Biosensors and Bioelectronics 82 (2016) 177–184

Contents lists available at ScienceDirect

Biosensors and Bioelectronics

http://d0956-56

n CorrE-m1 Eq

journal homepage: www.elsevier.com/locate/bios

Highly sensitive enzyme-free immunosorbent assay for porcinecircovirus type 2 antibody using Au-Pt/SiO2 nanocomposites as labels

Long Wu1, Wenmin Yin 1, Kun Tang, Kang Shao, Qin Li, Pan Wang, Yunpeng Zuo,Xiaomin Lei, Zhicheng Lu, Heyou Han n

State Key Laboratory of Agricultural Microbiology, College of Food Science and Technology, College of Science, Huazhong Agricultural University, Wuhan430070, PR China

a r t i c l e i n f o

Article history:Received 20 January 2016Received in revised form18 March 2016Accepted 2 April 2016Available online 6 April 2016

Keywords:Enzyme-free immunosorbent assayAu-Pt/SiO2 nanocompositesFe3O4 magnetic beadsHuman IgGPorcine circovirus type 2

x.doi.org/10.1016/j.bios.2016.04.00163/& 2016 Elsevier B.V. All rights reserved.

esponding author.ail address: hyhan@mail.hzau.edu.cn (H. Han)ual contribution.

a b s t r a c t

Improving the performance of conventional enzyme-linked immunosorbent assay (ELISA) is of greatimportance to meet the demand of early clinical diagnosis of various diseases. Herein, we report a fea-sible enzyme-free immunosorbent assay (EFISA) system using antibody conjugated Au-Pt/SiO2 nano-composites (APS NCs) as labels. In this system, Au-Pt/SiO2 nanospheres (APS NPs) were first synthesizedby wet chemical method and exhibited intrinsic peroxidase and catalase-like activity with excellentwater-solubility. Then APS NCs were utilized as labels to replace HRP conjugated antibody, and Fe3O4

magnetic beads (MBs) to entrap the analyte. To discuss the performance of EFISA system, Human IgG wasserved as a model analyte, and porcine circovirus type 2 (PCV2) serums as real samples. The systemboosted the detection limit of HIgG to 75 pg mL�1 with a RSD below 5%, a 264-fold improvement ascompared with conventional ELISA. This is the first time that APS NCs have been used and successfullyoptimized for the sensitive dilution detection of PCV2 antibody (5:107) in ELISA. Besides, APS NCs haveadvantages related to low cost, easy preparation, good stability and tunable catalytic activity, which makethem a potent enzyme mimetic candidate and may find potential applications in bioassays and clinicaldiagnostics.

& 2016 Elsevier B.V. All rights reserved.

1. Introduction

Enzyme linked immunosorbent assay (ELISA), a widely ac-cepted and powerful immunoassay method, has received commoninterests due to its simplicity, good specificity, easy operation andinstrumentation (Jayasena et al., 2015; Liang et al., 2015). Since itwas firstly constructed and applied for immunoglobulin assay in1971 (Engvall and Perlman, 1971), ELISA has become the goldstandard for experimental and clinical analysis, especially in theanimal disease diagnosis. Porcine circovirus type 2 (PCV2) is acommon virus known to infect mammals and has caused en-ormous economic losses in the swine industry worldwide since itwas firstly discovered in 1998 (Wu et al., 2015; Opriessnig et al.,2007). Thus, it is critical to construct an accurate and sensitivemethod to implement the early diagnosis for PCV2 antibody. Theconventional ELISA with HRP enzyme conjugated secondary anti-body and coloring substrate solution (TMB) plays an importantrole in immunoassay (Frey et al., 2000; Ambrosi et al., 2010), but it

.

is still a challenge to find new approaches that could improve thesimplicity, selectivity, and sensitivity (Laube et al., 2011; Tang et al.,2010). Therefore, it is of utmost importance to establish a methodwith high sensitivity and low cost for the detection of PCV2antibody.

To enhance the conventional ELISA, great efforts have beenmade in boosting the stability, detection limit and range (Gaoet al., 2015; Qu et al., 2014; Diaz-Amigo and Popping, 2013).Herein, the introduction of nanoparticles (NPs) has largely im-proved the performance of conventional ELISA (Jia et al., 2009),especially the sensitivity (Chen et al., 2014a). For example, a greatmany of NPs have been adopted as carriers for the recognitionantibody and/or HRP to obtain signal amplification owing to thestrong adsorption ability and high surface areas (Pérez-López andMerkoçi, 2011; Chen et al., 2014b). Besides, some NPs are used toserve as activatable fluorescence probes and chromogenic sub-strates (Liu et al., 2013; de la Rica and Stevens, 2012). In addition,some peroxidase- or oxidase-like NPs such as FeS nano-sheets andgraphene oxide, known as mimetic peroxidase, can take place ofHRP in conventional ELISA with further modification (Dai et al.,2009; Song et al., 2010). By taking advantages of good enzyme-likeactivity, low cost and high stability (Gao et al., 2007; Wu et al.,2014), the NPs family is expected to be a promising candidate as

L. Wu et al. / Biosensors and Bioelectronics 82 (2016) 177–184178

mimetic enzymes.Pt NPs, especially hybridized with other metals, have gained

extensive attention in electrochemical catalysis (Zuo et al., 2015),catalytic hydrogenation (Bratlie et al., 2007) and air purification(Zhou et al., 2005). Theses Pt NPs possess excellent oxidase-like(Bai et al., 2011), peroxidase-like (Cai et al., 2013) and catalase-likeactivity. To further improve the performance of Pt NPs, the designof supported Pt has been proposed as high-efficiency catalysts. Forexample, Pt NPs supported on grapheme (Xin et al., 2011), carbonnanotubes (Zhao et al., 2007), carbon (Song et al., 2007) and TiO2

(Drew et al., 2005) have been widely employed for electrocatalyticoxidation of methanol. However, it is inconvenient to furthermodify the Pt NPs and apply them to biological analysis sincethere is no extra functional group on their surface except the high-density surfactant such as cetyltrimethylammonium bromide(CTAB). Taking the convenient surface modification into account,SiO2 nanospheres are ideal carries in the construction of bio-sensors coupled with other techniques and widely used in im-munoassay (Mao et al., 2012; Shao et al., 2014; Huang et al., 2014).

In this work, we proposed a facile enzyme-free immunosorbentassay (EFISA) system using APS NCs as labels. A simple wet che-mical method was first explored to prepare SiO2 loaded Au-Pt NPsbased on the previous work (Guo et al., 2008) with some changes(see Scheme 1a). Afterwards, peroxidase-like activity of APS NPswas investigated and the HRP conjugated secondary antibody(Ab2) was replaced by APS NCs (see Scheme 1b). Even without theaddition of any H2O2, APS NCs could exhibit good catalytic prop-erties to TMB in the presence of oxygen and Hþ . Moreover, the APSNCs could provide both the mimetic peroxidase (Au-Pt NPs) andbinding sites (SiO2 nanospheres), suggesting their potentials infuture ELISA application. What is more, Fe3O4 magnetic beads(MBs) were fixed on the bottom of the well plates by simply ap-plying an external magnetic field to entrap antigens or Ab1 forconvenient separation (Gao et al., 2013) (see Scheme 1b). As aproof of concept, Human IgG was detected as a model analyte, andporcine circovirus type 2 (PCV2) serums as real samples in theEFISA system. This method behaved easier operation, lower costand higher sensitivity compared with previous work (Wu et al.,

Scheme 1. Schematic illustration of the enzyme-free immunosorbent assay system, (a)antibody using APS NCs as labels.

2015).

2. Material and methods

2.1. Chemicals and materials

Human IgG antigen (HIgG, 50 mg mL�1), rabbit anti-human IgGantibody (HAb1, 1 mg mL�1), goat anti-human IgG antibody(HAb2, 2 mg mL�1), rabbit anti-pig IgG antibody (PAb2,1 mg mL�1) were obtained from Sangon Biotech Co., Ltd. (Shang-hai, China); PCV2 antibody (PAb1), positive and negative serum ofPCV packaged in the PPA-ELISA kits were from Shandong Lvdu Bio-science & Technology Co. Ltd; Recombinant PCV2 Cap 2 antigenwas from Puhuashi Sci-Tech Development Co., Ltd. (Beijing, China).Human IgG ELISA kit was obtained from Wuhan Booute Bio-technology Co., Ltd. The porcine pseudorabies virus (PrV), porcinereproductive and respiratory syndrome virus (PRRSV) positive andnegative serum samples (ELISA kit) were obtained from WuhanKeqian Animal Biological Products Co.Ltd. (China).

Tetraethyl orthosilicate (TEOS, 99%), 3-aminopropyl tri-methoxysilane (APTMS, 97%), bovine serum albumin (BSA), glu-taraldehyde solution (GA, 25%) were purchased from Sigma-Al-drich; Potassium hexachloroplatinate (K2PtCl6, AR), tetra-chloroauric(III) acid hydrate (HAuCl4 �4H2O, AR), cetyl-trimethylammonium bromide (CTAB), poly(sodium-p-styr-enesulfonate) (PSS, MW¼70 K), polyvinyl pyrrolidone (PVP,MW¼58 K), o-Phenylenediamine (OPD), sodium citrate (SCT) andother relevant reagents were obtained from Sinopharm ChemicalReagent Co. Ltd.; Fe3O4 magnetic beads with amino group(MBs-NH2, 0.5%, w/w) were from Tianjin Baseline Chromtech Re-search Centre. All chemicals and solvents were of analytical gradeand used as received without further purification. Ultrapure waterobtained from a Millipore water purification system (Milli-Q,Millipore, 18.2 MΩ resistivity) was used throughout theexperiment.

the preparation procedures of APS NPs, and (b) the immunoassay process for PCV2

L. Wu et al. / Biosensors and Bioelectronics 82 (2016) 177–184 179

2.2. Instrumentation

Fluorescence and optical density (OD) measurements wereoperated on Perkin Elmer 1420 Multilabel Counter; The UV–visabsorption spectra were from Nicolet Evolution 300 UV–vis spec-trometer (Thermo Nicolet, America); Fouriertransform infrared(FT-IR) spectra were acquired on a Nicolet Avatar-330 spectro-meter with 4 cm�1 resolution using the KBr pellet technique;Transmission electron microscopy (TEM) images were collected bya JEM-2010 transmission electron microscope (JEOL, Japan); Hy-drodynamic diameters were measured by using dynamic lightscattering (DLS) technique on Malvern Zetasizer Nanoseries(Malvern, England).

2.3. Preparation of APS NPs and APS NCs labels

The preparation procedures of APS NPs were illustrated inScheme 1a. Firstly, Au NPs, SiO2 nanospheres and Au–SiO2 hybridswere synthesized according to the previous work with somemodification (Wu et al., 2014; Guo et al., 2008). Then, APS NPswere simply prepared by reducing H2PtCl6 on the surface ofAu–SiO2 hybrids under room temperature. To obtain APS NCs la-bels for the specific identification of antigen, Ab2 was conjugatedto APS NPs via electrostatic interaction and Schiff base structures.The details were all described in Supplementary information.

2.4. Kinetic analysis

To further understand the performance of APS NPs, the cata-lytic oxidation reaction kinetics for the two most common coloringsubstrate (TMB and OPD) were studied by absorption spectra.Unless otherwise stated, the kinetic experiments were carried outat room temperature with APS NPs (1.8 mg mL�1) in 1.0 mL of PBSbuffer (pH¼5.5) with different concentrations of TMB or OPD. Thekinetic parameters were calculated on the basis of Michaelis–Menten equation v¼Vmax� [S]/(Kmþ[S]), where v is the initialvelocity, Vmax is the maximal reaction velocity, Km is the Michaelisconstant and [S] is the concentration of substrate (He et al., 2011).

2.5. Construction of EFISA system

The goal of this work is to construct an enzyme-free im-munosorbent assay (EFISA) system for the highly sensitive detec-tion of PCV2 antibody. Scheme 1b showed the Ag–Ab1–Ab2 in-direct immunoassay model of EFISA system. Three main modifiedparts are described in the scheme: (i) APS NCs took the place of theHRP-conjugated Ab2 and acted as detection labels; (ii) Ag-con-jugated Fe3O4 MBs were fixed on the bottom of plates through anexternal magnetic field; (iii) TMB was used in the color reactionstep in the absence of H2O2. The detailed immunoassay for HIgGand PCV2 antibody was consistent with our previous work (Wuet al., 2015) in Supplementary information.

3. Results and discussion

3.1. Characterization of the APS NPs and APS NCs

FT-IR spectroscopy was first used to investigate the functiona-lization process of APS NPs. As shown in Fig. S1A (Supplementaryinformation), the wide and strong absorption band at 1100 cm–1

(Si-O-Si) was corresponding to the characteristic bands of silicananospheres. Meanwhile, the absorption bands of amino-functio-nalized silica at 1640, 960, 800, and 560 cm–1 (N–H) confirmed thesuccessful modification of silica nanospheres, Fe3O4 MBs, and APSNPs. As depicted in Fig. S1B, a typical UV–vis absorption spectrum

of Au NPs was exhibited with a sharp peak located at 508 nm(curve a), which was consistent with the local plasmon resonanceof Au NPs (�5 nm). However, the peak disappeared after the de-position of Pt on the surface of Au–SiO2 hybrids (curve b). For APSNCs, two weak absorption peaks (curve c) were observed, whichcan be ascribed to the amide bond (220 nm) and the tryptophanand tyrosine residues present in the protein (280 nm) (Bhainsaand D’Souza, 2006). Fig. S1C described the zeta potential of dif-ferent modification steps in the synthesis of APS NCs. It is note-worthy that the zeta potential of APS NPs increased dramaticallyfrom �38.67 to �5.57 mV after combining with Ab2, which re-vealed that Ab2 were successfully conjugated with APS NPs. Toprove the presence of Au-Pt, Fig. S1D showed the energy-dis-persive X-ray spectroscopy (EDX) spectrum of the APS NPs. Twomain peaks (Au and Pt) were observed (other peaks originatedfrom the carbon-supported copper grid), indicating that the APSNPs were made up of metallic gold, platinum, and silica.

TEM images of silica nanospheres, Au�SiO2 hybrids, and APSNCs were taken as shown in Fig. 1. Fig. 1A indicated that thesynthesized silica nanospheres are well-distributed and uniformwith an average diameter of approximately 90 nm. The re-presentative TEM images (Fig. 1B) showed numerous and in-dividual particles are dotted on SiO2 nanospheres, indicating thehomogeneous distribution of Au NPs on the SiO2 nanospheres.From Fig. 1C and E, it could be clearly observed that each silicananosphere is composed of small Au/Pt hybrid nanoparticles witha rough surface. These small Au/Pt hybrid nanoparticles (o5 nm)are spread out on the surface of silica nanospheres with somevacancy left, which provides extra binding sites for Ab2. In addi-tion, a thin layer of Ab2 conjugated shell could be recognized inFig. 1D. Furthermore, the features could be verified through themagnified image as shown in Fig. 1F. More detailed characteristicswere analyzed by XPS spectrum (Fig. S2), HRTEM image (Fig. S3),SEM image (Fig. S4) and elemental analysis (Table S1) in Supple-mentary information.

3.2. Characterization of oxidase- and peroxidase-like activity of APSNPs

To discuss the catalytic activity, TMB (Asati et al., 2009) andOPD (Gao et al., 2007) as substrate, were chosen to demonstratethe oxidase-/peroxidase-like activities of APS NPs in the absenceand presence of H2O2, respectively. We found that APS NPs cancatalyze the oxidation of TMB or OPD by dissolved oxygen in waterin the acid environment, generating the typical blue color for TMBand yellow color for OPD within 10 min (Fig. S5B). Meanwhile, nocolor change was observed in the solutions without APS NPs. Theresults verified our initial presumption that APS NPs possess oxi-dase-like activity. Next, the effect of H2O2 on TMB and OPD oxi-dation was discussed. With the addition of H2O2, the solution withTMB and OPD exhibited much faster and deeper color change atthe same conditions (Fig. S5A). In contrast, the solutions withoutAPS NPs showed negligible color change. The results revealed theperoxidase-like activities of APS NPs toward peroxidase substrates.It is well known that Pt NPs behaved good catalytic performancefor the oxygen reduction reaction (Beyhan et al., 2015) and hy-drogen peroxide reduction (Katsounaros et al., 2012) in electro-catalytic processes. The oxidation pathway of oxygen or hydrogenperoxide to TMB and OPD was believed to be similar with that ofthe electrochemical reductions (He et al., 2011). In this work, westudied the electrochemical performance of APS NPs as shown inFig. S6. Compared with the air-saturated solution (curve c), noreduction peak of oxygen was observed in the N2-purged solution(curve b). Also, the effects of oxygen and light on the oxidation ofTMB were explored (Fig. S7). The results indicated that the dis-solved oxygen is the electron acceptors for the oxidation in the

Fig. 1. TEM images of (A) silica nanospheres, (B) Au–SiO2 hybrids, (C) APS NPs, (D) APS NCs, (E) a magnified version of APS NPs, and (F) APS NCs.

L. Wu et al. / Biosensors and Bioelectronics 82 (2016) 177–184180

absence of H2O2, which is in agreement with the previous work(He et al., 2011).

3.3. Catalytic activity dependence on surface structure andmodification

Three main factors, such as the size and distribution of Au-Ptdots (Fig. S8), the density of Pt on SiO2 nanospheres (Figs. S9,S10A) and the surface capping molecules (Fig. S11) were in-vestigated on catalytic activity. Also, we explored the stability ofAPS NCs (Fig. S10B) and analyzed the elemental composition(Table S1). The related details were discussed in Supplementaryinformation.

3.4. Stability and catalytic activity of APS NPs against temperature,pH and time

The ability of nanoparticles against aggregation in differentharsh conditions is vital to promote their application. Herein, thethree main factors, temperature, pH and time (Fig. S12), werediscussed on the stability of APS NPs and APS NCs. Besides, thecatalytic stability of APS NPs was characterized by UV–vis ab-sorption spectra (Fig. S13). All the details were described in Sup-plementary Information.

3.5. The optimization of H2O2 and TMB concentrations and pH onEFISA system

The effects of H2O2 and TMB were first discussed before theEFISA immunoassay. As depicted in Fig. 2A, the initial reaction rateincreased with the increasing of H2O2 concentration and showed alinear relationship in range of 0 to 20 mM (see the red line). In-creasing H2O2 concentration further, the increase in reaction rateslowly dropped down. The whole developing trend is similar to

that of Au@Pt peroxidase mimetics (He et al., 2011). Consideringthe coloring time (�5 min) during the ELISA tests, we finally chose10 mM of H2O2 to react with the substrate solution. Meanwhile,we also discussed the effect of TMB concentration on the reactioncatalyzed by APS NCs in the presence of H2O2. Fig. 2B showed theabsorbance evolution over time at several TMB concentrations,indicating that the reaction rates gradually increased with TMBconcentration. It can also be observed from Fig. 2C that the reac-tion rate reached the maximum at a high TMB concentration(0.6 mM). Thus, the substrate solution containing 0.6 mM of TMBwas used in the following assays. Fig. 2D displayed the reactionrate in the absence of H2O2 with a slower increment over thewhole concentration range compared with Fig. 2C. The resultsrevealed that the feasibility of EFISA tests in the absence of H2O2 ata higher TMB concentration. Moreover, the effect of pH on therelative reaction rate of TMB oxidation was investigated (Fig. S14).Considering the stability of Ab2, the optimal pH of PBS buffer so-lution was chosen as 5.5 though the APS NPs behave higher ac-tivity at a lower pH value. Finally, a comparison between APS NCsand HRP was made by calculating the apparent kinetic parametersfor TMB oxidation in the presence and absence of H2O2 (Fig. S15).The results were discussed in Supplementary Materials and com-pared with other reported materials (Table S2).

3.6. Immunoassay for HIgG with EFISA system

Under the above optimized conditions, we conducted the im-munoassay for HIgG with Fe3O4 MBs based ELISA system (Fig. 3A)and EFISA system (Fig. 3B), respectively. Both systems exhibitedgood linearity between optical signal and the logarithmic value ofthe tested ranges of HIgG concentration. The linear regressionequation in Fig. 3A was presented as Y1¼�0.2266þ0.2098 lg X(X: pg mL�1) over the range from 100 to 30,000 pg mL�1 withR2¼0.9882, and the detection limit was 100 pg mL�1 (S/N¼3).

Fig. 2. (A) Effect of H2O2 concentration on the reaction rate of TMB oxidation catalyzed by APS NCs. The straight line is a linear regression of relative reaction rate upon H2O2

(from 0–20 mM). The inset is the evolution of A650 value over time at different H2O2 concentrations. (B) The time-dependent absorbance (A650) of different TMB con-centrations with H2O2. (C) Effect of TMB concentration on the reaction rate catalyzed by APS NCs with H2O2 and (D) without H2O2. All the error bars were calculated based onthe standard deviation of three measurements. (For interpretation of the references to color in this figure, the reader is referred to the web version of this article.)

L. Wu et al. / Biosensors and Bioelectronics 82 (2016) 177–184 181

Fig. 3B exhibited a regression equation of Y2¼�0.4112þ0.2893 lgX (X: pg mL�1) over the range from 100 to 30,000 pg mL�1 withR2¼0.9974, and showed a detection limit of 75 pg mL�1 (S/N¼3).Obviously, both the two systems possessed a higher sensitivitythan that of the amplified ELISA system (3.9 ng mL�1) previouslyreported (Wu et al., 2015). Besides, the EFISA system performed264-fold enhancements in sensitivity than that (19.8 ng mL�1) ofthe conventional ELISA system (Wu et al., 2015). This level ofsensitivity was comparable to that of other methods in previouslypublished reports for HIgG determination (Table 1).

Moreover, the reproducibility of the EFISA system was testedwith intra-assay and inter-assay by determining one HIgG level(20 ng mL�1) for three measurements and the variation coeffi-cients were under 5%, indicating the acceptable reproducibility ofthe proposed EFISA system. To further investigate the feasibilityand specificity of the EFISA system to real samples, HIgG in humanserum (Fig. 3C) and the interference such as BSA, hemoglobulin, L-cysteine and glucose (Fig. 3D) were investigated and comparedwith commercial kit. Fig. 3C behaved the results of HIgG in fivehuman serum samples after 106-fold dilution. The EFISA systemfollowed the same tendency with commercial kit but exhibitedhigher signal to HIgG. The facts reflected the acceptable feasibilityof EFISA system in real samples. Besides, the recovery and RSD ofEFISA system are in the range of 89.5–110.8% and 3.6–4.8%, re-spectively (Table S3), which was satisfactory for quantitative as-says in biological samples. It can be seen from Fig. 3D that both the

two systems behaved unobvious difference (RSD o5%) by in-troducing the four interferences but EFISA system showed higherresponse to HIgG, suggesting the good specificity of the EFISAsystem. Furthermore, it was found that the EFISA system retainednearly the same original optical response (OD450) even after6 weeks of storage at 4 °C (Fig. S10B), revealing the good stabilityof the system.

3.7. Detection of PCV2 antibody in swine serum samples with EFISAsystem

In order to evaluate the performance of the EFISA system inactual swine serum samples, a series of positive serum sampleswith different dilution ratios were tested. As shown in Fig. 4A, theEFISA system possessed higher sensitivity and the signal wasamplified by nearly 2 order of magnitude even judged by nakedeye as compared with conventional ELISA. The results demon-strated that the system had an excellent capability in response tochanges of the actual serum samples. After that, we discussed theEFISA system without H2O2, which shows comparable detectionresults with the presence of H2O2 (Fig. S16).

Fig. 4B showed the results of PCV2 antibody in positive swineserum samples at different dilution ratios. Inset calibration curvedisplayed a linear regression equation Y3¼1.150þ0.1636 lg X (X:dilution ratio) ranging from 1:102 to 5:107 with R2¼0.9883 (S/N¼3). Compared with the results as previously reported in the

Fig. 3. (A) Fe3O4 MBs based ELISA system (Fe3O4-Ag–Ab1–Ab2) and (B) EFISA system (Fe3O4-Ag–Ab1–Au-Pt/SiO2-Ab2) for the detection of HIgG (Inset: linear calibration plotbetween optical density and CHIgG from 100 to 30,000 pg mL�1). EFISA system and commercial kit for the immunoassay of (C) HIgG in five human serum samples and (D)20 ng mL�1 HIgGþ200 ng mL�1 interference: (1) HIgG only, (2) BSAþHIgG, (3) hemoglobulinþHIgG, (4) L-cysteineþHIgG and (5) glucoseþHIgG. All the error bars werecalculated based on the standard deviation of at least three measurements.

Table 1Comparison of the proposed EFISA system with other methods for HIgG detection.

Detectiontechnique

Linear range(ng mL�1)

Detection limit (ngmL�1)

Reference

NER-LISAa – ∼0.5 Eum et al., 2014ELISA-like system 0.7–100 0.3 Wang et al., 2016Amplified ELISA 5–20,000 3.9 Wu et al., 2015NSETb method 4–220 0.83 Tao et al., 2014CVc method 30–1000 25 Zarei et al., 2012EFISA system 0.1– 30 0.075 This work

a Nanoscale enzyme reactors-linked immunosorbent assay.b Nanoparticle surface energy transfer.c Cyclic voltammograms.

L. Wu et al. / Biosensors and Bioelectronics 82 (2016) 177–184182

conventional ELISA system (Wu et al., 2015) (range from 1:102 to5:105), the proposed EFISA system provides higher sensitivity andwider linear range. The results indicated that the detection sen-sitivity of PCV2 antibody was enhanced 100-fold by the EFISAsystem, which were consistent with the colorimetric detection(Fig. 4A).

We also investigated the probable causes of signal amplifica-tion for PCV2 antibody detection. In contrast, we discussed theindirect immunoassay in conventional ELISA (control), ELISA onlywith APS NCs, ELISA only with Fe3O4 MBs and the EFISA system,respectively. As shown in Fig. 4C, the signal got significantly en-hanced with the integration of Fe3O4 MBs and APS NCs, while

ELISA only with Fe3O4 MBs or APS NCs behaved limited signalamplification, suggesting the superiority of EFISA system. More-over, the relatively low absorption values without PCV2 antibodyalso illustrated that the nonspecific binding was successfully in-hibited. The facts revealed that the enhanced signal mainly causedby two factors: Fe3O4 MBs contribute to entrap the substrate ef-fectively, and APS NCs behave as excellent mimetic peroxidase,which is consistent with our proposed strategy (Scheme 1).

Furthermore, the specificity of the EFISA system for PCV2 po-sitive serum was also investigated by comparing with BSA solu-tion, porcine reproductive and respiratory syndrome (PRRS) posi-tive serum, pseudorabies virus (PrV) positive serum, PCV2 nega-tive serum, pooled serum (positive serum/negative serum ¼1:1),respectively. As shown in Fig. 4D, the stronger optical densitieswere acquired for PCV2 positive serum and the pooled serum.However, nearly no signals appeared for BSA, PRRS positive serum,PrV positive serum and PCV2 negative serum. The results testifiedthat the PCV2 antibody can be effectively recognized by the pro-posed EFISA system with high specificity.

4. Conclusions

In summary, a facile EFISA systemwas constructed in this studyusing two steps: Fe3O4 MBs entrapped antigen as substrate andAb2 conjugated APS NCs as labels. Verification of the indirect

Fig. 4. (A) Colorimetric ELISA detection of PCV2 antibody with (a) EFISA system and (b) conventional ELISA. (B) Amplified assay using EFISA system (Inset: the amplifiedresults of different dilution ratios of PCV2 positive serum: 1:102, 5:103, 1:103, 5:104, 1:104, 5:105, 1:105, 5:106, 1:106, 5:107). (C) Verification of the indirect immunoassay forPCV2 antibody with EFISA system. (D) Specificity of BSA, PRRS, PrV, PCV2 negative serum, pooled serum (PCV2 positive serum/PCV2 negative serum ¼1:1), and PCV2positive serum using EFISA system. All the error bars were calculated based on the standard deviation of three measurements.

L. Wu et al. / Biosensors and Bioelectronics 82 (2016) 177–184 183

immunoassay indicated that the EFISA system exhibited highersensitivity for both HIgG and PCV2 antibody than that of theconventional ELISA. The observations demonstrate that the APSNCs behaved favorable stability against temperature and pH aswell as excellent peroxide-like activities towards peroxidase sub-strates. Besides, they can be cheaply prepared by wet chemicalmethod at room temperature in a short time and showed tunablecatalytic activity by controlling the structure. These advantagesexactly overcome the shortcomings of natural enzymes. The sen-sitivity of HIgG is lowered to 75 pg mL�1 with a 264-fold im-provement as compared to conventional ELISA. Also, a 100-foldimprovement was obtained in the detection limit for PCV2 anti-body immunoassay. Furthermore, the EFISA system could behavecomparable detection results even without H2O2. All the resultsindicate that the proposed EFISA system can provide a promisingplatform in bioassays and clinical diagnostics.

Acknowledgements

We gratefully acknowledge the financial support from NationalNatural Science Foundation of China (21375043, 21175051).

Appendix A. Supporting information

Supplementary data associated with this article can be found inthe online version at http://dx.doi.org/10.1016/j.bios.2016.04.001.

References

Ambrosi, A., Airo, F., Merkoc, A., 2010. Anal. Chem. 82, 1151–1156.Asati, A., Santra, S., Kaittanis, C., Nath, S., Perez, J.M., 2009. Angew. Chem. Int. Ed. 48,

2308–2312.Bai, F., Sun, Z., Wu, H., Haddad, R.E., Xiao, X., Fan, H., 2011. Nano Lett. 11, 3759–3762.Beyhan, S., Sahin, N.E., Pronier, S., Léger, J.M., Kadirgan, F., 2015. Electrochim. Acta

151, 565–573.Bhainsa, K.C., D’Souza, S.F., 2006. Colloids Surf. B Biointerfaces 47, 160–164.Bratlie, K.M., Lee, H., Komvopoulos, K., Yang, P., Somorjai, G.A., 2007. Nano Lett. 7,

3097–3101.Cai, K., Lv, Z., Chen, K., Huang, L., Wang, J., Shao, F., Wang, Y., Han, H., 2013. Chem.

Commun. 49, 6024–6026.Chen, F., Hou, S., Li, Q., Fan, H., Fan, R., Xu, Z., Zhala, G., Mai, X., Chen, X., Chen, X., Liu,

Y., 2014a. Anal. Chem. 86, 10021–10024.Chen, K., Wu, L., Jiang, X., Lu, Z., Han, H., 2014b. Biosens. Bioelectron. 62, 196–200.Dai, Z., Liu, S., Bao, J., Ju, H., 2009. Chem. Eur. J. 15, 4321–4326.de la Rica, R., Stevens, M.M., 2012. Nat. Nanotechnol. 7, 821–824.Diaz-Amigo, C., Popping, B., 2013. J. Agric. Food Chem. 61, 5681–5688.Drew, K., Girishkumar, G., Vinodgopal, K., Kamat, P.V., 2005. J. Phys. Chem. B 109,

11851–11857.Engvall, E., Perlman, P., 1971. Immunochemistry 8, 871–874.Eum, J.Y., Hwang, S.Y., Ju, Y., Shim, J.M., Piao, Y., Lee, J., Kim, H.S., Kim, J., 2014. Chem.

Commun. 50, 3546–3548.Frey, A., Meckelein, B., Externest, D., Schmidt, M.A., 2000. J. Immunol. Methods 233,

47–56.Gao, L., Zhuang, J., Nie, L., Zhang, J., Zhuang, Y., Gu, N., Wang, T., Feng, J., Yang, D.,

Perrett, S., Yan, X., 2007. Nat. Nanotechnol. 2, 577–583.Gao, Z., Xu, M., Hou, L., Chen, G., Tang, D., 2013. Anal. Chem. 85, 6945–6952.Gao, Z., Xu, M., Lu, M., Chen, G., Tang, D., 2015. Biosens. Bioelectron. 70, 194–201.Guo, S., Zhai, J., Fang, Y., Dong, S., Wang, E., 2008. Chem. Asian J. 3, 1156–1162.He, W., Liu, Y., Yuan, J., Yin, J.J., Wu, X., Hu, X., Zhang, K., Liu, J., Chen, C., Ji, Y., Guo, Y.,

2011. Biomaterials 32, 1139–1147.Huang, L., Wu, Q., Wang, J., Foda, M., Liu, J., Cai, K., Han, H., 2014. Chem. Commun.

50, 2896–2899.Jayasena, S., Smits, M., Fiechter, D., Jong, A., Nordlee, J., Baumert, J., Taylor, S.L.,

Pieters, R.H., Koppelman, S.J., 2015. J. Agric. Food Chem. 63, 1849–1855.Jia, C.P., Zhong, X.Q., Hua, B., Liu, M.Y., Jing, F.X., Lou, X.H., Yao, S.H., Xiang, J.Q., Jin, Q.

H., Zhao, J.L., 2009. Biosens. Bioelectron. 24, 2836–2841.Katsounaros, I., Schneider, W.B., Meier, J.C., Benedikt, U., Biedermann, P.U., Auer, A.

L. Wu et al. / Biosensors and Bioelectronics 82 (2016) 177–184184

A., Mayrhofer, K.J., 2012. J. Phys. Chem. Chem. Phys. 14, 7384–7391.Laube, T., Kergaravat, S.V., Fabiano, S.N., Hernández, S.R., Alegret, S., Pividori, M.I.,

2011. Biosens. Bioelectron. 27, 46–52.Liang, J., Yao, C., Li, X., Wu, Z., Huang, C., Fu, Q., Lan, C., Cao, D., Tang, Y., 2015.

Biosens. Bioelectron. 69, 128–134.Liu, D., Huang, X., Wang, Z., Jin, A., Sun, X., Zhu, L., Wang, F., Ma, Y., Niu, G., Hight

Walker, A.R., Chen, X., 2013. ACS Nano 7, 5568–5576.Mao, Y., Bao, Y., Han, D.X., Li, F.H., Niu, L., 2012. Biosens. Bioelectron. 38, 55–60.Opriessnig, T., Meng, X.J., Halbur, P.G., 2007. J. Vet. Diagn. Investig. 19, 591–615.Pérez-López, B., Merkoçi, A., 2011. Anal. Bioanal. Chem. 399, 1577–1590.Qu, Z., Xu, H., Xu, P., Chen, K., Mu, R., Fu, J., Gu, H., 2014. Anal. Chem. 86, 9367–9371.Shao, K., Wang, J., Jiang, X., Shao, F., Li, T., Ye, S., Chen, L., Han, H., 2014. Anal. Chem.

86, 5749–5757.Song, Y., Li, Y., Xia, X., 2007. Electrochem. Commun. 9, 201–205.Song, Y., Qu, K., Zhao, C., Ren, J., Qu, X., 2010. Adv. Mater. 22, 2206–2210.Tang, J., Su, B., Tang, D., Chen, G., 2010. Biosens. Bioelectron. 25, 2657–2662.

Tao, H., Liao, X., Xu, M., Xie, X., Zhong, F., Yi, Z., 2014. Anal. Methods 6, 2560–2565.Wang, S., Chen, Z., Choo, J., Chen, L., 2016. Anal. Bioanal. Chem. 408, 1015–1022.Wu, L., Chen, K., Lu, Z., Li, T., Shao, K., Shao, F., Han, H., 2014. Anal. Chim. Acta 845,

92–97.Wu, L., Li, X., Shao, K., Ye, S., Liu, C., Zhang, C., Han, H., 2015. Anal. Chim. Acta 887,

192–200.Xin, Y., Liu, J., Zhou, Y., Liu, W., Gao, J., Xie, Y., Yin, Y., Zou, Z., 2011. J. Power Sources

196, 1012–1018.Zarei, H., Ghourchian, H., Eskandari, K., Zeinali, M., 2012. Anal. Methods 421,

446–453.Zhao, Y., Fan, L., Zhong, H., Li, Y., Yang, S., 2007. Adv. Funct. Mater. 17, 1537–1541.Zhou, S., Varughese, B., Eichhorn, B., Jackson, G., McIlwrath, K., 2005. Angew. Chem.

117, 4615–4619.Zuo, Y., Wu, L., Cai, K., Li, T., Yin, W., Li, D., Li, N., Liu, J., Han, H., 2015. ACS Appl.

Mater. Interfaces 7 17725�1773.