Modified screen printed electrode for development of a highly sensitive label free impedimetric immunosensor to detect amyloid beta peptides

Method A: sensor development on carbon DEP chip First, a volume of 1.5mL of 1-pyrenebutanoic acid, succinimidyl ester 100 mM was placed onto carbon ink working electrode of DEP chip for

Modi fied screen printed electrode for development of a highly

sensitive label-free impedimetric immunosensor to detect amyloid

beta peptides

Truong T.N Liena,b, Yuzuru Takamuraa, Eiichi Tamiyac, Mun'delanji C Vestergaarda,d,*

a School of Materials Science, Japan Advanced Institute of Science and Technology (JAIST), 1-1 Asahidai, Nomi, Ishikawa, 923-1292, Japan

b School of Engineering Physics, Hanoi University of Science and Technology (HUST), No.1 Dai Co Viet, Hai Ba Trung, Hanoi, Viet Nam

c Department of Applied Physics, Graduate School of Engineering, Osaka University, 2-1 Yamadaoka, Suita, Osaka, 565-0871, Japan

d Department of Biochemical Science and Technology, Faculty of Agriculture, Kagoshima University, Korimoto-1-21-24, Kagoshima City, Kagoshima,

890-0065, Japan

h i g h l i g h t s g r a p h i c a l a b s t r a c t

A label-free impedimetric

immuno-assay for amyloid beta was

developed

Sensitivity enhanced by elaborate

surface chemistry manipulation

us-ing SAM of AuNPs

Immobilized Protein G enhanced

sensitivity by directing optimal

anti-body orientation

Lack of interference from

high-abundant high-molecular weight

BSA demonstrated

a r t i c l e i n f o

Article history:

Received 24 October 2014

Received in revised form

10 August 2015

Accepted 11 August 2015

Available online xxx

Keywords:

Amyloid beta (Ab)

Immunosensor

Electrochemical impedance spectroscopy

(EIS)

Screen-printed electrode (SPE)

Disposable electrochemical printed (DEP)

chip

a b s t r a c t

Alzheimer's disease (AD) is a fatal neurodegenerative disease affecting approximately 26 million people world-wide, and the number is increasing as life expectancy increases Since the only reliable diagnosis for the pathology is histochemical post-mortem examination, there is a rather urgent need for reliable, sensitive and quick detection techniques Amyloid beta, being one of the established and widely accepted biomarkers of AD is a target biomolecule

Herein, we present fabrication of a labelless impedimetric amyloid beta immunosensor on carbon DEP (disposable electrochemical printed) chip Three types of amyloidbimpedimetric immunosensors were fabricated in a systematic step-wise manner in order to understand the effects that each surface modification chemistry had on detection sensitivity We found that compared to a bare electrode, surface modification through formation of SAM of AuNPs increased sensitivity by approximately three orders of magnitude (LoD from 2.04mM to 2.65 nM) A further modification using protein G, which helps orientate antibodies to an optimum position for interaction with antigen, lowered the LoD further to 0.57 nM We have demonstrated that the presence of one of the most abundance proteins in biologicalfluids, bovine serum albumin (BSA), did not interfere with the sensitivity of the sensor Since the DEP chips are disposable and the detection platform label-free, the developed sensor is relatively fast and cheap These methods could easily be applied for detection of other antigens, with selection of the detection platform based on the desired for sensitivity

© 2015 Elsevier B.V All rights reserved

* Corresponding author School of Materials Science, Japan Advanced Institute of Science and Technology (JAIST), 1-1 Asahidai, Nomi, Ishikawa, 923-1292, Japan E-mail address: munde@agri.kagoshima-u.ac.jp (M.C Vestergaard).

Contents lists available atScienceDirect Analytica Chimica Acta

j o u rn a l h o m e p a g e : w w w e l s e v i e r c o m / l o c a t e / a c a

http://dx.doi.org/10.1016/j.aca.2015.08.036

0003-2670/© 2015 Elsevier B.V All rights reserved.

Analytica Chimica Acta xxx (2015) 1e8

1 Introduction

Alzheimer's disease (AD) is a fatal neurodegenerative disease

affecting approximately 26 million people world-wide and the

number is increasing as life expectancy increases [1] However,

diagnosis remains in the hands of medical doctors who can only at

best propose‘probable Alzheimer's or dementia of the Alzheimer

type’ since there is no current test or procedure that is diagnostic

Although not unilaterally agreed upon, the progressive decline of

patients with AD has been correlated with extracellular deposition

of amyloid plagues, of which amyloid beta is the major constituent

[2] Amyloid beta (Ab) has therefore become an important

biomarker for the pathology The main detection method is by

enzyme-linked immunosorbent assay (ELISA) techniques, which

are lessflexible, costly, and labour-intensive[3,4] The past decade

or so has seen tremendous effort put into development of sensitive

and selective detection techniques for this and other peptides/

proteins They include FRET-based assays [5]; surface Raman

enhanced spectroscopy (SERS) [6]; and several electrochemical

platforms[7,8]

Electrochemical impedance spectroscopy (EIS) recently has

attracted much interest because it has some important advantages

over number of electrochemical methods such as amperometry and

potentiometry With EIS, developed sensing platforms are (i)

label-free with detection based on direct specific binding events, (ii) less

destructive to the activities of biomolecule due to the small voltage

excitation used during detection, (iii) a simple operation and very

sensitive, with comparable detection limits to optical-based

sen-sors[9e11] EIS biosensors have been successfully employed for

detection of various biomolecules and biological processes

including DNA hybridization, at very low (femtomolar) detection

limits[12] Previously, we reported on an impedimetric

immuno-sensor development using DEP chips, and demonstrated its

selec-tive detection using a model protein, chorionic gonadotropin

hormone (hCG) (limit of detection (LoD) of 33 pg/mL)[13] Lien and

colleagues also modified DEP chips using a conducting co-polymer,

polypyrrole-pyrolecarboxylic acid for hCG detection The LoD was

lowered by an order of magnitude, to 2.3 pg/mL[14] Most recently,

Rushworth and colleagues have reported on specific detection of

oligomeric amyloid beta using biotylated peptide of prion protein

as the recognition element The authors have reported an

impres-sive detection limit of 0.5 pM[15]

In this work, we fabricated a labelless EIS immunosensor for

amyloid beta peptide, isoforms 40 and 42 We have used disposable

electrochemical printed (DEP) chips, which have been used in

development of various DNA- and immuno-biosensors, giving very

good reproducibility [13,14,16] We developed this sensor in a

systematic step-wise fashion so that we could also better

under-stand the effects of surface chemistry modification on sensor

sensitivity The developed sensors were very reproducible (coef

fi-cient of variation<8%) Although the immunosensor's sensitivity

(LOD ~ 0.57 nM) is still lower than the recently reported

prion-based sensor[15], it is a good proof-of-principle antibody-based

EIS With further improvement in surface chemistry modification, it

offers much promise Besides, since the antibodyeantigen

chem-istry is well-understood and the fabrication relatively simple and

rapid, this sensor can easily be adaptable for application to other

antigens

2 Experimental

2.1 Reagents

1-pyrenebutanoic acid, succinimidyl ester was supplied from

Eugene, Oregon (USA) Chloroauric acid (HAuCl4), Bovine serum albumin (BSA) and Dimethyl Sulfoxide Dehydrated (DMSO) were purchased from Sigma Aldrich 1-Ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC) hydrochloride was supplied by Dojindo (USA) N-hydroxysuccinimide (NHS) and 16-mercaptohexadecanoic acid (MHDA) were provided by Wako Amyloid beta (Ab) peptides (tri-fluoroacetate salt) were purchased from Peptide Institute Inc., (Osaka, Japan) N-terminal human monoclonal Abantibody (anti

mAb) was from Calbiochem (CA, USA) All other reagents used were

of the analytical grade or the highest commercially available purity and used as supplied without further purification All solutions were prepared with deionized water of resistivity no less than

18 MUcm

2.2 Electrodes Commercial disposable electrochemical printed (DEP) chips were obtained from BioDevice Technology Ltd., Japan (http://www biodevicetech.com) The chips were fabricated by screen-printing technology and designed as system with three electrodes con-taining carbon ink working, carbon ink counter and Ag/AgCl ink reference electrodes The carbon ink contained 75% (w/w) graphite powder and 25% (w/w) mineral oil (Sigma) The surface area of the working electrode is 2.64 mm2

2.3 Instrumentation

An AutoLab PGSTAT 30 system (EcoChemie B.V., Ultrecht, The Netherlands) was used to perform electrochemical impedance spectroscopy measurements The spectra was recorded in 0.1 M KCl solution containing 5 mM of K3[Fe(CN6)]/K4[Fe(CN6)] within fre-quency range from 100 kHz to 50 mHz An ac probe amplitude of

10 mV was applied to the system around the open circuit potential 2.4 Gold nanoparticles fabrication

The carbon ink electrode of DEP chip was modified first by deposition of gold nanoparticles (AuNPs) on working electrode using both potential step voltammetry (PV) and cyclic voltammetry (CV) Tetrachloroauric acid (HAuCl4) was diluted in 100 mM phos-phate buffer solution (PBS) to afinal concentration of 1 mM Then

35mL of the 1 mM HAuCl4 solution was dropped onto DEP chip electrode surface covering all three electrodes (including Ag/AgCl, counter and working electrodes) With the PV method,0.4 V was applied for different time periods: 5 s, 20 s and 90 s With the CV method,600 to þ500 mV vs Ag/AgCl were cycled for 5, 10 and 20 cycles at scan rate of 50 mV/s Following deposition, the AuNPs-coated electrodes were washed several times in 10 mM PBS, pH 7.4 solution containing 0.05% Tween 20 followed by deionized water, and drying under nitrogen (N2) stream This AuNPs-coated electrode was then ready for immobilization of anti mAb 2.5 Electrode fabrication and sensor development Immunosensors detect signals resulting from specific immu-noreactions between antibodies immobilized on a transducer and the target antigens In order to have good sensitivity for detection, the concentration of immobilized antibodies as well as their orientation on the transducer surface should be as optimal as possible in order to interact with as many target antigens as possible Therefore, surface modification for best possible antibody immobilization is desirable In our work, three methods were used

to fabricate the immunosensor based on the physio-chemical moiety of carbon DEP chip (methods A, B and C) The schematic diagram of these methods is presented in Fig 1 In method A, T.T.N Lien et al / Analytica Chimica Acta xxx (2015) 1e8

2

N-terminal human monoclonal Ab antibody (anti mAb) was

directly immobilized on carbon DEP chip surface via pyrenyl

groups In methods B and C, carbon DEP chips werefirst modified

by an in situ AuNPs synthesis using method described by Lien et al

[13] Briefly, AuNPs were electrodeposited on the DEP chips using

either PV or CV Then anti mAbwas immobilized on the

AuNP-modified carbon DEP chip via COOH group of self-assembled

monolayer (SAM) of 16-mercaptohexadecanoic acid (MHDA),

which can serve as a linker for covalent biomolecule

immobiliza-tion Method C is different from B in that the latter had a

pre-deposition of a protein G layer on SAM AuNPs before

immobiliza-tion of anti mAb Since protein G binds specifically to the

non-antigenic (Fc) regions of an antibody, it helps immobilize the

antibody in a way that favours its antigen binding sites (Fab) to be

oriented away from the solid phase, towards the antigen[18] A

schematic diagram of these methods is presented inFig 1

2.5.1 Method A: sensor development on carbon DEP chip

First, a volume of 1.5mL of 1-pyrenebutanoic acid, succinimidyl

ester (100 mM) was placed onto carbon ink working electrode of

DEP chip for 1 h followed by rinsing several times with deionized

water to wash away excess reagent followed by drying over a

stream N2gas In this step, the pyrenyl groups interacted strongly

with the basal plane of carbon graphite viap-stacking This leads to

the functionalization of carbon surface with succinimidyl ester

groups that are highly reactivated to nucleophilic substitution by

primary and secondary amines[17,18]that exist in abundance on

the surface of most proteins After, 1.5mL of 100mg/mL anti mAbwas

dropped onto surface of these electrodes and incubated for 1 h at

room temperature (RT), followed by further washing with 10 mM

PBS to remove any loosely bound antibodies The surface was then

dried over a gentle stream of N2gas Last, the anti mAb-modified

electrodes were subjected to 1.5mL of 100 mM ethanolamine for 1 h

in order to block the remaining nonspecific binding sites, that is,

any sites that were not antibody-immobilized [12,19] Last, the

electrodes also were rinsed with PBS followed by deionized water and then dried over a gentle stream N2gas The immunosensors were ready to use at this point

2.5.2 Method B: sensor development on Au-NP-modified carbon DEP chip

In this method, the anti mAb was immobilized onto AuNPs-modified electrode surface via COOH group of self-assembled monolayer (SAM) of 16-mercaptohexadecanoic acid (MHDA), which can serve as a linker for covalent biomolecular immobiliza-tion MHDA solution was prepared in deionized water at 0.5 mM andfiltered through a 0.20mm pore size mesh To form SAMs, 1.5mL

of 0.5 mM MHDA was dropped onto AuNPs-modified working electrode surface for 45 min at RT In this step, the thiol groups of the mercaptoalkanes interacted with the AuNPs, forming AueS bond The carboxyl groups of MHDA layer need to be activated before they can react with the amino groups of anti mAb The activation is carried out using N-(3-dimethylaminopropyl)- N0 -ethylcarbodiimide hydrochloride (EDC, 0.2 M), and N-hydrox-ysuccinimide (NHS, 0.1 M), which were prepared in deionized water A volume of 1.5 mL of the solution was placed onto the AuNPs-modified working electrode surface for 15 min at RT The reason for using both EDC hydrochloride and NHS for activation of MHDA layer has been explained in more detail in our previous work

[14] After activation, the electrodes were rinsed with 10 mL of

10 mM PBS containing 0.05% Tween 20 followed by deionized water, and dried under a gentle stream N2 gas Then, 1.5 mL of

100mg/mL anti mAbsolution was placed onto the activated SAM-AuNPs modified electrode for 1 h at RT In this procedure, anti

mAb was immobilized successfully onto the AuNPs-modified electrode Following this, the electrode was washed with 10 mM PBS containing 0.05% Tween 20 followed by deionized water to remove the loosely bound antibodies and dried over a stream N2 gas In order to block non-specific adsorption, 100 mM ethanol-amine was used as described in method A The electrode was also

Fig 1 Schematic diagram showing the fabrication of impedimetric immunosensors based on disposable electrochemical printed (DEP) chip following A) Method A, B) Method B and C) Method C In method A, carbon ink electrode of the DEP chip was modified first by a functional molecule, 1-pyrenebutanoic acid succinimidyl ester and thence anti mAb

immobilization via succinimidyl ester groups In methods B and C, the carbon ink electrode of the DEP chip was modified first by deposition of gold nanoparticles (AuNPs) using cyclic voltammetry method After that, anti mAb was immobilized onto AuNPs-modified electrodes via COOH group of self-assembled monolayer (SAM) of 16-mercaptohexadecanoic acid (MHDA), which can serve as a linker for covalent biomolecule immobilization Furthermore, in the method C, in order to orientate immobilized an-tibodies optimally for enhanced antigen detection, protein G was immobilized before anti mAbimmobilization In all sensors ethanolamine was used to block the remaining non-specific binding sites after antibody immobilization.

T.T.N Lien et al / Analytica Chimica Acta xxx (2015) 1e8 3

rinsed with 10 mM PBS solution containing 0.05% Tween 20

fol-lowed by deionized water, then dried over a gentle stream N2gas

and used immediately

2.5.3 Method C: immobilization of protein G on AuNPs-modified

carbon DEP chip for enhanced immuno-sensor sensitivity

In this method, in order to help orientate the immobilized

an-tibodies in an optimum position for Abinteraction, a volume of

1.5mL of 100 mM protein G was applied onto SAM-AuNPs modified

electrode for 1 h at RT before anti mAb immobilization The

immobilized protein G helps provide a desirable orientation of

immobilized antibodies for increased sensitivity This has been

demonstrated before[20] In the preparation procedure,

ethanol-amine was also used to block the remaining non-specific binding

sites after antibody immobilization

2.6 Detection of amyloidbpeptides

Amyloid b-peptides were prepared in 0.02% (v/v) ammonia

water at 200mM concentration by brief vortexing and stored in

0.1 mL vials at80C During preparation, the vials and all the

reagents and chemicals were kept on ice Just before analysis, the

vials were left to equilibrate to RT (~24± 1C) and subsequently

diluted to the required concentration In this case, a concentration

range between 10 nM and 200mM was used A volume of 2 mL

amyloid b-peptides was dropped on sensor surface, and left for

30 min at RT to let the peptide attach to the immobilized anti mAb

Then, sensors were rinsed with 10 mM PBS followed by deionized

water and dried over a gentle stream N2gas Finally, all sensors

were subjected to electrochemical impedance spectroscopy (EIS)

measurement The impedance spectra was recorded in 0.1 M KCl

solution containing redox probe solution (5 mM of K3[Fe(CN)6]/

K4[Fe(CN)6]) within frequency range from 100 kHz to 50 mHz

around the open circuit potential with an ac probe amplitude of

10 mV

2.7 Label-free impedimetric immunosensor

In impedimetric sensors, detection is based on the principle that

any substance attached on the electrode will change the measured

impedance Therefore, any change in the impedance spectra can be

related to the change in interface properties In this case, the

binding of amyloidbpeptides with anti mAbcan be considered as a

coating film, which is expected to affect the sensor impedance

signal In the present work, Nyquist plots were used to investigate

the change in electron transfer resistance at the interface between

the sensor and the redox probe solution after the binding of

amy-loidbpeptides to the immunosensor chip, and also the change in

electron transfer resistance with changing concentration of

amy-loidbpeptides.Fig S1 in supporting informationshows the typical

Nyquist plot of faradaic impedance spectrum In general, the

complex impedance is displayed in two parts including a semicircle

(at high frequency region corresponding to the electron transfer

limited process) and a linear part (at lower frequencies resulting

from the diffusion limiting step of the electrochemical process) The

impedance result can be clarified by fitting with Randles equivalent

circuit (insetFig S1) This equivalent circuit model consists of the

Ohmic resistance of the electrolyte RS, the double layer capacitance

Cdl, and Warburg impedance ZW The Warburg resistance describes

the normal diffusion of the redox probe from bulk of solution to the

electrode surface through the complex layer The last parameter is

the electron transfer resistance RCTwhich, controls the interfacial

electron transfer rate between the redox probe in solution and the

electrode surface Ideally, Cdland RCTare both affected by modi

fi-cation occurring on the electrode surface[9,21] Thus, Cdland RCT

are parameters that are mainly used as signals in impedance sen-sors However, the value of RCTwas found to be strongly affected by modification occurring on the electrode surface [22e28] In this work, RCTparameter was therefore chosen to measure the amyloid beta concentration

Based on the principle of an EIS immunosensor, the impedance signal will be changed due to biomolecular interaction events at the electrode surface Through simple systematic surface chemistry modification of the working electrode affect the impedimetric measurements, a more sensitive and selective immunosensor can

be developed In this research, we applied carbon DEP chip with different modification strategies as our transducer for detecting amyloid beta Following our previous work[13,14], carbon DEP chip surface was modified through an in situ electrochemical synthesis using HAuCl4(sensors B and C) The obtained RCTvalue of in situ AuNPs-modified electrode was much lower than that of bare DEP chip (in the supportingFig S2) The result suggests that the charge transfer was mainly performed through AuNPs formed by elec-trolysis of HAuCl4 Concurrently, hydrochloric acid production after electrolysis of HAuCl4 could activate the oxidation of free AuNPs on the surface, under an electric potential As a whole, AuNPs and oxidized carbon surface acted as parallel resistors to reduce the total electrode impedance

3 Results and discussion 3.1 AuNPs-modified electrode surfaces Typical scanning electron microscope (SEM) images of AuNPs on carbon ink electrode are shown inFig 2 Using the PV method where the potential was applied for different time periods, AuNPs have angular shape and the particle morphology is not so well

defined When the duration of the applied potential was increased (up to 90 s), the particle morphology completely changed and the AuNPs formed Au nano-clusters or Au nano-islands For our pur-poses, AuNPs were not successfully formed using the PV method However, using the CV method, the AuNPs were formed success-fullyFig 2 Therefore, we selected the CV method to prepare our AuNPs-modified electrode for biosensor fabrication The size of the AuNPs averaged around 40e50 nm The uniform characteristic and density distribution of AuNPs significantly depended on the num-ber of cycles The SEM images showed that for both uniform characteristics and density distribution of AuNPs formation on carbon electrode, using 10 and 20 cycles was much better than using 5 cycles However, between 10 or 20 cycles, it was difficult to differentiate between the two Therefore, the investigation of anti

mAbimmobilization efficiency was implemented using both cycles For good measure, we also investigated using 5 cycles

3.2 Evaluation of the fabricated immunosensors Anti mAb was immobilized onto AuNPs-modified electrodes

Fig 3shows the impedance spectra obtained before and after anti

mAbimmobilization onto AuNPs-modified electrode via SAM layer, with and without protein G A significant difference in the impedance spectra of anti mAb immobilization electrodes compared with bare AuNPs-modified electrodes (RCT values observed after fitting the experimental spectra using Randle equivalent circuit) was observed The diameter of the semicircle is equal to the electron transfer resistance RCT, which denotes the blocking behaviour of the electrode surface against the redox probe, K3[Fe(CN)]6/K4[Fe(CN)]6 As seen from the results, the per-centage change in the RCT(%DRCT), which is obtained before and after anti mAb immobilization, is highest for 20 CVs-AuNPs-modified electrode (61%) and is lowest for 5 CVs-AuNPs-modified T.T.N Lien et al / Analytica Chimica Acta xxx (2015) 1e8

4

electrode (24%) This result shows that the anti mAbimmobilization

efficiency of 20 CVs-AuNPs-modified electrode was the most

opti-mum Therefore, the impedimetric biosensor used within this work

was fabricated from electrodes that were modified by AuNPs

deposited using CV method with 20 cycles

3.3 Performance of the impedimetric amyloid beta immunosensor

We evaluated the performance of the immunosensors

fabri-cated using methods A, B and C discussed in previous sections The

sensors were termed Sensors A, B, and C, respectively

3.3.1 Detection of amyloidb(1-40) peptide at anti mAb/pyrenyl groups-modified carbon DEP chip: Sensor A

To evaluate the performance of anti mAb/pyrenyl groups-modified carbon DEP chip, sensor A was exposed to various con-centrations of amyloidb1-40 peptide (from 1 nM to 200mM) The corresponding Nyquist plots of impedance spectra are shown inFig

4a, and thefitted impedance parameters are presented inTable S1

in supporting information SI It was observed that the Nyquist semicircle diameter (equal to the electron transfer resistance RCT) negligibly changed when the peptide concentration was increased from 1 nM to 0.1 mM However, when the concentration of the

Fig 2 Scanning electron microcopy (SEM) images of SAM gold nanoparticles on carbon ink electrode formed using (a) a potential step voltammetry (PV) method, where potential was applied for 5, 20 and 90 s (Top); and (b) a cyclic voltammetry (CV) method for 5, 10 and 20 cycles (5, 10 and 20 CV) (Bottom).

Fig 3 Electrochemical impedance spectra after immobilization of anti mAbon protein G-immobilized AuNPs-modified electrode.

T.T.N Lien et al / Analytica Chimica Acta xxx (2015) 1e8 5

peptide increased from 1mM to 200mM, the diameter of Nyquist

semicircle increased dramatically due to the binding of a significant

amount of peptide molecules to immobilized anti mAbin higher

concentration of peptide Thus, the interfacial electron transfer was

hindered significantly, resulting in a correspondingly increased

electron transfer resistance [23e29] However, the double layer

capacitance Cdl parameter which is expected to decrease when

peptide concentration increases, was found to be less sensitive to

the change of peptide concentration than RCT This is often observed

in protein detection[23e30]and cell detection[31,32]methods

We imagine that this could be because the sensitivity of capacitance

depends on obtaining the proper thickness of the original sensing

layer, while measurement of RCT only requires the presence of

redox-active species in the electrolyte[21] A calibration curve was

obtained by plotting the RCTcorrelated to the logarithm of peptide

concentration (Fig 4b) As can be seen, RCTdid not change with

increase in peptide concentration within the detected

concentra-tion from 1 nM to 100 nM, but increased quickly from 1 mM to

200mM Thus, the linear range was obtained from 1mM to 200mM

with linear equation of RCT (kU) ¼ 4.06 þ 2.17*log C (mM)

(R2¼ 0.9914), that is the correlation between RCTand Log if amyloid

beta concentration is 0.9914 The limit of detection (LoD) for

amyloid b peptide (1-40) of this sensor was determined to be 2.04mM with the sensor area of 2.64 mm2 The LoD was obtained based on the standard deviation (STDEV) of the blank sample and slope of calibration curve as:

LOD¼3xSTDEV

The coefficient of variation, which is defined as the ratio of the STDEV to the mean, changed between 1% and 8% for three repetitive measurements, indicating good reproducibility As far as we are aware, this is one of the few studies on fabrication and detection of amyloidbpeptide impedimetric immunosensor, but the detection limit was higher than the reported sensors[15,24]

3.3.2 Detection of amyloidb(1-40) peptide at anti mAb/SAM/ AuNPs-modified carbon DEP chip: Sensor B

It has been reported that the immobilization of antibody mol-ecules is a decisive factor for successful fabrication of immuno-sensor [13,14] The immobilization method must maintain the activity and maintain the stability of biomolecules, and must be controllable over the distribution and orientation of the immobi-lized species In the method A, the DEP chip carbon ink electrode was modified first by a functional molecule, 1-pyrenebutanoic acid succinimidyl ester and thence anti mAbimmobilization via succi-nimidyl ester groups The 1-pyrenebutanoic acid succisucci-nimidyl ester was dissolved in solution containing 70% of DMSO and 30% of deionized water with 100 mM concentration In our previous work

[13], we showed that a large amount of DMSO might influence the carbon ink electrode surface due to its ability to dissolve adhesives

in the carbon ink Furthermore, the excess DMSO on the electrode surface can interact with amino group of antibody, which decreases the anti mAb immobilization efficiency Therefore, the obtained LoD of this sensor (Sensor A) was relatively high In order to overcome such deficiencies, there are some protocols that are typically utilized They include physical absorption, chemical cross-linking and entrapment Among them, biomolecule immobilization methods through cross-linking can improve the stability of the biorecognition component[25e29] In the case of Sensor B, the DEP chip carbon electrode was modified first by deposition of AuNPs on working electrode using cyclic voltammetry over 20 cycles After, anti mAb was immobilized on to AuNPs-modified electrodes via COOH group of self-assembled monolayer of MHDA, which served as a linker for covalent biomolecular immo-bilization.Fig 5a illustrates the Nyquist plots of impedance spectra obtained upon a gradual increase of amyloidb1-40 peptide con-centration (from 1 nM to 200mM) and thefitted impedance pa-rameters are presented inTable S2in SI As can be clearly observed, the RCTincreases with increase in amyloidbpeptide concentration This result indicates that an insulating layer was formed on the electrode surface due to the binding of peptide molecules to immobilized anti mAb receptor This insulating layer acts by blocking the electron transfer between the redox probe and the electrode surface The calibration curve was obtained by plotting the RCT value against the logarithm of peptide concentration (Fig 5b) which shows a linear region from 1 nM to 1mM The RCT increased slowly with increasing peptide concentration from 1 nM

to 1mM, and a significant increase was observed between 1mM and

200mM peptide concentration

We suggest the reason why the calibration curve exhibits two regions Thefirst is a linear region from 1 nM to 1mM and the second starts at a concentration of 1mM and above We imagine that this is due to concentration of the immobilized antibody As discussed above, the RCTvalue denotes the blocking behaviour of electrode surface for redox probe The phenomenon of blocking an

Fig 4 (a) Impedance spectra of anti mAb/pyrenyl groups-modified electrodes exposed

to difference concentration of Amyloidb1-40 peptide and (b) the calibration curve of

R CT against Log amyloidb1-40 antigen concentration The impedance results were

obtained in solution containing 0.1 M KCl and 5 mM K 3 [Fe(CN)] 6 /K 4 [Fe(CN)] 6 at OCP

and frequency range was from 100 kHz to 50 mHz with an ac probe amplitude of

10 mV All data points are mean response values of three independent electrodes The

error bars (calculated as standard deviation) provide a measure of the reproducibility

of the system.

T.T.N Lien et al / Analytica Chimica Acta xxx (2015) 1e8 6

electrode surface is due to occupation of the surface by peptide

molecules when they bind with antibody molecules on the

elec-trode surface Therefore, the concentration of the antibody on the

surface plays important role If the concentration of antibody is

relatively high when the amount of antigen is relatively low, there

is a small amount of antibody enough to react with antigen Thus

the competitive reaction between the antigen and antibody leads

to the change of RCTthat is not high Meanwhile, at high

con-centration of the peptide, because of large amount of antibody to

react with antigen, the competition of peptide to occupy the space

leads to the dramatically increase of RCT In our study, a high

concentration (100 mg/mL) of immobilized antibody was used

This phenomenon was not observed when relatively low

con-centrations of immobilized antibody was used [19,24,27,29,30]

The advantage of our sensor is that it could work at both low and

high concentration of antigen The obtained linear equation from

1 nM to 1mM for this sensor is RCT(kU)¼ 1.56 þ 0.20*log C (mM)

(R2¼ 0.9662) Based on equation(1), the LoD of the sensor B was

determined to be 2.65 nM The LoD of Sensor B is much lower than

Sensor A In other words, the anti mAbimmobilization efficiency

of sensor B was significantly increased due to electrode surface

modification This dramatically increased the specific area of electrode surface which permitted a larger a mount of anti mAb

immobilization Sensor B also has a good reproducibility with a coefficient of coefficient of variation between 2.5% and 5.3% for three repetitive measurements

3.3.3 Detection of amyloidb(1-42) peptide at anti mAb/protein G/ SAM/AuNPs-modified carbon DEP chip: Sensor C

Protein G is an antibody binding protein, which specifically binds to the Fc region of an antibody In other words, protein G only binds antibodies through the non-antigenic regions, leaving the antigen binding sites of antibodies available to bind to their target antigen Thus, it has been widely used to immobilize anti-bodies in immunoassays offering important advantages such as controllable immobilization of the antibodies, resulting in high sensitivity and low detection limit[30,33,34] In Sensor C, AuNPs-thiolated protein G was used to immobilize anti mAb.Fig 6 illus-trates the Nyquist plots of impedance spectra obtained upon in-crease in amyloid b1-42 peptide concentration from 10 pM to

200mM, andfitted impedance parameters (Please seeTable S3in SI) Similar to Sensor B, it can be seen that the semicircle diameter

in the Nyquist plots increased with increasing concentration of amyloidb1-42, and there are also two regions in the calibration curve Notably, the value of RCTincreased rapidly even at very low peptide concentration (10 pMe100 nM), indicating that using AuNPs-thiolated protein G might offer much more large specific area to immobilize a larger amount of anti mAb than sensor B Based on equation of RCT (kU) ¼ 2.82 þ 0.11*log C (mM) (R2¼ 0.9969), the LoD for amyloidbpeptide 1-42 of this sensor was determined to be 0.57 nM (using equation(1)) Sensitivity of the developed sensor could be further improved by modifying the carbon electrode with graphene, for example, in order to increase conductivity of the electrode

Human serum albumin is one of the most abundant proteins in biologicalfluids and can interfere with detection of low-abundance biomarkers[35] Using bovine serum albumin (BSA) we evaluated the effect that different concentrations of BSA, co-incubated with amyloidbpeptide, had on the sensitivity of the immuno-sensor The results show that BSA 2.5 mg/mL did not have any discern-able effect on the sensor performance

4 Conclusion

In this work, we present a novel approach for modifying screen-printed carbon ink electrode for development of a highly sensitive labelless impedimetric immunosensor for amyloidbpeptide Am-yloid beta is an important potential biomarker of Alzheimer's dis-ease It is produced from a transmembrane protein, amyloid precursor protein Above a certain concentration, it starts to self-assemble (aggregate) into oligomers and then into either amor-phous aggregates or maturefibrils[36] The degree of amyloid beta bioactivity is dependent on its aggregation species [37,38] In particular, it is now almost widely accepted that the most potent neurotic species are the oligomers [39] Veloso and colleagues developed a species-specific EIS immunosensor for monitoring the effect of fibril inhibitor/dissociator using a species-specific antibodies[40] We have developed a relatively sensitive quanti-tative EIS immunosensor for amyloid beta Three types of amyloidb

impedimetric immunosensors were fabricated in a step-wise manner in order to understand the effects that each surface modification chemistry had on detection sensitivity We found that (i) immobilization of AuNPs, to improve stability of the recognition element and also increase the surface area for immobilization, lowered the LOD by both ~ three orders of magnitude (from 2.04 mM to e2.65 nM) A further modification using protein G

Fig 5 (a) Impedance spectra of anti mAb/SAM/AuNPs-modified electrodes exposed to

difference concentrations of amyloidb1-40 peptide and (b) the calibration curve of R CT

as against Log amyloidb1-40 concentration All data points are mean values of three

independent electrodes The error bars (calculated as standard deviation) provide a

measure of the reproducibility of the system.

T.T.N Lien et al / Analytica Chimica Acta xxx (2015) 1e8 7

towards a desired orientation of the antibody lowered the LOD

further to 0.57 nM The sensor also demonstrated reduced cost

(cost of carbon ink printed electrode but can use as gold ink printed

electrode), detection platform simplicity, and high reproducibility

Besides, based on current study, it was found that EIS is an

impressive method for monitoring the interaction of antigen with

antibody that occurred on the electrode surface These methods

could easily be applied for detection of other antigens, with

selec-tion of the detecselec-tion platform based on the desired for sensitivity

We are now planning on developing a biosensor for quantitative

detection of oligomeric amyloid beta species

Acknowledgements

Dr Truong TN Lien gratefully acknowledges receipt of a grant

from the Japan Society for the Promotion of Science (JSPS) This

work was supported by a Grant-in-Aid for Scientific Research C

from JSPS We would like to thank Dr Nguyen Xuan Viet for SEM

imagines

Appendix A Supplementary information Supplementary information related to this article can be found

athttp://dx.doi.org/10.1016/j.aca.2015.08.036

References [1] C.P Ferri, M Prince, C Brayne, H Brodaty, L Fratiglioni, M Ganguli, K Hall,

K Hasegawa, H Hendrie, Y Huang, A Jorm, C Mathers, P.R Menezes,

E Rimmer, M Scazufca, Lancet 366 (2005) 2112e2117 [2] M Vestergaaard, K Kerman, E Tamiya, Nanobiotechnology 2 (2006) 5e16 [3] K Blennow, J Intern Med 256 (2004) 224e234

[4] S.D Schimidt, R.A Nixon, P.M Mathews, Methods Mol Biol 299 (2005) 279e297 [5] T Takahashi, H Mihara, Chem Commun 48 (2012) 1568e1570

[6] I.H Chou, M Benford, H.T Beier, G.L Cote, M Wang, N Jing, J Kameoka, T.A Good, Nano Lett 8 (2008) 1729e1735

[7] M Vestergaard, K Kerman, M Saito, N Nagatani, Y Takamura, E Tamiya,

J Am Chem Soc 127 (2005) 11892e11893 [8] S Prabhulkar, R Piatyszek, J.R Cirrito, Z.Z Wu, C.Z Li, J Neurochem 122 (2012) 374e381

[9] B Pejcic, R.D Marco, Electrochim Acta 51 (2006) 6217e6229 [10] A Ramanavicius, A Finkelsteinas, H Cesiullis, A Ramanavicience, Bio-electrochemistry 79 (2010) 11e16

[11] C Esseghaier, S Helali, H.B Fredj, A Tilili, A Abdelghani, Sens Actuators C Chem 131 (2008) 584e589

[12] H Fayazfar, A Afshar, M Dolati, A Dolati, Anal Chim Acta 836 (2014) 34e44

[13] T.T.N Lien, N.X.V Viet, M Chikae, Y Ukita, Y Takamura, J Biosens Bio-electron 2 (2011) 107

[14] T.T.N Lien, M Chikae, Y Ukita, Y Takamura, Talanta 85 (2011) 2576e2580 [15] J.V Rushworth, A Ahmed, H.H Griffiths, N.M Pollock, N.M Hooper, P.A Millner, Biosens Bioelectron 56 (2014) 83e90

[16] V Dounin, A.J Veloso, H Schulze, T.T Bachmann, K Kerman, Anal Chim Acta

669 (2010) 63e67 [17] E Katz, J Electroanal Chem 365 (1994) 157e164 [18] H Jaegfeldt, T Kuwana, G Johansson, J Am Chem Soc 105 (1983) 1805e1814

[19] R Elshafey, C Tlili, A Abulrob, A.C Tavares, M Zourob, Biosens Bioelectron.

39 (2013) 220e225 [20] M Vestergaard, D.-K Kim, K Kerman, M.H Ha, E Tamiya, Talanta 74 (2008) 1030e1042

[21] I.I Suni, Trends Anal Chem 27 (2008) 604e611 [22] G Tsekenis, G.Z Garifallou, F Davis, P.A Millner, T.D Gibson, et al., Anal Chem 80 (2008) 2058e2062

[23] Y Huang, M.C Bell, I.I Suni, Anal Chem 80 (2008) 9157e9161 [24] E.C Rama, M.B Gonzalez-García, A Costa-García, Sens Actuators B 201 (2014) 567e571

[25] T.N.T Lien, T.T Nguyen, A.L.T Luu, Y Ukita, Y Takamura, in: Proceeding of 16th International Conference on Miniaturized Systems for Chemistry and Life Sciences, Okinawa, Japan, 2012, pp 1912e1914

[26] R Elshafey, C Tlili, A Abulorb, A.C Tavares, M Zourob, Biosens Bioelectron.

39 (2013) 220e225 [27] E Asav, M.K Sezginturk, Int J Biol Macromol 66 (2014) 273e280 [28] E Arkan, R Saber, Z Karimi, A Mostafaie, M Shamsipur, J Pharm Biomed Anal 92 (2014) 74e81

[29] S.K Arya, S Bhansali, Biosens J 1 (2012) Article ID H110601 [30] I Hafaiedh, H Chammem, A Abdelghani, E Ait, L Feldman, O Meilhac,

L Mora, Talanta 116 (2013) 84e90 [31] S Bayoudh, A Othmane, L Ponsonnet, H.B Ouada, Coloids Surf A 318 (2008) 291e300

[32] J Paredes, S Becerro, F Arizti, A Aguinag, J.L Del Pozo, S Arana, Sens Actu-ators B 178 (2013) 663e670

[33] J.M Fowler, M.C Stuart, D.K Wong, Anal Chem 79 (2007) 350e354 [34] P Xiong, N Gan, Y Cao, F Hu, T Li, L Zheng, Materials 5 (2012) 2757e2772 [35] M Vestergaard, E Tamiya, Anal Sci 23 (2007) 1443e1446

[36] C Haass, D Selkoe, Nat Rev Mol Cell Biol 8 (2007) 101e112 [37] M Morita, T Hamada, Y Tendo, T Hata, M.C Vestergaard, M Takagi, Soft Matter 8 (2012) 2816e2819

[38] M.C Vestergaard, M Morita, T Hamada, M Takagi Biochim, Biophys Acta Biomembr (2013) 1314e1321

[39] Y Gong, L Chang, K.L Viola, P.N Lacor, M.P Lambert, C.E Finch, G.A Krafft, W.L Klein, Proc Nat Acad Sci U S A 100 (2003) 10417e10422

[40] A.J Veloso, A.M Chow, H.V Ganesh, N Li, D Dhar, D.C Wu, S Mikhaylichenko, I.R Brown, K Kerman, Anal Chem 86 (2014) 4901e4909

Fig 6 (a) Impedance spectra of anti mAb/Protein G-SAM/AuNPs-modified electrodes

exposed to difference concentration of amyloidb1-42 peptide and (b) the calibration

curve of R CT as against Log amyloidb1-42 concentration All data points are mean

values of three independent electrodes The error bars (calculated as standard

devia-tion) provide a measure of the reproducibility of the system.

T.T.N Lien et al / Analytica Chimica Acta xxx (2015) 1e8 8