of angular reflectivity spectra
R(θ) as well as the kinetics of the reflectivity signal at an angle θ
set at the highest slope of the GC-LRSP or GC-SP resonance
dip. For the MNP-enhanced assays, an external magnetic field
with a gradient perpendicular to the surface of ∇B= 0.10 T
mm−1was applied by using a 1.4T NdFeB cylindrical magnet
from Neotexx (Berlin, Germany) that was placed at the
distance of 2 mm from the sensor surface. The magnet
diameter and length were of 10 and 25 mm, respectively.
Modification of MNPs with dAb. MNPs were decorated
with dAb against E. coli O157:H7 (see Figure 2b) according to
the protocol provided by the supplier with several modifica-
Figure 1. (a) Preparation of the layer structure supporting GC-LRSP
with an imprinting of a relief corrugation to a low-refractive index
Cytop fluoropolymer layer. (b) Schematic of the sensor chip for the
optical excitation of and interrogation of GC-LRSPs.
Analytical Chemistry Article
dx.doi.org/10.1021/ac301904x |Anal. Chem. 2012, 84, 8345−83508346
tions. Briefly, 20 mg of MNPs-ARA were activated in 2-(N-
morpholino)ethanesulfonic acid (MES) buffer with EDC and
NHS dissolved at a concentration of 11 mg mL−1. Afterward,
MNPs were washed with pure MES buffer and incubated with
100 μg of dAb for 2 h at room temperature. Then, the MNPs-
dAb conjugates were deactivated by the reaction with 1 M
ethanolamine (pH 8.5) for 20 min followed by the washing
with PBST. The antibody-to-MNP ratio was estimated as 10:1,
assuming 90% of antibodies were immobilized on the MNPs
surface during the labeling process.
24
Detection Formats. All samples were prepared by spiking
PBST with target E. coli O157:H7 or reference E. coli K12
analytes at concentrations between 103and 107cfu mL−1.In
direct detection format, the samples were flowed over the
sensor surface with immobilized cAbs for 15 min followed by
the rinsing with PBST. In the MNPs-enhanced assay, MNPs
conjugated with dAb were mixed with a sample containing
target analyte E. coli O157:H7 or control analyte E. coli K12 and
incubated for 15 min. Afterward, the mixture was circulated
through the sensor for 10 min with the magnetic field gradient
of ∇B= 0.10 T mm−1applied through the sensor chip. Then,
the surface was rinsed for 1 min with PBST, the magnet was
removed from the sensor surface leading to ∇B= 0, and the
surface was rinsed for additional 5 min. In order to use the
sensor chip for multiple experiments, 10 mM NaOH was
flowed over the sensor surface for several minutes to release the
captured analyte and leave the unoccupied cAb binding sites
available for the next detection cycle. Let us note that dAb
conjugated to MNPs is specific to analyte epitopes different
from that for cAb.
■RESULTS AND DISCUSSION
Sensitivity of GC-SP and GC-LRSP Sensor Chips.
Grating structures for the excitation of regular surface plasmons
(GC-SP) and long-range surface plasmons (GC-LRSP) were
designed for the first diffraction order coupling of an optical
wave incident at the surface. The diffraction on the periodically
modulated gold surface allows one to enhance the parallel
component of the light beam propagation constant 2π/λsin(θ)
by the grating momentum 2π/Λto match that of SP or LRSP.
On the GC-LRSP sensor chip, the light beam is partially
reflected and partially transmitted through the corrugated gold
film which hinders the coupling strength to LRSPs. In order to
increase the coupling efficiency, an additional flat metal film
between the fluoropolymer buffer layer and the glass substrate
was used. This layer reflects the transmitted beam back toward
to LRSP-guiding gold layer and prevents the leaking of LRSPs
into the substrate, see Figure 1b. The thickness of the
fluoropolymer layer of db= 633 nm was adjusted to provide
constructive interference between the (normal) incident and
back-reflected beams at the corrugated gold surface. Simu-
lations based on the finite element method (FEM) were carried
out in order to optimize the design of the prepared structure as
described in the Supporting Information (Figure S1).
As Figure 3a shows, the grating coupling to long-range
surface plasmons (GC-LRSP) and regular surface plasmons
(GC-SP) manifests itself as a narrow dip in the reflectivity
spectrum centered at an angle θwhere the incident and
plasmon waves are diffraction phase-matched. GC-SP reso-
nance occurs at an angle (θ=10°) that is higher than the one
where the GC-LRSP excitation occurs (θ=5°) due to higher
momentum of SPs. If the refractive index of a sample on the
gold surface nsis increased, the resonance dips shift to higher
angles θ. The angular sensitivity to refractive index changes
defined as S=δθ/δnsis lower for GC-LRSP (S=38°RIU−1)
Figure 2. (a) Optical setup utilizing GC-LRSPs and (b) used surface
architecture for the detection of bacterial pathogens by MNP-
enhanced assay.
Figure 3. (a) Measured angular reflectivity spectra from GC-LRSP and
GC-SP sensor chips brought in contact with a series of aqueous
samples spiked with ethylenglycol. Refractive indices of samples were
(1) ns= 1.3326, (2) ns= 1.3330, and (3) ns= 1.3334. Dashed lines
represent simulated reflectivity. (b) Simulated profile of electric field
intensity perpendicular to the sensor surface upon the resonant
coupling to GC-SPs (black curve) and LRSPs (red curve).
Analytical Chemistry Article
dx.doi.org/10.1021/ac301904x |Anal. Chem. 2012, 84, 8345−83508347
than that for regular GC-SP (S=92°RIU−1). The reason is
that LRSP field propagates on both sides of the thin corrugated
metallic film, while that of regular SP is fully confined in the
sample, see Figure 3b. However, the figure of merit (FOM,
defined as the ratio of the angular sensitivity Sand width of the
resonance Δθ) is about 5 times better for GC-LRSP (FOM =
127) compared to that for GC-SP (FOM = 24) as the coupling
to LRSPs is associated with a narrower resonance. This leads to
increased changes in the reflectivity Rdue to refractive index
changes δnsat an angle fixed at the highest slope of the
resonance dip (θ= 4.77°for GC-LRSP and θ= 9.85°for GC-
SP). The sensitivity of the reflectivity changes ΔRto refractive
index variations δnswas enhanced by a factor of 8.5 (see Figure
S2 in Supporting Information) which is better than that
predicted by the FOM, and it is caused by the nonsymmetrical
shape of the resonant dip. The electric field intensity
distribution |E/E0|
2
was calculated by using FEM for the
resonant coupling to LRSPs and regular SPs (see respective
simulated reflectivity curves in Figure 3a). Results in Figure 3b
predict that LRSP probes to a 3-fold higher distance from the
metal surface than regular surface plasmons. Therefore, the
combined higher probing depth and increased refractive index
sensitivity makes GC-LRSP a better suited platform than
regular GC-SP for the detection of specifically captured large
bacterial pathogens on the sensor surface that is functionalized
with cAb.
Magnetic Nanoparticle-Enhanced Immunoassay for E.
coli O157:H7. First, the angular reflectivity spectra were
measured from GC-LRSP and GC-SP sensor chips before and
after the affinity capture of E. coli O157:H7 analyte that was
reacted with MNP-dAb and pulled to the sensor surface by
applied magnetic field gradient ∇B. As Figure 4a shows, the
GC-SP and GC-LRSP resonances shift to higher angles as
increasing the concentration of E. coli O157:H7 in the sample,
indicating the binding of E. coli O157:H7 to cAb. In addition,
the overall reflectivity decreases for higher analyte concen-
trations due to the strong scattering and absorption of MNP
aggregates on the sensor surface. In further detection
experiments, the maximum reflectivity changes ΔRwere
measured at the angle of incidence of θ=5°for GC-LRSP
and θ= 11.5°for GC-SP. These changes were determined as
the difference in the reflectivity signal Rbefore the injection of
a sample and after the rinsing with PBST. From Figure 4a
follows that the capture of target analyte on the sensor surface
is accompanied with a significantly higher reflectivity change
ΔRon GC-LRSP sensor chip than on that supporting regular
GC-SPs. The enhancement of the reflectivity change (defined
as the ratio of ΔRfor GC-LRSP and GC-SP) is increasing when
decreasing the analyte concentration, and values of 2.4 and 4
were obtained for the concentrations 107cfu mL−1and 103cfu
mL−1, respectively. These factors are lower than those observed
in previous refractometric experiments (factor of 8.5), which is
probably due to the effect of absorption changes. In addition,
the dependence of the enhancement factor on the E. coli
O157:H7 concentration indicate that the density distribution of
MNP perpendicular to the surface plays an important role for
probing the sensor surface to different depths by LRSPs and
regular SPs (see Figure 3b) .
In order to maximize the enhancement of the sensor
response ΔRby MNP-assisted pulling of target E. coli O157:H7
analyte to the surface, the concentration of MNPs conjugated
with dAb and preincubated with a sample was optimized. The
sensor response ΔRwas measured by using GC-SPs for the
MNP-dAb concentrations of 3, 30, and 300 pM and analyte
concentrations of 104,10
5, and 106cfu mL−1. The results
presented in Figure 4b indicate that the MNP-dAb concen-
tration providing maximum sensor response decreases with
decreasing the concentration of target analyte. The reason for
this observation is that the magnetic field gradient ∇Bpulls to
the sensor surface both MNPs bound to the analyte and those
not reacted with the analyte. Therefore, a balance between the
efficient analyte delivery to the surface and avoiding blocking of
cAb active sites by the excess of free MNP-dAb which does not
contribute to the sensor response needs to be established. In
further experiments, the concentration of MNP-dAb was fixed
at 30 pM, which provided the highest sensor response for the
concentration of analyte around 104cfu mL−1.
Performance Characteristics. The sensor response to the
binding of target analyte E. coli O157:H7 was compared for the
direct detection format on a regular GC-SP sensor chip with
that for the MNP-enhanced assay combined with the readout
based on GC-LRSP. The measured reflectivity kinetics for the
direct GC-SP assay is presented in Figure 5a (measured at the
angle of incidence θ=9°). It shows a gradual increase in the
reflectivity signal Rupon the binding of E. coli O157:H7 to the
surface and reveals that the reflectivity change ΔRincreases
with the analyte concentration. For instance, the reflectivity
change reached ΔR= 1.2 ×10−3for the E. coli O157:H7
concentration of 106cfu mL−1. For the MNP-enhanced assay
performed on the GC-LRSP sensor chip that is shown in Figure
Figure 4. (a) A comparison of angular reflectivity spectra measured
after the capture of E. coli O157:H7 at concentrations of 0, 103,10
7cfu
mL−1on GC-SP and GC-LRSP sensor chips (MNP-dAb concen-
tration was 30 pM). (b) Reflectivity changes measured by GC-SP
resonance after the affinity capture of E. coli O157:H7 dissolved at
concentrations of 104,10
5, and 106cfu mL−1and incubated with
MNP-dAb conjugates at concentrations of 3, 30, and 300 pM.
Analytical Chemistry Article
dx.doi.org/10.1021/ac301904x |Anal. Chem. 2012, 84, 8345−83508348
5b, the reflectivity signal gradually decreases after the injection
of a sample preincubated with MNP-dAb. This observation is
due to the scattering and absorption of MNPs aggregates that
are pulled to the surface by the magnetic field gradient ∇B=
0.10 T mm−1. After switching offthe magnetic field (∇B=0T
mm−1) and rinsing the sensor surface with a buffer, a fast
increase of reflectivity signal Ris observed due to the release of
MNP-dAb conjugates that were not bound to the captured E.
coli O157:H7. For the sample spiked with E. coli O157:H7 at a
concentration of 106cfu mL−1, the reflectivity change of ΔR=
0.12 was determined, which is 2 orders of magnitude higher
than that measured by direct GC-SP resonance detection
format. Let us note that this signal enhancement originates
from combined effects of the MNP-enhanced mass transfer of
the analyte to the surface (more than an order of magnitude
higher rate is estimated based on the theory presented in the
Supporting Information), increased refractive index changes
associated with the analyte affinity binding to the sensor
surface, and higher sensitivity of the GC-LRSP platform. Figure
5b also demonstrates that the target analyte captured by cAb on
a surface can be fully released by the regeneration with NaOH
allowing for multiple detection cycles on a single sensor chip.
The E. coli O157:H7 biosensor calibration curves were
measured for each detection format in triplicate, and error bars
were determined as the standard deviation. The limit of
detection (LOD) was obtained as the concentration of E. coli
O157:H7 in a sample for which the response ΔRreached 3
times the standard deviation (3σR) of the reflectivity baseline R.
From results presented in Figure 6a, the LOD of 6.5 ×105cfu
mL−1was estimated for the GC-SP sensor chip with direct
measurement of E. coli O157:H7 binding-induced refractive
index changes. The same figure shows that the implementation
of the MNP-enhanced assay allowed decreasing the LOD by a
factor of 650 to 103cfu mL−1on the GC-SP sensor chip. As
seen in Figure 6b, the LOD was further improved to around 50
cfu mL−1by using the same assay and the readout based on
GC-LRSP, which corresponds to an enhancement by 4 orders
of magnitude with respect to that observed for GC-SP and
direct detection format. A control experiment in which a
different but closely related bacterium (E. coli K12) was
detected reveals excellent specificity of the developed biosensor,
see Figure 6b. In general, the improvements in LOD are better
than the enhancement of the sensor response ΔRobserved for
the binding of target analyte at concentration of 106cfu mL−1
and from refractometric measurements. This may be ascribed
to the nonlinear dependence of the sensor response ΔRon the
concentration of the target analytes for two key reasons. First,
we assume that the amount of MNP-dAb bound to target
analyte is increasing when decreasing the concentration of E.
coli O157:H7, which is associated with its more efficient
delivery to the surface. This leads to larger enhancement of the
sensor response at lower analyte concentrations and further
improved the LOD. Second, the binding of MNPs to the
surface is accompanied with scattering, effective refractive index
change, and an increase in absorption. These multiple effects
alter both the resonance angle and shape of the resonance dip
which results in nonlinear dependence of the reflectivity change
ΔRon the amount of MNP adhered to the surface. As
compared to other techniques such as polymerase chain
reaction (PCR, LOD = 6−10 cells),
25,26
enzyme linked
immunosorbent assays (ELISA, LOD 103−105cfu mL−1),
27
quartz crystal microbalance (QCM, 103cfu mL−1),
28
and
surface-enhanced Raman spectroscopy (SERS, LOD 5 cfu
mL−1,10
6cells mL−1),
29,30
the reported method shows superior
or similar sensitivity with shorter analysis time (30 min
including incubation), which otherwise requires several hours
or days.
■CONCLUSIONS
A novel approach to SPR biosensors based on grating-coupled
long-range surface plasmons (GC-LRSPs) advanced by MNP
immunoassays was reported and applied for the detection of
bacterial pathogen E. coli O157:H7. The sensor chip for
accurate measurements of refractive index variations was
developed by using nanoimprint lithography, and it provided
about a 8.5-fold better refractive index resolution compared to
regular grating-coupled surface plasmons (GC-SPs). With
Figure 5. Kinetics of reflectivity signal for (a) direct detection assay
with probing of E. coli O157:H7 binding by GC-SPs and (b) MNPs
immunoassay-enhanced assay with probing by GC-LRSPs.
Figure 6. Calibration curves for (a) GC-SP with direct and MNP-
enhanced detection of target E. coli O157:H7 and (b) for GC-LRSP
combined MNPs-enhanced assay for detection of target analyte E. coli
O157:H7 and a control analyte E. coli K12.
Analytical Chemistry Article
dx.doi.org/10.1021/ac301904x |Anal. Chem. 2012, 84, 8345−83508349
respect to the direct detection format, the MNP-enhanced
assay provided additional enhancement of the sensor signal
through a more efficient collecting of target analyte on the
sensor surface and by the amplified refractive index contrast of
the analyte. The assay based on GC-LRSP allowed for the
detection of a target analyte, E. coli O157:H7, with the limit of
detection of 50 cfu mL−1, which was about 4 orders of
magnitude better than that provided by regular GC-SP
resonance with direct detection format.
■ASSOCIATED CONTENT
*
SSupporting Information
FEM simulation of the angular spectra for GC-LSP,
refractometric experiments, and theoretical analysis of the
diffusion rate and the MNPs velocity. This material is available
free of charge via the Internet at http://pubs.acs.org.
■AUTHOR INFORMATION
Corresponding Author
*Fax: +43(0)50550-4450. E-mail: jakub.dostalek@ait.ac.at.
Notes
The authors declare no competing financial interest.
■ACKNOWLEDGMENTS
The authors would like to thank Marlies Czetina and Angela
Sessitsch (AIT, Vienna) for the preparation of the E. coli K12
samples. In addition, we are grateful to Prof. Bernhard Schuster
and Angelika Schrems (Department of Nanobiotechnology,
University of Natural Resources and Life Sciences, Vienna) for
the DLS measurements. The authors would like to acknowl-
edge the partial support for this work provided by the Austrian
NANO Initiative (FFG and BMVIT) through the NILPlas-
monics Project within the NILAustria Cluster (www.
NILAustria.at) and the Austrian Science Fund (FWF) through
the Project ACTIPLAS (Grant P 244920-N20).
■REFERENCES
(1) Homola, J. Chem. Rev. 2008,108, 462−493.
(2) Subramanian, A.; Irudayaraj, J.; Ryan, T. Biosens. Bioelectron.
2006,21, 998−1006.
(3) Taylor, A. D.; Ladd, J.; Yu, Q. M.; Chen, S. F.; Homola, J.; Jiang,
S. Y. Biosens. Bioelectron. 2006,22, 752−758.
(4) Koubova, V.; Brynda, E.; Karasova, L.; Skvor, J.; Homola, J;
Dostalek, J.; Tobiska, P.; Rosicky,J.Sens. Actuators, B 2001,74, 100−
105.
(5) Yanik, A. A.; Huang, M.; Kamohara, O.; Artar, A.; Geisbert, T.
W.; Connor, J. H.; Altug, H. Nano Lett. 2010,10, 4962−4969.
(6) Fang, S. P.; Lee, H. J.; Wark, A. W.; Corn, R. M. J. Am. Chem. Soc.
2006,128, 14044−14046.
(7) He, L.; Musick, M. D.; Nicewarner, S. R.; Salinas, F. G.; Benkovic,
S. J.; Natan, M. J.; Keating, C. D. J. Am. Chem. Soc. 2000,122, 9071−
9077.
(8) Wark, A. W.; Lee, H. J.; Qavi, A. J.; Corn, R. M. Anal. Chem.
2007,79, 6697−6701.
(9) Mitchell, J. S.; Wu, Y. Q.; Cook, C. J.; Main, L. Anal. Biochem.
2005,343, 125−135.
(10) Teramura, Y.; Arima, Y.; Iwata, H. Anal. Biochem. 2006,357,
208−215.
(11) Tseng, P.; Di Carlo, D.; Judy, J. W. Nano Lett. 2009,9, 3053−
3059.
(12) Wang, J. L.; Munir, A.; Zhu, Z. Z.; Zhou, H. S. Anal. Chem.
2010,82, 6782−6789.
(13) Wang, Y.; Dostalek, J.; Knoll, W. Anal. Chem. 2011,83, 6202−
6207.
(14) Soelberg, S. D.; Stevens, R. C.; Limaye, A. P.; Furlong, C. E.
Anal. Chem. 2009,81, 2357−2363.
(15) Dostalek, J.; Knoll, W. Biointerphases 2008,3,12−21.
(16) Stewart, M. E.; Anderton, C. R.; Thompson, L. B.; Maria, J.;
Gray, S. K.; Rogers, J. A.; Nuzzo, R. G. Chem. Rev. 2008,108, 494−
521.
(17) Sepulveda, B.; Angelome, P. C.; Lechuga, L. M.; Liz-Marzan, L.
M. Nano Today 2009,4, 244−251.
(18) Sarid, D. Phys. Rev. Lett. 1981,47, 1927−1930.
(19) Nenninger, G. G.; Tobiska, P.; Homola, J.; Yee, S. S. Sens.
Actuators, B 2001,74, 145−151.
(20) Dostalek, J.; Kasry, A.; Knoll, W. Plasmonics 2007,2,97−106.
(21) Vala, M.; Etheridge, S.; Roach, J. A.; Homola, J. Sens. Actuators,
B2009,139,59−63.
(22) Huang, C. J.; Sessitsch, A.; Dostalek, J.; Knoll, W. Anal. Chem.
2011,83, 674−677.
(23) Cheng, L. J.; Kao, M. T.; Meyhofer, E.; Guo, L. J. Small 2005,1,
409−414.
(24) Koh, I.; Wang, X.; Varughese, B.; Isaacs, L.; Ehrman, S. H.;
English, D. S. J. Phys. Chem. B 2006,110, 1553−1558.
(25) Belgrader, P.; Benett, W.; Hadley, D.; Richards, J.; Stratton, P.;
Mariella, R.; Milanovich, F. Science 1999,284, 449−450.
(26) Dharmasiri, U.; Witek, M. A.; Adams, A. A.; Osiri, J. K.; Hupert,
M. L.; Bianchi, T. S.; Roelke, D. L.; Soper, S. A. Anal. Chem. 2010,82,
2844−2849.
(27) Blais, B. W.; Leggate, J.; Bosley, J.; Martinez-Perez, A. Lett. Appl.
Microbiol. 2004,39, 516−522.
(28) Varshney, M; Li, Y. B. Biosens. Bioelectron. 2007,22, 2408−2414.
(29) Knauer, M.; Ivleva, N. P.; Liu, X. J.; Niessner, R.; Haisch, C.
Anal. Chem. 2010,82, 2766−2772.
(30) Temur, E.; Boyaci, I. H.; Tamer, U.; Unsal, H.; Aydogan, N.
Anal. Bioanal. Chem. 2010,397, 1595−1604.
Analytical Chemistry Article
dx.doi.org/10.1021/ac301904x |Anal. Chem. 2012, 84, 8345−83508350
