Enzyme-functionalized electrochemical immunosensor based on electrochemically reduced graphene oxide and polyvinyl alcohol-polydimethylsiloxane for the detection of Salmonella pullorum & Salmonella gallinarum

Abstract

A novel and ultrasensitive electrochemical enzyme immunosensor based on electrochemically reduced graphene oxide (ERGO) and a polyvinyl alcohol (PVA)-multilayer polydimethylsiloxane (PDMS) film modified screen-printed carbon electrode (SPCE) for the rapid detection of Salmonella pullorum (S. pullorum) & Salmonella gallinarum (S. gallinarum) was proposed in this study. The electrochemical characteristics of the stepwise modified electrodes and the determinations of S. pullorum & S. gallinarum were investigated by cyclic voltammetry (CV). PVA-PDMS was characterized by Fourier transform infrared spectroscopy (FTIR). The conductivity of the electrode was promoted by electrochemically reduced graphene oxide, and the sensitivity of the immunosensor was enhanced and strengthened. The binding capacity and biocompatibility of the ERGO modified electrode was elevated considerably by the PVA-PDMS film. Under optimized working conditions, the sensor showed a good performance with a linear response range from 10¹ CFU mL⁻¹ to 10⁹ CFU mL⁻¹, and the detection limit was 1.61 × 10¹ CFU mL⁻¹ (S/N = 3). The proposed enzyme immunosensor had high sensitivity, good specificity, acceptable accuracy and reproducibility, and low detection limit characteristics and could be a promising analytical tool in the detection of S. pullorum & S. gallinarum in practical samples.

---

1. Introduction

Salmonella pullorum (S. pullorum) & Salmonella gallinarum (S. gallinarum) generally causes pullorum and fowl typhoid, which is a kind of epidemic septicemia in chickens and causes heavy economic losses to the poultry industry.^1,2 These two kinds of bacteria have the same antigen of O_1,9,12, and exhibit high cross-reactivity with each other. It was difficult and unnecessary to differentiate S. pullorum & S. gallinarum strictly; therefore, they were often detected together.³ Minga et al. used the soluble bacterial proteins of Salmonella gallinarum strain 9 as the detection antigen to detect the chicken serological response against S. pullorum using ELISA.⁴ There are some commonly used detection methods for S. pullorum & S. gallinarum such as the standard method⁵ and a variety of rapid detection methods.^6,7 However, the methods introduced above are reliable but complicated and time-consuming; moreover, they require expensive, precise and cumbersome instruments and a great demand of the operator's professional skills, thus involving certain limitations in practical applications.⁵

An electrochemical immunosensor is a kind of rapid detection technology that began in the 1980s.¹ For its high sensitivity, simplicity and rapidity, electrochemical immunosensors are being rapidly developing.² However, many problems also exist such as poor reproducibility, which affects the accuracy of the test results. Although many studies have been reported, it is still difficult to popularize its application in the practical field.^1,8 In recent years, the screen-printed carbon electrode (SPCE) was introduced as an effective alternative for solid electrodes, which can fill the gap of the solid electrode for must be polished before each test and hardly be used in mass production.^9,10 Firstly, since the combination between the detection electrode and bioactive substances is not strong enough, it is easy to cause the separation of the modifier and the electrode after repeated washing and soaking procedures of the incubation steps in the test.^6,7,11 Secondly, the working electrode area is small, only about 1–2 mm², and the detection method is very sensitive; therefore, a small change between the modifier and electrode can cause large differences, which leads to the poor accuracy of the test results guarantee.^12,13 This problem largely affect the accuracy of the technology; therefore, the combination between the antibodies and the electrode becomes a key factor for the popularization and application of a sensor in practice.^14–16

The sol–gel polyvinyl alcohol (PVA)-polydimethylsiloxane (PDMS) is the most common liquid polymeric coating and has been prepared and studied for several applications.^17,18 PVA has a good biocompatibility and stability; moreover, it is the only kind of vinyl polymer that can be slowly used as a carbon source of the bacteria. Furthermore, PDMS has a lot of flexible siloxane chains, which can provide a high flux for the membrane channel, and establish a hydrophobic environment for the surroundings; moreover, its hydrophobic surface allows the high adsorption of protein molecules¹⁹ and PVA-PDMS has many advantages compared with PDMS. The morphology of the sol–gel PVA-PDMS is a highly porous structure, and this porous structure provides high surface area,²⁰ allowing for faster and more effective adsorption of proteins. By controlling the concentration of the PVA-PDMS polymer, the membrane hydrophilic/hydrophobic properties can be adjusted; moreover, PVA-PDMS can not only reduce the blocking of electron transfer, but also increase the stability and biocompatibility of a composite membrane. The thermogravimetric data for the PVA-PDMS coating showed that this material is more stable towards heating than conventional PDMS, and it supported a temperature that was higher than 300 °C without appreciable decomposition.²¹ Therefore, the PVA-PDMS films have a remarkable long lifetime. Based on these advantages mentioned above, PVA-PDMS is very suitable for immobilizing biomolecules, which is a critical step in the manufacture of electrochemical immunosensors.

Graphene, one of the allotropes of elemental carbon, is an infinite two-dimensional layer consisting of sp² hybridized carbon atoms.²² It is an ideal material in the fabrication of sensors.^23,24 Graphene-based biosensors have a high sensitivity because of the large surface-to-volume ratio and conductivity of graphene.^25–27 Recently, a facile electrochemical reduction method for the synthesis of high quality graphene from exfoliated graphite oxide (GO) was reported.^28,29 In comparison with the chemical reduction of GO, the electrochemical reduction method is green in nature and fast, and it does not result in the contamination of the reduced material.^30,31 Furthermore, the electrochemical application of electrochemically reduced graphene oxide (ERGO) has been investigated and it showed enhanced electron transfer properties.^32,33

Horseradish peroxidase (HRP) is widely used as a labeled enzyme in the design of electrochemical enzyme immunosensors.^34,35 Electron transfer from the reduction active center of an enzyme to the electrode is important for the sensitivity of the constructed enzyme immunosensor.³⁶ It is often difficult to buy HRP labeled antibodies and most of them need to be customized. This increases the sensor's manufacturing costs and is not conducive to its industrialization in the future. The HRP labeled anti-antibody is cheap and easy to obtain. By changing the antibody serum, this platform can be used to detect different bacteria, which is very suitable for the demand of industrialization.

In this study, we desired to construct an electrochemical enzyme immunosensor for the detection of S. pullorum & S. gallinarum. At first, graphene oxide was electrochemically reduced onto the SPCE as graphene. The graphene acted as an electron transfer agent and increased the electrochemical signals. Then, the sol–gel PVA-PDMS was adsorbed onto the modified electrode. By an indirect modification method, HRP-labeled Goat Anti-Rabbit second antibody was subsequently incorporated onto the PVA-PDMS modified electrode using a nonchemical bond interaction for unspecific immobilisation. Next, the anti-S. pullorum antibody isolated from rabbit serum and the S. pullorum antigen were incubated onto the electrode by the specific immobilisation between the antigen and antibody, respectively. CV was employed to determine S. pullorum & S. pullorum via changes of the reduction peak current in the substrate solution of H₂O₂ and the electron mediator of thionine.^37,38 This kind of immunosensor provides a model for pathogenic bacteria detection in many fields, and it has a very good scientific significance and practical application value.

2. Experimental

2.1 Reagents and apparatus

A variety of bacteria were employed in this work, including S. pullorum & S. gallinarum (ATCC 79201) as the target bacteria, and Escherichia coli (E. coli, ATCC 8739), Staphylococcus aureus (S. aureus, ATCC 27217), Enterobacter sakazakii (E. sakazakii, ATCC 29544), Vibrio parahaemolyticus (V. parahaemolyticus, ATCC 17802), Bacillus Subtilis (B. subtilis, ATCC 6633), S. pullorum negative serum and PBS (0.01 M, pH 7.4) as the control group. All of the bacteria was conserved in our laboratory. Anti-S.pullorum was obtained from the China Institute of Veterinary Drug Control (China, Beijing); HRP-labeled Goat Anti-Rabbit second antibody was purchased from Shangai Bo Gu Biotechnology Co., Ltd. (China, Shanghai); polymethoxysiloxanes (PDMS) was obtained from Hangzhou Kangchen Silicone Co., Ltd., (China, Hangzhou); polyvinyl alcohol (PVA) was supplied by Sinopharm Chemical Reagent Co., Ltd. (China, Shanghai); Bovine serum albumin (BSA) was supplied by Sino-American Biotechnology Co., Ltd. (China, Shanghai); graphite was obtained from Shanghai Colloid Chemical Plant; thionine (Thi, AR) was obtained from Shanghai Zhongtai Chemical Reagent Co., Ltd.; KMnO₄ was obtained from Shanghai Reagent Factory one; and other reagents were all of analytical grade and the water used was doubly distilled.

The CHI660C electrochemical workstation was provided by Shanghai Chen Hua Instruments, Inc., and the Screen-printed carbon electrode (SPCE) was jointly developed by our research team and Rong Bin Biotechnology Co., Ltd. The SPCE was a three-electrode system, comprising a working electrode (carbon electrode), auxiliary electrode (carbon electrode), and reference electrode (Ag/AgCl electrode). The pH was measured using a S20-K SevenEasy pH Meter (Mettler-Toledo, China). The magnetic stirrer was purchased from Hangzhou Electric Instrument Factory. All electrochemical experiments were performed at 22 ± 2 °C.

2.2 Immunosensor fabrication procedures

2.2.1 Preparation of graphene oxide. Graphite oxide (GO) was prepared according to the Hummers method with a slight modification.^39,40 Then, the homogeneous GO dispersions with a concentration of 1 mg mL⁻¹ were exfoliated in deionized water by ultrasonication (40 kHz, 150 W) for 5 h.^41,42

2.2.2 Preparation of the electrochemically reduced graphene oxide modified electrodes. Prior to the electrochemical reduction of graphene oxide, dispersions containing 1 mg mL⁻¹ GO were prepared. The electrochemical reduction was performed in the dispersion with magnetic stirring and N₂ bubbling, with the SPCE by CV. The scan was performed between −1.5 and 0.5 V at a rate of 25 mV s⁻¹ and the loading amount of the deposits was controlled by one potential cycle. Following that, the resulting electrode was washed with deionized water and dried by blowing N₂.

2.2.3 Preparation and characterization of PVA-PDMS. The PVA-PDMS sol was prepared by the following steps: a mixture of 4.4 mL of 0.45% PVA solution, 100 μL of PDMS, and 500 μL of 95% TFA aqueous solution were mixed for 2 min in a silanized glass tube, utilizing a vortex mixer, placed under ultrasonic conditions for 30 min, then reacted at 85 °C in a water bath for 7 h. Note that a uniform dispersion was obtained. The spectroscopic characterization of the PVA-PDMS was carried out by FTIR, which was recorded in a dry KBr pellet in the range of 400–4000 cm⁻¹.⁴³ The FTIR analysis revealed that the chemically synthesized PVA-PDMS well adhered to each other (for the FTIR data please refer to the ESI†).

2.2.4 Preparation of immunosensor. 2 μL of the abovementioned mixture solution was dropped onto the ERGO/SPCE using a pipette and dried slightly at room temperature. Subsequently, the electrode was carefully rinsed by deionized water to remove the excess PVA-PDMS. After that, the electrode further was coated with the contents of an HRP-labeled Goat Anti-Rabbit second antibody solution, which was combined with the PVA-PDMS. After that, the electrode was further coated with the contents of a 0.2% BSA solution to block the nonspecific active sites of the HRP-labeled Goat Anti-Rabbit second antibody and PVA-PDMS. The subsequent electrode was incubated with 3 μL of anti-S. pullorum antibody isolated from rabbit serum for 30 min at 30 ± 0.5 °C. The electrode was carefully rinsed with PBS to remove excess antibodies, which were not combined with the electrode. The modified electrode was stored in a humid, sterile and airtight container at 4 °C. The preparation of the immunosensor and the mechanism of the rapid detection of S. pullorum & S. gallinarum are shown in Scheme 1.

Scheme 1 Schematic illustration of the immunosensor preparation and electrochemical assay for the S. pullorum & S. gallinarum detection based on the ERGO and PVA-PDMS amplification strategy, which used a site-directed antibody immobilization method. (a) Bare SPCE; (b) electrochemically reduced graphene oxide to bare SPCE; (c) PVA-PDMS coated onto the electrode surface; (d) HRP-labeled Goat Anti-Rabbit second antibody incubation; (e) BSA block the nonspecific active sites of HRP-labeled Goat Anti-Rabbit second antibody and PVA-PDMS; (f) anti-S. pullorum incubation; (g) S. pullorum incubation (10⁹ CFU mL⁻¹); (h) the principle of electrochemical detection.

2.2.5 Preparation of food-borne pathogen detection samples. The bacterial suspension in the study was made by the following kinds of common food-borne pathogens: S. pullorum & S. gallinarum, E.coli, S. aureus, E. sakazakii, B. subtilis, V. parahaemolyticus. They were vaccinated into the nutrient broth medium and cultured at 37 °C for 24 h. After centrifugation (6000 rpm for 10 min) of the culture medium, 5 mL of physiological saline was added into the bacteria sludge, and uniformly oscillated. We prepared different concentrations of the bacterium suspension for colony counting and they were stored at 4 °C for later use.

2.3 Analytical procedure

Cyclic voltammetry (CV) is the most commonly used method in electrochemical immunoassays. The data are then plotted as current (i) vs. potential (E), and the oxidation peak usually has a similar shape to the reduction peak. As a result, information about the redox potential and electrochemical reaction rates of the compounds is obtained. We can judge the current size according to the reduction peak value of the curve.

The test method referred to in Zhao's article, with a few changes, is described briefly as follows:⁴⁴ 3 μL of S. pullorum antigen was dropped onto the previously modified electrode, incubated at 30 ± 0.5 °C for 30 min and rinsed carefully with PBS to remove any unbound bacterial antigen. The modified electrode was immersed in an electrolyte containing 10 mM Hac-NaAc (pH = 6.5), 1 mM Thi and 0.6 mM H₂O₂, and CV was conducted at a scan rate of 0.1 V s⁻¹ between −0.6 V and −0.1 V. The detection of S. pullorum & S. gallinarum was performed by measuring the reduction peak current shift (ΔIp) of the CV before and after the immunoreaction. Three successive CV scans were performed for each measurement, and the last cycle was recorded.

3. Results and discussion

3.1 Electrochemical characterization of the stepwise modified electrodes

3.1.1 Electrochemically reduced graphene oxide. The electrochemical method can equivalently, uniformly, firmly and steadily reduce graphene oxide to graphene and combine it on an electrode, which is the key to a sensor's excellent properties.^45–48 From Fig. 1, it can be clearly seen that the peak shape is the typical reduction peak of GO, and the ERGO modified electrodes reductive peak current value significantly increased compared to that of the bare electrode from the inset. Therefore, by this method, GO can be reduced to graphene and tightly combined on the electrode, and it also has a good effect on the enhancement of the electrochemical signal.

Fig. 1 CV curve of the electrochemical reduction of 1 mg mL⁻¹ GO aqueous solution in the dispersion with magnetic stirring and N₂ bubbling (inset: the CV curve of the ERGO/SPCE and Bare SPCE in 10 mM Hac-NaAc (pH 6.5) containing 1 mM thionine (Thi) and 0.6 mM H₂O₂).

3.1.2 The function of the graphene layer. Graphene has become an ideal material in the fabrication of sensors due to its high conductivity. It was introduced into the fabrication of the immunosensor to enhance the electrochemical signals in order to ensure the sensitivity of the test results. To justify this function, a side-by-side performance of the electrode sensor, upon which the PVA/PDMS films were deposited, both with and without the ERGO is shown in Fig. 2. We can see clearly that without the ERGO, the signal of the PVA-PDMS modified electrode is much smaller than that of the electrode with ERGO. The thin layer of graphene deposited on the carbon electrode of the SPCE resulted in an improved performance, because the conductivity of graphene is much higher than that of graphite; thus, graphene is an ideal material in the fabrication of sensors due to its high conductivity.

Fig. 2 CV current curves of the PVA-PDMS and ERGO/PVA-PDMS modified electrodes in a basic solution (1.0 M HAc-NaAc buffer solution pH = 6.5 with 1.0 mM thionine).

3.1.3 Cyclic voltammetry behavior of the stepwise modified electrodes. CV is a simple, valuable, and commonly used tool that was used to probe the packing structure of the modified electrode through the redox behavior of a reversible redox couple.⁴⁹ Fig. 3 shows the CV characterization of the electrode after each modifying step, to characterize if the electrodes modified successfully or not, S. pullorum & S. gallinarum included in the solution caused current value decrease to achieve the detection.

Fig. 3 CV current curves of different modified electrodes in a basic solution (1.0 M HAc-NaAc buffer solution pH = 6.5 with 1.0 mM thionine). (a) Bare SPCE; (b) ERGO/SPCE; (c) ERGO/PVA-PDMS/SPCE; (d) ERGO/PVA-PDMS/HRP-labeled Goat Anti-Rabbit second antibody SPCE; (e) BSA closed nonspecific sites of ERGO/PVA-PDMS/HRP-labeled Goat Anti-Rabbit second antibody SPCE; (f) immunoelectrode incubated with Anti-S. pullorum; (g) immunoelectrode incubated with S. pullorum & S. gallinarum (10⁹ CFU mL⁻¹).

The bare electrode presents a pair of reversible redox peaks in the test substrate solution (Fig. 3a). After the electrochemical reduction of graphene oxide, the peak currents of the reduction and oxidation significantly increased (Fig. 3b), which indicated that the graphene was successfully combined with the electrode. After the adsorption of the PVA-PDMS film (Fig. 3c), the reduction peak current decreased slightly, due to the insulating layer of PVA-PDMS coated onto the electrode surface. The HRP-labeled Goat Anti-Rabbit second antibody was firmly linked to the electrode surface through the high affinity of the noncovalent interactions between PVA-PDMS and proteins, and the HRP catalyzed reduction of H₂O₂ with the assistance of an electron mediator, which promoted electron transfer between the enzyme and the electrode and greatly increased the reduction peak current value (Fig. 3d). The peak current was slightly reduced when BSA (Fig. 3e) was used to block the remaining nonspecific active sites of HRP-labeled Goat Anti-Rabbit second antibody and PVA-PDMS. And BSA is excessive for it really needed, so we should carefully rinse the modified electrode with PBS to remove the excessive BSA. All these procedures can ensure that the available nonspecific binding sites in the PVA/PDMS film can be occupied. As shown in Fig. 3f, the anti-S. pullorum was irreversibly and specifically linked to the electrode surface through the high affinity of the noncovalent interactions of the antibody and antigen. After S. pullorum & S. gallinarum was bound to anti-S. pullorum, the formed antigen–antibody immune-complex on the electrode surface hindered the electron transfer toward the electrode surface, resulting in a further decrease of electrochemical signal (Fig. 3g).

3.1.4 Effect of the scanning rate. In order to study the electrochemical behavior of the modified electrodes, we investigated the influence of the scanning rate on the electrochemical signals. The electrode was modified by ERGO; PVA-PDMS and HRP-labeled Goat Anti-Rabbit second antibody SPCE. As shown in Fig. 4, in the potential from −0.6 V to −0.1 V with a scan rate of 100 mV s⁻¹, the reduction of hydrogen peroxide was catalyzed by HRP and reached a cathodic peak current at −0.38 V (Fig. 4a). The peak values of the forward and reverse reactions are linearly proportional to the scan rates in the range from 50 to 200 mV s⁻¹. The good linearity indicates that the rate of the redox reactions is controlled by the surface diffusion of electrons, which produces a concentration gradient of the electroactive species in the HRP-labeled antibody and Thi on the electrode (Fig. 4b).

Fig. 4 Electrochemistry behavior of ERGO/PVA-PDMS/HRP-labeled Goat Anti-Rabbit second antibody SPCE: (A) cyclic voltammogram of the modified electrodes at scan rates from 50 to 200 mV s⁻¹ in buffer solution; (B) plot of the anodic and cathodic currents versus the square root of the scan rates. The regression equation of the linear fit for anodic response is as follows: y = −1.2574x − 2.599 (R² = 0.9978), and for cathodic response is as follows: y = 0.7388x + 0.8804 (R² = 0.9935).

3.2 Optimization of the experimental conditions

To achieve the optimal immunoassay performance, the concentration of PVA, the concentration of HRP-labeled Goat Anti-Rabbit second antibody and H₂O₂, pH of the electrolyte and incubation temperature and time were optimized as important factors that influenced the sensitivity of the proposed immunosensor.⁴⁴

The aqueous solution of PVA has a good film-forming ability, good biocompatibility and a strong adhesion capacity.^50,51 By controlling the concentration of PVA, the polymer membrane hydrophilic/hydrophobic property can be adjusted, which reduces the blocking of electron transfer and the ability of membrane resistance to swelling is retained. As shown in Fig. 5A, with the increase of PVA concentration from 0.1% to 0.9%, the reduction peak current value increases slowly, and then decreases after reaching the maximum value when the concentration was 0.45%. Due to the excessive concentration of PVA, the increase in viscosity also hinders electrons through film thickening. The PVA-PDMS film's hydrophilic increase reduces the stability, and thus the optimal concentration of PVA, which is 0.45%, was chosen.

Fig. 5 Optimization of the experimental conditions including: (A) the effect of the PVA concentration; (B) dilution of the HRP-labeled Goat Anti-Rabbit second antibody; (C) H₂O₂ concentration; (D) electrolyte pH value on the reduction peak current value of immunosensor array; effect of immune-reaction conditions on the decrease of the reduction peak current value: (E) incubation time and (F) incubation temperature. In all of the above experiments, the other parameters were kept fixed at the optimum condition: 0.45% PVA, 1 : 700 antibody solution, 0.6 mM H₂O₂, pH 6.8, 25 min, 30 °C.

As shown in Fig. 5B, different enzyme-labeled antibody concentrations of immunosensor also has an effect on the peak currents of reduction and oxidation. Moreover, with the increase in the antibody dilution ratio, the reduction peak current value decreases, when diluted ratio is 1 : 800, the reductive peak current value has no significance for detection. This is because as the enzyme-labeled antibody concentration decreases, the concentration of HRP combined with the antibody is also gradually reduced. The catalytic ability of H₂O₂, as well as reduction peak current, is greatly reduced. By comparison with the electrochemical response signal of the Bare SPCE, select the optimal antibody dilution multiple ratio is 1 : 700.

The effect of the concentration of H₂O₂ on the response signal was examined in the range from 0 to 1.0 mM. The reduction peak current increased with increase in concentration of H₂O₂ from 0 to 0.6 mM, after which the reduction peak current decreased slightly. With increase in the concentration of H₂O₂, there are more and more redox reactions occurring; thus, the peak current increased with increase in concentration of H₂O₂ from 0 to 0.6 mM. However, a high concentration of peroxide would also inactivate the enzyme and the peak current will decrease slightly. In order to ensure an adequate response and the activity of enzyme-labeled antibody, an H₂O₂ concentration of 0.6 mM was selected (Fig. 5C).

Electrolyte pH has a great effect on both the electrochemical characteristics of the mediator and the activity of enzyme and antibody. The immunoelectrode was tested over a range of pH from 5 to 8 with increments of 0.5; moreover, the electrical signal increased rapidly with increase in pH value from 4 to 6.8, and then rapidly decreased from 6.8 to 8. The modified electrode showed a maximum response at pH 6.8. With the pH continuing to increase, the reduction peak current decreased rapidly, indicating that a weak acidic environment was more conducive for the enzyme-labeled antibody to be in operation. Because the modified electrode had the best response at pH 6.8, this pH value was chosen as the optimal one for the determination of S. pullorum & S. gallinarum (Fig. 5D).

The formation of the immune-complex on the electrode surface depends on the incubation time and incubation temperature. S. pullorum & S. gallinarum was incubated with the modified electrode for different time periods (10 to 60 min). With increase in incubation time, the current signal increased and then reached a constant value when the incubation time was longer than 25 min. It was possibly because S. pullorum & S. gallinarum was adsorbed to some degree with time. Thus, the incubation time of 25 min was selected as the optimal incubation condition for the immunoassay (Fig. 5E). The influence of the incubation temperature on the performance of the immunosensor was studied in temperatures ranging from 24 to 40 °C, it was found that the maximum current of the sensor occurred at a temperature 30 °C; hence, 30 °C was selected as the optimal incubation temperature (Fig. 5F). The concentration of S. pullorum & S. gallinarum used in this experiment was 10⁹ CFU mL⁻¹.

3.3 Calibration curve of the immunosensor

Under the optimal detection conditions, the immunosensor was subjected to a standard S. pullorum & S. gallinarum solution with various concentrations. After incubation with the standard S. pullorum & S. gallinarum culture, the reduction peak currents decreased with increase in S. pullorum & S. gallinarum concentration, which may be ascribed to the fact that more antigen–antibody immunocomplexes were formed on the surface of the electrode, and S. pullorum & S. gallinarum hindered the electron transfer. The corresponding calibration curve is plotted in Fig. 6.

Fig. 6 Calibration curves of the log[S. pullorum & S. gallinarum] concentrations ranging from 10¹ to 10¹⁰ CFU mL⁻¹ under the optimized conditions. The error bars indicate the relative standard deviations of three time measurements, and the linear relation between the reduction peak current change (ΔIp) and logarithm of concentrations of S. pullorum is 0.9961.

The reduction peak currents were linearly related to the logarithm of the S. pullorum and S. gallinarum concentration in the range of 10¹ to 10⁹ CFU mL⁻¹, respectively, and the regression equation was ΔIpc (μA) = 0.5245 lg[C CFU⁻¹ mL⁻¹] − 0.0375 (μA) with a correlation coefficient of 0.9961. The detection limit corresponding to three times the standard deviation of the blank solution was estimated at 1.66 × 10¹ CFU mL⁻¹ (S/N = 3), which is much higher than the detection limit of our previous work for S. pullorum & S. gallinarum (1.95 × 10²).⁴⁴ This highly sensitive detection method may depend on the following factors: PVA-PDMS provided a large number of binding sites and was employed as a substrate for the immobilization of a considerable amount of HRP-labeled Goat Anti-Rabbit second antibody. On the other hand, the ERGO and HRP amplified the electrochemical signal, which could further enhance the sensitivity of immunosensor.^51,52

3.4 Specificity, reproducibility and stability of the immunosensor

The specificity of the immunosensor is related to its ability to distinguish the target bacteria from other bacteria in a sample. The specificity of a method is a key measure of its applicable value. In order to test, as much as possible, the impact of other common pathogens, we selected six potential disruptors: E. coli, S. aureus, E. sakazakii, V. parahaemolyticus, B. subtilis, and S. pullorum negative serum as the control group, together with PBS as a blank control (Fig. 7A), and all of the bacteria solution concentrations were 10⁹ CFU mL⁻¹. As shown in Fig. 7A, the current increase induced by S. pullorum & S. gallinarum was much larger than that induced by PBS or the other pathogens (n = 7), which indicated that the immunosensor was highly specific for S. pullorum & S. gallinarum. These demonstrated results that the disruptors did not cause an observable interference to the S. pullorum & S. gallinarum detection and the reduction peak current varied slightly, which was attributed to the highly specific antigen–antibody immunoreaction. The experimental results indicated that the proposed immunosensor specifically recognized the S. pullorum & S. gallinarum antigen and exhibited good selectivity.^53,54

Fig. 7 (A) Specificity of the S. pullorum & S. gallinarum detection immunosensor. The immunosensor incubated with S. pul & S. gal, E. coli, S. aureus, E. sakazakii, B. subtilis, V. parahaemolyticus, S. pullorum negative serum and PBS (0.01 M, pH 7.4), respectively, for 30 min at 30 °C (10⁹ CFU mL⁻¹). (B) The immunosensor incubated with S. pullorum & S. gallinarum suspension (5 × 10⁴ CFU mL⁻¹) containing contaminating E. coli (5 × 10⁸ CFU mL⁻¹)and S. aureus (5 × 10⁴ CFU mL⁻¹), and S. pullorum (5 × 10⁴ CFU mL⁻¹) & PBS as control group, respectively.

In order to investigate the influence of other bacteria on the detection of S. pullorum & S. gallinarum, we compared the reduction peak current decrease (ΔIp) in the presence and absence of contaminating microorganisms. The data in Fig. 7B show the electrochemical signals of S. pullorum & S. gallinarum containing the contaminating microorganisms E. coli and S. aureus. It can be seen that E. coli and S. aureus do not interfere with the detection of S. pullorum, and the reduction peak current varies slightly. There were no significant current change differences both with and without the addition of the other species. The specificity of the modified sensors towards S. pullorum & S. gallinarum was good.

3.5 Storage stability study

A long-term storage assay was used to examine the stability of the proposed immunosensor. Before the immunosensor was in use, it was stored in a sterile, humid and airtight container at 4 °C, and was measured every five days, by means of cyclic voltammetry (Fig. 8). The reduction peak current was slightly reduced during the storage, and 94.7% of initial response remained after storage for thirty days (n = 7), demonstrating that the developed immunosensor had good storage stability. It is much higher than 88.3% of the initial response that remained after storage for thirty days in our previous reported immunosensor for S. pullorum & S. gallinarum.⁴⁴ In this experiment, the stability of the modified electrode without dispensing PVA-PDMS composite film was also measured. The reduction peak current was reduced much more quickly compared with the above-mentioned electrode during the storage, and 84.0% of the initial response remained after storage for 4 weeks (n = 7). This result proved well that the PVA-PDMS film could effectively maintain the activity of HRP and increase the stability of the immunosensor. The acceptable stability of the immunosensor may be due to the fact that the PVA-PDMS film on the electrode surface could make protein molecules be more firmly attached on electrode and keep the activity of the modified layer for its good biocompatibility.

Fig. 8 The long-time storage stability of different proposed immunosensors: (a) ERGO/PVA-PDMS/HRP-labeled Goat Anti-Rabbit second antibody SPCE, (b) ERGO/HRP-labeled Goat Anti-Rabbit second antibody SPCE.

The reproducibility of the proposed immunosensor was evaluated by seven equally proposed immunosensors incubated with the same concentration S. pullorum & S. gallinarum (10⁹ CFU mL⁻¹). The seven electrodes exhibited similar electrochemical responses and a relative standard deviation (RSD) of 6.30% was obtained, indicating a satisfying reproducibility.

3.6 Food sample detection

In order to verify the accuracy of the immunosensor practically, a series of food samples: eggs, chicken meat, heart, liver, and intestines were bought from a market, and were tested for the target antigens by the China national food safety standard (GB/T 17999.8-2008) for the detection of S. pullorum & S. gallinarum, involving the following basic steps: pre-enrichment, selective enrichment, selective plating, biochemical screening and serological confirmation. The comparison with the immunosensor detection method is shown in group A (Table 1). We found that all of them were not affected by S. pullorum & S. gallinarum except for one chicken sample.

Table 1 Accuracy of the experimental results of the A group immunosensor (n = 60)

Sample		National standard	Immunosensor	Accordance rate
Chicken	Positive	0	1	96.67%
	Negative	30	29	96.67%
Egg	Positive	0	0	100%
	Negative	30	30	100%
Total		60	60	98.33%

---

Group B (Table 2) used the blind method, and was performed by two teams. One team randomly selected the S. pullorum & S. gallinarum negative samples, whereas a proper dose of S. pullorum & S. gallinarum was spiked into the negative samples. Note that 30 new samples of chicken and eggs were prepared, and another team used the new type of sensor for testing. The testers of the two teams did not exchange views in this experiment; however, the results exhibited were consistent, which revealed a good accuracy of the proposed immunosensor, indicating the acceptable agreement between the two methods, which revealed that the immunosensor held great promise as a reliable tool for the detection of S. pullorum & S. gallinarum in real samples.⁵⁵

Table 2 Accuracy of the experimental results of the B group immunosensor (n = 60)^a

		1			2			3			4			5
a Note: +: positive sample, −: negative sample.
Chicken	Sample	−	−	+	−	+	+	+	−	+	+	+	−	+	+	−
	Result	−	+	+	−	+	+	+	−	+	+	+	−	+	+	−
Egg	Sample	+	+	−	−	−	+	+	−	+	−	+	+	+	−	+
	Result	+	+	−	−	−	+	+	−	+	−	+	+	+	−	+

---

4. Conclusions

In summary, a simple and sensitive immunosensor was developed for S. pullorum & S. gallinarum detection. ERGO and PVA-PDMS were employed to develop this electrochemical enzyme immunosensor. As can be seen in characterization of the cyclic voltammetry, ERGO increased the electron transfer and enhanced the conductivity of the electrode. In addition, the immobilization amount of the incorporating bio-molecules of the immunosensor was also greatly improved by the PVA-PDMS film, which promoted the adsorption of HRP-labeled Goat Anti-Rabbit second antibody molecules onto the surface and the antibody–antigen system, which contributed to integrating more of the subsequently corresponding antigen. More HRP-labeled Goat Anti-Rabbit second antibody molecules were well bonded to the electrode surface so that more active sites, which could implement catalysis, were accessible to the substrate. Note that the limit of detection of this immunosensor for S. pullorum & S. gallinarum was 1.66 × 10¹ CFU mL⁻¹ (S/N = 3). The stability of the enzyme immunosensor constructed from the PVA-PDMS system was much higher than only the ERGO assay enzyme immunosensor, which mainly originated from the good adsorption and biocompatibility of the PVA-PDMS film. The immunosensor exhibited the following characteristics: rapidity, sensitivity, specificity, accuracy, reproducibility and low detection limit, which could be a model for the development of immunosensors for other bacteria of interest.

Acknowledgements

The authors gratefully acknowledge the financial support from the National Natural Science Foundation of China (30571623), the University-level Innovation Fund of Zhejiang Gongshang University (1110XJ1511104 & 1110XJ1513128), and the Higher Education Research Project fund of Zhejiang Gongshang University (No. 1110KU212062).

Notes and references

1. K. Muzyka, Biosens. Bioelectron., 2014, 54, 393.
2. F. R. R. Teles and L. R. Fonseca, Mater. Sci. Eng., C, 2008, 28, 1530.
3. P. A. Barrow and O. C. Freitas Neto, Avian Pathol., 2011, 40, 1.
4. U. M. Minga, C. Wray and P. S. S. Gwakisa, Scand. J. Immunol., Suppl., 1992, 11, 157.
5. V. Velusamy, K. Arshak, O. Korostynska, K. Oliwa and C. Adley, Biotechnol. Adv., 2010, 28, 232.
6. P.-Y. Cheung and K. Kai Man, Food Res. Int., 2012, 45, 802.
7. Z. Xiaoyan, S. Zhongyue, C. Zhimin and Li. Haibing, Sens. Actuators, B, 2014, 191, 313.
8. N. Thiyagarajan, J.-L. Chang, K. Senthilkumar and J.-M. Zen, Electrochem. Commun., 2014, 38, 86.
9. O. D. Renedo, M. A. Alonso-Lomillo and M. J. A. Martinez, Talanta, 2007, 73, 202.
10. M. A. Alonso-Lomillo, O. Domínguez-Renedo and M. J. Arcos-Martínez, Talanta, 2010, 82, 1629.
11. X. Yuan, D. Fabregat, K. Yoshimoto and Y. Nagasaki, Colloids Surf., B, 2012, 99, 45.
12. J.-J. Shi, T.-T. He, F. Jiang, E. S. Abdel-Halim and J.-J. Zhu, Biosens. Bioelectron., 2014, 55, 51.
13. A. P. Washe, P. Lozano-Sánchez, D. Bejarano-Nosas and I. Katakis, Electrochim. Acta, 2013, 91, 166.
14. M. Nayak, A. Kotian, S. Marathe and D. Chakravortty, Biosens. Bioelectron., 2009, 25, 661.
15. C. Chao, X. Qingji, Y. Dawei, X. Hualing, F. Yingchun, T. Yueming and Y. Shouzhuo, RSC Adv., 2013, 3, 4473.
16. W. Siangproh, W. Dungchai, P. Rattanarat and O. Chailapakul, Anal. Chim. Acta, 2011, 690, 10.
17. L. B. Carneiro, J. Ferreira, M. J. L. Santos, J. P. Monteiro and E. M. Girotto, Appl. Surf. Sci., 2011, 257, 10514.
18. A. Kreider, K. Richter, S. Sell, M. Fenske, C. Tornow and V. Stenzel, Appl. Surf. Sci., 2013, 273, 562.
19. D. S. Pito, Hydrolysis of sucrose over composite catalysts, Chem. Eng. J., 2012, 184, 347.
20. T. C. O. Mac Leod, R. S. Marques, M. A. Schiavonc and M. D. Assisa, Appl. Catal., B, 2010, 100, 55.
21. J. Li, M. Wang and Y. Shen, Surf. Coat. Technol., 2012, 206, 2161.
22. A. K. Geim, Science, 2009, 324, 1530.
23. S. Sadeghi, M. Hemmati and A. Garmroodi, Electroanalysis, 2013, 25, 316.
24. X.-H. Zhou, L.-H. Liu, X. Bai and H.-C. Shi, Sens. Actuators, B, 2013, 181, 661.
25. L. Ruiyi, X. Qianfang, L. Zaijun, S. Xiulan and L. Junkang, Biosens. Bioelectron., 2013, 44, 235.
26. W. Huiwen, R. Fangfang, W. Caiqin, Y. Beibei, B. Duan, Z. Ke and D. Yukou, RSC Adv., 2012, 00, 1.
27. J. Wen, S. Zhou and Y. Yuan, Biosens. Bioelectron., 2014, 52, 44.
28. J. H. Warner, E. R. Margine, M. Mukai, A. W. Robertson, F. Giustino and A. I. Kirkland, Science, 2012, 337, 209.
29. N. N. Klimov, S. Jung, S. Zhu, T. Li, C. A. Wright and S. D. Solares, Science, 2012, 336, 1557.
30. S. Kim, S. Zhou, Y. Hu, M. Acik, Y. J. Chabal and C. Berger, Nanostruct. Mater., 2012, 11, 544.
31. L. Britnell, R. V. Gorbachev, R. Jalil, B. D. Belle, F. Schedin and A. Mishchenko, Science, 2012, 335, 947.
32. Y. Liang, H. Wang, J. Zhou, Y. Li, J. Wang and T. Regier, J. Am. Chem. Soc., 2012, 134, 3517.
33. Y. Li, W. Zhou, H. Wang, L. Xie, Y. Liang and F. Wei, Nat. Nanotechnol., 2012, 7, 394.
34. D. Sun, C. Cai, X. Li, W. Xing and T. Lu, J. Electrochem. Soc., 2004, 566, 415.
35. Z. Wang, M. Li, P. Su, Y. Zhang, Y. Shen and D. Han, Electrochem. Commun., 2008, 10, 306.
36. S. Chen, R. Yuan, Y. Chai and N. Li, Electroanalysis, 2008, 20, 418.
37. Q. Gao, X. Cui, F. Yang, Y. Ma and X. Yang, Biosens. Bioelectron., 2003, 19, 277.
38. T. Komura, G. Y. Niu, T. Yamaguchi and M. Asano, Electrochim. Acta, 2003, 48, 631.
39. J. Chen, B. Yao, C. Li and G. Shi, Carbon, 2013, 64, 225.
40. N. Kovtyukhova, E. Buzaneva and A. Senkevich, Carbon, 1998, 36, 549.
41. M. H. Rümmeli, C. G. Rocha, F. Ortmann, I. Ibrahim, H. Sevincli and F. Börrnert, Adv. Mater., 2011, 23, 4471.
42. A. Moradi Golsheikh, N. M. Huang, H. N. Lim, R. Zakaria and C.-Y. Yin, Carbon, 2013, 62, 405.
43. U. Subramanyam and J. P. Kennedy, J. Polym. Sci., Part A: Polym. Chem., 2009, 47, 5272.
44. H. Chunmei, D. Wenchao and Z. Guangying, Electrochim. Acta, 2014, 117, 239.
45. A. Florea, Z. Taleat, C. Cristea, M. Mazloum-Ardakani and R. Săndulescu, Electrochem. Commun., 2013, 33, 127.
46. V. Serafín, L. Agüí, P. Yáñez-Sedeño and J. M. Pingarrón, J. Electroanal. Chem., 2011, 656, 152.
47. X. Zeng, X. Li, L. Xing, X. Liu, S. Luo and W. Wei, Biosens. Bioelectron., 2009, 24, 898.
48. W. Shi, N. Lin, Y. Song, C. Liu, S. Zhou and X. Cai, Biosens. Bioelectron., 2014, 51, 244.
49. G. Zhao, X. Zhan and W. Dou, Anal. Biochem., 2011, 408, 53.
50. C. C. DeMerlis and D. R. Schoneker, Food Chem. Toxicol., 2003, 41, 319.
51. C. Luo, Y. Lei, L. Yan, T. Yu, Q. Li and D. Zhang, Electroanalysis, 2012, 24, 1186.
52. X. Kang, G. Pang, Q. Chen and X. Liang, Sens. Actuators, B, 2013, 177, 1010.
53. K.-F. Low, K. Chuenrangsikul, P. Rijiravanich, W. Surareungchai and Y.-Y. Chan, World J. Microbiol. Biotechnol., 2012, 28, 1699.
54. Y. Li, L. Fang, P. Cheng, J. Deng, L. Jiang and H. Huang, Biosens. Bioelectron., 2013, 49, 485.
55. C. Cheng, Y. Peng, J. Bai, X. Zhang, Y. Liu and X. Fan, Sens. Actuators, B, 2014, 190, 900.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra09901j

---