Development of an Amperometric Biosensor Based on Peroxidases from Brassica napus for the Determination of Ochratoxin a Content in Peanut Samples

Copyright: © 2011 Ramírez EA, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Development of an Amperometric Biosensor Based on Peroxidases from Brassica napus for the Determination of Ochratoxin a Content in Peanut Samples


Introduction
The ochratoxins are a group of highly toxic fungal secondary metabolites that are produced by several species of the Aspergillus and Penicillium genera [1,2]. Ochratoxin A (OTA) is the most toxic and prevalent among these toxins, and is receiving increasing attention because of the hazard imposed on both human and animal health [3]. OTA or 7-carboxyl-5-chloro-8-hydroxil-3,4-dihydro-3R-methylisocumarin-7-L-b-phenylalanine, is a colorless crystalline compound that belongs to a group of closely related derivatives of isocumarin linked to L-phenylalanine and classified as pentaketides [4] (Strcture 1). [10]. Each year, a significant portion of the peanuts produced can not be marketed because of fungal disease at the postharvest stage and mycotoxin contamination [11]. With imports and exports of agricultural commodities increasing in recent years, monitoring of OTA is essential for food safety in the world-wide market.
Many research groups have been studying methods for monitoring naturally occurring OTA in a number of agricultural commodities [9,12]. Many different analytical methods are now available for the determination of OTA in a variety of matrices [9]. Chromatographic methods such as thin-layer chromatography (TLC), gas chromatography (GC) and high-performance liquid chromatography (HPLC) with immunoaffinity columns and fluorescence detection [13][14][15][16][17] provide sensitive and specific techniques, but generally require multiple steps prior to detection. Today, the electroanalytical methods are very valuable alternatives to other methods because generally no separation procedure is required prior to substrate determination. Furthermore, the required instrumentation is typically less expensive, smaller amounts of solvent are necessary and analysis time is shorter for electroanalytical methods over chromatographic measurements.
Here, we report on the development of an amperometric biosensor based on peroxidases obtained from Brassica napus hairy roots (PBHR) to determine OTA content in solutions composed by comercial reagent and peanut samples.
The method employs carbon paste (CP) electrodes filled up with PBHR and ferrocene (Fc) at a given composition (PBHR-Fc-CP). The biosensor was covered externally with a dialysis membrane, which was fixed at the electrode body side part with a Teflon laboratory film and an O-ring. It is well known that phenolic and/or polyphenolic compounds can work as electron-donors for peroxidases in the catalytic reduction of H 2 O 2 [26,27]. This approach allows detecting the decrease in H 2 O 2 concentration in a solution after the oxidation of phenolic and/or polyphenolic compounds produced by the PBHR or horse radish peroxidase (HRP) in the presence of H 2 O 2 , given that HRP acts in cascade in the solution and PBHR at the electrode surface. The separation of the electrode surface from the solution by a semipermeable membrane allows minimizing the electrical noises as well as the fouling of the electrode surface.

Reagents and materials
Brassica napus hairy roots, obtained "in vitro" in our laboratory according to a procedure previously described by Agostini et al. [28] were used as the enzyme source.
Hydrogen peroxide (30%, v/v), pH 8.00 phosphate buffer solutions (PBS), sodium acetate, acetic acid, NaCl, HCl, NaOH and Na 2 CO 3 were Merck p.a.; ochratoxin A (OTA), ferrocene (Fc) and o-dianisidine were purchased from Sigma and used as received. All solutions were prepared using water purified by a Labconco WaterPro Mobile System, Model 90901-01 (HPLC grade water). The concentration of H 2 O 2 was determined spectrophotometrically at l max = 240nm (ε = 43.6 mol -1 cm -1 ) [29]. Stock solutions of OTA were prepared in PBS and kept at a temperature of 40C. The electrode surface was covered with a dialysis membrane (Spectrum Co., Houston, TX, cut off molecular weight 100), which was fixed at the electrode body side part with a Teflon laboratory film and an O-ring.

Total extraction and purification of enzymes and determination of de peroxidase activity
Hairy roots were homogenized in a mortar with 10 mmol L -1 pH 4.00 sodium acetate/acetic acid buffer, containing 1 mol L -1 NaCl (1 g fresh roots per 3 mL of buffer) at 4 ºC. Homogenates were centrifuged at 5000 rpm for 5 min. The supernatants were considered as total peroxidase extracts (TPE). They were used in order to purify peroxidases and determine total peroxidase activity.
The purification of peroxidases was performed by molecular exclusion chromatography on Sephacryl S-200-HR (SIGMA) columns. Samples of 2.5 mL of TPE, previously dialyzed, were loaded on a Sephacryl S-200-HR column (2 cm × 32 cm) equilibrated with a 20 mmol L -1 pH 7.00 phosphate buffer. The column was washed with 100 mL of buffer at a flow rate of 1 mL min -1 controlled with a peristaltic pump (LKB 2232 Microperpex). Fractions of 5 mL were collected and monitored to determine their peroxidase activity using a qualitative method proposed by Forchetti and Tigier [30]. Fractions containing most peroxidases were mixed and lyophilized before they were used. These enzymes presented a peroxidase activity of 280 IU per solid mg and they were labeled as PBHR.
The total peroxidase activity was determined with o-dianisidine as substrate [31] and expressed in international unit (IU), defined as the amount of enzyme forming 1 mmol of product in 1 min under the experimental conditions employed.

Real samples
OTA production in peanut kernels: Fungal Strains: Four Aspergillus section Nigri strains were evaluated. A. carbonarius strains RCPG and RCP203 and A. niger aggregates RCP42 and RCP191. All strains were isolated from peanut kernels in Argentina and were identified on morphological characters base [32]. OTA production was assayed on YES medium (2% yeast extract and 15% sucrose) [33]. Strains were maintained in glicerol (15%) at -80ºC and kept in the cultura collection at the Department of Microbiology and Immunology of the National University of Río Cuarto (Córdoba, Argentina).
Substrate: Peanut kernels (Runner variety) that had been irradiated (7 kGy) and retained their germinative capacity were used. Kernels were checked for sterility and the absence of OTA [34] and were kept at 4ºC until used. Irradiated peanut kernels were rehydrated to 0.93, 0.95, and 0.98 a w by adding sterile distilled water. The a w of representative samples of each treatment was checked at the beginning of the experiment with an AquaLabSeries 3 water activity meter (Decagon Devices, Inc., Pullman, WA).

Inoculation and Incubation conditions:
Peanuts were conditioned with an appropriate amount of water far from the a W levels. The substrate was kept at 4ºC for 48 h with manual periodic shaking to allow absorption and to reach equilibrium. 20 g of peanuts were placed as a monolayer into sterile Petri dishes. Peanuts were inoculated centrally with 5 µL of a spore suspension (1 x 10 6 spores/mL) from a 7-dayold culture growing on 2% malt extract agar (MEA) [35]. Inoculated peanuts plates with the same aW were sealed in plastic containers. Each container had beakers with a glycerol/water solution at the same aW as the peanuts, to maintain a constant relativity humidity. Temperatures tested were 18 and 25°C for a period of 30 days.
Ochratoxin A extraction: OTA production was analyzed after 30 days of incubation. Five grams of a finely ground peanut sample was added to a 250 mL Erlenmeyer flask along with 50 mL of an acetonitrile: water mixture (84:16). The mixture was shaken in an orbital shaker for 30 min and filtered through filter paper (Whatman No. 1). The extract (7 mL) was added to a clean up column (MycoSep ®229 Ochra column, MFC, Romer Labs®, Inc., MO., USA), and acidified with 70 μL of acetic acid. 4 mL of the purified extract was removed, evaporated to dryness, re-dissolved in the 200 µL of mobile phase (acetonitrile-water-acetic acid, 57: 41: 2), and injected into the HPLC.
OTA levels in each treatment were measured as previously described with minor modifications [34]. High performance liquid chromatography (HPLC) with fluorescence detection (lexc 330nm; lem 460nm) was applied. A C 18 column (Supelcosil LC-ABZ, Supelco; 150 x 4.6 mm, 5 mm particle size) was used. The mobile phase was pumped at 1.0 mL/min. The injection volume was 100 mL and the retention time was 4 min. The detection limit was 1 ng/g.

Electrochemical instruments and experimental measurements
Cyclic voltammograms and amperometric measurements were performed with an Epsilon (BAS) potentiostat controlled by electrochemical analysis software. Electrochemical measurements were carried out in a 2 mL Pyrex cell. The working electrode was a CP disk of 1.6 mm diameter obtained from Bioanalytical System (BAS), Inc. The counter electrode was a platinum foil of large area (~2 cm 2 ). An aqueous SCE was used as the reference electrode. Aliquots of 5 to 25 mL were added to the electrochemical cell for the determination of OTA content in solutions composed by comercial reagent and peanut samples contamined with the mycotoxin, respectively. Amperometric measurements were performed at a potential of −0.050 V vs SCE in solutions stirred at 1600 rpm. This operational applied potential was previously optimized by Granero et al. [36].
The Fc was used as a redox mediator. Biosensors were constructed by using the following procedure: PBHR, Fc and CP in different weight ratios were checked in order to obtain the best biosensor composition. Then, the optimal composition was used to fill up the CP electrodes. The biosensors were stored at 4ºC in a dry container when they were not in use. Experiments performed with and without bubbling pure nitrogen in solutions did not show any significant difference. Therefore, measurements were carried out in non-deoxygenated solutions. Experiments were performed at 25 ± 0,2ºC.

Response of PBHR-FC-CP biosensor versus H 2 O 2 concentration
Paste electrodes were prepared by mixing different amounts of PBHR, Fc and CP. A cyclic voltammogram recorded in unstirred pH 8.00 PBS (blank solution) for the PBHR-Fc-CP biosensor showed an increase in the anodic current at potentials higher than 0.16 V vs SCE, which corresponds to the oxidation of Fc to Fc+. A cathodic peak with a peak potential at about 0.2 V was observed when the potential sweep direction was reversed at 0.6 V, which can be assigned to the reduction of Fc + to Fc at the electrode surface ( Figure 1). On the other hand, a cyclic voltammogram recorded when H 2 O 2 was added to the reaction médium showed a cathodic current at potentials more negative than -0.3 V vs SCE, which corresponds to the reduction of H 2 O 2 at the biosensor surface. Therefore, the base current was practically zero between -0.3 and 0.2 V in pH 8.00 PBS. This behavior was similar to that previously found by our group when a similar biosensor was employed to determine t-Res using the comercial reagent [36].
On the other hand, four biosensors were constructed to optimize de amount of PBHR in the biosensor, for which the ratio of the other two components remained constant, i.e., 3 mg Fc and 45 mg CP. These biosensors were used to study their responses towars H 2 O 2 . Therefore, when only H 2 O 2 was added to the stirred reaction medium, the enzimatic reaction took place between PBHR on the electrode surface and H 2 O 2 penetrated the inner layer between the semi permeable membrane and the electrode surface. H 2 O 2 was reduced to H 2 O by PBHR and the enzyme was reduced to its native form by Fc, which was oxidized to Fc+. The Fc+ was then immediately reduced to Fc at the electrode surface held at a potential of -0.050 V vs SCE. Steady-    (Figure 3). Plots of I ss vs I ss / (Eadie-Hofstee plots) were linear and from them we calculate the kinetic parameters of the PBHR-Fc-CP biosensor for H 2 O 2 which are summarized in Table 1 ) were calculated from the intercept and the slope of Eadie-Hofstee plots for five replicate measurements. Therefore, from results shown in Table 2, biosensors constructed using 8-6-45 mg of PBHR, Fc and CP, respectively, were employed in the next experiments, considering that this electrode composition gave the maximum current value. It should be noted that when the biosensor was built with HRP instead of PBHR as the recognition element, it showed less affinity by H 2 O 2 , as has been previously reported [36]. For this reason PBHR was selected as the recognition element for the assembly of the biosensor.

Responses of PBHR-Fc-CP biosensors versus OTA concentration
When both HRP and OTA were added to the stirred reaction médium composed by pH 8.00 PBS and a given H 2 O 2 bulk concentration, the enzimatic catalytic cycle also took place in the solution bulk. The oxidized HRP, produced as a result of its reaction with H 2 O 2 , is reduced back to its native state by OTA. The decrease of H 2 O 2 in the solution bulk was detected as a decrease in the Fc+ reduction current at the biosensor surface [36]. The optimum HRP concentration in solution was 10 nmol L -1 . Thus, the addition of H 2 O 2 to the pH 8.00 PBS + 10 nmol L -1 HRP reaction medium produced a steady limiting current ( ,lim   This means that at these peroxide concentrations, the bioelectrode responses are limited by the enzymatic kinetics rate. Therefore, a concentration of 100 mmol L -1 of H 2 O 2 was chosen as the best peroxide level in the reaction medium. It should be noted that when PBHR is used as biological material in solution, after the addition of OTA there were no apreciables changes in the amperometric response, which would indicate that OTA is not a good substrate for this enzyme. On the other hand, it has been found that OTA behave as a good substrate for HRP. Thus, we decided to use HRP as biological material in solution.

Biosensor statistical parameters
The PBHR-Fc-CP biosensor reproducibility was tested by measuring the calibration curve slope for OTA of five different bioelectrodes in a solution of pH 8.00 PBS + 10 nmol L -1 HRP + 100 mmol L -1 H 2 O 2 as the reaction media. Percent relative standard deviations (%RSD) of 7.7% were obtained for OTA. The repeatability assays were performed carrying out six consecutive amperometric measurements on the same biosensor. In this case, %RSD of calibration curve slopes was 5.4%, when a PBHR-Fc-CP biosensor was used in PBS pH 8.00 + 10 nmol L -1 HRP + 100 mmol L -1 H 2 O 2 . The stability of PBHR-Fc-CP biosensor was tested by using the same biosensor to determined the slopes from several calibration curves (n = 5) for OTA. The slopes obtained were practically constant in the order of the experimental error until about five days, showing a good stability of PBHR biosensor. A noticeable decrese in the slope started from the fifth day and about 40% decrease was obtained on the seventh day. A linear variation was found for ∆I s,lim vs OTA concentration ( * OTA c ) in the range from 1x10 -8 to 1.4x10 -4 mol L −1 . The lowest concentration value measured with PBHR-Fc-CP biosensor for a signal to noise ratio of 3:1 was 5.7 x 10 -9 mol L -1 (2.3 ppb).
In order to achieve another criterion to test the accuracy of the biosensor method for OTA quantification in peanut samples the results were also investigated by the HPLC method [34]. Five samples were used to perform this analysis. The concentrations determined by HPLC-immunoaffinity column cleanup were plotted vs those obtained from the biosensor method. We found a linear variation with slope = 0.931 and correlation coefficient close to 1 (r = 0.9973) indicating a good correspondence between methods (Figure 7).
Experimental results obtained demonstrate that our enzymatic bioelectrode method is a very useful technique to detect and quantify OTA in peanut samples. The correlation between biosensor and HPLC method [34] indicates the effectiveness of the enzimatic electrode for OTA determination. While the detection limit determined is higher than values in some previous reports [38,39] still this electrode shows important advantages by using it for screening purposes due to its very simple construction and rapid measurements.

Conclusions
It is concluded that the PBHR biosensor exhibited a good performance, stability, reproducibility, repeatability, detection limit and linear range for the quantification of OTA. This good analytical performance allowed us to estimate the OTA in peanut samples using a very simple experimental procedure. The contents of OTA in samples of peanuts determined with the biosensor show a good correlation with values obtained with HPLC method. In addition, this method offers some advantages over the HPLC method, such as: short detection time (~1 min), small sample volumes, reduced use of solvents and low cost. These advantages indicate that a PBHR biosensor can be used as a useful tool for a rapid screening in the determination of OTA in peanuts samples.