Color online FTIR spectra of antibody immo-bilized electrode’s surface: a ATPS/CeO2 NWSs modified electrode, b absorption, c EDC/NHS - activated antibody, and d protein A-mediated immobil
Trang 1Detection of Vibrio cholerae O1 by Using Cerium Oxide Nanowires - Based
Immunosensor with Different Antibody Immobilization Methods
Phuong Dinh Tam,∗ Nguyen Luong Hoang, Hoang Lan and Pham Hung Vuong
Advanced Institute for Science and Technology (AIST), Hanoi University of Science and Technology (HUST), No 1 Dai Co Viet St., Hanoi, Viet Nam
Ta Thi Nhat Anh
Vinhphuc Technology Economic Colleges, No 10 Hung Vuong St., Vinhphuc, Viet Nam
Tran Quang Huy and Nguyen Thanh Thuy
National Institute of Hygiene and Epidemiology (NIHE), No.1 Yersin St., Hanoi, Viet Nam
(Received 18 March 2016, in final form 18 April 2016)
In this work, we evaluated the effects of different antibody immobilization strategies on the
response of a CeO2-nanowires (NWs)-based immunosensor forV ibrio cholerae O1 detection
Ac-cordingly, the changes in the electron-transfer resistance (R et) from before to after cells bind to an
antibody-modified electrode prepared by using three different methods of antibody immobilization
were determined The values were 16.2%, 8.3%, and 6.65% for the method that utilized protein A,
antibodies activated by 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride
(EDC)/N-hydroxysuccinimide (NHS), and absorption, respectively Cyclic voltammetry confirmed that the
change in the current was highest for the immunosensors prepared using protein A (11%), followed
by those prepared with EDC/NHS-activated antibodies (9%), and finally, those prepared through
absorption (7.5%) The order of the antibody immobilization strategies in terms of resulting
im-munosensor detection limit and sensitivity was as follows order: absorption (3.2 × 103 CFU/mL;
45.1 Ω/CFU·mL −1)< EDC/NHS-activated antibody (1.0 × 103 CFU/mL; 50.6 Ω/CFU·mL −1)<
protein A (1.0 × 102 CFU/mL; 65.8 Ω/CFU·mL −1) Thus, we confirmed that the protein A
-mediated method showed significantly high cell binding efficiencies compared to the random
immo-bilization method
PACS numbers: 81.05.Cy, 82.45.Tv, 81.16.-c
Keywords: Antibody, Protein A, Immunosensor, Nanowire, CeO 2
DOI: 10.3938/jkps.68.1235
I INTRODUCTION
Immunosensors are currently attracting much
atten-tion because of their promising applicaatten-tions, particularly,
in clinical diagnostics [1–5], the food industry [6–8], and
environmental monitoring [9–12] Many kinds of
im-munosensors, such as electrochemical, optical, and
me-chanical immunosensors, are used for various purposes
However, lectures report that immunosensor
perfor-mance is significantly affected by antibody
immobiliza-tion approaches Therefore, the selectivity of the
immo-bilization methods used to improve immunosensor
per-formance is very important To date, several techniques
for antibody immobilization have been reported,
includ-ing physicochemical absorption [13], covalent attachment
∗E-mail: tam.phuongdinh@hust.edu.vn, phuongdinhtam@gmail.
com; Fax: +84-4-3623-0293
[14, 15] or Langmuir Blodgett method [16], and other methods The physicochemical adsorption strategy relies
on weak binding, such as van der Waals, hydrophobic, or electrostatic interactions, to attach antibody molecules For example, Buijs and coworkers [13] studied the effect
of adsorption on antigen binding by IgG and its F(ab) fragments The group performed antibody absorption
on hydrophilic silica and hydrophobic methylated sil-ica surfaces The electrostatic interactions were studied
by varying pH values and ionic strength Experimen-tal results showed that the orientation of the adsorbed antibodies could be strongly influenced by electrostatic interactions Similar results were attained by Chen et
al [17] An ultrasensitive microcantilever
immunosen-sor based on antibody chemical adimmunosen-sorption was devel-oped by Sungkanaket al [18] In the study, linkers,
in-cluding 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC) and sulfo-N-hydroxysuccinimide
Trang 2-1235-(sulfo-NHS), were used to activate carboxylic groups to
form peptide bonds with the primary amines of the
anti-bodies Monoclonal antibodies in the PBS solution were
spread onto the sensor surface, and were achieved a
de-tection limit of 1.0 × 103 CFU/mL and a mass
sensi-tivity of 146.5 pg/Hz As demonstrated, the adsorption
approach was used to attach antibodies onto the
trans-ducer surface This method is simple, easy to perform,
and cost effective
However, fabricating high-performance
immunosen-sors is difficult because of the random antibody
orienta-tions achieved from this method, leading to a low density
of antigen-binding sites [19] To address this issue,
sen-sor performance was improved by using the covalent
at-tachment methods developed by various research groups
[14, 15, 20] These techniques rely on linker materials
to fasten antibodies onto electrode surfaces Wang et
al [20] reported an orientation-controlled
immobiliza-tion method based on protein A for immunosensor
de-sign Their study showed the good binding ability of
pro-tein A with gold nanoparticles The inclusion of
amine-terminated plasma polymerization also led to enhanced
antibody binding capabilities, which improved sensor
performance In another study, Franco and coworkers
[21] implemented an oriented immobilization of
antibod-ies on gold surfaces by using protein A to fabricate a
surface plasmon resonance immunosensor The protein
A-gold binding domain consist of a gold-binding
pep-tide coupled with the immunoglobulin-binding domains
of staphylococcal protein A This coupling facilitated the
oriented immobilization of antibodies The fabricated
immunosensor achieved a detection limit of 90 ng/mL,
with an interchip variability of lower than 7%
Many other studies used different linkers, such
as (3-glycidyloxypropyl)trimethoxysilane (GPS) [14],
graphene paper [15], carboxylmethyl dextran [22], and
semiconductor metal oxide [23–30], to immobilize
anti-bodies on electrode surfaces
To date, studies comparing antibody
immobiliza-tion techniques have been conducted Babacan et al.
[31] evaluated different antibody immobilization
meth-ods for piezoelectric immunosensor application In
the study, two linkers (protein A and
polyethylen-imine/glutaraldehyde, PEI/GA) were assessed In the
PEI/GA method, the antibodies were immobilized in
random orientations through the surface aldehyde group
of GA on a PEI-coated quartz crystal Otherwise,
antibodies were bound to protein A, which was
di-rectly immobilized on the quartz crystal’s surface by
van der Waals interactions The group showed that
antibody immobilization onto piezoelectric quartz
crys-tals through protein A showed better results than
im-mobilization through PEI/GA Meanwhile, Danczyk et
al [32] investigated three different antibody
immo-bilization methods: adsorption, aminosilance, and
N-gamma-maleimidobutyryloxy-succinimide ester (GMBS)
linkers They demonstrated that the presence of
pro-tein A improved the antigen capture ability of the
ad-sorbed antibodies and the GMBS surface However, protein A did not increase the antigenic capture of the aminophase surfaces Additionally, the aminosilance sur-face exhibited the highest level of nonspecific binding Vashist et al [33] investigated antibody
immobiliza-tion by using EDC, EDC/N-hydroxysuccinimide (NHS), and EDC/sulfo-NHS for immunosensor applications At
pH 7.4, EDC crosslinks antibodies to the 3-aminopropyl triethoxy-silane (APTS)-modified surface of SPR more efficiently than EDC/NHS and EDC/sulfo-NHS Re-cently, Lee and co-workers [34] compared two antibody-oriented immobilization methods that adopted thiol-conjugated secondary antibodies and thiolated-protein A/G linkers The secondary antibody-mediated attach-ment method provided better antigen-binding efficiency compared with the other strategy
As discussed, studies of various antibody immobiliza-tion approaches have been extensively reported in the literature However, no consensus has been achieved on recommendations for antibody attachment approaches Furthermore, the use of antibody immobilization to im-prove immunosensor performance continues to challenge the immunosensor fabrication process Therefore, a suit-able immobilization method for each immunosensor ap-plication needs to be found In this paper, we report three different approaches for immobilizing antibodies
on the surface of electrode modified by using CeO2 nanowires to detect V ibrio cholerae O1 These
meth-ods include adsorption, the use of EDC/NHS-activated antibodies, and the use of protein A Our main aim was
to evaluate the effects of the different antibody immo-bilization strategies on the response of the CeO2 NWs-based immunosensor The results of this study would provide insight into the best immobilization method for the design of CeO2 NWs-based immunosensors, as well
as immunosensors in general
II EXPERIMENTS
CeO(NO3 3·6H2O, H2O2, toluene, and antibodies against V cholerae O1 (anti-V cholerae O1) were
provided by Invitrogen Co Phosphate buffered saline PBS (0.01 M, pH 7.4), EDC, NHS, bovine serum albu-min (BSA), 98% H2SO4, KCr2O7, protein A, and were purchased from Sigma-Aldrich Potassium ferrocyanide and potassium derricyanide were acquired from Beijing Chemical Reagent (China) All solutions were prepared with de-ionized (DI) water (18.2 MΩ·cm).
In this work, a microelectrode was utilized as a sen-sor for electrochemical measurements Briefly, the sensen-sor was fabricated by sputtering 10 nm Cr and 200 nm Pt onto a∼150 nm thick silicon-dioxide (SiO2) layer ther-mally grown on top of a silicon wafer The sensor sur-face was initially cleaned with KCr2O7 in 98% H2SO4, followed by cyclic voltammograms sweep from −1 V to
+2.1 V, at a scan rate of 25 mV/s in 0.5 M H2SO4
Trang 3Fig 1 (Color online) (a) schematic of an electrode
struc-ture with a photograph of the electrodes, (b) fingers of the
electrode image with higher magnification, (c) the FE-SEM
image of CeO2nanowires deposited on the electrode, and (d)
the FE-SEM image of CeO2 nanowires with higher
magnifi-cation
to activate the sensor’s surface CeO2 NWs were
pre-pared by a reaction with 1% (v/v) APTS solution in
ethanol for 36 h at room temperature Afterward, a
mix-ture of APTS/CeO2NWs was centrifuged for 60 min at
3500 rpm and washed thrice with ethanol Subsequently,
APTS/CeO2 NWs were sprayed onto the activated
elec-trode’s surface (Fig 1)
The APTS/CeO2-modified electrode was incubated in
PBS solution (PBS 0.01 M, pH 7.4) containing 12μg/mL
of anti-V cholerae O1 and then stored at 4 ◦C overnight.
Continuously, the electrode’s surface was rinsed with
double-distilled water and dried under nitrogen flow To
block nonspecific sites, we immersed the immunosensor
in PBS solution containing 1% BSA for 30 min, washed
it with DI water, and dried under nitrogen flow
Fig-ure 2(a) shows a schematic illustration of antibody
im-mobilization on the surface of the CeO2 NWs-deposited
electrode by absorption
For the EDC/NHS - activated antibody method, first
12μg/mL of anti-V cholerae O1 was incubated in 50 μL
of solution containing 5 mg/mL of EDC and 10 mg/mL
of NHS in 0.01 M PBS buffer for 20 min at pH 7.4 and
room temperature The EDC binds carboxyl groups to
primary amines, forming an O-acylisourea intermediate
This product is unstable in aqueous solutions Thus,
NHS is required for stabilization by converting the
inter-mediate to an amine-reactive NHS ester [33,35] Second,
the electrode modified with APTS/CeO2 NWs was
im-mersed in a solution of EDC/NHS-activated antibodies
for 24 h at room temperature In this step, the activated
antibodies crosslinked with the free amino groups on the
surface of the electrode modified with APTS/CeO2NWs
Afterward, the immunosensor was rinsed with
double-distilled water and dried under nitrogen flow Finally,
1% BSA was added to the modified nanowire’s surface
to block nonspecific sites as previously described The immunosensor was rinsed with DI water and dried under nitrogen flow When not in use, the immunosensors were stored at 4 ◦C in a refrigerator Figure 2(b) displays a schematic illustration of the antibody immobilization on the surface of the APTS/CeO2NWs-deposited electrode using EDC/NHS-activated antibodies
Similarly, protein A was activated with EDC/NHS
as mentioned in Ref [22] Then, an electrode modi-fied with APTS/CeO2 NWs was dipped in a solution of EDC/NHS-activated protein A for 24 h at room tem-perature Subsequently, 50 μL of anti-V cholerae O1
(12μg/mL) was dropped on the electrode modified with
protein A/APTS/CeO2 NWs Finally, the electrode was treated with 1% BSA for 30 min to prevent nonspecific binding, as previously described In each step, the elec-trode was rinsed with double-distilled water and dried under nitrogen flow Figure 2(c) shows a schematic of the antibody immobilization on the surface of APTS/CeO2 NWs-deposited electrode using protein A
For the electrochemical impedance spectroscopy (EIS) measurements, an impedance analyzer with IM6-THALES software was used to detect the cell concen-trations of V cholerae O1 In this work, the electrode
modified with anti-V cholerae O1 was immersed in a
measuring cell that was filled with 5 mL of 0.01 M PBS solution (pH 7.4) containing a defined cell concentration
of V cholerae O1 for 90 min at room temperature to
form an antibody-antigen complexes The immunosen-sor was rinsed with buffer solution to remove the non-specifically adsorbed cells The immunosensor responses were monitored by dipping the modified sensor in 2 mL of 0.01 M PBS solution containing 20 mM [Fe(CN)6]3−/4−
as an indicator probe The detected immunosensor was connected to the test and sense probes, and the Pt elec-trode was connected to the counter elecelec-trode on the IM6-impedance analyzer An Ag/AgCl electrode was used
as a reference electrode All tests were conducted in an open circuit The tested frequency range was 1 Hz to 100 kHz, with an amplitude of ±5 mV The Nyquist plots
were recorded The differences in the electron-transfer resistance (R et) were considered as the signal produced
by the interaction between the antibodies and the cells For cyclic voltammetry (C-V) measurements, an IM6-impedance analyzer with IM6-THALES software was used under the C-V program The electrode modified with anti-V cholerae O1 was immersed in a measuring
cell that was filled with 5 mL of 0.01 M PBS solution (pH 7.4) containing a defined concentration ofV cholerae O1
cells for 90 min at room temperature to form antibody-antigen complexes The immunosensor was rinsed with buffer solution to remove nonspecifically adsorbed cells Immunosensor responses were monitored by dipping the modified sensor in 2 mL of 0.01 M PBS solution con-taining 20 mM [Fe(CN)6]3−/4− as an indicator probe The detected immunosensor was connected to the test and sense probes, and the Pt electrode was connected to
Trang 4Fig 2 (Color online) Schematic illustration of antibody immobilization on the surface of a CeO2 - nanowire - modified electrode: (a) antibody absorption immobilization, (b) EDC/NHS-activated antibody method and (c) antibody immobilization via protein A
the counter electrode on the IM6-impedance analyzer
An Ag/AgCl electrode was used as a reference electrode
The potential was scanned from −0.2 V to 0.67 V at
a scan rate of 100 mV·s −1 The differences in the
cur-rent peak were regarded as the signal produced by the
interaction between the antibodies and the cells
III RESULTS
In this work, Fourier transforms infrared (FTIR) spectroscopy was used to verify the existence of CeO2 nanowires and anti-V cholerae O1 on the sensor’s
sur-face Figure 3 shows the FTIR spectra of (a) the
Trang 5Fig 3 (Color online) FTIR spectra of antibody
immo-bilized electrode’s surface: (a) ATPS/CeO2 NWSs modified
electrode, (b) absorption, (c) EDC/NHS - activated antibody,
and (d) protein A-mediated immobilization method
ATPS/CeO2 nanowires, (b) immobilization via
absorp-tion antibody, (c) immobilizaabsorp-tion via EDC/NHS
acti-vated antibody and (d) protein A-mediated
immobiliza-tion The FTIR spectral features of the CeO2 nanowires
sample displays in Fig 3(a) The intense band observed
round 825 cm−1 is due to the Ce−O−C stretching
vi-bration The peak at 3410 cm−1 is related to hydrogen
bond of − OH groups of water molecules or surface −
OH groups The peak around 542 cm−1is assigned to the
Ce−O stretching band The presence of APTS was
con-firmed by the aliphatic C-N character around 1230 cm−1,
and the peak at 1589 cm−1confirmed primary N-H
bend-ing When the antibody is absorbed on the electrode’s
surface, peaks corresponding to C=O stretching and
N-H bending of amin I form at 1671 cm−1, 1695 cm−1,
respectively The band at 1548− 1586 cm −1 is assigned
to the N-H bending of amin II vibration Meanwhile, the
peak around 467 cm−1 is assigned to the Ce−O
stretch-ing band, and the peak at 1430 cm−1 is assigned to the
bending vibration of C-H stretching of CeO2(Fig 3(b))
Figure 3(c) shows the FTIR spectrum for antiboby
im-mobilization via EDC/NHS-activated antibody It can
be seen that the peak at 1673 cm−1corresponding to the
C=O stretching of amin I The peak around 610 cm−1is
assigned to the Ce−O stretching band, the intense band
round 825 cm−1is due to the Ce−O−C stretching
vibra-tion In the case of protein A-mediated immobilization
(Fig 3(d)), the presence of a 825 cm−1 peak, which is
clearly seen in the spectrum corresponds to Ce−O−C
Fig 4 (Color online) Fluorescence images of antibody im-mobilized electrode: (a) ATPS/CeO2 NWSs modified elec-trode, (b) absorption, (c) EDC/NHS - activated antibody, and (d) protein A-mediated immobilization methods
stretching vibration, the peak at 695 cm−1 is assigned to the Ce−O stretching band, and peak at 556 cm −1 is
as-signed to the Ce−O stretching band The peak at 1675
cm−1corresponds to the C=O stretching of amin I, con-firming the immobilization of antibodies Additionally, the peak at 3435 cm−1 is related to the O-H stretching vibration of H2O in sample
The density of antibodies on an APTS-CeO2 NWs modified electrode was studied by using fluorescence microscopy, and the results are shown in Fig 4 APTS/CeO2 NWs deposited electrode’s surface was clearly black (Fig 4(a)) while the surface of the elec-trode with antibodies immobilized by using absorption (Fig 4(b)), EDC/NHS activated antibody (Fig 4(c)), and protein A-mediated immobilization (Fig 4(d)) clearly shown green fluorescence spots This could con-firm that antibodies were immobilized on the electrode’s surface
In this work, we used EIS to determine the im-munosensor response defined by the interaction between the antibodies and the V cholerae O1 cells The
re-sponses of the immunosensors prepared by using the three-immobilization methods were compared and eval-uated to determine the best immobilization method for the CeO2NWs-based immunosensor As mentioned in a previous study [36], EIS is an effective technique for de-veloping biosensors that detect bacteria The principle underlying the function of the EIS-based immunosensor depends on measurements of electrochemical Faradaic impedance with [Fe(CN)6]3−/4− as the indicator probe The electrode was immersed in PBS solution containing [Fe(CN)6]3−/4−, and an alternating current potential of
5 V was applied to the electrode Consequently, oxi-dation and reduction of the [Fe(CN)6]3−/4− occurred The electrons were transferred between the two fingers
of the electrode array When linkers modified the
Trang 6elec-Fig 5 (Color online) Nyquist diagrams for impedance measurements of immunosensors in the presence of 20 mM [Fe(CN)6]3−/4− in a 1 mM PBS solution: (A) Antibody absorption method: (a) bare electrode, (b) APTS-CeO2 nanowire-modified electrode, (c) antibodies/APTS-CeO2 NWs-modified electrode, (d) BSA/antibodies/APTS-CeO2 NWs-modified elec-trode, and (e) cells/BSA/antibodies/APTS-CeO2 NWs-modified electrode (B) Immobilization method via EDC/NHS acti-vated antibody: (a) bare electrode, (b) APTS-CeO2 NWs-modified electrode, (c) antibodies-EDC-NHS/APTS-CeO2 NWs-modified electrode, (d) BSA/antibodies-EDC-NHS/APTS-CeO2NWs-modified electrode, and (e) cells/ BSA/antibodies-EDC-NHS/APTS-CeO2 NWs-modified electrode (C) Immobilization method via protein A: (a) bare electrode, (b) APTS-CeO2 NWs-modified electrode, (c) protein A/APTS-CeO2 NWs-modified electrode, (d) antibodies/protein A/APTS-CeO2 NWs-modified electrode, (e) BSA/antibodies/protein A/APTS-CeO2NWs-modified electrode, and (f) cells/BSA/antibodies/Protein A/APTS-CeO2 NWs-modified electrode (D) Comparison of the immunosensor responses for the three different antibody immobilization methods
trode’s surface, a thin film layer was formed, leading to
the inhibition of electron transfer between the fingers
Thereby, an increase in electron-transfer resistance was
detected Figure 5 presents the Nyquist plots for
bacte-rial cell detection by immunosensors prepared by using
the three different immobilization methods: (A)
absorp-tion, (B) EDC/NHS-activated antibody, and (C) protein
A-mediated methods Meanwhile, Fig 5(D) compares
the immunosensor output signals for the three
differ-ent immobilization methods As observed in Fig 5(A),
when the bare electrode was immersed in PBS
solu-tion containing the redox probe, the reducsolu-tion process
of the redox probe was initiated, and electrons were
transferred between the two electrodes through the
re-dox probe [Fe(CN)6]3−/4− The electron transfer was
not blocked by any monolayer on the electrode’s surface
Thus, the electron-transfer resistance (Ret) was
deter-mined to be 745 Ω, as shown in Fig 5A(a), thereby
indi-cating high electron-transfer kinetics of the redox probe
at the electrode interface After the APTS/CeO2 NWs modified electrode surface, a thin film layer that could have hindered electron transfer from [Fe(CN)6]3−/4− to the conductive electrode’s surface was formed However, the positively-charged APTS/CeO2 nanowire promoted transfer of the negative redox probe to the electrode’s surface by electrostatic attraction Thus, the value of Ret decreased to 695 Ω, as shown in Fig 5A(b) When the antibodies were absorbed directly onto the electrode’s surface modified with APTS/CeO2 nanowires, a layer of antibodies was formed on the elec-trode surface This layer inhibited the electron trans-fer between the fingers of the electrode An increase
in the electron-transfer resistance of 812 Ω was noted,
as shown in Fig 5A(c) To block the unreacted and nonspecific sites, the electrode surface modified with antibodies/APTS/CeO2 nanowires was treated with 1%
Trang 7Table 1 Comparison of the analytical parameters of the developed immunosensors forV cholera O1 detection by using the
three different immobilization methods
BSA for 30 min AR etof 858 Ω was obtained, as shown
in Fig 5A(d) When the cells are bound to the
modi-fied electrode’s surface, a reaction between the antibodies
and the cells potentially occurs Consequently,
immuno-complexes would then attach to the electrode’s surface,
forming an additional barrier that would further increase
the electron-transfer resistance The Ret was 915 Ω, as
shown in Fig 5A(e)
The changes in electron-transfer resistance before and
after cell detection were calculated by the following
equa-tion:
%ΔR et= 100(R et cell/antibody −R et antibody)/R et antibody
(1) WhereR et antibody andR et cells/antibody are theR et
val-ues before and after the cells bind to the
antibody-modified electrode’s surface, respectively A 6.65%
in-crease in the electron-transfer resistance was observed
after cell binding Antibodies were confirmed to be
im-mobilized successfully on the electrode’s surface by
us-ing the adsorption method Figure 5B illustrates the
Nyquist plots for the immunosensors prepared by using
the EDC/NHS-activated antibody approach As
previ-ously mentioned, the value of Ret for the bare electrode
and the electrode surface modified with APTS/CeO2
NWs were 745 and 695 Ω, respectively The Ret
in-creased to 847 Ω when EDC/NHS-activated antibodies
were immobilized on the surface of the electrode
mod-ified with APTS/CeO2 NWs, as shown in Fig 2B(c)
When the modified electrode’s surface was blocked with
1% BSA for 30 min, the value ofR et increased to 892 Ω,
as shown in Fig 5B(d) The value ofR etincreased
signif-icantly (965 Ω) when cells were bound to the
antibody-modified surface of the electrode, as shown in Fig 5B(e)
By using Eq (1), we calculated the change in R et to be
about 8.3%
Figure 5C shows the response of the immunosensor
prepared by antibody immobilization through protein A
In this work, protein A was immobilized on the electrode
modified with APTS/CeO2 nanowires, as described in a
previous study [22] The terminal carboxyl groups of
protein A were activated by incubation in EDC/NHS
solution for 20 min to produce an active intermediate of
the NHS ester (Fig 2(c)) Afterward, the electrode
mod-ified with APTS/CeO2NWs was immersed in a solution
of EDC/NHS-activated protein A for 24 h at room tem-perature This step led to a crosslinking of the activated protein A with the free amino groups on the electrode’s surface modified with APTS/CeO2 nanowires A Ret of
870 Ω was obtained for protein A immobilization on the modified electrode (Fig 5C(c)) When anti-V cholerae
O1 was dropped on the protein A/APTS/CeO2 -NW-modified electrode, the value of Ret increased to 910 Ω,
as shown in Fig 5C(d)
To avoid the binding of other proteins to the unreacted NHS ester, the electrode was blocked with 1% BSA for
30 min, which resulted in an increase in the value ofR et
to 955 Ω, as shown in Fig 5C(e) When the cells inter-acted with the antibodies immobilized on the electrode surface, aR etof 1110 Ω was noted By using Eq (1), we calculated the change in Ret to be 16.2% A comparison among the responses of the immunosensors prepared by using the three different antibody immobilization meth-ods is shown in Fig 5D The highest and the lowest re-sponses were achieved by the immunosensors prepared by antibody immobilization with protein A (16.2% accord-ing to Eq (1)) and with adsorption (6.65%) The re-sponse of the immunosensor prepared using EDC/NHS-activated antibodies was 8.3% The analytical param-eters of the developed immunosensors for V cholera
O1 detection were also compared (Table 1) As shown in Table 1, the order of the antibody immobilization strate-gies in terms of resultant immunosensor detection limit and sensitivity is as follows: adsorption immobilization
< immobilization using EDC/NHS-activated antibodies
< protein A-mediated immobilization Meanwhile, the
linearity range changed insignificantly
C-V studies on the immunosensor detection of bacte-rial cells were performed in PBS solution containing 20
mM [Fe(CN)6]3−/4−at a scan rate of 100 mV·s −1versus
a Ag/AgCl reference electrode Figure 6 presents the C-V characterizations of the immunosensors prepared
by antibody immobilization through (A) absorption, (B) the use of EDC/NHS-activated antibodies, and (C) the use of protein A Meanwhile, Fig 6D displays a com-parison of the responses of the immunosensors prepared
by using the three different methods As presented in Fig 6A(a, b), the C-V response of the APTS/CeO2 NWs-modified electrode is shifted relative to that of the bare electrode The peak current obviously increased
up to 84 μA because the APTS/CeO2 NWs film could
Trang 8Fig 6 (Color online) Cyclic voltammetry of the sensor in the presence of 20 mM [Fe(CN)6]3−/4− in a 1mM PBS solu-tion, at a scan rate 100 mV/s (A) Antibody absorption method: (a) bare electrode, (b) APTS-CeO2 NWs-modified elec-trode, (c) antibodies/APTS-CeO2 NWs-modified electrode, (d) BSA/antibodies/APTS-CeO2 NWs-modified electrode, and (e) cells/BSA/antibodies/APTS-CeO2 NWs-modified electrode (B) Immobilization method via EDC/NHS - activated antibody: (a) bare electrode, (b) APTS-CeO2 NWs-modified electrode, (c) antibodies-EDC-NHS/APTS-CeO2 NWs-modified electrode, (d) BSA/antibodies-EDC-NHS/APTS-CeO2 NWs-modified electrode, and (e) cells/BSA/antibodies-EDC-NHS/APTS-CeO2 NWs-modified electrode (C) Immobilization method via protein A: (a) bare electrode, (b) APTS-CeO2 NWs-modified elec-trode, (c) protein A/APTS-CeO2NWs-modified electrode, (d) antibodies/protein A/APTS-CeO2NWs-modified electrode, (e) BSA/antibodies/protein A/APTS-CeO2NWs-modified electrode, and (f) cells/BSA/antibodies/Protein A/APTS-CeO2 NWs-modified electrode (D) Comparison of the immunosensor responses for the three different antibody immobilization methods
promote electron transfer of [Fe(CN)6]3−/4− A slight
decrease in the peak current (77μA) and a separation of
the peak potential (0.38 V) were observed for antibody
absorption on the APTS/CeO2NWs-modified electrode
surface, as shown in Fig 6A(c)
When the antibody-modified electrode’s surface was
blocked with 1% BSA for 30 min, the current value
de-creased continuously by 75 ΩA An insignificant
sepa-ration of the peak potential (0.36 V) was observed, as
shown in Fig 6A(d) When the cells interacted with
the modified electrode’s surface, a decrease in the peak
current of 70 μA was observed, and this decrease
cor-responded to a signal variation of about 7.5%
com-pared with that of the modified electrode, as shown in
Fig 6A(e)
In the case of EDC/NHS activated antibody
immobi-lization method, a decrease in the peak current of 76.5
μA was observed when EDC/NHS-activated
antibod-ies were bound to the APTS/CeO2-NW-modified elec-trode’s surface, as shown in Fig 6B(c) A change of 69
μA was observed continuously after blocking with 1%
BSA for 30 min (Fig 6B(d)) When cells were bound
on the electrode’s surface, a further decrease in the peak current was found at 63 μA, as shown in Fig 6B(e).
Therefore, the current value varied by about 9% com-pared with the signal for the antibody-modified elec-trode
For the immunosensors prepared using protein A,
a slight decrease in the peak current (70 μA) was
observed under protein A immobilization on the APTS/CeO2-NWs-modified electrode surface, as shown
in Fig 6C(c) When antibodies attached onto the pro-tein A/APTS/CeO2-NWs-modified electrode, a further decrease of 68 μA in peak current was noted, as shown
Trang 9in Fig 6C(d) A drop in the peak current of 66.7 μA
was obtained continuously when the sensor’s surface was
blocked with 1% BSA for 30 min, as shown in Fig 6C(e)
When cells bound to the electrode surface, a decrease in
the peak current (11%) was clearly observed (Fig 6C(f))
Thus, the voltammetric behavior of the redox probe
was demonstrated to be affected by electrode’s surface
modification Changes in the current peak were observed
because of electrode’s surface modification, which
cated another barrier layer that blocked access of the
re-dox probe to the electrode’s surface within the applied
potential The change in current was highest for the
im-munosensors with antibodies immobilized using protein
A (11%), followed by those prepared using
EDC/NHS-activated antibodies (9%), and finally, by those prepared
by direct adsorption (7.5%), as shown in Fig 6D
The stability of the immunosensors prepared by using
different immobilization methods was studied and
eval-uated The nine immunosensors that utilized different
antibody immobilization methods were stored in 0.1 M
PBS (pH 7.4) at 4 ◦C for 135 days and subsequently
analyzed at different times (45 day/times), as shown in
Fig 7(a) Repeatable signals were observed for up to
45 days for all the immunosensors At day 90, the
sig-nal response of the stored immunosensors was changed
by approximately 11.4% for those prepared by using
di-rect antibody absorption, 6.1% for those prepared by
using EDC/NHS-activated antibodies, and 5.6% for
pro-tein A-mediated immobilization No response signal was
detected for all of the immunosensors after they had been
stored for 135 days, indicating a loss of biological activity
of the anti-V cholerae O1, which could have been
dena-tured by that time; thus, no binding with the bacterial
cells was exhibited By these results, we can conclude
that the immunosensors exhibited acceptable stability
and that the immunosensor using protein A showed the
greatest stability
The specificity of the immunosensors was also
stud-ied by usingSalmonella and Escherichia coli O157:H7
bacteria as control samples No response signals of the
immunosensors were detected for either bacterial species
Shifts in the value ofR et of 900, 975, and 1010 Ω
corre-sponding to 5.9%, 14.7%, and 18.8%, respectively, after
V cholerae O1 detection were noted after antibody
im-mobilization by adsorption, EDC/NHS activation, and
protein A, respectively (Fig 7(b)) These results show
that the specificities of the immunosensors prepared
un-der the three approaches were all high
Regeneration is a significant aspect in the
develop-ment of immunosensors for in-field/on-site detection To
study the regeneration of the immunosensors,
antibody-immobilized electrodes were immersed in a buffer
solu-tion containingV cholerae O1 for 90 min Then, the
im-munosensors were washed with PBS buffer solution and
DI water and dried with nitrogen gas The V cholerae
O1 cell concentration was determined from the change
in the measure value of Ret After detection of cells of
V cholerae O1, the immunosensor was dipped into a
Fig 7 (Color online) (a) Stability of the immunosensor at different times for 135 day, (b) specificity of the immunosen-sor toward E coli O157: H7, salmonella, and V cholerae O1 bacterium, and (c) the regeneration performance of the immunosensor
glycine-HCl buffer (pH 2.8) for about 10 min to remove cells Subsequently, the immunosensor was washed with PBS buffer solution, DI water and dried with nitrogen gas The immunosensor was again measured with cells
of V cholerae O1 under the same conditions The
ob-tained results indicate that the signal responses had de-creased by approximately 25%, 18%, and 11% for the im-munosensors prepared by using adsorption,
Trang 10EDC/NHS-activated antibodies, and protein A-mediated
immobi-lization, respectively (Fig 7(c))
IV DISCUSSION
The three-immobilization methods were compared to
determine the best immobilization method to prepare
CeO2-NWs-based immunosensors for V cholerae O1
detection As mentioned above, the changes in the
immunosensor’s response after the use of the three
different immobilization methods were approximately
6.65%, 8.3%, and 16.2% for the antibody adsorption,
the EDC/NHS-activated antibody, and the protein
A-mediated immobilization methods, respectively The
variation in the immunosensor’s signal was also
deter-mined by using cyclic voltammetry The peak current
changed by about 7.5% for antibody adsorption, 9% for
EDC/NHS-activated antibody immobilization, and 11%
for antibody immobilization via protein A Antibody
ad-sorption exerted no significant effect on the
immunosen-sor’s response In contrast, the immunosenimmunosen-sor’s response
increased significantly when antibodies were immobilized
by using EDC/NHS activation and protein A As
previ-ously discussed, the change in the immunosensor’s
re-sponse could be attributed to several factors including
the following: 1) the random orientation of antibodies
and steric hindrance caused by improper antibody
ori-entation with respect to the electrode’s surface, 2)
de-naturation of immobilized antibodies by the
environmen-tal pH and temperature, and 3) biocompatibility
Sev-eral researchers [31, 37–39] have reported that protein
A could specifically bind with the Fc fragment of
anti-bodies and control antibody orientation This attribute
does not block the antibody’s active sites for antigen
binding, resulting in high cell-binding capacities
Addi-tionally, protein A’s biocompatibility conferred the
im-munosensor with enhanced detection performance In
the immunosensors prepared with EDC/NHS-activated
antibodies, the antibodies could be crosslinked to the
amino groups on the surface of the APTS/CeO2
-NWs-modified electrode This occurrence would remove some
biological activity and prevent some of the random
ori-entations of the antibodies Several binding sites of
an-tibodies could be blocked because of steric hindrance of
the antigen-binding domains For this reason, the
munosensor’s response was lower than that of the
im-munosensors prepared via immobilization using protein
A
For direct antibody absorption, antibodies were
ab-sorbed on the sensor’s surface by weak binding, such as
van der Waals or electrostatic interactions However,
these interactions can be denatured by thermal and
acid-base environmental or analytical chemistry conditions
Furthermore, antibody orientation was random,
lead-ing to low cell-bindlead-ing capacity and low cell detection
Therefore, the immunosensor response was the lowest
among those of all methods tested Thus, CeO2 -NWs-based immunosensors should be prepared by using the protein A-mediated immobilization method to enhance immunosensor performance
V CONCLUSION
Our study mainly aims to evaluate the effects of differ-ent antibody immobilization strategies on the response
of CeO2-NWs-based immunosensors We determined the best immobilization method to apply for the design of CeO2-NWs-based immunosensors and immunosensors in general In this study, we compared the performances of immunosensors prepared by using three antibody immo-bilization methods: adsorption, the use of EDC/NHS-activated antibodies, and the use of protein A-mediated antibodies Both the adsorption and the EDC/NHS-activated antibody methods are simple, cost effective, and appropriate for large-scale production However, im-proving sensor performance using these approaches was difficult because of low cell-binding capacity from ran-dom antibody orientations and denaturation under en-vironmental conditions In contrast, protein A-mediated immobilization yielded the best orientation and antigen-binding efficiency among the tested methods This ap-proach was then confirmed as the most suitable strat-egy for preparing CeO2-NWs-based immunosensors for
V cholerae O1 detection.
ACKNOWLEDGMENTS
This work was supported by the Ministry of Edu-cation and Training under the research project code B2014.01.78
REFERENCES
[1] S Weng, M Chen, C Zhao, A Liu, L Lin, Q Liu, J Lin and X Lin, Sensors and Actuators B184, 1 (2013).
[2] C Song, G Xie, L Wang, L Liu, G Tian and H Xiang, Biosensors and Bioelectronics 58, 68 (2014).
[3] X Wang, J Miao, Q Xia, K Yang, X Huang, W Zhao and J Shen, Electrochimica Acta112, 473 (2013).
[4] M.-J Song and S W Hwang, J Korean Phys Soc.54,
1612 (2009)
[5] H.-K Seo, D J Park and J Y Park, J Korean Phys Soc.51, S284 (2007).
[6] R C Alves, F B Pimentel, H P A Nouws, R C
B Marques, M B Gonz´alez-Garc´ıa, M Beatriz, P
P Oliveira and Cristina Delerue-Matos, Biosensors and Bioelectronics64, 19 (2015).
[7] G Yang, W Jin, L Wu, Q Wang, H Shao, A Qin, B
Yu, D Li and B Cai, Analytica Chimica Acta706, 120
(2011)