Biosensors and Bioelectronics 63 (2015) 190–195
Contents lists available at ScienceDirect
Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios
Electrochemical impedance immunosensor for sub-picogram level detection of bovine interferon gamma based on cylinder-shaped TiO2 nanorods Zhanjun Yang a,n, Zhiqin Jian a, Xiang Chen b, Juan Li a, Piya Qin a, Jie Zhao a, Xin'an Jiao b, Xiaoya Hu a,n a Jiangsu Key Laboratory of Environmental Material and Environmental Engineering, College of Chemistry and Chemical Engineering, Yangzhou University, Yangzhou 225002, PR China b Jiangsu Key Lab of Zoonosis/Jiangsu Co-Innovation Center for Prevention and Control of Important Animal Infectious Diseases and Zoonoses, Yangzhou University, Yangzhou 225002, PR China
art ic l e i nf o
a b s t r a c t
Article history: Received 8 May 2014 Received in revised form 4 July 2014 Accepted 16 July 2014 Available online 21 July 2014
Bovine interferon gamma (BoIFN-γ) released by T cells plays very important roles in early diagnosis of Mycobacterium tuberculosis (MTB) infections and control of bovine tuberculosis. In this work, a labelfree electrochemical impedance immunosensor is for the first time developed for highly sensitive determination of BoIFN-γ. Cylinder-shaped TiO2 nanorods are synthesized by a facile hydrothermal method, and show high surface area and good hydrophicility. The immunosensor is fabricated by the immobilization of BoIFN-γ monoclonal antibody on the TiO2 nanorods modified glassy carbon electrode. The prepared TiO2 and immunosensor are characterized using transmission electron microscopy, scanning electron microscopy, X-ray diffraction, contact angle measurement, cyclic voltammetry, and electrochemical impedance spectra. The BoIFN-γ concentration is measured through the relative increase of impedance values in corresponding specific binding of BoIFN-γ antigen and BoIFN-γ antibody. The relative increased impedance values are proportional to the logarithmic value of BoIFN-γ concentrations in a wide range of 0.0001 to 0.1 ng/mL with a low detection limit of 0.1 pg/mL. The developed BoIFN-γ immunosensor shows a 249-fold decrease in detection limit in comparison with current enzyme-linked immunosorbent assay. This study provides a new, simple, and highly sensitive approach for very potential application in early diagnosis of MTB infections and control of bovine tuberculosis. & Elsevier B.V. All rights reserved.
Keywords: Bovine interferon gamma Electrochemical impedance immunosensor Cylinder-shaped TiO2 nanorods Label-free Cytokine
1. Introduction Bovine tuberculosis (TB) is a chronic, infectious, and zoonotic disease in cattle caused by the pathogen Mycobacterium tuberculosis (MTB), and seriously affects the healthy of cow and quality of milk. Statistically, about ten percent of TB infections progress to active TB disease, and the others are asymptomatic and latent. However, this latent infection will be active, and eventually develop into TB disease if the organism of bovine has a low resistance to diseases. Interferon gamma (IFN-γ) is a cytokine produced primarily by activated T lymphocytes and natural killer cells in response to mitogen or antigen stimulation (Farrar and Schreiber, 1993; Billiau, 1996; Liu et al., 2010). At the early phase of TB infection, bovine IFN-γ (BoIFN-γ) level released in response to n
Corresponding authors. Tel./fax: þ86 514 87972034. E-mail addresses: [email protected] (Z. Yang), [email protected] (X. Hu).
http://dx.doi.org/10.1016/j.bios.2014.07.032 0956-5663/& Elsevier B.V. All rights reserved.
specific antigen increases significantly (Andersen et al., 2000; Walravens et al., 2002). Therefore, BoIFN-γ produced by T cells can be used as a good indicator for early diagnosis of TB infection (Wood et al., 1990a). In 1990, an in vitro cellular assay was developed for the detection of BoIFN-γ released in a culture of whole blood (Wood et al., 1990a). However, this bioassay takes 4 days to perform, is laborious and not specific for BoIFN-γ (Wood et al., 1990b). Since the production of monoclonal antibodies specific for BoIFN-γ (Wood et al., 1990b), sandwich enzyme-linked immunosorbent assay (ELISA) was later developed to replace the bioassay method for the determination of BoIFN-γ (Rothel et al., 1990), which can detect less than 25 pg/mL of BoIFN-γ and is specific for biologically active BoIFN-γ. Unfortunately, ELISA suffers from low detection sensitivity as well as labor-intensive and timeconsuming manipulations. Thus it is important to explore new routes for highly sensitive detection of BoIFN-γ. Due to the high sensitivity and selectivity as well as the simple and convenient operation, electrochemical immunosensors have
Z. Yang et al. / Biosensors and Bioelectronics 63 (2015) 190–195
been received tremendous attention (Li et al., 2008; Lai et al., 2009; Du et al., 2010; Zhang et al., 2010; Wu et al., 2011; Yuan et al., 2011; Wang et al., 2013a, 2013b). As most antibody and antigen molecules are electrochemically inert, the label-free technique of electrochemical impedance spectroscopy (EIS) is developed for a direct detection of immunospecies by measuring the change of impedance (Chen et al., 2008; Li et al., 2009). This reduces assay costs because no additional materials, such as labeled antibodies, are required for the detection. In addition, EIS provides a nondestructive means for the characterization of electrical properties in biological interface (Chen et al., 2006). In recent years, electrochemical impedance immunosensors have been widely explored for a broad set of biological studies (Dijksma et al., 2001; Bart et al., 2004; Grossi et al., 2008; Derkus et al., 2013; Elshafey et al., 2013; Montrose et al., 2013; Wang et al., 2013a, 2013b). Nanomaterials, especially oxide nanoparticles, have recently attracted considerable interest in various fields for their unique properties (Wu et al., 2009; Qi et al., 2010; Tu et al., 2011; Dong et al., 2012; Jiang et al., 2012; Xu et al., 2012). Titanium dioxide (TiO2), the best known semiconductor with band gap energy of 3.2 eV, has gained increasing attention in various fields such as pigment, cosmetic, catalyst, photovoltaic materials and sensors (Koo et al., 2006; Wang et al., 2008; Liang and Li, 2009; Tu et al., 2009, 2010; Zhang et al., 2009; Li et al., 2011). In recent years, TiO2 nanomaterials have been used to immobilize biomolecules for electrochemical biosensing due to their high surface area and electron transfer rate promoting properties (Bao et al., 2008; Khan and Dhayal, 2008; Derkus et al., 2013; Yang et al., 2014). Onedimensional (1D) TiO2 nanostructures, especially nanorods, nanowires and nanotubes, are superior to their spherical and planar counterparts due to their high surface-to-volume ratio, increased number of delocalized carriers, and improved charge transport afforded by their dimensional anisotropy (Koo et al., 2006; Wang et al., 2008). Here, we reported a facile hydrothermal route for the large scale preparation of 1D cylinder-shaped TiO2 nanorods. Based on the immobilization of BoIFN-γ monoclonal antibody on cylinder-shaped TiO2 nanorods modified electrode, a novel electrochemical impedance immunosensor is for the first time developed for highly sensitive determination of BoIFN-γ (Scheme 1). The prepared TiO2 nanorods and immunosensor were characterized with different means. The results indicated that cylinder-shaped TiO2 provided a hydrophilic, favorable and high capacity platform for loading of proteins, which thus improves the sensitivity of the resultant immunosensor. The BoIFN-γ detection is performed by measuring the relative increase of impedance values in corresponding specific reaction of BoIFN-γ antigen and BoIFN-γ antibody. Compared with the conventional assay, this method shows remarkably increased sensitivity, lower cost and
191
simpler operation. This research provides a label-free and promising approach for early diagnosis of MTB infections and control of bovine tuberculosis.
2. Materials and methods 2.1. Materials and reagents The purified recombinant BoIFN-γ antigen (0.63 mg/mL) and purified monoclonal BoIFN-γ antibody named 2G5 (0.919 mg/mL) were made by Jiangsu Key Lab of Zoonosis of Yangzhou University. Titanium tetrachloride, chloroform, and absolute ethanol were bought from Sinopharm Chemical Reagent Co. Ltd. (China). Nafion and bovine serum albumin (BSA) were obtained from Sigma (St. Louis, MO). Phosphate buffer saline (PBS buffer, 0.1 M, pH 7.0) was prepared by varying the ratio of NaH2PO4 and Na2HPO4. The standard BoIFN-γ antigen solutions were prepared daily in 0.1 M PBS (pH 7.0). All other chemicals and reagents are analytical grade and were prepared using distilled water. 2.2. Apparatus Electrochemical measurements were carried out on an Autolab Electrochemical Analyzer (Ecochemie, Netherlands) and a CHI 852C electrochemical workstation (Co., CHI, Shanghai Chenhua, China). All experiments were performed with a three-electrode system using a glassy carbon electrode (GCE, D¼ 3 mm) as the working electrode, a platinum wire as the auxiliary electrode and a saturated calomel electrode (SCE) as reference electrode. The cyclic voltammetry and EIS experiments were carried out in PBS solution containing 0.1 M KCl and 5.0 mM K3[Fe(CN)6]/K4[Fe(CN)6] at a scan rate of 100 mV/s. The EIS was recorded within the frequency ranging from 10 1 to 105 Hz, and the amplitude of the applied sine wave potential was 10 mV. X-ray diffractometer (XRD) pattern was recorded on D8 Advance X-ray diffractometer (Bruker Co., Germany) at room temperature (RT). Transmission electron micrograph (TEM) was obtained with a Philips Tecnai-12 transmission electron microscope (Holland) at an acceleration voltage of 120 kV. Scanning electron micrographs (SEM) were obtained with a Hitachi S-4800 scanning electron microscope (Japan) at an acceleration voltage of 15 kV. The static water contact angles were measured with a contact angle meter (Rame-Hart-100) using droplets of the distilled water at RT. 2.3. Preparation of cylinder-shaped TiO2 nanorods 19 mL of water was transferred into a 50 mL glass beaker, and the beaker was then kept in an ice-water bath. 1.8 mL of titanium tetrachloride was slowly added into the distilled water. After the solution was stirred vigorously for 10 min, a white suspension was obtained. Subsequently, 1.8 mL of chloroform was added into the white suspension, stirred for 10 min and then transferred into Teflon-lined stainless-steel autoclave. The autoclave was placed in an electrothermal oven and maintained at 160 °C for 12 h. The asformed precipitate was collected, washed carefully with distilled water and absolute ethanol until the pH reached about 7.0 and then dried in a vacuum at 60 °C for 24 h. 2.4. Preparation of BoIFN-γ immunosensor
Scheme 1. Schematic illustration for the fabrication of the immunosensor and immunoassay process for BoIFN-γ.
The GCE was firstly polished successively with 0.03 and 0.05 μm alumina slurry (Buhler) followed by rinsing thoroughly with distilled water, and then sonicated in 1:1 nitric acid–water (v/v), ethanol and distilled water and finally dried in air. 2.0 mg of TiO2 nanorods were dispersed in 1.0 mL distilled water under
192
Z. Yang et al. / Biosensors and Bioelectronics 63 (2015) 190–195
ultrasonic stirring, and then 0.5 mL of TiO2 nanorods suspension was mixed with equivalent volume of 400 μg/mL monoclonal BoIFN-γ antibody. 5.0 μL of the resulting mixture was dropped on the pretreated GCE and followed at 4 °C for 12 h. To prevent the leakage of BoIFN-γ antibody from TiO2 nanorods modified electrode, 5.0 μL of 1.0% Nafion was dropped on the surface of antibody embedded GCE. Next, the non-specific sites were blocked with 10 mg/mL of BSA solution for 30 min at 37.5 °C. After washing three times with PBS, the BoIFN-γ immunosensor was obtained. The fabricated immunosensor was stored at 4 °C when not in use. 2.5. Immunoassay procedure The detailed process for the detection of BoIFN-γ is illustrated in Scheme 1. The prepared BoIFN-γ immunosensor was incubated with different concentrations of BoIFN-γ antigen at 37.5 °C for 80 min, and then washed carefully with PBS. The relative change in impedance of the immunosensor was measured in 0.1 M PBS (pH 7.0) and 0.1 M KCl solution containing 5 mM K3[Fe(CN)6]/ K4[Fe(CN)6].
3. Results and discussion 3.1. Characterization of cylinder-shaped TiO2 and BoIFN-γ immunosensor The TEM and SEM images of the synthesized cylinder-shaped TiO2 are shown in Fig. 1a and b, which display an anisotropic and uniform cylinder-like morphology with about 180 nm length and 20 nm thickness. Fig. 1c shows the XRD pattern of TiO2
nanorods, the as-prepared product displayed only the XRD peaks characteristic of rutile TiO2 (JCPDS card no. 65-0190), indicating that phase-pure TiO2 powders have been successfully prepared. Fig. 1d shows the SEM image of BoIFN-γ antibody immobilized TiO2, which exhibits obviously different surface morphology from TiO2 film (Fig. 1b). The regularly distributed aggregates of the loaded proteins were clearly observed, indicating the successful immobilization of BoIFN-γ antibody on TiO2 nanorods. The hydrophilicity of an interface is very important for the bioactivity of loaded biomolecules (Zhu et al., 2002). It is generally characterized by measuring the contact angle of the substrate. Fig. S1 (Supporting information) shows the contact angles of the bare GCE, TiO2 nanorods/GCE, and BoIFN-γ antibody/TiO2 nanorods/ GCE. Compared with bare electrode, the TiO2 nanorods modified electrode shows a lower contact angle. This demonstrates that the synthesized TiO2 nanorods have excellent hydrophilicity, and provides a favorable microenvironment for proteins immobilization. The antibody/TiO2 nanorods/GCE displays the smallest contact angle, indicating the successful loading of BoIFN-γ monoclonal antibody in the TiO2 film. The excellent hydrophilicity of the TiO2 nanorods is believed to be highly advantageous to the preparation of bioreactors and biosensors. The cyclic voltammetric experiment was chosen to investigate the electrode behavior after each assembly step. Fig. 2 shows the cyclic voltammograms of Fe(CN)63 /4 at the bare GCE (curve a), TiO2 nanorods/GCE (curve b), anti-BoIFN-γ/TiO2 nanorods/ GCE (curve c), BSA/anti-BoIFN-γ/TiO2 nanorods/GCE (curve d), and BoIFN-γ/BSA/anti-BoIFN-γ/TiO2 nanorods/GCE (e), respectively. As seen from Fig. 2, the redox currents decrease gradually with stepwise modifications on the bare electrode, suggesting that the electrode transfer process between Fe(CN)63 /4 probe and
Fig. 1. TEM image of TiO2 nanorods (a), SEM image of TiO2 nanorods (b), X-ray diffraction pattern of TiO2 nanorods (c), and anti-BoIFN-γ antibody immobilized TiO2 nanorods (d).
Z. Yang et al. / Biosensors and Bioelectronics 63 (2015) 190–195
Fig. 2. Cyclic voltammograms of bare GCE (a), TiO2 nanorods/GCE (b), antiBoIFN-γ/TiO2 nanorods/GCE (c), BSA/anti-BoIFN-γ/TiO2 nanorods/GCE (d), and BoIFN-γ/BSA/anti-BoIFN-γ/TiO2 nanorods/GCE (e) recorded in PBS solution (0.1 M, pH 7.0) including 5.0 mM K3[Fe(CN)6]/K4[Fe(CN)6] and 0.1 M KCl at a scan rate of 100 mV/s.
electrode surface was blocked. In particular, BoIFN-γ antibody and BSA immobilized electrodes produced bigger inhibition of the electron transfer, showing obvious decreases of the current of the anodic and cathonic peaks. EIS was also used to examine detailed information on the impedance changes in the modification process. The electron transfer resistance (Rct) can be quantified using the semicircle diameter in the Nyquist plot of impedance spectra. Fig. 3A displays Faradaic impedance spectra obtained upon the stepwise modification process. A very small semicircle can be observed at bare GCE,
193
indicating a low transfer resistance (curve a). After TiO2 nanorods were modified on the GCE surface, the Rct increased slightly (curve b), implying the formation of TiO2 film on the electrode surface. When BoIFN-γ antibody was loaded on the surface of TiO2 nanorods modified GCE, the Rct increased greatly (curve c), suggesting successful loading the antibody on the TiO2 nanorods. After BoIFN-γ antibody immobilized electrode was blocked with BSA, the Rct further increased (curve d). The similar increase also could be observed when the immunosensor was used to detect BoIFN-γ antigen (curve e). The possible reason is that the formed protein layer on the electrode surface significantly hinders the diffusion of ferricyanide probe toward the electrode surface. The impedance data were fitted according to a Randles equivalent circuit based on FRA software of Autolab. A modified Randles equivalent circuit and the fitting of one measured spectrum to the equivalent circuit (solid line) are shown in Fig. 3B, demonstrating the good agreement with the circuit model and the measurement system over the whole measurement frequency range. The circuit consists of the solution resistance (Rs), the electron transfer resistance (Rct), interfacial double layer capacitance (Cdl) between an electrode and a solution, and Warburg impedance (ZW). Usually, Rs and ZW represent the bulk properties of the electrolyte solution and diffusion features of the redox probe in solution, which are thus not affected by modifications on the electrode surface. Since Cdl electrode/electrolyte interface is not simple capacitance in this study, a constant phase element (CPE) was used instead of the classical capacitance to fit the impedance data in Randles equivalent circuit. The fitting value for various modification layers on the electrode are shown in Table S1 (Supporting information), indicating that Rct is a suitable signal for sensing the interfacial properties of the prepared immunosensor during all these modification steps. The relative change in Rct (%ΔRct) is calculated by the following equation:
%ʁ R ct =
R ct(Ag ℬAb) ℬ R ct(BSA) R ct(BSA)
¬ 100
where Rct(Ag Ab) is the value of the electron transfer resistance after BoIFN-γ-Ag coupling to the immobilized Ab on TiO2 nanorods modified electrode. Rct(BSA) represents the impedance value after blocking the remaining adsorption-reactive sites by BSA.
3.2. Optimization of the experimental conditions
Fig. 3. (A) Nyquist plots of bare GCE (a), TiO2 nanorods/GCE (b), anti-BoIFN-γ/TiO2 nanorods/GCE (c), BSA/anti-BoIFN-γ/TiO2 nanorods/GCE (d), and BoIFN-γ/ BSA/anti-BoIFN-γ/TiO2 nanorods/GCE (e) recorded in PBS solution (0.1 M, pH 7.0) including 5.0 mM K3[Fe(CN)6]/K4[Fe(CN)6] and 0.1 M KCl at a scan rate of 100 mV/s. The inset is the plots of (a) and (b). (B) Fitted (solid line) and experimental (scattered line) Nyquist plots of impedance spectra. The inset is the equivalent circuit applied to fit the impedance spectroscopy in the presence of the redox couple of Fe(CN)63 /4 .
The amount of antibody immobilized on the TiO2 nanorods modified GCE could affect the analytical property of the immunosensor. Thus the influence of BoIFN-γ antibody concentration was examined here (data not shown). It was found that the ΔRct value increased with the increasing antibody concentration (from 0.1 to 0.9 mg/mL), and trended a plateau at a concentration of 0.4 mg/mL. Therefore, 0.4 mg/mL of BoIFN-γ antibody concentration was chosen for the immobilization in the experiment. The capture of target proteins on the TiO2 nanorods modified GCE could change the interface properties of electrodes, thus leading to a change of Rct. Therefore, the factors affecting antibody-antigen reaction, such as pH of the solution, incubation temperature and incubation time, were optimized. Fig. S2A (Supporting information) shows the effect of solution pH on the relative change in Rct for antigen-antibody reaction at a concentration of 0.01 ng/mL BoIFN-γ. With the increasing solution pH from 4.0 to 9.0, the ΔRct value increased and trended to the maximum value at pH 7.0, indicating the optimal pH for the immunoreaction. The influence of incubation temperature for the immunoassay was also investigated in a range of 25–45 °C at a
194
Z. Yang et al. / Biosensors and Bioelectronics 63 (2015) 190–195
3.4. Specificity of the proposed BoIFN-γ immunosensor Nonspecific adsorption is usually a major problem in label-free immunosensing, and greatly affects the specificity of an immunosensor. In order to confirm that the relative changes result from specific interaction between BoIFN-γ and the BoIFN-γ antibody loaded on the electrode, nonspecific interaction between the proposed immunosensor and the analogous cytokine was examined in Fig. S3 (Supporting information). After the resultant immunosensor was incubation with 0.01 ng/mL bovine interleukin-4 (BoIL-4), the relative impedance value is very small. However, the relative impedance value remarkably increased after the resultant immunosensor was incubated with 0.01 ng/mL target BoIFN-γ. Moreover, the relative impedance value shows no obvious changes upon the addition of different concentrations of interferent BoIL-4 to 0.01 ng/mL BoIFN-γ solution, indicating that nonspecific adsorption between BoIFN-γ antibody and the other analogous analyte can be negligible. These results demonstrate that the developed label-free immunosensor has high specificity and can be used for the selective detection of BoIFN-γ. 3.5. Reproducibility, stability and accuracy of the BoIFN-γ immunosensor
Fig. 4. (A) Faradaic impedance spectra of the immunosensor after incubating with 0.0001, 0.0005, 0.001, 0.01 and 0.1 ng/mL BoIFN-γ in PBS solution (0.1 M, pH 7.0) including 5.0 mM K3[Fe(CN)6]/K4[Fe(CN)6] and 0.1 M KCl. (B) Calibration curve for BoIFN-γ (n¼5 for each point).
concentration of 0.01 ng/mL BoIFN-γ. As seen from Fig. S2B (Supporting information), the maximal relative change in Rct was observed at an incubation temperature of 37.5 °C. Additionally, the incubation time greatly affected the Ag–Ab combination, thus the effect of reaction time on the ΔRct value was examined at a concentration of 0.01 ng/mL BoIFN-γ and shown in Fig. S2C (Supporting information). With the increase of incubation time, the electrochemical response of binding reaction increased and reached a plateau beyond 80 min incubation time. To obtain the optimal analytical performance, a pH value of 7.0, incubation temperature of 37.5 °C and 80 min of incubation time were selected for the immunoassay of BoIFN-γ. 3.3. Calibration curve for BoIFN-γ Under the optimum conditions, the fabricated immunosensor was incubated with different concentrations of BoIFN-γ. Fig. 4A illustrates the corresponding Nyquist plots of impedance spectra. It could be found that the semicircle diameter in the Nyquist plot increased with the increasing concentration of the BoIFN-γ. The possible reason is that more antigen molecules are bound to the immobilized antibody in higher concentration of BoIFN-γ, resulting in bigger inhibitions of the electron transfer of the redox couple. Fig. 4B shows a linear relationship between the relative change in Rct and logarithm of BoIFN-γ in the range of 0.0001 to 0.1 ng/mL, and the linear regression equation was ΔRct(%) ¼ 43.9 þ16.6 log CBoIFN-γ (R2 ¼0.9981). The detection limit was calculated to be 0.1 pg/mL at a signal/noise ration of 3, which is 249-flod lower than that of 25 ng/mL obtained for BoIFN-γ by conventional microplate ELISA (Rothel et al., 1990).
The reproducibility of the proposed BoIFN-γ immunosensor was evaluated by the intra- and inter-assay coefficients of variation (CV). The intra-assay CV at a concentration of 0.01 ng/mL BoIFN-γ for six replicative measurements was 3.7%, demonstrating good detection reproducibility. The inter-assay CV at a concentration of 0.01 ng/mL BoIFN-γ for six immunosensors fabricated in six batches was 6.3%, indicating good fabrication reproducibility. The immunosensor could be stored in a refrigerator at 4 °C when it was not in use. The results showed that the relative change in Rct retained no obvious change after a storage period of 60 days. Evidently, the prepared TiO2 nanorods can provide a good microenvironment, which is efficient to retain the bioactivity of the immobilized antibody molecules. To evaluate the accuracy and application potential of the developed electrochemical impedance BoIFN-γ immunosensor, the recovery was tested by spiking 0.0002, 0.0005, 0.001, 0.005 and 0.01 ng/mL BoIFN-γ in bovine serum samples. The obtained recoveries were shown in Table 1, demonstrating good accuracy of the fabricated immunosensor for label-free detection of BoIFN-γ in practical samples.
4. Conclusions In this work, a first electrochemical impedance immunosensor was developed for label-free and highly sensitive determination of BoIFN-γ based on cylinder-shaped TiO2 nanorods, which was synthesized by a facile hydrothermal method. The efficient immunosensor was fabricated by immobilizing BoIFN-γ monoclonal antibody on the TiO2 nanorods modified GCE. The as-prepared Table 1 Recoveries for BoIFN-γ by the proposed electrochemical impedance immunosensor (n ¼ 5). Sample
Added (ng/mL)
Found (ng/mL)
Recovery (%)
RSD (%)
1 2 3 4 5
0.0002 0.0005 0.001 0.005 0.01
0.00022 0.00054 0.00109 0.00510 0.00990
110 108 109 102 99
1.6 2.1 2.7 3.5 1.5
Z. Yang et al. / Biosensors and Bioelectronics 63 (2015) 190–195
TiO2 nanorods provide large surface area and friendly microenvironment for the immobilization of protein biomolecules while retaining the biological activity of the immunosensor. Compared with traditional biological assay and ELISA, this strategy shows ultrahigh sensitivity, simple manipulation and wide linear range. The resultant BoIFN-γ immunosensor also displayed excellent specificity, good reproducibility and acceptable stability. The recovery experiment demonstrates that the proposed immunosensor has high accuracy and can be applied in the detection of BoIFN-γ in practical samples. This study opens new avenues for highly sensitive detection of BoIFN-γ and further early diagnosis of MTB infections.
Acknowledgments This work was financially supported by National Natural Science Foundation of China (21275124 and 21103148), National Basic Research and Development Program of China (2012CB518805), The Priority Academic Program Development of Jiangsu Higher Education Institution, Postdoctoral Science Foundation of Jiangsu Province (1101020B) and University Natural Science Foundation of Jiangsu Province (13KJB150039). Also, thanks to The Testing Center of Yangzhou University for the characterization data.
Appendix A. Supplementary information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.bios.2014.07.032.
References Andersen, P., Munk, M.E., Pollock, J.M., Doherty, T.M., 2000. Lancet 356, 1099–1104. Bao, S.J., Li, C.M., Zang, J.F., Cui, X.Q., Qiao, Y., Guo, J., 2008. Adv. Funct. Mater. 18, 591–599. Bart, M., Stigter, E.C. A., Stapert, H.R., de Jong, G.J., van Bennekom, W.P., 2004. Biosens. Bioelectron. 21, 49–59. Billiau, A., 1996. Adv. Immunol. 62, 61–130. Chen, H., Jiang, J.H., Huang, Y., Deng, T., Li, J.S., Shen, G.L., Yu, R.Q., 2006. Sens. Actuators B 117, 211–218.
195
Chen, X.J., Wang, Y.Y., Zhou, J.J., Yan, W., Li, X.H., Zhu, J.J., 2008. Anal. Chem. 80, 2133–2140. Derkus, B., Emregul, E., Yucesan, C., Emregul, K.C., 2013. Biosens. Bioelectron. 46, 53–60. Dijksma, M., Kamp, B., Hoogvliet, J.G., van Bennekom, W.P., 2001. Anal. Chem. 73, 901–907. Dong, H.F., Lei, J.P., Ju, H.X., Zhi, F., Wang, H., Guo, W.J., Zhu, Z., Yan, F., 2012. Angew. Chem. Int. Ed. 51, 4607–4612. Du, D., Zou, Z.X., Shin, Y.S., Wang, J., Wu, H., Engelhard, M.H., Liu, J., Aksay, I.A., Lin, Y.H., 2010. Anal. Chem. 82, 2989–2995. Elshafey, R., Tlili, C., Abulrob, A., Tavares, A.C., Zourob, M., 2013. Biosens. Bioelectron. 39, 220–225. Farrar, M.A., Schreiber, R.D., 1993. Annu. Rev. Immunol. 11, 571–611. Grossi, M., Lanzoni, M., Pompei, A., Lazzarini, R., Matteuzzi, D., Ricco, B., 2008. Biosens. Bioelectron. 23, 1616–1623. Jiang, H.D., Kim, S.K., Chang, H., Roh, K.M., Choi, J.W., Huang, J.X., 2012. Biosens. Bioelectron. 38, 184–188. Khan, R., Dhayal, M., 2008. Electrochem. Commun. 10, 492–495. Koo, B., Park, J., Kim, Y., Choi, S.H., Sung, Y.E., Hyeon, T., 2006. J. Phys. Chem. B 110, 24318–24323. Lai, G.S., Yan, F., Ju, H.X., 2009. Anal. Chem. 81, 9730–9736. Li, H.B., Li, J., Yang, Z.J., Xu, Q., Hu, X.Y., 2011. Anal. Chem. 83, 5290–5295. Li, X.H., Dai, L., Liu, Y., Chen, X.J., Yan, W., Jiang, L.P., Zhu, J.J., 2009. Adv. Funct. Mater. 19, 3120–3128. Li, X.M., Yang, X.Y., Zhang, S.S., 2008. Trends Anal. Chem. 27, 543–553. Liang, H.T., Li, X.Z., 2009. J. Hazard. Mater. 162, 1415–1422. Liu, Y., Tuleouva, N., Ramanculov, E., Revzin, A., 2010. Anal. Chem. 82, 8131–8135. Montrose, A., Cargou, S., Nepveu, F., Manczak, R., Gue, A.M., Reybier, K., 2013. Biosens. Bioelectron. 49, 305–311. Qi, W.J., Huang, C.Z., Chen, L.Q., 2010. Talanta 80, 1400–1405. Rothel, J.S., Jones, S.L., Corner, L.A., Cox, J.C., Wood, P.R., 1990. Aust. Vet. J. 67, 134–137. Tu, W.W., Dong, Y.T., Lei, J.P., Ju, H.X., 2010. Anal. Chem. 82, 8711–8716. Tu, W.W., Lei, J.P., Ding, L., Ju, H.X., 2009. Chem. Commun. 28, 4227–4229. Tu, W.W., Lei, J.P., Wang, P., Ju, H.X., 2011. Chem. Eur. J 17, 9440–9447. Walravens, K., Wellemans, V., Weynants, V., Boelaert, F., deBergeyck, V., Letesson, J. J., Huyen, K., Godfroid, J., 2002. Vet. Immunol. Immunopathol. 84, 29–41. Wang, L.S., Lei, J.P., Ma, R.N., Ju, H.X., 2013a. Anal. Chem. 85, 6505–6510. Wang, Y.X., Ping, J.F., Ye, Z.Z., Wu, J., Ying, Y.B., 2013b. Biosens. Bioelectron. 49, 492–498. Wang, Z.Y., Xia, D.G., Chen, G., Yang, T., Chen, Y., 2008. Mater. Chem. Phys. 111, 313–316. Wood, P.R., Corner, L.A., Plackett, P., 1990a. Res. Vet. Sci. 49, 46–49. Wood, P.R., Rothel, J.S., McWaters, P.G., Jones, S.L., 1990b. Vet. Immunol. Immunopathol. 25, 37–46. Wu, Y.F., Chen, C.L., Liu, S.Q., 2009. Anal. Chem. 81, 1600–1607. Wu, Y.F., Liu, S.Q., He, L., 2011. Analyst 136, 2558–2563. Xu, X., Wei, W., Huang, M.H., Yao, L., Liu, S.Q., 2012. Chem. Commun. 48, 7802–7804. Yang, Z.J., Tang, Y., Li, J., Zhang, Y.C., Hu, X.Y., 2014. Biosens. Bioelectron. 54, 528–533. Yuan, L., Hua, X., Wu, Y.F., Pan, X.H., Liu, S.Q., 2011. Anal. Chem. 83, 6800–6809. Zhang, J., Lei, J.P., Xu, C.L., Ding, L., Ju, H.X., 2010. Anal. Chem. 82, 1117–1122. Zhang, S., Liu, C.Y., Liu, Y., Zhang, Z.Y., Mao, L.J., 2009. Mater. Lett. 63, 127–129. Zhu, A.P., Zhang, M., Wu, J., Shen, J., 2002. Biomaterials 23, 4657–4665.
Contents lists available at ScienceDirect
Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios
Electrochemical impedance immunosensor for sub-picogram level detection of bovine interferon gamma based on cylinder-shaped TiO2 nanorods Zhanjun Yang a,n, Zhiqin Jian a, Xiang Chen b, Juan Li a, Piya Qin a, Jie Zhao a, Xin'an Jiao b, Xiaoya Hu a,n a Jiangsu Key Laboratory of Environmental Material and Environmental Engineering, College of Chemistry and Chemical Engineering, Yangzhou University, Yangzhou 225002, PR China b Jiangsu Key Lab of Zoonosis/Jiangsu Co-Innovation Center for Prevention and Control of Important Animal Infectious Diseases and Zoonoses, Yangzhou University, Yangzhou 225002, PR China
art ic l e i nf o
a b s t r a c t
Article history: Received 8 May 2014 Received in revised form 4 July 2014 Accepted 16 July 2014 Available online 21 July 2014
Bovine interferon gamma (BoIFN-γ) released by T cells plays very important roles in early diagnosis of Mycobacterium tuberculosis (MTB) infections and control of bovine tuberculosis. In this work, a labelfree electrochemical impedance immunosensor is for the first time developed for highly sensitive determination of BoIFN-γ. Cylinder-shaped TiO2 nanorods are synthesized by a facile hydrothermal method, and show high surface area and good hydrophicility. The immunosensor is fabricated by the immobilization of BoIFN-γ monoclonal antibody on the TiO2 nanorods modified glassy carbon electrode. The prepared TiO2 and immunosensor are characterized using transmission electron microscopy, scanning electron microscopy, X-ray diffraction, contact angle measurement, cyclic voltammetry, and electrochemical impedance spectra. The BoIFN-γ concentration is measured through the relative increase of impedance values in corresponding specific binding of BoIFN-γ antigen and BoIFN-γ antibody. The relative increased impedance values are proportional to the logarithmic value of BoIFN-γ concentrations in a wide range of 0.0001 to 0.1 ng/mL with a low detection limit of 0.1 pg/mL. The developed BoIFN-γ immunosensor shows a 249-fold decrease in detection limit in comparison with current enzyme-linked immunosorbent assay. This study provides a new, simple, and highly sensitive approach for very potential application in early diagnosis of MTB infections and control of bovine tuberculosis. & Elsevier B.V. All rights reserved.
Keywords: Bovine interferon gamma Electrochemical impedance immunosensor Cylinder-shaped TiO2 nanorods Label-free Cytokine
1. Introduction Bovine tuberculosis (TB) is a chronic, infectious, and zoonotic disease in cattle caused by the pathogen Mycobacterium tuberculosis (MTB), and seriously affects the healthy of cow and quality of milk. Statistically, about ten percent of TB infections progress to active TB disease, and the others are asymptomatic and latent. However, this latent infection will be active, and eventually develop into TB disease if the organism of bovine has a low resistance to diseases. Interferon gamma (IFN-γ) is a cytokine produced primarily by activated T lymphocytes and natural killer cells in response to mitogen or antigen stimulation (Farrar and Schreiber, 1993; Billiau, 1996; Liu et al., 2010). At the early phase of TB infection, bovine IFN-γ (BoIFN-γ) level released in response to n
Corresponding authors. Tel./fax: þ86 514 87972034. E-mail addresses: [email protected] (Z. Yang), [email protected] (X. Hu).
http://dx.doi.org/10.1016/j.bios.2014.07.032 0956-5663/& Elsevier B.V. All rights reserved.
specific antigen increases significantly (Andersen et al., 2000; Walravens et al., 2002). Therefore, BoIFN-γ produced by T cells can be used as a good indicator for early diagnosis of TB infection (Wood et al., 1990a). In 1990, an in vitro cellular assay was developed for the detection of BoIFN-γ released in a culture of whole blood (Wood et al., 1990a). However, this bioassay takes 4 days to perform, is laborious and not specific for BoIFN-γ (Wood et al., 1990b). Since the production of monoclonal antibodies specific for BoIFN-γ (Wood et al., 1990b), sandwich enzyme-linked immunosorbent assay (ELISA) was later developed to replace the bioassay method for the determination of BoIFN-γ (Rothel et al., 1990), which can detect less than 25 pg/mL of BoIFN-γ and is specific for biologically active BoIFN-γ. Unfortunately, ELISA suffers from low detection sensitivity as well as labor-intensive and timeconsuming manipulations. Thus it is important to explore new routes for highly sensitive detection of BoIFN-γ. Due to the high sensitivity and selectivity as well as the simple and convenient operation, electrochemical immunosensors have
Z. Yang et al. / Biosensors and Bioelectronics 63 (2015) 190–195
been received tremendous attention (Li et al., 2008; Lai et al., 2009; Du et al., 2010; Zhang et al., 2010; Wu et al., 2011; Yuan et al., 2011; Wang et al., 2013a, 2013b). As most antibody and antigen molecules are electrochemically inert, the label-free technique of electrochemical impedance spectroscopy (EIS) is developed for a direct detection of immunospecies by measuring the change of impedance (Chen et al., 2008; Li et al., 2009). This reduces assay costs because no additional materials, such as labeled antibodies, are required for the detection. In addition, EIS provides a nondestructive means for the characterization of electrical properties in biological interface (Chen et al., 2006). In recent years, electrochemical impedance immunosensors have been widely explored for a broad set of biological studies (Dijksma et al., 2001; Bart et al., 2004; Grossi et al., 2008; Derkus et al., 2013; Elshafey et al., 2013; Montrose et al., 2013; Wang et al., 2013a, 2013b). Nanomaterials, especially oxide nanoparticles, have recently attracted considerable interest in various fields for their unique properties (Wu et al., 2009; Qi et al., 2010; Tu et al., 2011; Dong et al., 2012; Jiang et al., 2012; Xu et al., 2012). Titanium dioxide (TiO2), the best known semiconductor with band gap energy of 3.2 eV, has gained increasing attention in various fields such as pigment, cosmetic, catalyst, photovoltaic materials and sensors (Koo et al., 2006; Wang et al., 2008; Liang and Li, 2009; Tu et al., 2009, 2010; Zhang et al., 2009; Li et al., 2011). In recent years, TiO2 nanomaterials have been used to immobilize biomolecules for electrochemical biosensing due to their high surface area and electron transfer rate promoting properties (Bao et al., 2008; Khan and Dhayal, 2008; Derkus et al., 2013; Yang et al., 2014). Onedimensional (1D) TiO2 nanostructures, especially nanorods, nanowires and nanotubes, are superior to their spherical and planar counterparts due to their high surface-to-volume ratio, increased number of delocalized carriers, and improved charge transport afforded by their dimensional anisotropy (Koo et al., 2006; Wang et al., 2008). Here, we reported a facile hydrothermal route for the large scale preparation of 1D cylinder-shaped TiO2 nanorods. Based on the immobilization of BoIFN-γ monoclonal antibody on cylinder-shaped TiO2 nanorods modified electrode, a novel electrochemical impedance immunosensor is for the first time developed for highly sensitive determination of BoIFN-γ (Scheme 1). The prepared TiO2 nanorods and immunosensor were characterized with different means. The results indicated that cylinder-shaped TiO2 provided a hydrophilic, favorable and high capacity platform for loading of proteins, which thus improves the sensitivity of the resultant immunosensor. The BoIFN-γ detection is performed by measuring the relative increase of impedance values in corresponding specific reaction of BoIFN-γ antigen and BoIFN-γ antibody. Compared with the conventional assay, this method shows remarkably increased sensitivity, lower cost and
191
simpler operation. This research provides a label-free and promising approach for early diagnosis of MTB infections and control of bovine tuberculosis.
2. Materials and methods 2.1. Materials and reagents The purified recombinant BoIFN-γ antigen (0.63 mg/mL) and purified monoclonal BoIFN-γ antibody named 2G5 (0.919 mg/mL) were made by Jiangsu Key Lab of Zoonosis of Yangzhou University. Titanium tetrachloride, chloroform, and absolute ethanol were bought from Sinopharm Chemical Reagent Co. Ltd. (China). Nafion and bovine serum albumin (BSA) were obtained from Sigma (St. Louis, MO). Phosphate buffer saline (PBS buffer, 0.1 M, pH 7.0) was prepared by varying the ratio of NaH2PO4 and Na2HPO4. The standard BoIFN-γ antigen solutions were prepared daily in 0.1 M PBS (pH 7.0). All other chemicals and reagents are analytical grade and were prepared using distilled water. 2.2. Apparatus Electrochemical measurements were carried out on an Autolab Electrochemical Analyzer (Ecochemie, Netherlands) and a CHI 852C electrochemical workstation (Co., CHI, Shanghai Chenhua, China). All experiments were performed with a three-electrode system using a glassy carbon electrode (GCE, D¼ 3 mm) as the working electrode, a platinum wire as the auxiliary electrode and a saturated calomel electrode (SCE) as reference electrode. The cyclic voltammetry and EIS experiments were carried out in PBS solution containing 0.1 M KCl and 5.0 mM K3[Fe(CN)6]/K4[Fe(CN)6] at a scan rate of 100 mV/s. The EIS was recorded within the frequency ranging from 10 1 to 105 Hz, and the amplitude of the applied sine wave potential was 10 mV. X-ray diffractometer (XRD) pattern was recorded on D8 Advance X-ray diffractometer (Bruker Co., Germany) at room temperature (RT). Transmission electron micrograph (TEM) was obtained with a Philips Tecnai-12 transmission electron microscope (Holland) at an acceleration voltage of 120 kV. Scanning electron micrographs (SEM) were obtained with a Hitachi S-4800 scanning electron microscope (Japan) at an acceleration voltage of 15 kV. The static water contact angles were measured with a contact angle meter (Rame-Hart-100) using droplets of the distilled water at RT. 2.3. Preparation of cylinder-shaped TiO2 nanorods 19 mL of water was transferred into a 50 mL glass beaker, and the beaker was then kept in an ice-water bath. 1.8 mL of titanium tetrachloride was slowly added into the distilled water. After the solution was stirred vigorously for 10 min, a white suspension was obtained. Subsequently, 1.8 mL of chloroform was added into the white suspension, stirred for 10 min and then transferred into Teflon-lined stainless-steel autoclave. The autoclave was placed in an electrothermal oven and maintained at 160 °C for 12 h. The asformed precipitate was collected, washed carefully with distilled water and absolute ethanol until the pH reached about 7.0 and then dried in a vacuum at 60 °C for 24 h. 2.4. Preparation of BoIFN-γ immunosensor
Scheme 1. Schematic illustration for the fabrication of the immunosensor and immunoassay process for BoIFN-γ.
The GCE was firstly polished successively with 0.03 and 0.05 μm alumina slurry (Buhler) followed by rinsing thoroughly with distilled water, and then sonicated in 1:1 nitric acid–water (v/v), ethanol and distilled water and finally dried in air. 2.0 mg of TiO2 nanorods were dispersed in 1.0 mL distilled water under
192
Z. Yang et al. / Biosensors and Bioelectronics 63 (2015) 190–195
ultrasonic stirring, and then 0.5 mL of TiO2 nanorods suspension was mixed with equivalent volume of 400 μg/mL monoclonal BoIFN-γ antibody. 5.0 μL of the resulting mixture was dropped on the pretreated GCE and followed at 4 °C for 12 h. To prevent the leakage of BoIFN-γ antibody from TiO2 nanorods modified electrode, 5.0 μL of 1.0% Nafion was dropped on the surface of antibody embedded GCE. Next, the non-specific sites were blocked with 10 mg/mL of BSA solution for 30 min at 37.5 °C. After washing three times with PBS, the BoIFN-γ immunosensor was obtained. The fabricated immunosensor was stored at 4 °C when not in use. 2.5. Immunoassay procedure The detailed process for the detection of BoIFN-γ is illustrated in Scheme 1. The prepared BoIFN-γ immunosensor was incubated with different concentrations of BoIFN-γ antigen at 37.5 °C for 80 min, and then washed carefully with PBS. The relative change in impedance of the immunosensor was measured in 0.1 M PBS (pH 7.0) and 0.1 M KCl solution containing 5 mM K3[Fe(CN)6]/ K4[Fe(CN)6].
3. Results and discussion 3.1. Characterization of cylinder-shaped TiO2 and BoIFN-γ immunosensor The TEM and SEM images of the synthesized cylinder-shaped TiO2 are shown in Fig. 1a and b, which display an anisotropic and uniform cylinder-like morphology with about 180 nm length and 20 nm thickness. Fig. 1c shows the XRD pattern of TiO2
nanorods, the as-prepared product displayed only the XRD peaks characteristic of rutile TiO2 (JCPDS card no. 65-0190), indicating that phase-pure TiO2 powders have been successfully prepared. Fig. 1d shows the SEM image of BoIFN-γ antibody immobilized TiO2, which exhibits obviously different surface morphology from TiO2 film (Fig. 1b). The regularly distributed aggregates of the loaded proteins were clearly observed, indicating the successful immobilization of BoIFN-γ antibody on TiO2 nanorods. The hydrophilicity of an interface is very important for the bioactivity of loaded biomolecules (Zhu et al., 2002). It is generally characterized by measuring the contact angle of the substrate. Fig. S1 (Supporting information) shows the contact angles of the bare GCE, TiO2 nanorods/GCE, and BoIFN-γ antibody/TiO2 nanorods/ GCE. Compared with bare electrode, the TiO2 nanorods modified electrode shows a lower contact angle. This demonstrates that the synthesized TiO2 nanorods have excellent hydrophilicity, and provides a favorable microenvironment for proteins immobilization. The antibody/TiO2 nanorods/GCE displays the smallest contact angle, indicating the successful loading of BoIFN-γ monoclonal antibody in the TiO2 film. The excellent hydrophilicity of the TiO2 nanorods is believed to be highly advantageous to the preparation of bioreactors and biosensors. The cyclic voltammetric experiment was chosen to investigate the electrode behavior after each assembly step. Fig. 2 shows the cyclic voltammograms of Fe(CN)63 /4 at the bare GCE (curve a), TiO2 nanorods/GCE (curve b), anti-BoIFN-γ/TiO2 nanorods/ GCE (curve c), BSA/anti-BoIFN-γ/TiO2 nanorods/GCE (curve d), and BoIFN-γ/BSA/anti-BoIFN-γ/TiO2 nanorods/GCE (e), respectively. As seen from Fig. 2, the redox currents decrease gradually with stepwise modifications on the bare electrode, suggesting that the electrode transfer process between Fe(CN)63 /4 probe and
Fig. 1. TEM image of TiO2 nanorods (a), SEM image of TiO2 nanorods (b), X-ray diffraction pattern of TiO2 nanorods (c), and anti-BoIFN-γ antibody immobilized TiO2 nanorods (d).
Z. Yang et al. / Biosensors and Bioelectronics 63 (2015) 190–195
Fig. 2. Cyclic voltammograms of bare GCE (a), TiO2 nanorods/GCE (b), antiBoIFN-γ/TiO2 nanorods/GCE (c), BSA/anti-BoIFN-γ/TiO2 nanorods/GCE (d), and BoIFN-γ/BSA/anti-BoIFN-γ/TiO2 nanorods/GCE (e) recorded in PBS solution (0.1 M, pH 7.0) including 5.0 mM K3[Fe(CN)6]/K4[Fe(CN)6] and 0.1 M KCl at a scan rate of 100 mV/s.
electrode surface was blocked. In particular, BoIFN-γ antibody and BSA immobilized electrodes produced bigger inhibition of the electron transfer, showing obvious decreases of the current of the anodic and cathonic peaks. EIS was also used to examine detailed information on the impedance changes in the modification process. The electron transfer resistance (Rct) can be quantified using the semicircle diameter in the Nyquist plot of impedance spectra. Fig. 3A displays Faradaic impedance spectra obtained upon the stepwise modification process. A very small semicircle can be observed at bare GCE,
193
indicating a low transfer resistance (curve a). After TiO2 nanorods were modified on the GCE surface, the Rct increased slightly (curve b), implying the formation of TiO2 film on the electrode surface. When BoIFN-γ antibody was loaded on the surface of TiO2 nanorods modified GCE, the Rct increased greatly (curve c), suggesting successful loading the antibody on the TiO2 nanorods. After BoIFN-γ antibody immobilized electrode was blocked with BSA, the Rct further increased (curve d). The similar increase also could be observed when the immunosensor was used to detect BoIFN-γ antigen (curve e). The possible reason is that the formed protein layer on the electrode surface significantly hinders the diffusion of ferricyanide probe toward the electrode surface. The impedance data were fitted according to a Randles equivalent circuit based on FRA software of Autolab. A modified Randles equivalent circuit and the fitting of one measured spectrum to the equivalent circuit (solid line) are shown in Fig. 3B, demonstrating the good agreement with the circuit model and the measurement system over the whole measurement frequency range. The circuit consists of the solution resistance (Rs), the electron transfer resistance (Rct), interfacial double layer capacitance (Cdl) between an electrode and a solution, and Warburg impedance (ZW). Usually, Rs and ZW represent the bulk properties of the electrolyte solution and diffusion features of the redox probe in solution, which are thus not affected by modifications on the electrode surface. Since Cdl electrode/electrolyte interface is not simple capacitance in this study, a constant phase element (CPE) was used instead of the classical capacitance to fit the impedance data in Randles equivalent circuit. The fitting value for various modification layers on the electrode are shown in Table S1 (Supporting information), indicating that Rct is a suitable signal for sensing the interfacial properties of the prepared immunosensor during all these modification steps. The relative change in Rct (%ΔRct) is calculated by the following equation:
%ʁ R ct =
R ct(Ag ℬAb) ℬ R ct(BSA) R ct(BSA)
¬ 100
where Rct(Ag Ab) is the value of the electron transfer resistance after BoIFN-γ-Ag coupling to the immobilized Ab on TiO2 nanorods modified electrode. Rct(BSA) represents the impedance value after blocking the remaining adsorption-reactive sites by BSA.
3.2. Optimization of the experimental conditions
Fig. 3. (A) Nyquist plots of bare GCE (a), TiO2 nanorods/GCE (b), anti-BoIFN-γ/TiO2 nanorods/GCE (c), BSA/anti-BoIFN-γ/TiO2 nanorods/GCE (d), and BoIFN-γ/ BSA/anti-BoIFN-γ/TiO2 nanorods/GCE (e) recorded in PBS solution (0.1 M, pH 7.0) including 5.0 mM K3[Fe(CN)6]/K4[Fe(CN)6] and 0.1 M KCl at a scan rate of 100 mV/s. The inset is the plots of (a) and (b). (B) Fitted (solid line) and experimental (scattered line) Nyquist plots of impedance spectra. The inset is the equivalent circuit applied to fit the impedance spectroscopy in the presence of the redox couple of Fe(CN)63 /4 .
The amount of antibody immobilized on the TiO2 nanorods modified GCE could affect the analytical property of the immunosensor. Thus the influence of BoIFN-γ antibody concentration was examined here (data not shown). It was found that the ΔRct value increased with the increasing antibody concentration (from 0.1 to 0.9 mg/mL), and trended a plateau at a concentration of 0.4 mg/mL. Therefore, 0.4 mg/mL of BoIFN-γ antibody concentration was chosen for the immobilization in the experiment. The capture of target proteins on the TiO2 nanorods modified GCE could change the interface properties of electrodes, thus leading to a change of Rct. Therefore, the factors affecting antibody-antigen reaction, such as pH of the solution, incubation temperature and incubation time, were optimized. Fig. S2A (Supporting information) shows the effect of solution pH on the relative change in Rct for antigen-antibody reaction at a concentration of 0.01 ng/mL BoIFN-γ. With the increasing solution pH from 4.0 to 9.0, the ΔRct value increased and trended to the maximum value at pH 7.0, indicating the optimal pH for the immunoreaction. The influence of incubation temperature for the immunoassay was also investigated in a range of 25–45 °C at a
194
Z. Yang et al. / Biosensors and Bioelectronics 63 (2015) 190–195
3.4. Specificity of the proposed BoIFN-γ immunosensor Nonspecific adsorption is usually a major problem in label-free immunosensing, and greatly affects the specificity of an immunosensor. In order to confirm that the relative changes result from specific interaction between BoIFN-γ and the BoIFN-γ antibody loaded on the electrode, nonspecific interaction between the proposed immunosensor and the analogous cytokine was examined in Fig. S3 (Supporting information). After the resultant immunosensor was incubation with 0.01 ng/mL bovine interleukin-4 (BoIL-4), the relative impedance value is very small. However, the relative impedance value remarkably increased after the resultant immunosensor was incubated with 0.01 ng/mL target BoIFN-γ. Moreover, the relative impedance value shows no obvious changes upon the addition of different concentrations of interferent BoIL-4 to 0.01 ng/mL BoIFN-γ solution, indicating that nonspecific adsorption between BoIFN-γ antibody and the other analogous analyte can be negligible. These results demonstrate that the developed label-free immunosensor has high specificity and can be used for the selective detection of BoIFN-γ. 3.5. Reproducibility, stability and accuracy of the BoIFN-γ immunosensor
Fig. 4. (A) Faradaic impedance spectra of the immunosensor after incubating with 0.0001, 0.0005, 0.001, 0.01 and 0.1 ng/mL BoIFN-γ in PBS solution (0.1 M, pH 7.0) including 5.0 mM K3[Fe(CN)6]/K4[Fe(CN)6] and 0.1 M KCl. (B) Calibration curve for BoIFN-γ (n¼5 for each point).
concentration of 0.01 ng/mL BoIFN-γ. As seen from Fig. S2B (Supporting information), the maximal relative change in Rct was observed at an incubation temperature of 37.5 °C. Additionally, the incubation time greatly affected the Ag–Ab combination, thus the effect of reaction time on the ΔRct value was examined at a concentration of 0.01 ng/mL BoIFN-γ and shown in Fig. S2C (Supporting information). With the increase of incubation time, the electrochemical response of binding reaction increased and reached a plateau beyond 80 min incubation time. To obtain the optimal analytical performance, a pH value of 7.0, incubation temperature of 37.5 °C and 80 min of incubation time were selected for the immunoassay of BoIFN-γ. 3.3. Calibration curve for BoIFN-γ Under the optimum conditions, the fabricated immunosensor was incubated with different concentrations of BoIFN-γ. Fig. 4A illustrates the corresponding Nyquist plots of impedance spectra. It could be found that the semicircle diameter in the Nyquist plot increased with the increasing concentration of the BoIFN-γ. The possible reason is that more antigen molecules are bound to the immobilized antibody in higher concentration of BoIFN-γ, resulting in bigger inhibitions of the electron transfer of the redox couple. Fig. 4B shows a linear relationship between the relative change in Rct and logarithm of BoIFN-γ in the range of 0.0001 to 0.1 ng/mL, and the linear regression equation was ΔRct(%) ¼ 43.9 þ16.6 log CBoIFN-γ (R2 ¼0.9981). The detection limit was calculated to be 0.1 pg/mL at a signal/noise ration of 3, which is 249-flod lower than that of 25 ng/mL obtained for BoIFN-γ by conventional microplate ELISA (Rothel et al., 1990).
The reproducibility of the proposed BoIFN-γ immunosensor was evaluated by the intra- and inter-assay coefficients of variation (CV). The intra-assay CV at a concentration of 0.01 ng/mL BoIFN-γ for six replicative measurements was 3.7%, demonstrating good detection reproducibility. The inter-assay CV at a concentration of 0.01 ng/mL BoIFN-γ for six immunosensors fabricated in six batches was 6.3%, indicating good fabrication reproducibility. The immunosensor could be stored in a refrigerator at 4 °C when it was not in use. The results showed that the relative change in Rct retained no obvious change after a storage period of 60 days. Evidently, the prepared TiO2 nanorods can provide a good microenvironment, which is efficient to retain the bioactivity of the immobilized antibody molecules. To evaluate the accuracy and application potential of the developed electrochemical impedance BoIFN-γ immunosensor, the recovery was tested by spiking 0.0002, 0.0005, 0.001, 0.005 and 0.01 ng/mL BoIFN-γ in bovine serum samples. The obtained recoveries were shown in Table 1, demonstrating good accuracy of the fabricated immunosensor for label-free detection of BoIFN-γ in practical samples.
4. Conclusions In this work, a first electrochemical impedance immunosensor was developed for label-free and highly sensitive determination of BoIFN-γ based on cylinder-shaped TiO2 nanorods, which was synthesized by a facile hydrothermal method. The efficient immunosensor was fabricated by immobilizing BoIFN-γ monoclonal antibody on the TiO2 nanorods modified GCE. The as-prepared Table 1 Recoveries for BoIFN-γ by the proposed electrochemical impedance immunosensor (n ¼ 5). Sample
Added (ng/mL)
Found (ng/mL)
Recovery (%)
RSD (%)
1 2 3 4 5
0.0002 0.0005 0.001 0.005 0.01
0.00022 0.00054 0.00109 0.00510 0.00990
110 108 109 102 99
1.6 2.1 2.7 3.5 1.5
Z. Yang et al. / Biosensors and Bioelectronics 63 (2015) 190–195
TiO2 nanorods provide large surface area and friendly microenvironment for the immobilization of protein biomolecules while retaining the biological activity of the immunosensor. Compared with traditional biological assay and ELISA, this strategy shows ultrahigh sensitivity, simple manipulation and wide linear range. The resultant BoIFN-γ immunosensor also displayed excellent specificity, good reproducibility and acceptable stability. The recovery experiment demonstrates that the proposed immunosensor has high accuracy and can be applied in the detection of BoIFN-γ in practical samples. This study opens new avenues for highly sensitive detection of BoIFN-γ and further early diagnosis of MTB infections.
Acknowledgments This work was financially supported by National Natural Science Foundation of China (21275124 and 21103148), National Basic Research and Development Program of China (2012CB518805), The Priority Academic Program Development of Jiangsu Higher Education Institution, Postdoctoral Science Foundation of Jiangsu Province (1101020B) and University Natural Science Foundation of Jiangsu Province (13KJB150039). Also, thanks to The Testing Center of Yangzhou University for the characterization data.
Appendix A. Supplementary information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.bios.2014.07.032.
References Andersen, P., Munk, M.E., Pollock, J.M., Doherty, T.M., 2000. Lancet 356, 1099–1104. Bao, S.J., Li, C.M., Zang, J.F., Cui, X.Q., Qiao, Y., Guo, J., 2008. Adv. Funct. Mater. 18, 591–599. Bart, M., Stigter, E.C. A., Stapert, H.R., de Jong, G.J., van Bennekom, W.P., 2004. Biosens. Bioelectron. 21, 49–59. Billiau, A., 1996. Adv. Immunol. 62, 61–130. Chen, H., Jiang, J.H., Huang, Y., Deng, T., Li, J.S., Shen, G.L., Yu, R.Q., 2006. Sens. Actuators B 117, 211–218.
195
Chen, X.J., Wang, Y.Y., Zhou, J.J., Yan, W., Li, X.H., Zhu, J.J., 2008. Anal. Chem. 80, 2133–2140. Derkus, B., Emregul, E., Yucesan, C., Emregul, K.C., 2013. Biosens. Bioelectron. 46, 53–60. Dijksma, M., Kamp, B., Hoogvliet, J.G., van Bennekom, W.P., 2001. Anal. Chem. 73, 901–907. Dong, H.F., Lei, J.P., Ju, H.X., Zhi, F., Wang, H., Guo, W.J., Zhu, Z., Yan, F., 2012. Angew. Chem. Int. Ed. 51, 4607–4612. Du, D., Zou, Z.X., Shin, Y.S., Wang, J., Wu, H., Engelhard, M.H., Liu, J., Aksay, I.A., Lin, Y.H., 2010. Anal. Chem. 82, 2989–2995. Elshafey, R., Tlili, C., Abulrob, A., Tavares, A.C., Zourob, M., 2013. Biosens. Bioelectron. 39, 220–225. Farrar, M.A., Schreiber, R.D., 1993. Annu. Rev. Immunol. 11, 571–611. Grossi, M., Lanzoni, M., Pompei, A., Lazzarini, R., Matteuzzi, D., Ricco, B., 2008. Biosens. Bioelectron. 23, 1616–1623. Jiang, H.D., Kim, S.K., Chang, H., Roh, K.M., Choi, J.W., Huang, J.X., 2012. Biosens. Bioelectron. 38, 184–188. Khan, R., Dhayal, M., 2008. Electrochem. Commun. 10, 492–495. Koo, B., Park, J., Kim, Y., Choi, S.H., Sung, Y.E., Hyeon, T., 2006. J. Phys. Chem. B 110, 24318–24323. Lai, G.S., Yan, F., Ju, H.X., 2009. Anal. Chem. 81, 9730–9736. Li, H.B., Li, J., Yang, Z.J., Xu, Q., Hu, X.Y., 2011. Anal. Chem. 83, 5290–5295. Li, X.H., Dai, L., Liu, Y., Chen, X.J., Yan, W., Jiang, L.P., Zhu, J.J., 2009. Adv. Funct. Mater. 19, 3120–3128. Li, X.M., Yang, X.Y., Zhang, S.S., 2008. Trends Anal. Chem. 27, 543–553. Liang, H.T., Li, X.Z., 2009. J. Hazard. Mater. 162, 1415–1422. Liu, Y., Tuleouva, N., Ramanculov, E., Revzin, A., 2010. Anal. Chem. 82, 8131–8135. Montrose, A., Cargou, S., Nepveu, F., Manczak, R., Gue, A.M., Reybier, K., 2013. Biosens. Bioelectron. 49, 305–311. Qi, W.J., Huang, C.Z., Chen, L.Q., 2010. Talanta 80, 1400–1405. Rothel, J.S., Jones, S.L., Corner, L.A., Cox, J.C., Wood, P.R., 1990. Aust. Vet. J. 67, 134–137. Tu, W.W., Dong, Y.T., Lei, J.P., Ju, H.X., 2010. Anal. Chem. 82, 8711–8716. Tu, W.W., Lei, J.P., Ding, L., Ju, H.X., 2009. Chem. Commun. 28, 4227–4229. Tu, W.W., Lei, J.P., Wang, P., Ju, H.X., 2011. Chem. Eur. J 17, 9440–9447. Walravens, K., Wellemans, V., Weynants, V., Boelaert, F., deBergeyck, V., Letesson, J. J., Huyen, K., Godfroid, J., 2002. Vet. Immunol. Immunopathol. 84, 29–41. Wang, L.S., Lei, J.P., Ma, R.N., Ju, H.X., 2013a. Anal. Chem. 85, 6505–6510. Wang, Y.X., Ping, J.F., Ye, Z.Z., Wu, J., Ying, Y.B., 2013b. Biosens. Bioelectron. 49, 492–498. Wang, Z.Y., Xia, D.G., Chen, G., Yang, T., Chen, Y., 2008. Mater. Chem. Phys. 111, 313–316. Wood, P.R., Corner, L.A., Plackett, P., 1990a. Res. Vet. Sci. 49, 46–49. Wood, P.R., Rothel, J.S., McWaters, P.G., Jones, S.L., 1990b. Vet. Immunol. Immunopathol. 25, 37–46. Wu, Y.F., Chen, C.L., Liu, S.Q., 2009. Anal. Chem. 81, 1600–1607. Wu, Y.F., Liu, S.Q., He, L., 2011. Analyst 136, 2558–2563. Xu, X., Wei, W., Huang, M.H., Yao, L., Liu, S.Q., 2012. Chem. Commun. 48, 7802–7804. Yang, Z.J., Tang, Y., Li, J., Zhang, Y.C., Hu, X.Y., 2014. Biosens. Bioelectron. 54, 528–533. Yuan, L., Hua, X., Wu, Y.F., Pan, X.H., Liu, S.Q., 2011. Anal. Chem. 83, 6800–6809. Zhang, J., Lei, J.P., Xu, C.L., Ding, L., Ju, H.X., 2010. Anal. Chem. 82, 1117–1122. Zhang, S., Liu, C.Y., Liu, Y., Zhang, Z.Y., Mao, L.J., 2009. Mater. Lett. 63, 127–129. Zhu, A.P., Zhang, M., Wu, J., Shen, J., 2002. Biomaterials 23, 4657–4665.
