Electrochemical Immunosensor Based on Polythionine/Gold Nanoparticles for the Determination of Aflatoxin B1

An aflatoxin B1 (AFB1) electrochemical immunosensor was developed by the immobilisation of aflatoxin B1-bovine serum albumin (AFB1-BSA) conjugate on a polythionine (PTH)/gold nanoparticles (AuNP)-modified glassy carbon electrode (GCE). The surface of the AFB1-BSA conjugate was covered with horseradish peroxidase (HRP), in order to prevent non-specific binding of the immunosensors with ions in the test solution. The AFB1 immunosensor exhibited a quasi-reversible electrochemistry as indicated by a cyclic voltammetric (CV) peak separation (ΔEp) value of 62 mV. The experimental procedure for the detection of AFB1 involved the setting up of a competition between free AFB1 and the immobilised AFB1-BSA conjugate for the binding sites of free anti-aflatoxin B1 (anti-AFB1) antibody. The immunosensor's differential pulse voltammetry (DPV) responses (peak currents) decreased as the concentration of free AFB1 increased within a dynamic linear range (DLR) of 0.6 - 2.4 ng/mL AFB1 and a limit of detection (LOD) of 0.07 ng/mL AFB1. This immunosensing procedure eliminates the need for enzyme-labeled secondary antibodies normally used in conventional ELISA–based immunosensors.


Introduction
Aflatoxin B 1 (AFB 1 ) is an example of a group of highly toxic difurancoumarin derivatives that are produced by many strains of Aspergillus flavus and A. parasiticus which often contaminate a variety of food and animal feed stored under temperate and humid conditions favourable to mould growth [1]. The four major aflatoxins have been designated as B 1 , B 2 , G 1 and G 2 based on their fluorescence under UV light and their relative chromatographic mobility during thin layer chromatography. AFB 1 has been classified as a group 1 human carcinogen and aflatoxins G 1 , G 2 and B 2 belong to a group human carcinogens [2]. These toxins exhibit carcinogenic, teratogenic and mutagenic properties and have now been isolated from a wide variety of agricultural products [3]. AFB 1 can enter the food chain mainly through the ingestion of contaminated human or animal food. The intake of AFB 1 over a long period of time, even in low concentrations, may be very deleterious to health [4]. The Food and Agricultural Organization 2004 report on mycotoxins [5] revealed that as of December 2003, at least 99 countries worldwide had regulations in place for permitted mycotoxin levels in food/or feed, and have set limits for AFB 1 alone or for the sum of aflatoxins B 1 , B 2 , G 1 and G 2 . The maximum permissible level for AFB 1 in food was set at 2 µg/kg (2 ppb).
Many analytical methods have been developed for the determination of aflatoxins. These include thin-layer chromatography (TLC) [6] and high-performance liquid chromatography (HPLC) [7]. Though these techniques have excellent sensitivities they typically require skilled operators, extensive sample pre-treatment and expensive equipment [8]. The goal of more recent studies has been to simplify and expedite the method of detection while attempting to maintain or improve the sensitivity. Among the immunochemical approaches, the enzyme-linked immunosorbent assay (ELISA) method is the most widely applied. Spectrophotomeric ELISAs specific for AFB 1 [9,10], total aflatoxins [11,12] and AFM 1 [13,14] have been developed and their simplicity, adaptability and sensitivity have been demonstrated. In order to achieve higher sensitivity and move to the use of disposable probes, electrochemical immunosensors for aflatoxins based on indirect competitive ELISA format have been proposed [15][16][17]. These immunosensors require the use of labeled secondary antibodies for detection. To achieve label-free immunosensors, direct electrochemical immunosensors for AFB 1 based on electrochemical impedance spectroscopy [18], optical waveguide lightmode spectroscopy [19] and room temperature ionic liquids [20] have been reported.
The search for a simple and label-free amperometric immunosensor is of considerable interest. Among the various conducting polymers, thionine (phenothiazine) is a redox dye which has been studied extensively due to its potential utility in sensor applications [21,22]. Its electroactivity lies not only in the heterocyclic nitrogen atoms and nitrogen bridges, but also in its free amine groups [23]. In addition polythionine (PTH) can be easily functionalised due to the abundant amino groups which adsorb metal ions and various organic halogen substances, thus preventing proteins from damage [24].On the other hand, gold nanoparticles (AuNPs) have been extensively used as matrices for the immobilization of macromolecules such as proteins, enzymes and antibodies; as well as chemical labels for biomolecules [25][26][27]. Modification of electrode surfaces with AuNPs provides a microenvironment similar to what obtains under physiological conditions [28]. In this investigation, an electrochemical immunosensor for the detection of AFB 1 was developed by the drop-coating of AuNPs on PTH-modified glassy carbon electrode (GCE) surface. Subsequently AFB 1 -conjugate was adsorbed on to the gold nanoparticles surface. Details of the preparation, characterization and application of the immunosensor are described.

Reagents and materials
Analytical reagent grade chemicals from Sigma-Aldrich were used in all experiments. Phosphate buffer saline (PBS) solution pH 7.2 contained 0.1 M KH 2 PO 4 , 0.1 M Na 2 HPO 4 , 2.7 mM KCl and 0.137 M NaCl. Acetate buffer (pH 6.5) was prepared from 0.1 M CH 3 COONa, 0.1 M CH 3 COOH and 0.1 M KCl. 1 mg/mL Aflatoxin B 1 (AFB 1 ) solution was prepared by dissolving AFB 1 from Aspergillus flavus in methanol followed by dilution in PBS/10% methanol to give a series of standard solutions with a concentration range of 0.1-3.0 ng/mL. The antibody reagent was an immunoglobin (Ig) fraction of rabbit antiserum AFB 1 (anti-AFB 1 ) antibody that contained 6.8 mg/mL of total protein (which has reactivity with aflatoxins B 1 , G 1 and B 2 , but no cross reactivity with B 2a , G 2 , G 2a or M 1 ) and 0.15 M NaN 3 as preservative. The anti-AFB 1 antibody solution for immunosensing reactions was prepared by diluting the stock 6.8 mg/mL solution to 1:2000 v/v in PBS and stored at -20 o C when not in use. Aflatoxin B 1 -BSA conjugate (8-12 mol AFB 1 per mol BSA), horseradish peroxidase (HRP; 169 units/mL lyophilised powder; EC.1.11.1.7; M r 4000), thionine acetate (3,7-diamino-5-phenothiazinium acetate; C 12 H 9 N 3 S.C 2 H 4 O 2 ; M r 287.34), tetrachloro-auric acid (HAuCl 4 .3H 2 0; M r 393.83), sodium citrate and 30% hydrogen peroxide solution were other reagents used in the studies. Colloidal gold nanoparticles (AuNP) (diameter = 20 nm) were prepared according to the procedure of Yuan et a., 2004 [29] by adding 2 mL of 1% (w/w) sodium citrate to a boiling solution of 50 mL 0.01% (w/w) tetrachloro-auric acid. The AuNP solution was stored in a refrigerator in a dark-colored glass bottle. The production of colloidal AuNP was confirmed by UV-Vis measurement covering 200 -700 nm at room temperature using distilled water as the reference, which gave a maximum absorption at 520 nm. The particle size of the AuNP was determined with transmission electron microscopy (TEM), in which a sample of the colloidal AuNP was dropped on a carbon-coated copper grid and left to dry for 24 h, after which TEM images were recorded and analysed.

Instrumentation
All electrochemical experiments were performed with a BAS/50W integrated automated electrochemical workstation (Bioanalytical Systems Lafayette, IN, USA). Cyclic voltammetry and differential pulse voltammetry experiments were performed with Ag/AgCl (3 M NaCl type) and platinum wire as reference and auxiliary electrodes, respectively. A BAS 3-mm diameter GCE in bare or modified form, was used as the working electrode. Transmission electron microscopy studies were carried out with JEOL JEM-1200 EX II electron microscope. UV-Vis experiments were performed with GBC UV/Vis 920 spectrophotometer (GBC Scientific Instruments, Australia).

Immunosensor preparation
Before use the GCE was polished consecutively with 1.0, 0.3 and 0.05 micron aqueous slurry of alumina micropolish (Buehler, IL, USA), followed by sonication in double-distilled water and ethanol for 5 min and dried in air. 0.1 mM thionine was polymerised on the GCE by cyclic voltammetry (CV) in a 20-voltammetric cycle experiment covering a potential window of -400 to +1200 mV at a scan rate of 50 mV/s and a sensitivity of 0.001 A/V to produce the required PTH. Freshly prepared PTH film was drop-coated with 2 µL of colloidal AuNP and allowed to dry for 24 h. Subsequently the electrode was coated with 5 µL of AFB 1 -BSA conjugate (1 µg/mL) and incubated for 1 h at 37 o C. The resultant immunosensor was incubated in 1 mg/mL HRP contained in acetate buffer (pH 6.5) for 60 min at 4 o C in order to block any remaining active sites of the AuNP layer and avoid non-specific adsorptions. The immunosensor was stored at 4 o C when not in use. The schematic illustration of the stepwise immunosensor assembly procedure is shown in Scheme 1.

Scheme 1.
Reaction scheme for the preparation of aflatoxin B 1 immunosensor.

Procedure for electrochemical immunosensing
Competitive immunosensing was performed in the absence and presence of free AFB 1 . Experiment in the absence of free AFB 1 (i.e. 0.0 ng/mL AFB 1 ) were performed by placing 5 µL of anti-AFB 1 antibody on top of the immunosensor (i.e. GCE|AuNP|PTH|AFB 1 -BSA-conjugate HRP) and allowed to react for 15 min. The electrode system was immersed in a 1 mL cell solution containing acetate buffer pH 6.5 and 3.2 µM H 2 O 2 and DPV measurement was performed by scanning cathodically from 0 to -480 mV at 10 mV pulse amplitude, 50 ms pulse width and 20 mV/s potential scan rate. For experiments in the presence of free AFB 1 (i.e. 0.6 -3.0 ng/mL AFB 1 ), 10 µL of anti-AFB 1 antibody solution was mixed with 10 µL of 0.6, 1.2, 1.8, 2.4 or 3.0 ng/mL AFB 1 solutions. 5 µL of this mixture was placed on top of the immunosensor and allowed to react for 15 min. This was followed by DPV measurement as described above for 0.0 ng/mL AFB 1 .

Characterisation of gold nanoparticles
UV-Vis spectrum of the colloidal gold nano-particles prepared by the method of Yuan et al. [29] showed a maximum absorption, λ max , at 520 nm ( Figure 1). The formation of nano-particles and the particle size (20 nm diameter) were confirmed by the TEM data ( Figure 2). These results agreed with what has been reported for colloidal AuNP [29].

Electropolymerization of multiporous thionine film
The GCE was first reduced in acetate buffer solution (pH 6.5) containing thionine at -1500 mV to make it negatively charged and interact with positively charged thionine in the mildly acidic conditions. Subsequently cyclic voltammetry was performed at a bias voltage of -400 to +1200 mV at a scan rate of 50 mV/s. When the applied potential exceeded +1100 mV, the electropolymerization reaction proceeded and formed cationic radical species on the GCE [30]. This process resulted in a multiporous structure which facilitated the assembly of AuNP. The oxidation potential in the first step was the most important factor for electropolymerization of thionine and should not be less than +1100 mV. Two reasons suffice for this. Firstly in order to achieve the formation of polythionine (PTH) film, the electrode potential must be larger than the potential at which the oxidations of -NH 2 groups of thionine molecule occurs, and secondly the modified cation must be necessarily associated with the surface activation of GCE. During thionine electropolymerization, a pair of quasi reversible redox peaks with a cathodic peak potential (E pc ) of -250 mV and an anodic peak potential E pa of -100 mV, increased gradually and tended to become stable with increasing number of polymerization cycles (Figure 3a). On removal of the electrode from the dye-containing solution, a golden film was seen on the electrode surface. The CV of the GCE|PTH in acetate buffer (pH 6.5) (Figure 3b) confirmed the presence of surface-bound electroactive material. Figure 4 shows the scan rate dependence of the electrochemistry of GCE|PTH and GCE|PTH|AuNP|AFB 1 -BSA-conjugate electrodes in acetate buffer pH 6.5. The peak currents varied linearly with scan rate as is characteristic of the electrochemistry of a surface-bound thin film of electroactive material [31]. It is thus suggested that nearly all the reduced polythionine (PTH (red) ) was converted to the oxidized polythionine (PTH (ox) ) on the forward scan and vice versa.

Electrochemical characteristics of the electrode
The stepwise assembly of the immunosensor was monitored cyclic-voltammetrically in acetate buffer pH 6.5 and the results are shown in Figure 5. The CV of the bare GCE is shown as Figure 5a. The bare GCE did not exhibit any redox chemistry over the potential window (-500 to 0 mV) used in the study. The GCE|PTH gave a quasi reversible electrochemistry with a peak separation of ~62 mV at 10 mV/s (Figure 5b). The redox couple can be ascribed to the electrochemistry of the PTH film on the GCE. The peak currents increased after the modification of GCE|PTH with AuNP (Figure 5e). The reason for this increase is that nanometer-sized particles of colloidal gold behave like a conducting wire or electron conducting tunnel, which makes it easier for the electron transfer to take place. However, the peak currents decreased (Figure 5d) after AFB 1 -BSA conjugate was incorporated to produce GCE|PTH|AuNP|AFB 1 -BSA-conjugate, which indicated that the relatively insulating AFB 1 -BSA conjugate was immobilized on the electrode surface. The subsequent blocking of possible active sites with HRP to form GCE|PTH|AuNP|AFB 1 -BSA-conjugate|HRP immunosensor system decreased the CV peak current due to the adsorption of the protein, HRP (Figure 5c). Surface concentration (Γ * ) values of 7.06 x 10 -11 and 9.65 x 10 -10 mol/cm 2 were calculated for GCE|PTH and GCE|PTH|AuNP|AFB 1 -BSA-conjugate electrode systems, respectively.

Assay of the HRP enzymatic catalytic activity
Horseradish peroxidase was used in the preparation of the immunosensors to block possible uncovered active sites on the GCE|AuNP|PTH|AFB 1 -BSA-conjugate in order to avoid non-specific adsorption. Experiments were performed with hydrogen peroxide to ascertain if HRP was indeed immobilized on the electrode. Figure 6 depicts the CV's of the proposed immunosensor in the presence and absence of H 2 O 2 . A redox couple representing the electrochemistry of PTH was observed in the absence of H 2 O 2 ( Figure 6a). However, an enhancement of the cathodic peak and a concomitant decrease of the anodic peak current were observed in the presence of 3.2 µM H 2 O 2 (Figure 6b). It can thus be concluded that HRP was attached to the immunosensor surface and it retained its enzymatic catalytic activity [21], which is coupled to the PTH electron transfer process shown below.
The optimal amount of H 2 O 2 required for the immunosensing reaction with GCE|AuNP|PTH|AFB 1 -BSA-conjugate|HRP-blocked electrode system was determined from steadystate amperometry experiments performed at -175 mV. As shown in Figure 7, the sensor gave a maximum response at 3.2 µM H 2 O 2 . At H 2 O 2 concentrations higher than 3.2 µM, the sensor response decreased owing to the irreversible transition of the immobilized HRP to its inactive form [32]. As a result, 3.2 µM H 2 O 2 was chosen for the immunosensing experiments.  The effect of pH on the immunosensor was studied for pH's 4.5 to 6.5 at 25 o C. In order to retain the bioactivity of both AFB 1 -BSA conjugate and HRP, pH 6.5 was chosen as the optimal pH. Although 37 o C is the best incubation temperature for antibody-antigen reactions, immunoproteins and HRP do not maintain their activities for a long time at this temperature. As a result 25 o C was used in all experiments.

Performance of the immunosensor
The responses of the immunosensor to AFB 1 were recorded with DPV as shown in Figure 8 for 0 -3 ng/mL AFB 1 . As expected, the peak current was inversely proportional to the analyte concentration. The detection principle is based on the inhibition of the active centre of HRP by the formation of antigen-antibody complex. The observed current is attributed to the catalytic response of the adsorbed HRP to H 2 O 2 . The formation of the antigen-antibody complex introduces a local current change at the sites of the adsorbed HRP due to the inhibition of the H 2 O 2 reduction reaction. The attenuation of the DPV response is dependent on AFB 1 concentration as shown in Figure 8 and it is attributable to the increase in electron transfer resistance of adsorbed HRP [33] caused by the insulating properties of the complex formed by the binding of AFB 1 -BSA conjugate and anti-AFB 1 antibody.  The calibration graph of the AFB 1 immunosensor (Figure 9) plotted from DPV results gave DLR, sensitivity and LOD values of 0.6 -2.4 ng/mL, 1.23 x 10 -6 A/(ng/mL) and 0.07 ng/mL, respectively. The DPV measurements were carried out in triplicates. The relative standard deviations of the DPV responses were 3.1%, 2.9%, 2.4% and 5.1% for 0.6, 1.2, 1.8 and 2.4 ng/mL, respectively, indicating acceptable level of precision.

Conclusions
The DLR and LOD of the immunosensor were compared with those of other electrochemical AFB 1 immunosensors [15,18,20] as shown in Table 1. The values of these sensor parameters for the PTHbased immunosensor are in good agreement with those reported by other laboratories. In addition, the sensor exhibited high sensitivity and good reproducibility. These characteristics of the immunosensor show that it can be used to screen food products for AFB 1 , since the DLR and LOD values cover the FAO's limit of 2 ppb AFB 1 in food samples [5]. The immunosensing procedure reported in this study eliminated the requirement of secondary labeled antibodies as is the case in conventional electrochemical immunosensors based on ELISA techniques.