Broad-specificity immunoassay for simultaneous detection of

immunoassay is simple, sensitive and specific due to their time saving, reliability, low. 69 cost and high ... ochratoxin A.31-33 To our best knowledg...
1 downloads 0 Views 4MB Size
Article pubs.acs.org/JAFC

Broad-Specificity Immunoassay for Simultaneous Detection of Ochratoxins A, B, and C in Millet and Maize Yaqiong Zhang,† Lanteng Wang,† Xing Shen,† Xiaoqun Wei,† Xinan Huang,*,§ Yingju Liu,*,‡ Xiulan Sun,⊥ Zhanhui Wang,¶ Yuanming Sun,† Zhenlin Xu,† Sergei A. Eremin,#,∥ and Hongtao Lei*,† †

Guangdong Provincial Key Laboratory of Food Quality and Safety and ‡Department of Applied Chemistry, College of Materials & Energy, South China Agricultural University, Guangzhou 510642, China § Tropical Medicine Institute & South China Chinese Medicine Collaborative Innovation Center, Guangzhou University of Chinese Medicine, Guangzhou 510405, China ⊥ State Key Laboratory of Food Science and Technology, School of Food Science of Jiangnan University, Wuxi, Jiangsu 214122, China ¶ Department of Veterinary Pharmacology and Toxicology, College of Veterinary Medicine, China Agricultural University, Beijing 100094, China # Faculty of Chemistry, M.V. Lomonosov Moscow State University, Leninskie gory 1, Building 3, Moscow 119991, Russia ∥ A.N. Bach Institute of Biochemistry, Research Center of Biotechnology of the Russian Academy of Sciences, Moscow 119071, Russia S Supporting Information *

ABSTRACT: Ochratoxins A, B, and C (OTA, OTB, and OTC) can be found in cereals and feeds; the simultaneous detection of these ochratoxins holds a great need in food safety. In this study, four antibodies raised from two ochrotoxin haptens and two coating antigens were compared, and then a sensitive and broad-specificity enzyme-linked immunosorbent assay (ELISA) was established for the simultaneous determination of three ochratoxins, where the detection limits were 0.005, 0.001, and 0.001 ng/ mL for OTA, OTB, and OTC, respectively, and recoveries of three ochratoxins were between 84.3% and 111.7%. This developed method had been successfully applied to detect ochratoxins in both millet and maize. Molecular modeling revealed that the broad-specificity was related with the chlorine electronegativity on OTA and OTC and the potential of the acetyl ester group on OTC. The proposed ELISA can be used for simultaneous detection of three ochratoxins. KEYWORDS: ochratoxin, antibody, immunoassay, broad-specificity, molecular modeling



that of OTA.3 Moreover, OTC converting to OTA after both oral and intravenous administration was readily proved, which was believed that a comparable toxicity of two toxins was based upon this conversion.10 The European Union sets a maximum limit of OTA in unprocessed cereals as 5 μg/kg and in processed cereal products as 3 μg/kg.11 The maximum limit of OTA set by Codex Alimentarius Commission (CAC) and China in the cereal is 5 μg/kg.12 Australia, France, and Romania also set the limit of OTA at 5 μg/kg in foodstuffs.13 However, the maximum limits of OTB and OTC have not been set by any country or organization. Ochratoxins are widely found in cereals and feed products; OTA was observed in the range of 10.20−46.57 μg/kg in all 18 millet samples and 0−139.2 μg/kg in 17 maize samples,14 while OTA was as high as 17−204 μg/ kg in millet samples and 3−1738 μg/kg in maize.15 In addition, OTA and OTB could be simultaneously detected in sorghum and spices samples at the same time.16−18 Although the cooccurrence of OTA, OTB, and OTC in cereals was rarely reported, three ochratoxins had been simultaneously observed

INTRODUCTION The mycotoxin ochratoxins are secondary metabolites produced by various fungi such as Aspergillus and Penicillium.1,2 Ochratoxins A−C (OTA, OTB, and OTC, respectively) are main ochratoxin analogues, while OTB and OTC are the dechlorination derivative and the ethyl esterification of OTA, respectively (Figure 1). The toxicity of these molecules is

Figure 1. Structures of orchratoxins A, B, and C.

attributable to the isocoumarin moiety and the lactone carbonyl group.3 Although OTB showed lower toxicity than that of OTA, OTA and OTB both exhibited renal toxicity, 4 hepatotoxicity,5 immunotoxicity, carcinogenic,6 and teratogenic properties.7 Compared with OTA, OTC possesses a similar or even stronger toxicity including immunomodulatory, cytotoxic, and genotoxic effects.8,9 It was reported that the inhibition effect on the growth of bacterium of OTC was just lower than © 2017 American Chemical Society

Received: Revised: Accepted: Published: 4830

February 19, 2017 May 21, 2017 May 23, 2017 May 23, 2017 DOI: 10.1021/acs.jafc.7b00770 J. Agric. Food Chem. 2017, 65, 4830−4838

Article

Journal of Agricultural and Food Chemistry

sampler (SCL-10 Avp), and a fluorescence detector (RF-10 AL). An Inertsil ODS-SP HPLC column (C18, 4.6 × 150 mm, 5 μm, GL science, Japan) was used. Conjugate Preparation. The active ester method was used for the conjugate preparation.34 Briefly, 2 mg of OTA, 1 mg of NHS, and 2 mg of EDC were sequentially dissolved in 150 μL of DMF. After mixing at room temperature overnight, the activated hapten solution was added dropwise into 300 μL of stirred solution of 10 mg of KLH in PBS and stirred for 8 h at room temperature. The mixture was dialyzed against PBS to guarantee the purification quality (see dialysis in the Supporting Information).35 Finally, the OTA-KLH solution was collected and stored at −20 °C until use. The conjugates, OTA-BSA, OTA-OVA, OTB-KLH, OTB-BSA, and OTB-OVA, were prepared by the same method above. OTA-BSA, OTA-KLH, OTB-BSA, and OTB-KLH were used for immunogens, while OTA-OVA and OTB-OVA were used for coating antigens. Antibody Generation. Four kinds of immunogens, OTA-BSA, OTA-KLH, OTB-BSA, and OTB-KLH, were injected to the rabbits, respectively. Each rabbit was injected with mixture of 0.6 mL of immunogen (1 mg/mL) and 0.6 mL of Freund’s complete adjuvant. For booster injections, the same amount of immunogen was mixed with equal volume of Freund’s incomplete adjuvant. The rabbits were injected four times total in one month intervals. After the final injection, antisera were collected at the tenth day by ELISA.36 The antisera were purified by the ammonium sulfate precipitation method for the further method development.37 The antibody immunized by OTA-BSA will be called OTA-BSA-Ab, and the others are named OTA-KLH-Ab, OTB-BSA-Ab, and OTB-KLH-Ab, respectively. ELISA Procedure. The ELISA polystyrene plates were coated with OTA-OVA or OTB-OVA (100 μL/well) in carbonate buffer (CB, 0.1 mol/L, pH 9.6) and incubated at 37 °C overnight. The plates were washed twice with PBST and blocked with blocking solution (120 μL/ well) at 37 °C for 3 h. After blocking, the plates were dried at 37 °C for 1 h. Ochratoxins were diluted with PBST (1000, 100, 10, 1, 0.1, 0.001, and 0.001 ng/mL) and then added to the wells (50 μL/well) followed by purified antibodies (50 μL/well) and then incubated at 37 °C for 40 min. After washing five times, goat antirabbit IgG-HRP (0.5 mg/mL) was diluted by 4000-times with PBST and then added to wells (100 μL/well). After another 30 min of incubation at 37 °C and five times washing, TMB solution (100 μL/well) was then added and incubated for 10 min. Finally, stop regents (10% H2SO4, v/v) (50 μL/ well) were added to the reaction. The absorbance was recorded on the microplate reader at wavelength of 450 nm. To select the broad-specificity antibody, the inhibition rate (B/B0 × 100) was used to characterize the binding ability of antibodies, where B0 and B represent the absorbance values of the negative solution and OTA standard solution (1 μg/mL), respectively.38 Then the fourparameter logistic function plotted by Origin 8.5 (Origin Lab Corp., Northampton, MA) was fitted to establish calibration curves. The detection limit (LOD) was defined as inhibitory concentration at 10% (IC10), while the dynamic detectable range was defined as the values ranging from IC20−IC80. The maximal absorbance (Amax), IC50, and the Amax/IC50 ratio values were used for the performance evaluation to gain the optimal working conditions.39 Specificity. The specificity of the ELISA was determined using OTA, OTB, and OTC, citrinin, ZEN, and AFB1 under optimized conditions. The cross-reactivity (CR) was calculated according to the following equation:

in wines.19,20 Therefore, the occurrence of OTC in food and feeds could also be paid attention for a surveillance screening purpose as well as OTA and OTB.21 The main methods previously used for the detection of ochratoxins included thin layer chromatography (TLC), highperformance liquid chromatography (HPLC),22−24 and liquid chromatography tandem mass spectrometry (LC−MS/ MS).25,26 Although chromatography is sensitive and reliable, it requires expensive instruments, complex sample pretreatments, and large amounts of organic solvents. As an alternative, immunoassay is simple, sensitive, and specific due to their time saving, reliability, low cost, and high sensitivity.27 Immunoassay methods for the detection of ochratoxin A, for example, enzyme-linked immunosorbent assay (ELISA),28 colloidal gold immunochromatographic test strips,29 and chemiluminescent enzyme immunoassay,30 were reported. In these immunoassays, the antibodies usually resulted from the immunizing hapten OTA, and the raised antibodies were applied to detect single ochratoxin A.31−33 To our best knowledge, there is not yet a broad-specific antibody reported for the simultaneous detection of OTA, OTB, and OTC. In this study, two immunizing haptens OTA and OTB combined with two kinds of carriers, bovine serum albumin (BSA) and keyhole limpet hemocyanin (KLH), were prepared for the use of four immunogens, and four antibodies were then raised. Every resultant antibody was evaluated with both homologous and heterologous formats to obtain a broadspecificity. A broad-specificity ELISA was established for the simultaneous detection of OTA, OTB, and OTC in millet and maize samples. Molecule modeling provided an insight into the broad-specific recognition mechanism, and the HPLC confirmed the reliability of the proposed immunoassay.



MATERIALS AND METHODS

Chemicals and Reagents. OTA, OTB, OTC, citrinin, zearalenone (ZEN), and aflatoxin B1 (AFB1) were purchased from Puhuashi Technology Company. Keyhole limpet hemocyanin (KLH), bovine serum albumin (BSA), ovalbumin (OVA), 1-ethyl-3-(3dimethylaminopropy) carbodiimide hydrochloride (EDC), N-hydroxylsuccinamide (NHS), Freund’s complete adjuvant, Freund’s incomplete adjuvant, N,N-dimethylformamide (DMF), 3,3′,5,5′-tetramethylbenzidine (TMB), and goat antirabbit IgG-HRP were obtained from Sigma (America). Tween-20, vitamin B1, and agar were purchased from Guangzhou Chemical Reagent Factory, China. HPLC-grade chloroform, methanol (MeOH), and acetonitrile (MeCN) (ANPEL Instrument, Shanghai, China) were used in this study. All of other reagents were of analytical grade. New Zealand white rabbits 2−3 months old (about 2 kg) were purchased from Guangdong Experimental Animal Center. Penicillium verrucosum (AS3.4517) was bought from Guangdong Microbial Culture Center (GDMCC, China). Buffers and Solutions. The carbonate buffer (CB, 0.1 mol/L, pH 9.6) was used as coating buffer, the washing solution is PBST solution containing 0.5% Tween-20 in phosphate-buffered saline (PBS, 0.01 mol/L, pH 7.4), and the blocking solution and the stop regent are 5% BSA in PBST and 10% sulfuric acid, respectively. The potato dextrose agar medium (PDA) contains 1 L of 20% potato juice, 20 g of glucose, 3 g of KH2PO4, 1.5 g of MgSO4·7H2O, 8 mg of vitamin B1, and 20 g of agar. Ultrapure water was used throughout. Instrumentation. Ultraviolet−visible spectrum was recorded on a Nanodrop 2000C spectrophotometer (Thermo, America). Polystyrene 96-well plates (KE-96−8) were purchased from Yijiamei (China). Absorbance was performed on a Multiskan Spectrum (Thermo, America). The chromatography was manipulated on the HPLC system (LC20 A, Shimadzu, Japan) with a pump (LC-10 AD), an automatic

CR (%) = IC50(analyte,ng/mL)/IC50(analogs,ng/mL) × 100 Molecular Modeling. The “SKETCH” module of SYBYL-X 2.1.1 was used to optimize the stereochemical configurations of OTA, OTB, and OTC. Then, Gasteiger-Hückel method gave the atom charge in the corresponding molecule. The Powell method was selected to minimize the energy. The Simplex part was selected in initial optimization. The criteria of the termination and max iterations were set at 0.005 kcal/(mol × Å) and 1000. The standard Tripos force field was used to minimize the overall energy by using 8 Å as the cutoff 4831

DOI: 10.1021/acs.jafc.7b00770 J. Agric. Food Chem. 2017, 65, 4830−4838

Article

Journal of Agricultural and Food Chemistry

Figure 2. Real samples. OTA positive maize sample gifted by Dr. Yu Wang (A), six samples collected from the local markets (maize, B, C, D; millet, G, H, I), four samples artificially infected by Penicillium verrucosum (maize, E, F; millet, J, K).

Table 1. Crossreactivity of Homologous and Heterologous Assay Formats (n = 3) combinationsa

analogs

OTA-KLH-Ab (OTA-OVA)

OTA OTB OTC OTA OTB OTC OTA OTB OTC OTA OTB OTC Citrinin ZEN AFB1

OTA-KLH-Ab (OTB-OVA)

OTB-BSA-Ab (OTB-OVA)

OTB-KLH-Ab (OTB-OVA)

four combinations above

LOD ± SDb (ng/mL) 0.04 7.6 0.02 0.02 1.6 0.001 0.005 0.001 0.001 0.03 0.2 0.02 NDd ND ND

± ± ± ± ± ± ± ± ± ± ± ±

0.005 0.88 0.001 0.002 0.10 -c 0.004 0.03 0.002

IC50 ± SD (ng/mL)

dynamic working range (ng/mL)

CR%

± ± ± ± ± ± ± ± ± ± ± ±

0.2−11.7 22.6−944.7 0.05−1.9 0.07−4.1 5.0−223.4 0.002−0.5 0.03−10.4 0.02−15.4 0.01−0.9 0.1−6.8 0.5−26.4 0.06−1.4 ND ND ND

100 1 403 100 2 1700 113 100 667 415 100 1217 ND ND ND

1.3 146.1 0.3 0.5 33.3 0.03 0.5 0.6 0.09 0.9 3.5 0.3 ND ND ND

0.05 9.3 0.04 0.04 4.3 0.004 0.02 0.08 0.01 0.11 0.09 0.01

a

Combinations, antibody and coating antigen combinations, coating antigens were in brackets. bSD, standard deviation. cBelow 0.001. dND, not detected.

value of the noncovalent bonding phase’s interaction. The dielectric constant was set to 1.0. On the basis of the compound conformation of lowest energy, the view part, surfaces and ribbons, and create part were sequentially selected. Then the MOLCAD was used for the molecular surface construction. Finally, the fast Connolly method was used to create the molecular electrostatic potential surface. Sample Preparation. A previously reported extraction method was applied to prepare the samples.40 Briefly, 2 g of grinded millet or maize samples was extracted with 8 mL of a mixture of acetonitrile/ water/acetic acid (79:20:1, v/v/v). The mixture was mixed for 30 s on a vortex mixer and then sonicated for 15 min. The supernatant was collected by filtering through 0.22 μm nylon-66 membranes. This extraction process was repeated twice. Then 2.6 mL of chloroform was added into the filtrate and mixed for 30 s. The mixture was then rapidly injected into a screw-cap glass centrifuge tube containing 14 mL of a 2% NaCl solution (pH 3.0). In this ternary system, 79% acetonitrile containing 1% acetic acid served as a disperser solvent, and

CHCl3 acted as an extraction solvent. This ternary system was mixed for 15 s to form a cloudy solution and then centrifuged at 3000 rpm for 8 min. The upper aqueous phase was removed, and the sedimented CHCl3 phase was collected and dried with a gentle stream of nitrogen gas at room temperature. Finally, 2 mL of MeOH was added to redissolve the residues for HPLC analysis, and this solution could be further diluted to suitable concentration with PBST for the immunoassay. Matrix Effect. Blank millet or maize samples were prepared by the method mentioned above. The extracting solution of two kinds of samples was diluted to 5-, 10-, and 15-times with PBST, respectively. Then PBST and three different dilution extracts were used to dilute OTA (1000, 100, 10, 1, 0.1, 0.001, and 0.001 ng/mL) to establish calibration curves by ELISA. To ensure the matrix effect negligible, the t test was used to compare calibration curves, and the most suitable dilution was confirmed. Recovery. To evaluate the recovery of the developed ELISA, millet and maize samples were spiked with ochratoxin. Three concentration 4832

DOI: 10.1021/acs.jafc.7b00770 J. Agric. Food Chem. 2017, 65, 4830−4838

Article

Journal of Agricultural and Food Chemistry Table 2. Comparison with Reported Antibodies for Ochratoxin Detection IC50 (ng/mL)a antibody

ELISA method

coupling method

c

ci

AE

Mab-1e Mab-2 Pab Pab Mab Mab Mab Pab Mab

ci

AE AE EDCg AE/CDI EDC AE AE MAh EDC

Pab

cd cd ci ci cd ci

d

immunogen

competing reactant

OTB-BSA OTB-KLH OTA-KLH

OTB-OVA

OTA-BSA OTA-γ-globulin OTA-BSA OTA-BSA OTA-BSA OTA-BSA OTB-CHMC-KLH

3 H-OTA OTA-HRP OTA-HRP OTA-OVA OTA-KLH OTA-OVA OTA-BSA

OTA-BSA

CR%b

OTA

OTB

OTC

OTA

OTB

OTC

ref

0.5 0.9 27 28 3 0.9 1.2 0.38 1.7 0.07 41.2

0.6 3.5 17 13 400 110 3.8 7.1 10 3.5 1.4

0.09 0.3 -f 8 0.54 -

113.0 415.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

100.0 100.0 158.8 215.4 0.8 0.8 31.6 5.4 17.0 2.0 29.4

667.0 1217.0 37.5 166.7 -

present study 44 31 45 46 32 34 33 47

a

IC50, 50% inhibiting concentration. bCR, cross-reactivity. cPab, polyclonal antibody. dAE, activated ester method. eMab, monoclonal antibody. fNot given in the articles. gEDC, carbodiimide method. hMA, mixed anhydride method. levels of OTA (2.5, 5, and 10 μg/kg), OTB (2.5, 5, and 10 μg/kg), and OTC (2.5, 5, and 8 μg/kg) were added to the millet and maize samples, respectively. Three replicates were detected. Then the samples were conducted by the above extraction method. To make the method convenient and accomplish the detection in an action for real samples, the calibration curve of OTA was also used to calculate concentrations of OTA, OTB, and OTC, besides calculated by the calibration curves of themselves. All the detectable concentrations of OTA, OTB, and OTC were expressed as OTA, OTBOTA, and OTCOTA, which were regarded as the equivalent amount of OTA. The accuracy and precision were estimated by the recovery and coefficient of variation (CV), respectively.41 The standard OTA, OTB, and OTC were dissolved in MeOH with the concentration from 0.4−80 ng/mL, respectively. HPLC analysis was performed using acetonitrile and 1% acetic acid as mobile phase. The proportions of acetonitrile: 1% acetic acid were 45:55, 32:68, and 55:45 (v/v) for OTA, OTB, and OTC at a flow rate of 1 mL/min, respectively. The excitation and emission wavelengths of fluorescence detector were 333 and 460 nm, respectively. The injection volume was 20 μL; the column temperature was 35 °C. ELISA results were further verified by HPLC. Analysis of Real Samples. Eleven samples (Figure 2) were analyzed. One OTA positive maize sample (No. 1) was gifted by Dr. Yu Wang from the Food and Environment Research Agency, China, while six samples (3 maize samples No. 2−4 and 3 millet samples No. 7−9) were collected from the local markets in Guangzhou. Four real samples (2 maize samples No. 5−6 and 2 millet samples No. 10−11) were artificially infected by Penicillium verrucosum, which was cultivated on the PDA medium and incubated for 15 days at 25 °C. Then the strains were inoculated into 20 g of millet and maize samples with two copies, respectively, mixed evenly, incubated at 25 °C, and detected at the tenth day. All 11 samples were prepared by the above extraction method. After dilution in proper times, the solution was used for ELISA and HPLC confirmation. The obtained concentrations in ELISA were expressed as equivalent amount of OTA that was evaluated from the calibration curve of OTA.



RESULTS AND DISCUSSION Conjugate Preparation. The successful conjugations were confirmed by UV−visible spectra (Figure S1). Three carrier proteins, KLH, BSA, and OVA, showed characteristic absorption at 280 nm, while chratoxins A and B showed absorption peaks at 380 and 320 nm, respectively. However, it could be found that the characteristic absorption peaks of the conjugate OTA-proteins were moved almost at 360 nm (Figure S1A), while two peaks occurred at 340 and 390 nm for OTBproteins (Figure S1B). These differences in waveforms and the

Figure 3. Lowest energy conformations and molecular electrostatic potential isosurfaces of OTA, OTB, OTC and OTA-lysine: OTA (A, B), OTB (C, D), OTC (E, F), and OTA-lysine (G, H) (white for C, red for O, blue for H, green for Cl).

maximum absorption wavelength between analytes and their corresponding conjugate indicated that immunogens were synthesized successfully.42 4833

DOI: 10.1021/acs.jafc.7b00770 J. Agric. Food Chem. 2017, 65, 4830−4838

Article

Journal of Agricultural and Food Chemistry

of OTA-KLH-Ab, OTB-BSA-Ab, and OTB-KLH-Ab were used to construct calibration curves. The obtained four calibration curves relevant to OTA were conducted (Figure S2) and compared with the values of Amax, IC50, and Amax/IC50 to evaluate the working conditions.39 It was found that OTA-KLH-Ab diluted 4000 times with OTA-OVA (7.8 ng/mL), or OTB-OVA (125 ng/mL) demonstrated best sensitivity, while OTB-BSA-Ab diluted 4000 times with OTBOVA (0.125 μg/mL), and OTB-KLH-Ab diluted 8000 with OTB-OVA (0.5 μg/mL) demonstrated acceptable performance. The IC50 values of these four combinations were 0.5, 1.3, 0.5, and 0.9 ng/mL, respectively, which were all sensitive enough to detect OTA. Specificity. The antibody specificities were evaluated with the cross-reactivity (Table 1). OTA-KLH-Ab showed low crossreactivity to OTB, no matter which coating antigen was used (below 2%). However, regardless of carrier protein of immunogen, either OTB-BSA-Ab or OTB-KLH-Ab exhibited high cross-reactivity of 113% or 415% to OTA. All four antibodies demonstrated much higher cross-reactivity to OTC (403−1700%) than that to OTA and OTB. All four antibodies showed no cross-reactivity with citrinin, ZEN, and AFB1. This suggested that the OTB antibodies showed excellent broadspecific to OTA, OTB, and OTC. Furthermore, homologous assay format (OTB-BSA-Ab with OTB-OVA) exhibited the highest sensitivity (Figure 4 A); the IC50 values to OTA, OTB, and OTC were 0.5, 0.6, and 0.09 ng/mL, respectively. The dynamic working ranges to OTA, OTB, and OTC were 0.03− 10.4, 0.02−15.4, and 0.01−0.9 ng/mL, respectively. The LODs of OTA, OTB, and OTC were 5, 1, and 1 pg/mL, respectively. Hence, this homologous format (OTB-BSA-Ab with OTBOVA) was used to establish the broad-specific ELISA. To date, although there were several investigations about antibodies against ochratoxins, all these antibodies demonstrated different performances than that of OTB antibodies in this work (Table 2). The previous antibodies raised by OTA showed low sensitivity 44 or low cross-reactivity with OTB.31−34,45,46 Although two of these antibodies exhibited cross-reactivity to OTC, the cross-reactivity to OTB was below 1%.31,45 Another reported monoclonal antibody could recognize OTB at a low IC50 (1.4 ng/mL), but the sensitivity to OTA was unsatisfactory, which might be due to the different conjugation method or Ribi adjuvant system.47 Conformation/Potential Analysis. To elucidate the recognition mechanism of OTA, OTB, and OTC, the lowest energy conformational optimizations and the electrostatic potential surfacesof ochratoxins (Figure 3) were calculated by molecular modeling (SYBYL-X 2.1.1).48 It was found that the conformation and potential of OTA and OTB were almost similar except for a chlorine atom on OTA. As is known, chlorine atom can cause the electron-withdrawing induction in the benzene ring, which can increase the density of electron clouds, resulting in electronegativity enhancement. The electron-withdrawing induction of chlorine atom may form dipolar interaction between antigen and antibody, while the benzene ring of OTB cannot react with antibody. It indicted that the chlorine atom increased a recognition site for antibody to recognize OTA. Hence, the electronegativity of the chlorine atom should be the main cause responsible for increasing the antibody binding ability to OTA, antibody binding ability to OTB was weaker than that toward OTA, because OTB has no chlorine atom.

Figure 4. Calibration curves of ochratoxins (A), sample matrix effect relevant to OTA of the millet samples (B) and the maize samples (C).

ELISA Optimization. It is well-known that assay formats can bring differences in the sensitivity of an immunoassay.43 To gain high sensitivity, the competitive indirect ELISA was evaluated based on eight combinations with four antibodies and two coating antigens to construct homologous assay (in which the immunizing hapten and the coating hapten were identical), and heterologous assay (in which the immunizing hapten and the coating hapten were different). OTA-BSA-Ab showed low inhibition rates below 65% (Table S1), no matter which coating antigen was used. For OTA-KLHAb, although heterologous strategy can improve inhibition rate to OTB (3.2-times higher) compared with homologous strategy, homologous combination was also chosen for the further investigation due to its superior inhibition rate to OTA (85.6%). However, for OTB-BSA-Ab and OTB-KLH-Ab, homologous strategy demonstrated a significantly higher inhibition rate (5−10-times higher) for detecting OTA or OTB, which might result from the strong binding ability of OTB antibodies to OTA. Therefore, one heterologous assay format of OTA-KLH-Ab and three homologous assay formats 4834

DOI: 10.1021/acs.jafc.7b00770 J. Agric. Food Chem. 2017, 65, 4830−4838

Article

Journal of Agricultural and Food Chemistry Table 3. Recovery of Ochrotoxins in Millet and Maize by ELISA and HPLC (n = 3)a ELISA samples

analogs

spiked levels (μg/kg)

millet

OTA

2.5 5 10 2.5 5 10 2.5 5 8 2.5 5 10 2.5 5 10 2.5 5 8

OTB

OTC

maize

OTA

OTB

OTC

mean recovery (%) 100.5 91.0 85.1 97.6 94.9 92.4 95.6 100.6 93.2 107.5 86.4 92.5 94.9 90.8 99.5 86.2 96.6 110.2

(67.1)d (58.9) (50.0) (377.7) (485.5) (499.7)

(65.5) (58.4) (57.7) (347.5) (463.0) (574.3)

HPLC CVb (%) 6.6 4.3 8.4 4.9 8.0 6.7 2.8 5.5 5.4 9.6 4.0 3.7 4.7 6.5 5.9 3.4 2.2 7.2

(4.5) (3.7) (1.9) (4.2) (6.5) (3.0)

(4.2) (5.7) (5.0) (4.3) (2.6) (8.0)

mean recovery (%)

CV (%)

R2c (%)

98.3 99.9 89.1 95.5 88.1 85.8 94.8 93.6 98.6 101.9 98.0 106.5 93.9 87.9 107.1 110.9 100.8 111.7

0.8 1.7 2.2 6.5 10.4 7.6 2.6 2.3 3.6 9.3 12.4 4.3 0.7 4.6 3.8 3.6 6.2 3.2

0.986

0.965

0.966

0.982

0.986

0.967

a

For one concentration, three samples were spiked and determined by developed ELISA and HPLC. bCV, coefficient of variation. cR2, determination coefficient. dRecoveries and CVs of OTBOTA and OTCOTA, which were the equivalent amounts of OTA.

OTC, which contains a chlorine atom at the benzene ring in isocoumarin and an ethyl ester group. The chlorine atom was responsible for binding force increase. Besides, ethyl ester group could improve the steric hindrance and electropositivity of molecule, which might lead to increase the binding ability to OTC. The similar resultant effect of alkyl esterification could be found in another molecular recognition, where two antibodies raised R-ofloxacin and S-ofloxacin showed strong recognition to R-methyl esterified ofloxacin (R-OFLM) and S-OFLM at 1376% and 1800%, respectively.49 The methyl ester of OFLM replaced the carboxyl in ofloxacin, which was just like the ethyl ester of OTC. The accessorial ethyl ester group eliminated the strong positive electricity of hydrogen on the carboxyl group, and then the potential was close to the lysine’s carbon chain. In addition, the ethyl ester group might also increase the hydrophobic force. The antibody might recognize the haptencarrier linking group during the antigen−antibody interaction, and then the coeffect of electrical property and hydrophobic force caused high cross-reactivity to OTC. Thus, the ethyl ester group caused positive contribution to the antigen−antibody interaction. Furthermore, since the cross-reactivity of the antibody to OTC is much higher than that of OTA and OTB, this can be ascribed to the synergistic effect of the ethyl ester and chlorine atom of OTC. Matrix Effect. In this work, the calibration curve of OTA in PBST was used as control and then compared with those calibration curves of OTA in millet and maize extracts (Figure 4). The t test results showed that most of the texp values of 15times dilution were below tcrit[4,0.05] (tcrit[4,0.05] = 2.776), indicating that no significant differences were encountered between the two methods at the 0.05 level (Table S2). This indicated that 15-times dilution was enough to make the matrix effect negligible. Taking into account of the matrix effect, the LODs for OTA, OTB, and OTC were 0.075, 0.015, and 0.015 ng/mL, which were far below the specified maximum residue limit set by EU, CAC, and China.11,12 Hence, the proposed ELISA can be used to detect the spiked and real samples.

Figure 5. Chromatogram of OTA, OTB, and OTC.

Table 4. Analysis of Ochrotoxins in Real Millet and Maize Samples by ELISA and HPLC (n = 3)a ELISA

samples

no.

maize

1 2 3 4 5 6 7 8 9 10 11

millet

observed concentration ± SD (μg/kg) 4.7 79.3 96.9 42.6 50.0

± 0.4

± 5.5 ± 11.2

± 3.1 ± 2.7

HPLC OTA ± SD (μg/kg) 4.9 62.1 70.0 31.3 40.5

± 0.3

± 0.8 ± 0.8

± 1.8 ± 2.2

OTB ± SD (μg/kg) -b 31.7 40.1 20.4 26.7

± 1.5 ± 1.4

± 0.7 ± 1.0

OTC ± SD (μg/kg) -

a

For one sample, three parallels were determined by developed ELISA and HPLC. bObserved to be negative.

The cross-reactivity of four antibodies to OTC ranged from 403−1700%, which was higher compared with those to OTA and OTB. It should be ascribed to the unique structure of 4835

DOI: 10.1021/acs.jafc.7b00770 J. Agric. Food Chem. 2017, 65, 4830−4838

Journal of Agricultural and Food Chemistry



Recovery. The millet and maize samples were spiked at three concentration levels for each ochratoxin, OTA (2.5, 5, and 10 μg/kg), OTB (2.5, 5, and 10 μg/kg), and OTC (2.5, 5, and 8 μg/kg). When the recovery of each ochrotoxin was evaluated by its own calibration curve separately, the average recoveries (Table 3) for millet samples were among 85.1− 100.6% with the CVs ranged from 2.8−8.4%, and the average recoveries for maize samples were among 86.2−110.2% with the CVs ranged from 2.2−9.6%. These results indicated that the developed immunoassay possessed both acceptable accuracy and excellent reproducibility for each kind of ochratoxins. However, in practice it is best to conduct a simultaneous detection in one convenient assay, but not three times for three analyses. Therefore, OTA calibration curve was used to calculate OTA, OTBOTA, OTCOTA, which were defined as the equivalent amounts of OTA. It was found that the average recoveries of OTBOTA for millet and maize samples were among 50.0−67.1% with CVs ranged from 1.9−5.7%, and the average recoveries of OTCOTA for millet and maize samples were among 347.5−574.3% with CVs ranged from 2.6−8.0%. The obtained concentrations of OTBOTA, OTCOTA were 0.5− 0.7- or 3−6-times the real OTB and OTC concentrations. However, it is still usable and acceptable as a screening method due to its convenience with simultaneous assay capability in one action.50 The HPLC results of standard solution of OTA, OTB, and OTC were shown in Figure 5. The retention times of OTA, OTB, and OTC were 9.68, 13.93, and 12.52 min, respectively. Recoveries of HPLC ranged from 85.8−99.9% with CVs of 0.8−10.4% (millet samples) and 87.9−111.7% with CVs of 0.7−12.4% (maize samples), respectively. The determination coefficients (R2) of ELISA and HPLC in quantification were 0.986, 0.965, and 0.966 for OTA, OTB, and OTC in millet samples, respectively. The R2 of maize samples were 0.982, 0.986, and 0.967 for OTA, OTB, and OTC, respectively. This indicated the results of ELISA and HPLC were consistent. Analysis of Real Samples. All 11 samples were detected by both HPLC and ELISA (Table 4). Only OTA was found out in the No. 1 sample at 4.7 ± 0.3 μg/kg by HPLC and 4.7 ± 0.4 μg/kg by ELISA. Six samples from local markets did not detect ochratoxins by both HPLC and ELISA, which indicated all samples from local markets were negative. The HPLC results of samples artificially infected by Penicillium verrucosum found that OTA and OTB ranged from 31.1−70.0 μg/kg and 20.4−40.1 μg/kg, respectively. The equivalent amounts of OTA of ELISA were found to range from 42.6−96.9 μg/kg evaluated with the calibration curve of OTA. The ELISA results were lower than the real amount summation of OTA and OTB, but higher than the real amount of OTA, which was consistent with the recovery results of spiked samples evaluated by the OTA calibration curve. In conclusion, four antibodies raised from two immunizing ochrotoxin haptens were tested in this work. Homologous combination of OTB-BSA-Ab with OTB-OVA showed high sensitivity and broad-specificity to OTA, OTB, and OTC. A broad-specific ELISA was established and demonstrated a good correlation with HPLC in spiked and real millet and maize samples. The chlorine atom and ethyl ester group may play an important role in the broad-specific recognition. The proposed ELISA can be used as a rapid screening and high-throughput method for the simultaneous detection of ochratoxins A, B, and C.

Article

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jafc.7b00770. UV absorption curve of conjugates; optimization of ELISA working condition; inhibition for different combinations of antibody and coating antigens; t test results of millet and maize matrix effect (PDF)



AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. Phone: +86 20-36585475. Fax: +86 20-8637-3516. (Xinan Huang) *E-mail: [email protected]. Phone/Fax: +86 20-85282366. (Yingju Liu) *E-mail: [email protected]. Phone: +86 20-8528-3448. Fax: +86 20-8528-0270. (Hongtao Lei) ORCID

Zhanhui Wang: 0000-0002-0167-9559 Hongtao Lei: 0000-0002-1697-1747 Funding

This work was supported by Natural Science Foundation of China (U1301214, 21475047, 30700663, 31601555) and Guangdong and Guangzhou Planned Program in Science and Technology (S2013030013338, 2016201604030004, 2014TX01N250, 2013B051000072, and 2014A030306026). Notes

The authors declare no competing financial interest.



ABBREVIATIONS USED Ab, antibody; AFB1, aflatoxin B1; BSA, bovine serum albumin; DMF, N,N-dimethylformamide; ELISA, enzyme-linked immunosorbent assay; EDC, 1-ethyl-3-(3-(dimethylamino)propyl) carbodiimide hydrochloride; HPLC, high performance liquid chromatography; KLH, keyhole limpet hemocyanin; LOD, limit of detection; NHS, N-hydroxysuccinimide; OTA, ochratoxin A; OTB, ochratoxin B; OTC, ochratoxin C; OVA, ovalbumin; SD, standard deviation; ZEN, zearalenone



REFERENCES

(1) Stander, M. A.; Steyn, P. S.; Lübben, A.; Miljkovic, A.; Mantle, P. G.; Marais, G. J. Influence of halogen salts on the production of the ochratoxins by Aspergillus ochraceus Wilh. J. Agric. Food Chem. 2000, 48, 1865−1871. (2) Gareis, M.; Gareis, E. M. Guttation droplets of Penicillium nordicum and Penicillium verrucosum contain high concentrations of the mycotoxins ochratoxin A and B. Mycopathologia 2007, 163, 207−214. (3) Xiao, H.; Madhyastha, S.; Marquardt, R. R.; Li, S. Z.; Vodela, J. K.; Frohlich, A. A.; Kemppainen, B. W. Toxicity of ochratoxin A, its opened lactone form and several of its analogs: Structure-activity relationships. Toxicol. Appl. Pharmacol. 1996, 137, 182−192. (4) Heussner, A. H.; Dietrich, D. R.; O’Brien, E. In vitro investigation of individual and combined cytotoxic effects of ochratoxin A and other selected mycotoxins on renal cells. Toxicol. In Vitro 2006, 20, 332− 341. (5) Zheng, Z.; Hanneken, J.; Houchins, D.; King, R. S.; Lee, P.; Richard, L. J. Validation of an ELISA test kit for the detection of ochratoxin A in several food commodities by comparison with HPLC. Mycopathologia 2005, 159, 265−272. (6) Heussner, A. H.; O’Brien, E.; Dietrich, D. R. Effects of repeated ochratoxin exposure on renal cells in vitro. Toxicol. In Vitro 2007, 21, 72−80. 4836

DOI: 10.1021/acs.jafc.7b00770 J. Agric. Food Chem. 2017, 65, 4830−4838

Article

Journal of Agricultural and Food Chemistry

(27) Fujii, S.; Ono, E. Y. S.; Ribeiro, R. M. R.; Assunçaõ , F. G. A.; Takabayashi, C. R.; Oliveira, T. C. R. M.; Itano, E. N.; Ueno, Y.; Kawamura, O.; Hirooka, E. Y. A comparison between enzyme immunoassay and HPLC for ochratoxin a detection in green, roasted and instant coffee. Braz. Arch. Biol. Technol. 2007, 50, 349−359. (28) Thirumala-Devi, K.; Mayo, M. A.; Reddy, G.; Reddy, S. V.; Delfosse, P.; Reddy, D. V. Production of polyclonal antibodies against ochratoxin A and its detection in chilies by ELISA. J. Agric. Food Chem. 2000, 48, 5079−5082. (29) Liu, B. H.; Tsao, Z. J.; Wang, J. J.; Yu, F. Y. Development of a monoclonal antibody against ochratoxin A and its application in enzyme-linked immunosorbent assay and gold nanoparticle immunochromatographic strip. Anal. Chem. 2008, 80, 7029−7035. (30) Jo, E. J.; Mun, H.; Kim, S. J.; Shim, W. B.; Kim, M. G. Detection of ochratoxin A (OTA) in coffee using chemiluminescence resonance energy transfer (CRET) aptasensor. Food Chem. 2016, 194, 1102− 1107. (31) Chu, F. S.; Chang, F. C.; Hinsdill, R. D. Production of antibody against ochratoxin A. Appl. Environ. Microb. 1976, 31, 831−835. (32) Zhang, X.; Sun, M.; Kang, Y.; Xie, H.; Wang, X.; Song, H.; Li, X.; Fang, W. Identification of a high-affinity monoclonal antibody against ochratoxin A and its application in enzyme-linked immunosorbent assay. Toxicon 2015, 106, 89−96. (33) Wang, X. H.; Liu, T.; Xu, N.; Zhang, Y.; Wang, S. Enzymelinked immunosorbent assay and colloidal gold immunoassay for ochratoxin A: investigation of analytical conditions and sample matrix on assay performance. Anal. Bioanal. Chem. 2007, 389, 903−911. (34) Zhang, A.; Ma, Y.; Feng, L.; Wang, Y.; He, C.; Wang, X.; Zhang, H. Development of a sensitive competitive indirect ELISA method for determination of ochratoxin A levels in cereals originating from Nanjing, China. Food Control 2011, 22, 1723−1728. (35) Oplatowska, M.; Connolly, L.; Stevenson, P.; Stead, S.; Elliott, C. T. Development and validation of a fast monoclonal based disequilibrium enzyme-linked immunosorbent assay for the detection of triphenylmethane dyes and their metabolites in fish. Anal. Chim. Acta 2011, 698, 51−60. (36) Luo, L.; Xu, Z. L.; Yang, J. Y.; Xiao, Z. L.; Li, Y. J.; Beier, R. C.; Sun, Y. M.; Lei, H. T.; Wang, H.; Shen, Y. D. Synthesis of novel haptens and development of an enzyme-linked immunosorbent assay for quantification of histamine in foods. J. Agric. Food Chem. 2014, 62, 12299−12308. (37) Page, M.; Thorpe, R. Purification of IgG by precipitation with sodium sulfate or ammonium sulfate. In The Protein Protocols Handbook; Springer, 1996; pp 721−722. (38) Zou, X.; Chen, C.; Huang, X.; Chen, X.; Wang, L.; Xiong, Y. Phage-free peptide ELISA for ochratoxin A detection based on biotinylated mimotope as a competing antigen. Talanta 2016, 146, 394−400. (39) Oubiña, A.; Ballesteros, B.; Galve, R.; Barcelo, D.; Marco, M. P. Development and optimization of an indirect enzyme-linked immunosorbent assay for 4-nitrophenol. application to the analysis of certified water samples. Anal. Chim. Acta 1999, 387, 255−266. (40) Lai, X. W.; Sun, D. L.; Ruan, C. Q.; Zhang, H.; Liu, C. L. Rapid analysis of aflatoxins B1, B2, and ochratoxin A in rice samples using dispersive liquid-liquid microextraction combined with HPLC. J. Sep. Sci. 2014, 37, 92−98. (41) Zhang, Z.; Li, Y.; Li, P.; Zhang, Q.; Zhang, W.; Hu, X.; Ding, X. Monoclonal antibody-quantum dots CdTe conjugate-based fluoroimmunoassay for the determination of aflatoxin B1 in peanuts. Food Chem. 2014, 146, 314−319. (42) Wang, X. C.; Bao, M.; Wu, J. J.; Luo, Y.; Ma, L. Y.; Wang, Y.; Zhang, A. H.; He, C. H.; Zhang, H. B. Characterization and Comparison of Ochratoxin A-Ovalbumin (OTA-OVA) Conjugation by Three Methods. J. Integr. Agric. 2014, 13, 1130−1136. (43) Jinqing, J.; Zhang, H.; An, Z.; Xu, Z.; Yang, X.; Huang, H.; Wang, Z. Development of an Heterologous Immunoassay for Ciprofloxacin Residue in Milk. Phys. Procedia 2012, 25, 1829−1836. (44) Uchigashima, M.; Yamaguchi Murakami, Y.; Narita, H.; Nakajima, M.; Miyake, S. Development of an immuno-affinity column

(7) O’Brien, E.; Prietz, A.; Dietrich, D. R. Investigation of the teratogenic potential of ochratoxin A and B using the FETAX system. Birth Defects Res., Part B 2005, 74, 417−423. (8) Müller, G.; Rosner, H.; Rohrmann, B.; Erler, W.; Geschwend, G.; Gräfe, U.; Burkert, B.; Möller, U.; Diller, R.; Sachse, K.; Köhler, H. Influence of the mycotoxin ochratoxin A and some of its metabolites on the human cell line THP-1. Toxicology 2003, 184, 69−82. (9) Muller, G.; Burkert, B.; Moller, U.; Diller, R.; Rohrmann, B.; Rosner, H.; Kohler, H. Ochratoxin A and some of its derivatives modulate radical formation of porcine blood monocytes and granulocytes. Toxicology 2004, 199, 251−259. (10) Fuchs, R.; Hult, K.; Peraica, M.; Radic, B.; Plestina, R. Conversion of ochratoxin C into ochratoxin A in vivo. Appl. Environ. Microb. 1984, 48, 41−42. (11) Commision Regulation (EC) No 1881/2006 of 19 December 2006 Setting Maximum Levels for Certain Contaminants in Foodstuffs; European Union, (EU), 2006. (12) Codex General Standard for Contaminants and Toxins in Food and Feed (Codex Stan 193−1995); Codex Alimentarius Commission, (CAC), 1995. (13) Mazumder, P. M.; Sasmal, D. Mycotoxins − Limits and Regulations. Ancient Science of Life 2001, 1−19. (14) Makun, H. A.; Adeniran, A. L.; Mailafiya, S. C.; Ayanda, I. S.; Mudashiru, A. T.; Ojukwu, U. J.; Jagaba, A. S.; Usman, Z.; Salihu, D. A. Natural occurrence of ochratoxin A in some marketed Nigerian foods. Food Control 2013, 31, 566−571. (15) Sangare-Tigori, B.; Dem, A. A.; Kouadio, H. J.; Betbeder, A. M.; Dano, D. S.; Moukha, S.; Creppy, E. E. Preliminary survey of ochratoxin A in millet, maize, rice and peanuts in Côte d’Ivoire from 1998 to 2002. Hum. Exp. Toxicol. 2006, 25, 211−216. (16) Elbashir, A. A.; Ali, S. E. A. Aflatoxins ochratoxins and zearalenone in sorghum and sorghum products in Sudan. Food Addit. Contam., Part B 2014, 7, 135−140. (17) Scheuer, R.; Gareis, M. Occurrence of ochratoxin A and B in spices. Mycotoxin Res. 2002, 18, 62−66. (18) Han, Z.; Zheng, Y.; Luan, L.; Ren, Y.; Wu, Y. Analysis of ochratoxin A and ochratoxin B in traditional Chinese medicines by ultra-high-performance liquid chromatography-tandem mass spectrometry using [(13)C(20)]-ochratoxin A as an internal standard. J. Chromatogr. A 2010, 1217, 4365−4374. (19) Remiro, R.; González-Peñas, E.; Lizarraga, E.; López de Cerain, A. Quantification of ochratoxin A and five analogs in Navarra red wines. Food Control 2012, 27, 139−145. (20) Remiro, R.; Ibáñez-Vea, M.; González-Peñas, E.; Lizarraga, E. Validation of a liquid chromatography method for the simultaneous quantification of ochratoxin A and its analogues in red wines. J. Chromatogr. A 2010, 1217, 8249−8256. (21) Zimmerli, B.; Dick, R. Ochratoxin A in table wine and grapejuice: occurrence and risk assessment. Food Addit. Contam. 1996, 13, 655−668. (22) Sultan, Y.; Magan, N.; Medina, A. Comparison of five different C18 HPLC analytical columns for the analysis of ochratoxin A in different matrices. J. Chromatogr. B: Anal. Technol. Biomed. Life Sci. 2014, 971, 89−93. (23) Lai, X.; Ruan, C.; Liu, R.; Liu, C. Application of ionic liquidbased dispersive liquid-liquid microextraction for the analysis of ochratoxin A in rice wines. Food Chem. 2014, 161, 317−322. (24) Di Stefano, V.; Pitonzo, R.; Avellone, G.; Di Fiore, A.; Monte, L.; Ogorka, A. Z. T. Determination of aflatoxins and ochratoxins in Sicilian sweet wines by high-performance liquid chromatography with fluorometric detection and immunoaffinity cleanup. Food Anal. Method 2015, 8, 569−577. (25) Jeswal, P.; Kumar, D. Mycobiota and Natural Incidence of Aflatoxins, Ochratoxin A, and Citrinin in Indian Spices Confirmed by LC-MS/MS. Int. J. Microbiol. 2015, 2015, 242486. (26) Lau, B. P.; Scott, P. M.; Lewis, D. A.; Kanhere, S. R. Quantitative determination of ochratoxin A by liquid chromatography/electrospray tandem mass spectrometry. J. Mass Spectrom. 2000, 35, 23−32. 4837

DOI: 10.1021/acs.jafc.7b00770 J. Agric. Food Chem. 2017, 65, 4830−4838

Article

Journal of Agricultural and Food Chemistry for ochratoxin analysis using an organic solvent-tolerant monoclonal antibody. Methods 2012, 56, 180−185. (45) Yu, F. Y.; Chi, T. F.; Liu, B. H.; Su, C. C. Development of a sensitive enzyme-linked immunosorbent assay for the determination of ochratoxin A. J. Agric. Food Chem. 2005, 53, 6947−6953. (46) Cho, Y. J.; Lee, D. H.; Kim, D. O.; Min, W. K.; Bong, K. T.; Lee, G. G.; Seo, J. H. Production of a monoclonal antibody against ochratoxin A and its application to immunochromatographic assay. J. Agric. Food Chem. 2005, 53, 8447−8451. (47) Heussner, A. H.; Auslander, S.; Dietrich, D. R. Development and characterization of a monoclonal antibody against ochratoxin B and its application in ELISA. Toxins 2010, 2, 1582−1594. (48) Wang, Z.; Zhu, Y.; Ding, S.; He, F.; Beier, R. C.; Li, J.; Jiang, H.; Feng, C.; Wan, Y. P.; Zhang, S.; Kai, Z.; Yang, X.; Shen, J. Development of a monoclonal antibody-based broad-specificity ELISA for fluoroquinolone antibiotics infoods and molecular modeling studies of cross-reactive compounds. Anal. Chem. 2007, 79, 4471− 4483. (49) Mu, H.; Lei, H.; Wang, B.; Xu, Z.; Zhang, C.; Ling, L.; Tian, Y.; Hu, J.; Sun, Y. Molecular modeling application on hapten epitope prediction: an enantioselective immunoassay for ofloxacin optical isomers. J. Agric. Food Chem. 2014, 62, 7804−7812. (50) Xu, Z. L.; Zeng, D. P.; Yang, J. Y.; Shen, Y. D.; Beier, R. C.; Lei, H. T.; Wang, H.; Sun, Y. M. Monoclonal antibody-based broadspecificity immunoassay for monitoring organophosphorus pesticides in environmental water samples. J. Environ. Monit. 2011, 13, 3040− 3048.

4838

DOI: 10.1021/acs.jafc.7b00770 J. Agric. Food Chem. 2017, 65, 4830−4838