Construction of Electrochemical Immunosensor Based on Gold-Nanoparticles/Carbon Nanotubes/Chitosan for Sensitive Determination of T-2 Toxin in Feed and Swine Meat

T-2 toxin (T-2) is one of major concern mycotoxins acknowledged as an unavoidable contaminant in human foods, animal feeds and also agriculture products. Thus, a facile and sensitive method is essential for accurate T-2 toxin detection. In our work, a specific electrochemical immunosensor based on gold nanoparticles/carboxylic group-functionalized single-walled carbon nanotubes/chitosan (AuNPs/cSWNTs/CS) composite was established. The mechanism of the electrochemical immunosensor was an indirect competitive binding to a given amount of anti-T-2 between free T-2 and T-2-bovine serum albumin, which was conjugated on covalently functionalized cSWNTs decorated on the glass carbon electrode. Afterwards, the alkaline phosphatase labeled anti-mouse secondary antibody was bound to the electrode surface by reacting with the primary antibody. Lastly, alkaline phosphatase catalyzed the hydrolysis of the substrate α-naphthyl phosphate, which produced an electrochemical signal. Compared with conventional methods, the established immunosensor was more sensitive and simpler. Under optimal conditions, this method could quantitatively detect T-2 from 0.01 to 100 μg·L−1 with a detection limit of 0.13 μg·L−1 and favorable recovery 91.42–102.49%. Moreover, the immunosensor was successfully applied to assay T-2 in feed and swine meat, which showed good correlation with the results obtained from liquid chromatography-tandem mass spectrometry (LC-MS/MS).


Introduction
Species of the genus Fusarium embody the most destructive fungal plant pathogens and are responsible for major yield losses during the cultivation of wheat, maize, barley, and soybeans [1,2]. T-2 toxin is one of major concern mycotoxins, a group of type A trichothecenes secondary metabolite produced by several fungalgenera, including Fusarium fungi, when they infect grains, especially oats and wheat [3][4][5]. Reportedly, it also has been acknowledged as an unavoidable contaminant in human foods, animal feeds and agricultural products such as maize, wheat and oats among many regions and countries [6]. An acute exposure to high-dose trichothecenes triggers various clinical signs for animals such electrochemical impedance spectroscopy (EIS), optimized the conditions of T-2 detection by difference pulse voltammetry (DPV) and lastly determined the spiked samples by this method. The aim of this study is to provide a sensitive and selective analytical method for quantitative detection of T-2 in feed and animal-derived food.

Identification of the Hapten and Antigen
The hapten was identified by ion trap and time-offlight mass spectrometry coupled with an HPLC system (LC/MS-IT-TOF, Shimadzu, Kyoto, Japan). The calculated m/z for C28H38O12 was M = 566.58; for [M − H] − the m/z was 565.2407 which showed that almost all the T-2 toxin was converted to T-2HS ( Figure S1). The synthetic antigen was identified with an 8453 UV-Visible spectrophotometer. The results showed that the ultraviolet absorbance spectra of T-2HS-OVA (λmax, 278 nm) was different from that of OVA (λmax, 279 nm), which revealed that the antigen was successfully prepared ( Figure S2).

SEM and AFM Characterization
In the research, the morphological features of modified electrodes were characterized with the scanning electron microscopy (SEM; Figure 1A,B) and atomic force microscopy (AFM; Figure 1C,D). As shown in Figure 1A, the surface of electrodes was dotted with AuNPs uniformly. When modified with cSWNTs/CS, it seemingly shaped a three-dimensional network structure on the GCE surface ( Figure 1B). Those characteristics provided a film surface area with great biocompatiblity [31]. The analytical results of AuNPs/cSWNTs/CS and T-2-OVA-cSWNTs/CS/AuNPs by AFM colored graphic were presented in Figure 1C,D, where the location of all kinds of elements were marked by corresponding pixels. As the AFM image of Figure 1D showed, after the immobilization of carboxylated SWNTs/CS/AuNPs onto the electrode surface by diimide activation, several changes happened in the superficial morphology. When comparing the AFM image of T-2-OVA-cSWNTs/CS/AuNPs and that of AuNPs/cSWNTs/CS, the characterized result of AFM showed a tidy and closely image in Figure 1D, implying that T-2-OVA was successfully immobilized onto the cSWNTs/CS/AuNPs film.

Characterization of the Immunosensor
In this research, both cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS) methods were adopted to test the interface features of the stepwise modified immunosensor. Figure

Characterization of the Immunosensor
In this research, both cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS) methods were adopted to test the interface features of the stepwise modified immunosensor. Figure 2A showed the CV results of different fabricated electrodes performed in 5 mM [Fe(CN)6] 3−/4− containing 1 M KCl at 50 mV/s scan rate. When the AuNPs were deposited onto the GCE surface, the redox peak current went up dramatically (curve b), indicating that the AuNPs can accelerate electron transfer speed. After the cSWNTs were dropped on the electrode surface, the redox peak current continuously grew (curve c) because the cSWNTs could promote electron transfer and supplied a larger surface to enrich the loading amount of T-2-OVA. In sequence, when T-2-OVA was dropped onto the activated electrode surface, apparently the peak current decreased owing to the steric hindrance and an obstacle of insulated T-2-OVA for the electron transfer, explaining that T-2-OVA was successfully immobilized on the SWNTs/CS/GCE film surface (curve d). After the immunosensor was incubated with anti-T-2, a further decrease in peak current was produced for the increasing of electron transfer resistance (curve e), implying that we successfully obtained the fabricated electrochemical biosensor for specific recognition and binding with anti-T-2. Lastly when we dropped the diluted ALP-anti-antibody onto the surface of the established electrode, accordingly the redox peak current decreased, further implying that the ALP-anti-antibody was immobilized to the immunosensor successfully. In addition, to further research the properties of the desirous immunosensor, we also utilized the procedure of EIS to produce the Nyquist plots composed with a semicircle portion at higher frequencies level corresponding to the electrontransfer-limited process and a linear part at the lower frequency range corresponding to the diffusion-limited process to characterize the whole process of preparing modified electrodes [32,33]. Figure 2B [32][33][34][35]. However, a stepwise growth in semicircle diameters corresponding with the increased resistances happened after the sequential fabricating of T-2-OVA, Anti-T-2 and ALP-anti-anti-body (curves d, e and f separately). The reason for that was the protein layer acting as the inert electron, while the mass-transfer blocking layer significantly deterred the diffusion of ferricyanide towards the electrode surface [24,28].

Sensor Optimization
In order to achieve the best analytical property for T-2, we optimized several experimental key parameters. As displayed in Figure 3A, we recorded the peak current response, which increased to a In addition, to further research the properties of the desirous immunosensor, we also utilized the procedure of EIS to produce the Nyquist plots composed with a semicircle portion at higher frequencies level corresponding to the electrontransfer-limited process and a linear part at the lower frequency range corresponding to the diffusion-limited process to characterize the whole process of preparing modified electrodes [32,33]. Figure 2B [32][33][34][35]. However, a stepwise growth in semicircle diameters corresponding with the increased resistances happened after the sequential fabricating of T-2-OVA, Anti-T-2 and ALP-anti-anti-body (curves d, e and f separately). The reason for that was the protein layer acting as the inert electron, while the mass-transfer blocking layer significantly deterred the diffusion of ferricyanide towards the electrode surface [24,28].

Sensor Optimization
In order to achieve the best analytical property for T-2, we optimized several experimental key parameters. As displayed in Figure 3A, we recorded the peak current response, which increased to a stable value at the concentration of immobilized T-2-OVA of 10 µg·mL −1 . Therefore, we selected 10 µg·mL −1 as the optimal concentration. When the amounts of T-2-OVA domains on the electrode surface was fixed, the concentration of anti-T-2 in the reacting solution was another crucial influencing factor because the competitive immunoreaction exists between the fixed antigen domains and the free T-2 in solution to bind to anti-T-2. For getting the optimal concentration of anti-T-2, it was necessary to incubate the fabricated T-2-OVA-AuNPs/SWNTs/CS/GCE in different concentrations of anti-T-2 solutions. As displayed in Figure 3B, the signal response rose highest at 5 µg·mL −1 . Therefore, we chose 5 µg·mL −1 as the performed concentration in this assay. concentration of α-NP in DEA solution increasing. As the concentration of α-NP was over 0.75 μg·mL −1 , the further increase of response signal was not obvious, so 0.75 μg·mL −1 was adopted as the final work concentration.

Linear Range and LOD
To investigate the performance of this electrochemical immunoassay, different concentrations of T-2 accurately prepared with PBS (pH 7.4) were tested, and the resulting DPV responses were shown in Figure 4A. The peak current values exhibited a linear decrease with the logarithm of T-2 concentrations from 0.01 to 100 μg·L −1 , the correlation coefficient was 0.9986 ( Figure 4B), and the error bars represented standard deviations of three independent measurements. In addition, the dilution ratio of ALP-anti-antibody greatly affected the enzymatic reaction rate. In Figure 3C, the DPV response gradually increased with the reducing dilution ratio of ALP-anti-antibody from 1:3200 to 1:200 and gradually trended to level off at 1:800. Thus 1:800 was used in the experiment as the optimal ALP-anti-antibody ratio. At last, the concentration of substrate α-NP would be another important parameter influencing the enzyme-catalyzed reaction when the dilution ratio of ALP-anti-antibody was regular [36]. In Figure 3D, the signal enhanced continuously with the concentration of α-NP in DEA solution increasing. As the concentration of α-NP was over 0.75 µg·mL −1 , the further increase of response signal was not obvious, so 0.75 µg·mL −1 was adopted as the final work concentration.

Linear Range and LOD
To investigate the performance of this electrochemical immunoassay, different concentrations of T-2 accurately prepared with PBS (pH 7.4) were tested, and the resulting DPV responses were shown in Figure 4A. The peak current values exhibited a linear decrease with the logarithm of T-2 concentrations from 0.01 to 100 µg·L −1 , the correlation coefficient was 0.9986 ( Figure 4B), and the error bars represented standard deviations of three independent measurements.

Linear Range and LOD
To investigate the performance of this electrochemical immunoassay, different concentrations of T-2 accurately prepared with PBS (pH 7.4) were tested, and the resulting DPV responses were shown in Figure 4A. The peak current values exhibited a linear decrease with the logarithm of T-2 concentrations from 0.01 to 100 μg·L −1 , the correlation coefficient was 0.9986 ( Figure 4B), and the error bars represented standard deviations of three independent measurements. The LOD of this sensor was calculated using a linear regression curve based on the method reported in literature [33,37]: LOD = 3 × Sa/b. For this method, Sa is the standard deviation of the intercept and b is the slope of the calibration curve. Using the linear regression y = −54.838x + 137.57 presented in Figure 4B, the calculated LOD values for T-2 was 0.14 µg·L −1 (linear range: 0.01 to 100 µg·L −1 , R 2 = 0.9986).
As compared in Table 1, the LOD value was lower and the obtained linear range of toxin concentration was wider than those from other reported tests, such as ELISA and LC-MS/MS [30,38,39], further highlighting the merits of the suggested sensor.

Specificity, Reproducibility and Stability
The specificity of the immunosensors was tested by examining 10 µg·L −1 T-2 toxin and other three mycotoxins including Deoxynivalenol (DON), Nivalenol (NIV), Neosolaniol (NEO) respectively. As displayed in Figure S3, the DPV signal values of the three different mycotoxins were similar to that of the blank (without T-2). On the contrary, the DPV signal of T-2 showed a sharp reduction and the error bars represented the standard deviation for three measurements, which revealed that the assembled sensor could effectively analyze the target toxin owing to the specific recognition mechanism between antigen and antibody.
To evaluate the reproducibility of this developed immunoassay, carried out five reduplicative tests of T-2 at 0.1 µg·L −1 and 100 µg·L −1 with the same batch of immunosensors. The coefficients of variation were 5.73% and 8.26%, separately implying that the designed immunosensor using AuNPs/SWNTs/CS as a signal amplification strategy confirmed good reproducibility.
The stability was evaluated by recording DPV current signals of T-2 toxin after storing the electrode at 4 • C for up to two weeks. During the first week, the current response showed a slight change; however, a somewhat larger decrease was observed during the next week ( Figure S4). Overall, the catalytic current decreased to about 70% of its initial value and an RSD was less than 7.02% (n = 3), indicating good stability over two weeks. This result expressing optimistic stability of the immunosensor is due to the fact that the platform was relatively stable and both the attachment of antibodies onto the electrode surface and the affinity bonding between antibody and antigen were relatively firm [28]. On the other hand, the gradual decline in the current value may be due to the slow denature of biomolecules [40].

Sensor Performance in Extracts from Spiked Maize and Swine Meat Samples
To validate the applicability of the presented method for specifically and sensitively detecting T-2, the spiked feed and swine meat samples at concentrations of 10, 50, 200 µg/kg were assessed with the suggested method. The recovery of spiked toxins in feed and swine meat samples were from 91.59% to 102.49%, 91.42% to 100.80% and their RSD results were lower than 10% (Table 2). These results indicated very high acceptability of the designed sensor for detecting mycotoxin in food matrices.

Apparatus
All the electrochemical experiments including differential pulse voltammetry (DPV), cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS) were performed with Autolab PGSTAT128N (Vanton, Switzerland). All the measurements were carried out with a conventional three-electrode system that composed of an Ag/AgCl saturated with KCl as reference electrode, a platinum wire as auxiliary electrode and a 3-mm-diameter GCE modified with the AuNPs/SWNTs/CS as working electrode at room temperature. The scanning electron microscopic (SEM) images and atomic force microscope (AFM) images were photographed with the SU8010 scanning electron microscope (Hitachi, Japan) and Multimode 8 (Cambridge, MA, USA).

Electrochemical Deposition of AuNPs onto the GCE
Firstly, the glass carbon electrode (GCE, 3 mm diameters) was polished with 0.3 and 0.05 µm alumina powder respectively and then rinsed completely with absolute alcohol and distilled water in ultrasonic bath. After being dried at room temperature, the GCE was immersed in 1 g/L auric chloride acid aqueous solution and afterwards AuNPs were deposited by chronoamperometric method with a voltage of −0.2 V and a time of 200 s.

Preparation of GCE Modified with AuNPs/SWNTs/CS
5 mg carboxylic group-functionalized SWNTs fine powders and 10 mg CS were dispersed into 10 mL N-N-dimethylformamide and sonicated to black uniform mixture. We stored the SWNTs/CS dispersion liquid at 4 • C. 10 µL of the mixture was drop-casted onto the GCE surface modified with nano-gold and dried at room temperature overnight.

Synthesis of T-2HS Hapten and T-2HS-OVA Antigen Hemisuccinate
The T-2-hemisuccinate (T-2HS) hapten was synthesized from HS to create a reactive group by modifying the procedures from a previous study [41,42]. Briefly, 5 mg of T-2 was dissolved in 2 mL pyridine and then 80 mg of succinic anhydride was added. The mixture was in a oil bath at 60 • C for 48 h and was dried with a flow of nitrogen. The residue was redissolved in 2 mL of chloroform and washed with distilled water four times. The chloroform extract was subsequently dried with a flow of nitrogen to obtain the T-2HS conjugate.
The hapten was transferred into 500 µL dimethylformamide (DMF) containing 5 mg DCC and 3 mg NHS and the mixture was stirred at RT overnight. Then the reaction solution was added drop-wise into the 5 mL of 0.1 mol·L phosphate buffer (PBS, pH 7.4) containing 20 mg OVA in an ice-bath and under continuous stirring for 24 h. Lastly, put the solution into a dialysis bag and dialyzed in PBS buffer solution for 3 days. Afterwards, we centrifuged the mixture for 10 min at 5000 rpm and reserved the supernatant at −20 • C prior to use.

Preparation of Electrochemical Immunosensor
We prepared the electrochemical immunosensor using the following steps: (i) As 2.3 mentioned; (ii) As 2.4 mentioned; (iii) After PBS clearing, we immersed the Au/SWNTs/CS/GCE into 50 µL freshly prepared EDC and NHS solution (5 mM:2 mM) and incubated it in a 37 • C incubator for 1 h to activate the carboxylic group of SWNTs; (iv) After the activation, we thoroughly washed Au/SWNTs/CS/GCE with PBS. Following this, we immediately dropped 10 µL of 15 µg·mL −1 T-2-OVA onto the electrode surface and incubated it for 1 h at 37 • C; (v) Then, we rinsed the built electrode with PBS and next treated it with 10 µL 2% BSA solution for 1 h at room temperature to block the inactive sites; (vi) Finally, we thoroughly rinsed the electrode with PBS again. The electrochemical immunosensor based on SWNTs/CS/AuNPs/GCE was successfully prepared and stored at 4 • C prior to use. The fabrication process for this electrochemical immunosensor was displayed in Scheme 1.
with PBS. Following this, we immediately dropped 10 μL of 15 μg·mL −1 T-2-OVA onto the electrode surface and incubated it for 1 h at 37 °C; (v) Then, we rinsed the built electrode with PBS and next treated it with 10 μL 2% BSA solution for 1 h at room temperature to block the inactive sites; (vi) Finally, we thoroughly rinsed the electrode with PBS again. The electrochemical immunosensor based on SWNTs/CS/AuNPs/GCE was successfully prepared and stored at 4 °C prior to use. The fabrication process for this electrochemical immunosensor was displayed in Scheme 1.

Electrochemical Measurements
In order to perform the competition assay, firstly blended 5 μL of the diluted anti-T-2 (5 μg·mL −1 ) with 5 μL of T-2 standard solution, with concentrations from 0 to 500 ng·mL −1 in PBS. We added the miscible liquids onto the modified T-2-OVA electrochemical immunosensor surface and then incubated it at 37 °C for 1.5 h. In the incubation process, a competing reaction happened between the immobilized T-2-OVA and free T-2 toward a rationed anti-T-2 in the mixture. After being rinsed with PBS, we incubated the prepared immunosensor with 10 μL of diluted ALP-anti-antibody (1:800, v/v) for 1.5 h at 37 °C. We washed the resulting electrode again with PBS and finally immersed it into the freshly prepared DEA solution containing 0.75 mg·mL −1 α-NP. We performed the differential pulse voltammetry (DPV) measurements at the setting of 0.05 s modulation time, potential scan from −0.2 Scheme 1. The fabrication process of this electrochemical biosensor.

Electrochemical Measurements
In order to perform the competition assay, firstly blended 5 µL of the diluted anti-T-2 (5 µg·mL −1 ) with 5 µL of T-2 standard solution, with concentrations from 0 to 500 ng·mL −1 in PBS. We added the miscible liquids onto the modified T-2-OVA electrochemical immunosensor surface and then incubated it at 37 • C for 1.5 h. In the incubation process, a competing reaction happened between the immobilized T-2-OVA and free T-2 toward a rationed anti-T-2 in the mixture. After being rinsed with PBS, we incubated the prepared immunosensor with 10 µL of diluted ALP-anti-antibody (1:800, v/v) for 1.5 h at 37 • C. We washed the resulting electrode again with PBS and finally immersed it into the freshly prepared DEA solution containing 0.75 mg·mL −1 α-NP. We performed the differential pulse voltammetry (DPV) measurements at the setting of 0.05 s modulation time, potential scan from −0.2 to 0.8 V, pulse amplitude of 70 mV, a pulse period of 0.2 s and interval time of 0.017 s at room temperature.

Sample Preparation
We obtained swine feed from the Breeding Swine Testing Centre (Huazhong Agriculture University, Wuhan, China). Firstly, we sifted them through a 40-mesh sieve, weighed a total 2 g of every finely ground samples into 50 mL tubes and added 10 mL of methanol-water (7:3, v/v). After a 5 min vortex, we then filtered the samples through Whatman No.1 filter paper. After that, we diluted 1 mL of the filtrate into 4 mL of PBS (0.01 mol·L −1 , pH 7.4) and filtered it again for analysis.
We purchased minced and finely homogenized swine muscle samples from a local market. We weighed 2 g of every sample into 50 mL tubes, and added 10 mL of methanol-water (7:3, v/v), and then centrifuged at 4000 rpm for 10 min after a 5 min vortex. In sequence, we removed 1 mL of the supernatant, then diluted into 3 mL of PBS and lastly added 1 mL hexylhydride into the lower layer after vortexing and standing for the electrochemical analysis.

Conclusions
In conclusion, the electrochemical immunosensor based on AuNPs/SWNTs/CS integrated with an enzymatic signal readout mechanism displayed excellent properties for the detection of T-2. The double signal amplification strategy on the AuNPs/SWNTs/CS prominently improved the sensitivity of T-2 determination. Simultaneously, the ALP possessing excellent catalytic activity towards α-NP enhanced the sensitivity of this assembled immunosensor. The developed method showed optimistic establishment reproducibility and a relative excellent specificity and sensitivity for T-2 detection as well. It also additionally exhibited a wide liner range from 0.01 to 100 ng·mL −1 with the limit of detection of 0.14 µg·L −1 . In summary, the suggested strategy provides an immobilized and sensitized recognition platform for T-2, which could be a promising application for T-2 detection in food samples and hopefully could further furnish a versatile and forceful tool to guarantee food safety in the future.