Sandwich ELISA Assay for the Quantitation of Palytoxin and Its

Article pubs.acs.org/est

Sandwich ELISA Assay for the Quantitation of Palytoxin and Its Analogs in Natural Samples S. Boscolo,† M. Pelin,‡ M. De Bortoli,‡ G. Fontanive,§ A. Barreras,‡ F. Berti,§ S. Sosa,‡ O. Chaloin,∥ A. Bianco,∥ T. Yasumoto,⊥ M. Prato,# M. Poli,▽ and A. Tubaro*,† †

Department of Life Sciences, University of Trieste, Via L. Giorgieri 10, 34127 Trieste, Italy Department of Life Sciences, University of Trieste, Via A. Valerio 6, 34127 Trieste, Italy § Department of Chemical and Pharmaceutical Sciences, University of Trieste, Via L. Giorgieri 1, 34127 Trieste, Italy ∥ CNRS, Institut de Biologie Moléculaire et Cellulaire, Laboratoire d’Immunologie et Chimie Thérapeutiques, 67000 Strasbourg, France ⊥ Japan Food Research Laboratories, Tama Laboratory, 6-11-10 Nagayama, Tama-shi, Tokyo 206-0025 Japan # Department of Chemical and Pharmaceutical Sciences, University of Trieste, P.le Europa 1, 34127 Trieste, Italy ▽ U.S. Army Medical Research Institute of Infectious Diseases, Ft Detrick, Maryland, 21701-5011, United States ‡

S Supporting Information *

ABSTRACT: Palytoxins are potent marine biotoxins that have recently become endemic to the Mediterranean Sea, and are becoming more frequently associated with seafood. Due to their high toxicity, suitable methods to quantify palytoxins are needed. Thus, we developed an indirect sandwich ELISA for palytoxin and 42-hydroxy-palytoxin. An intralaboratory study demonstrated sensitivity (limit of detection, LOD = 1.1 ng/mL; limit of quantitation, LOQ = 2.2 ng/mL), accuracy (bias of 2.1%), repeatability (RSDr = 6% and 9% for intra- and interassay variability, respectively) and specificity: other common marine toxins (okadaic acid, domoic acid, saxitoxin, brevetoxin-3, and yessotoxin) do not cross-react in this assay. It performed well in three different matrices: observed LOQs were 11.0, 9.6, and 2.4 ng/mL for mussel extracts, algal net samples and seawater, respectively, with good accuracy and precision. The LOQ in seafood is 11 μg palytoxin/kg mussel meat, lower than that of the most common detection technique, LC-MS/MS.

■

INTRODUCTION Palytoxins are hydrophilic, nonproteinaceous polyether biotoxins of MW around 2680 Da, and include at least 10 congeners. Originally isolated from soft corals, they are produced by tropical and subtropical dinoflagellates of the genus Ostreopsis and tropical marine cyanobacteria of the genus Trichodesmium.1,2 In the past decades, climate changes and/or anthropogenic factors, such as eutrophication and ballast water transfer, have extended these organisms into Mediterranean Sea temperate waters, where they now thrive.3,4 The toxicological properties of PLTXs, which interact with the Na+/K+ ATPase, are well-described.1,2 Cutaneous and ocular toxicity of PLTXs, as well as their possible involvement in respiratory problems associated with marine aerosols exposure during Ostreopsis blooms, have been reported.5−8 However, the main human health concern is PLTXcontaminated seafood ingestion.9−12 Indeed, PLTXs can be found in tropical fish and crabs whose consumption causes a poisoning characterized by general malaise, weakness, myalgia, respiratory problems, impairment of the neuromuscular apparatus and/or of cardiac functions.2 © 2013 American Chemical Society

Human intoxications from PLTXs in the Mediterranean Sea have not been reported, so far. However, increasing reports of PLTXs in seafood (shellfish, sea urchins, octopi) cause serious concern, as well as raising possible ecotoxicological implications.4 Shellfish collected in the North Aegean Sea (Greece) were contaminated by PLTX-like compounds during 2004− 2006 Ostreopsis blooms.12 Similarly, wild mussels (Mytilus galloprovincialis) from the rocky Campania coast (Italy) were shown to contain PLTX-like compounds during O. ovata monitoring in 2007−2009.4,13 Moreover, as a precaution, France prohibited sea urchin collection from April to November due to the presence of PLTXs.14 In spite of such reports, no validated and accepted protocols for PLTXs detection and quantification are available, and the existing methods present limitations in sensitivity and specificity.15 Moreover, other factors, such as costs, speed of Received: Revised: Accepted: Published: 2034

October 16, 2012 January 21, 2013 January 22, 2013 January 22, 2013 dx.doi.org/10.1021/es304222t | Environ. Sci. Technol. 2013, 47, 2034−2042

Environmental Science & Technology

Article

Figure 1. Chemical structures of PLTX and of its analogues 42-hydroxy-palytoxin and ostreocin-D.

both enzyme linked immunosorbent assays (ELISA) and a radioimmunoassay have been reported, but never commercialized or characterized for their suitability in PLTXs monitoring in seafood.22,25−27 Recently, an immunoassay based upon surface plasmon resonance biosensor was proposed.28 Although rapid and sensitive, it requires sophisticated, expensive equipment and technical expertise not often available for monitoring programs. Thus, an indirect sandwich ELISA, based upon a mouse antipalytoxin monoclonal antibody (capture antibody) and a rabbit antipalytoxin polyclonal antibody (detection antibody), was developed and characterized by an intralaboratory study. The assay is able to detect PLTX and 42hydroxy-PLTX in a similar way. It is sensitive, specific, accurate and precise, with limits of detection (LOD) and quantitation (LOQ) suitable for monitoring seafood, below the limit suggested by European Food Safety Authority (EFSA).29

the analysis, equipment, and technical expertise limit the suitability of these methods for the specific purpose of seafood safety monitoring. Until recently, the simplest and most common method for palytoxin detection was the mouse bioassay,16 with or without modifications.17 However, in addition to ethical considerations, it suffers from significant variability due mainly to extraction issues, being originally developed for lipophilic toxins. Currently, a positive mouse bioassay result requires confirmatory analysis using liquid chromatography-mass spectrometry (LC-MS). However, methods for routine LC-MS analysis of PLTXs in natural samples have yet to be fully developed.18,19 One recently developed LCMS/MS method20 requires cleanup and oxidation steps that seem to limit its suitability for routine use. A method based on PLTX binding to its molecular target, Na+/K+-ATPase, and a fluorescence polarization technique have been recently reported,21 while a variety of cytotoxicity assays have been proposed. However, all require expertise and facilities not typically associated with monitoring programs, while cytotoxicity of many marine toxins makes specificity an issue. Another method is the hemolytic assay, based upon PLTXs ability to hemolyze mouse erythrocytes, but false positives by other natural compounds must be ruled out through hemolysis neutralization by ouabain or specific antibodies.15,22−24 In general, immunoassays have the advantages of speed, specificity, and ease-of-use in monitoring situations when large numbers of samples require rapid analysis. Anti-PLTX antibodies have been produced by several investigators and

■

MATERIALS AND METHODS Toxins and Other Materials. PLTX was purchased from Wako Chemicals (Neuss, Germany); okadaic and domoic acids were from Sigma Aldrich (Milan, Italy). 42-hydroxy-palytoxin was isolated from Palythoa toxica as previously reported.30 Yessotoxin, brevetoxin-3 and saxitoxin were kindly provided by Prof. T. Yasumoto (Japan Food Research Laboratories; Tokyo, Japan), Dr. M. Poli (U.S. Army Medical Research Institute of Infectious Diseases; Ft. Detrick, MD) or Dr. F. Van Dolah (National Oceanic and Atmospheric Administration; Charleston, SC). Multiwell strips were from Nunc (Langenselbold, Germany); Amicon Ultra-15 centrifugal filter devices were from 2035

dx.doi.org/10.1021/es304222t | Environ. Sci. Technol. 2013, 47, 2034−2042


Article

Surface Plasmon Resonance. All biosensor assays were performed with HEPES-buffered saline (HBS-P) as running buffer (10 mM HEPES, 150 mM sodium acetate, 3 mM magnesium acetate, 0.005% surfactant P20, pH 7.4). Immobilization of streptavidin was performed by injecting streptavidin (40 μL, 100 μg/mL in formate buffer, pH 4.3) onto the surface of a CM5 chip activated by EDC/NHS according to the manufacturer’s directions. This was followed by 20 μL of ethanolamine hydrochloride, pH 8.5, to saturate the free activated sites of the matrix. Biotinylated palytoxin (10 μM in formate buffer, pH 4.3) was allowed to interact with streptavidin until a response of 300 RU was obtained. Immobilization of PLTX and of an unrelated monoclonal Ab as negative control (mAb Sc433) were performed by injection onto the EDC/NHS-activated surface of a CM5 sensor chip, 35 μL of ligands (50 μg/mL in acetate buffer, pH 4.9) which gave a signal of 300 and 1000 RU, respectively, followed by 20 μL of ethanolamine hydrochloride, pH 8.5, to saturate the free activated matrix sites. Anti-PLTX mAb was immobilized through the cysteine thiol groups using 35 μL of 2-(2-pyridinyl-dithioethaneamine)(PDEA) in 50 mM borate buffer pH 8.3 on the NHS/EDCactivated matrix. Then 35 μL of mAb (100 μg/mL in formate buffer, pH 4.3) were injected until a response of 6000 RU was obtained. Finally, 20 μL of 50 mM cysteine, 1 M NaCl solution was used to saturate the unoccupied sites on the chip. All binding experiments were carried out at 25 °C with a constant flow rate (20 μL/min). Different concentrations of antibody, PLTX or analogues were injected for 3 min, followed by a 3 min dissociation phase. The sensor chip surface was regenerated after each experiment injecting 10 μL of 10 mM HCl. Indirect Sandwich ELISA. Anti-PLTX mAb was diluted in PBS (20 μg/mL) immediately before use. Multiwell strips coated with 100 μL/well of the antibody solution were incubated for 16 h at 4 °C. Each well was washed with PBS and blocked for 1 h at room temperature with 200 μL of 2% skimmed milk (w/v) dissolved in PBS containing 0.1% Tween 20 (PBS-Tw). The wells were then incubated for 2 h at room temperature with different concentrations of toxins (100 μL) diluted in PBS-Tw containing ovalbumin (1 mg/mL). After washing with PBS-Tw, followed by washing with PBS, 100 μL of purified pAb-PLTX (0.17 μg/mL) were added and the wells incubated for 2 h at room temperature. After washing, 100 μL of HRP-conjugated goat antirabbit pAbs (1:2000) were added to each well and incubated for 1 h at room temperature. All antibodies were diluted in blocking solution. After washing, 3,3′,5,5′-tetramethylbenzidine liquid substrate (60 μL) was added to each well and the developing reaction was stopped after 30 min by 30 μL of 1 M H2SO4. The absorbance was read at 450 nm (Spectra photometer; Tecan Italia; Milan, Italy). Evaluation of Matrix Effect and Recovery. To evaluate the applicability of the ELISA to quantify PLTX in shellfish, microalgae and seawater, three different matrices were prepared as described below. Each matrix was analyzed by LC-MS/MS to confirm PLTX absence before matrix effect and PLTX recovery evaluation. Mussels. Mussels (Mytilus galloprovincialis) were collected in the Gulf of Trieste (Italy). Shucked meat (200 g) was homogenized by Ultra-Turrax (Ika-Werk; Staufen, Germany). Mussel homogenate (1 g) was extracted three times by homogenization (14 000 rpm, 3 min) with 3 mL of 80% aqueous MeOH19 followed by centrifugation at 5500 rpm for

Millipore (Vimodrone, Italy); cellulose membranes for dialysis were from Spectrum Laboratories (Rancho Dominguez, CA); horseradish peroxidase (HRP)-conjugated polyclonal antirabbit goat antibodies were from DakoCytomation (Milan, Italy). The Biacore 3000 system, sensor chip CM5, surfactant P20, amine coupling kit containing N-hydroxysuccinimide (NHS) and Nethyl-N′-dimethylaminopropyl carbodiimide (EDC), were from GE-Healthcare (Uppsala, Sweden). Other materials and chemicals were from Sigma Aldrich. Biotinylation of Palytoxin. PLTX was biotinylated at the terminal primary amino group (Figure 1) by reaction with biotin via N-hydroxysuccinimide activation. PLTX (4.8 mg) was dissolved in 1 mL of water. Biotin N-hydroxysuccinimide ester (0.5 mg) was added and the solution was stirred at room temperature for 24 h. The solution was then washed with dichloromethane (3 × 0.5 mL) and the formation of biotinylated PLTX was confirmed by mass spectrometry. Analysis (ESI) demonstrated an m/z of 2929.2 [M+Na]+, which corresponds to the sodium-associated molecular ion and shows the expected increase of 226 mass units from the molecular weight of PLTX+Na (2703). Biotinylated PLTX was stored at −20 °C until use. Preparation of the Immunogenic PLTX Conjugate. PLTX conjugation was performed at the terminal primary amino group (Figure 1). Bovine serum albumin (BSA) conjugate was prepared via SMCC-IMT [succinimidyl-4-(Nmaleimidomethyl)cylclohexane-1-carboxylate-2-iminothiolane] two-step protocol,31−36 as previously reported.37 Polyclonal Anti-PLTX Antibody Production. Rabbit polyclonal anti-PLTX antibodies (pAb-PLTX) were produced according to Bignami et al.22 Rabbits were immunized by PLTX-BSA, as previously reported.37 Polyclonal anti-PLTX antibodies were affinity purified from rabbit sera diluted 1:10 (v/v) in 10 mM Tris buffer pH 7.5, using a 2 mL Sepharose CNBr (0.45 g) immunoaffinity column. The gel was swelled in 15 volumes of 1 mM HCl at 0 °C, washed exhaustively with 0.5 M phosphate buffer pH 7.5, and coupled with 3 mg of PLTX in 0.5 M phosphate buffer pH 7.5, overnight. After washing with two volumes of the same buffer, another washing step was carried out with 5 mM phosphate buffer and 1 M NaCl, pH 7.5. The gel was then saturated overnight with 100 mM ethanolamine buffer pH 7.5 at 0 °C. Subsequent washings were carried out with 10 volumes of each of the following buffers: 10 mM Tris pH 7.5, 100 mM glycine pH 2.5, 10 mM Tris pH 8.8, 100 mM triethylamine pH 11.5 and 10 mM Tris pH 7.5. Aliquots (2 mL) of rabbit immune serum were diluted 1:10 in 10 mM Tris buffer pH 7.5 and loaded on the Sepharose gel column by gravity. The column was then washed with 20 volumes of 10 mM Tris buffer pH 7.5 and 20 volumes of 10 mM Tris buffer pH 7.5 and 500 mM NaCl. Glycine (100 mM, pH 2.5) was used as elution buffer. Fractions (1 mL) were collected in 100 μL 1 M Tris buffer, pH 8.0, and dialyzed in 2 l PBS overnight (6−8000 Da cellulose membrane). Dialyzed fractions were concentrated using Amicon Ultra-15 centrifugal filter devices, following the manufacturer’s directions, and final antibody concentrations were evaluated spectrophotometrically (280 nm). Monoclonal Anti-PLTX Antibody Production. Mouse monoclonal anti-PLTX antibody 73D3 (mAb-PLTX) was produced and purified from a hybridoma cell culture as previously described22 at the U.S. Army Medical Research Institute of Infectious Diseases (Fort Detrick, MD, USA). 2036



Article

Table 1. Kinetic Parameters of SPR Interactions: ka, Kinetic Association Rate; kd, Kinetic Dissociation Rate; KD, Equilibrium Dissociation Constant; Rmax Maximal Response Expressed As Resonance Units; Chi,2 Chi Square Value ligand

analyte

ka (1/Ms)

PLTX PLTX-biotin mAb mAb

mAb mAb PLTX 42-OH-PLTX

1.97 × 105 1.51 × 105 3.22 × 105 256

kd (1/s) 3.83 3.72 9.47 1.75

× × × ×

KD (M)

10−5 10−5 10−4 10−3

1.94 2.46 2.94 6.84

× × × ×

10−10 10−10 10−9 10−6

Rmax (RU)

Chi2

17 52 102 67

0.1 1.2 0.4 5.9

the curves of its analogues were compared by two-way ANOVA statistical analysis and Bonferroni post test (GraphPad Prism, GraphPad Software, Inc.; San Diego, CA) and significant differences were considered at p < 0.05. EC50 (effective concentration giving 50% of the maximal response) was calculated by nonlinear regression using a four parameters curve-fitting algorithm of the SigmaPlot software (Jandel Scientific). Limits of detection (LOD) and quantitation (LOQ), accuracy, precision and repeatability parameters were calculated according to the international principles as described by Eurachem Guide.39 LOD and LOQ were expressed as the analyte concentration corresponding to the average of 10 blank values plus 3 or 10 times the standard deviations, respectively. Accuracy was measured as % Bias (n = 10), calculated as % difference between PLTX concentration measured by the assay and the theoretical concentration in the sample. Repeatability was expressed as relative standard deviation of repeatability (RSDr), measured as % ratio between the standard deviation of independent results and their mean value. Both independent results obtained by the same operator in one day (intra-assay RSDr; n = 10) and within a 6-month period by different operators (interassay RSDr; n = 10) were considered.

30 min. The supernatants were pooled, gently dried under vacuum, and the residue was dissolved in 1 mL 80% aqueous MeOH, obtaining 1 g meat equivalents/ml. Seawater and Microalgae. A seawater sample (8 l; 9.1 × 106 algal cells/l) free of Ostreopsis sp. and containing mainly diatoms (Proboscia alata, Guinardia f laccida, Chaetoceros sp., and Hemiaulus hauckii, among the most abundant) and dinoflagellates (Ceratium f usus, C.furca, Dinophysis sacculus, Hermesinum adriaticum, and Protoperidinium sp., among the most abundant) was collected by several vertical net hauls (20 μm mesh) in the Gulf of Trieste (Italy). After centrifugation at 5000 rpm for 20 min, the pellet was separated and used as microalgae matrix, while the supernatant was used as seawater matrix. Microalgae were extracted following the procedure reported by Ciminiello’s group.38 Microalgae (1 g) were sonicated with 4 mL of 50% aqueous MeOH for 3 min in pulse mode, while cooling in ice bath. After centrifugation at 5500 rpm for 30 min, the supernatant was decanted, and the pellet was washed twice with 3 mL of 50% aqueous MeOH. The extracts were combined, and the volume was adjusted to 10 mL with 50% aqueous MeOH, obtaining 0.1 g microalgae equivalents/ml. Seawater (500 mL) was extracted three times by liquid/ liquid partition with equal volumes of BuOH. The butanol phases were pooled, evaporated under vacuum, and the pellet was dissolved in 10 mL of 50% aqueous MeOH.38 Evaluation of Matrix effect. Dilutions (1:1, 1:10, 1:100) of PLTX-free extracts of mussels, microalgae and seawater were spiked with known PLTX amounts to prepare a series of matrix matched-samples at PLTX levels ranging from 1.25 to 80 ng/ mL. Each sample was then analyzed by the ELISA, as previously described. Evaluation of PLTX Recovery. Samples of PLTX-free mussels, microalgae and seawater were spiked with PLTX (before extraction), and then extracted as above to obtain samples with different degrees of contamination (mussels: 1 g at 125.0, 250.0, 375.0 ng PLTX/g; microalgae: 1 g at 31.2, 62.5, 125.0, 250.0 ng PLTX/g; seawater: 500 mL at 0.062, 0.125, 0.250, and 0.500 ng/mL). Each sample was then analyzed by the ELISA, as previously described. Statistical Analysis. SPR kinetic parameters were calculated using the BIAeval 4.1 software. Analysis was performed using the simple 1:1 langmuir binding model or separate ka/kd (kon/koff). The specific binding profiles were obtained after subtracting the response signal from the channel control (mAb Sc433 for antibody ligands or ethanolamine for PLTX ligands) and from blank buffer injection. The fitting to each model was judged by the reduced chi square and randomness of residue distribution compared to the theoretical model. Results of ELISA assay are presented as mean (±SEM) from at least three independent experiments performed in duplicate. Linearity (r2) of the calibration curve was estimated by linear regression analysis, using the SigmaPlot software (Jandel Scientific; Erkrath, Germany). PLTX calibration curve and

■

RESULTS

Anti-PLTX mAb Affinity for Palytoxins. The affinity of the mAb-PLTX for PLTX and its analogues was assessed by surface plasmon resonance (SPR). Two series of experiments were performed. Initially, PLTX was immobilized to the surface of the sensor chip in two ways: through its terminal amino group or by covalently linking streptavidin to the sensor chip and the subsequent complexation with PLTX-biotin. The binding constant of the mAb-PLTX was obtained by fluxing the antibody over the PLTX-functionalized SPR surfaces at different concentrations (Supporting Information (SI) Figure S1). The values of the association and dissociation rate constants and the affinity KD are reported in Table 1. The mAb-PLTX binds with a high affinity (KD ∼ 2 × 10−10 M) to PLTX immobilized on the chip, either directly or through streptavidin/biotin interaction. In the second series of experiments, the mAb-PLTX was immobilized on the sensor chip exploiting the available free cysteine residues, which were selectively allowed to react with a maleimido group previously introduced on the chip. In this case, the recognition capacity of the antibody toward PLTX and its analogues was measured. Different concentrations of PLTX and 42-OH-PLTX in the fluid phase were tested. PLTX bound with high affinity to the mAb with a KD in the nanomolar range (Table 1; SI Figure S2). 42-OH-PLTX bound to mAb with a KD in the micromolar range, with an affinity about 3 orders of magnitude lower than that of PLTX (Table 1; SI Figure S2). ELISA Calibration Curve for PLTX. Using the indirect sandwich ELISA for PLTX detection, based on the mAb-PLTX (capturing reagent) and the rabbit pAb-PLTX (detecting 2037



Article

Figure 2. Calibration curve for PLTX. Each point represents the mean ± SE of 3 different experiments (A). The working range was analyzed by linear regression plotting theoretic PLTX concentrations against PLTX concentrations measured by the assay; n = 10 (B).

reagent), the calibration curve reported in Figure 2A was obtained. The working range for PLTX detection was 1.25−80 ng/mL, with an EC50 (effective concentration giving 50% of the maximal response) of 7.6 ± 1.1 ng/mL. The calculated limits of detection (LOD) and quantification (LOQ) were 1.1 and 2.2 ng/mL, respectively. The working range has been analyzed by linear regression, plotting the theoretical PLTX concentrations against toxin concentrations measured by the assay, with an excellent correlation coefficient (r2 = 0.9933; n = 10)(Figure 2B). A mean Bias value of 2.1% (range: −2.8 to 7.0%; SI Table S1) denotes the accuracy of the measures. Intra-assay repeatability was estimated over 10 replicates (n = 10; SI Figure S3A) and interassay repeatability was evaluated over a six months period (n = 10; SI Figure S3B): very good correlations were obtained with linearity values of r2 = 0.9933 (SI Figure S3A) for intra-assay and of r2 = 0.9807 for interassay (SI Figure S3B), as well as very good repeatability coefficients (RSDr = 6 and 9%, respectively; SI Table S2). Cross-Reactivity with PLTX Analogues. The ability of the ELISA to detect the natural PLTX analogue 42-OH-PLTX and biotinylated PLTX was studied at concentrations within the working range of PLTX (1.25−80 ng/mL). The reactivity of biotinylated PLTX was similar to that of PLTX, whereas the calibration curve of 42-OH-PLTX was only slightly lower (Figure 3). Cross-Reactivity with Other Marine Toxins. Crossreactivity was evaluated analyzing other marine algal toxins structurally unrelated to PLTXs, which can contaminate seafood (okadaic acid, yessotoxin, domoic acid, brevetoxin-3, and saxitoxin). These toxins were tested at concentrations up to 100 μg/mL, that is, about 4 orders of magnitude higher than PLTX EC50. No significant cross-reactivity was detected to any toxin at any concentration. Matrix Effects. To evaluate the interference of different matrices on ELISA performance, extracts from mussels, microalgae and seawater were prepared. Preliminarily, extracts of toxin-negative mussels, microalgae and seawater were tested after 1:1, 1:10, and 1:100 dilutions to evaluate whether extracts matrix could produce false positive signals. No significant absorbance at 450 nm was detected at any dilution (SI Table S3). Thus, subsequent studies to verify the influence of matrix on PLTX detection were carried out at these dilutions. In particular, extracts dilutions (1:1, 1:10, 1:100) were spiked with PLTX at final concentrations ranging from 1.25 to 80 ng/mL,

Figure 3. Standard curves for PLTX and analogues. Each point represents the mean ± SE of 3 different experiments. Statistical differences with respect to PLTX: **p < 0.01; ***p < 0.001 (two-way ANOVA and Bonferroni post test).

and analyzed by the ELISA in comparison to the same PLTX concentrations, without matrices. The minimum extract dilution not interfering with the assay was 1:10 for all the matrices (SI Figure S4). Figure 4 shows the linear regression between the results obtained analyzing the 10-fold diluted mussels (A), microalgae (B) and seawater (C) extracts spiked with PLTX (1.25−80 ng/ mL) and those obtained analyzing the same PLTX concentrations without matrix. The LOQ for PLTX in mussels was 11 ng/mL, corresponding to 11 μg/kg meat. To characterize the accuracy of the ELISA for PLTX detection in mussels, the correlation coefficient and the Bias value were calculated from the linear regression analysis (r2 = 0.9663; Bias = 1.5%; SI Table S4). The LOQ for PLTX in microalgae and seawater was 9.6 and 2.4 ng/mL, respectively. The relevant correlation coefficients were 0.9878 and 0.9719, with mean Bias values of 5.0% and 8.0%, respectively (SI Table S4). Recovery of PLTX from Different Matrices. Recovery experiments were performed to determine the efficiency of the toxin extraction from different matrices. Aliquots of PLTX-free mussel meat homogenate were spiked with 125, 250, and 375 ng PLTX/g meat and extracted as above to obtain extracts containing 1 g meat equivalents/ml (12.5, 25.0, 37.5 ng PLTX/ mL). Extracts were diluted 1:10 and analyzed by ELISA. The 2038



Article

Table 2. Recovery of PLTX from Matrices Analyzed by Sandwich ELISA PLTX concentration

recovery (%)

coefficient of variability (%)

n° replica

Mussels 125.0 ng/g 250.0 ng/g 375.0 ng/g

95.7 98.2 97.2

31.2 ng/g 62.5 ng/g 125.0 ng/g 250.0 ng/g

96.6 105.5 101.7 95.6

1.0 1.1 2.6

4 4 4

12.3 4.9 0.9 7.4

4 4 4 4

29.2 19.5 4.7 9.1

4 4 4 4

Microalgae

Seawater 0.062 0.125 0.250 0.500

ng/mL ng/mL ng/mL ng/mL

72.6 59.2 85.3 57.3

PLTX recovery from seawater was evaluated spiking the samples (500 mL) with the toxin at 0.062, 0.125, 0.250, and 0.500 ng/mL that, after extraction, correspond to 3.12, 6.25, 12.5, and 25.0 ng PLTX/ml extract. The toxin recovery from seawater ranged from 57.3 to 85.3% (coefficient of variability: 4.7−29.2%)(Table 2).

■

DISCUSSION The recent appearance of Ostreopsis cf. ovata in the Mediterranean Sea and the consequent presence of PLTXs in edible shellfish, crustaceans, and echinoderms4,12,13 have highlighted new issues concerning the toxicological potential of these toxins. Exposure to PLTXs could occur through different routes, the most dangerous being the ingestion of contaminated seafood: several cases of oral human poisonings have been reported from tropical or subtropical regions, sometimes with fatal outcomes.2 Despite the extent of the contaminated area, few methods are currently available for PLTXs detection and quantitation in seafood. Moreover, among PLTXs, only palytoxin is commercially available, though expensive, and no certified standard material is currently sold. For monitoring purposes, a combination of screening methods followed by a chemical confirmatory analysis, such as LC-MS, is commonly used to detect PLTXs.15 Other methods for PLTX analysis include mouse bioassay, cytotoxicity assays, hemolysis assays, receptor binding assays, and immunoassays.15,21−28,40 However, most of them lack specificity or have other limitations. Thus, a method based on the well established techniques of indirect sandwich ELISA has been set up. The developed ELISA is based on a mouse mAb as capturing reagent, which was preliminarily evaluated for its affinity for PLTX and the available naturally occurring analog, 42-hydroxypalytoxin. As determined by SPR, the mAb binds to PLTX with high affinity (Kd about 2 × 10−10 M), either when the toxin is directly immobilized to the chip or when biotinylated PLTX is immobilized through streptavidin. This was confirmed by antibody immobilization to the chip and flushing PLTX over SPR surfaces. These results are in agreement with those reported by Yakes et al.28 In the same way, mAb affinity for the PLTX analog 42-hydroxy-PLTX was estimated to be 6.84 × 10−6 M. Thus, the mAb is still able to bind PLTX after biotinylation and its analogue 42-hydroxy-PLTX.

Figure 4. Matrix effects: linear regression analysis within the working range of the ELISA (1.25−80 ng PLTX/ml) performed on matrices containing 0.1 g mussel meat equivalents/ml (A), 0.01 g microalgae equivalents/ml (B) and 5 mL seawater equivalents/ml (C). Theoretical PLTX concentrations (spiked concentration) have been plotted against measured PLTX concentrations (six replicates per matrix).

recovery of PLTX ranged between 95.7% and 98.2%, with a coefficient of variability within the range of 1.0−2.6% (Table 2). The recovery of PLTX from algae was evaluated spiking the relevant samples with 31.5, 62.5, 125, and 250 ng PLTX/g that, after extraction, correspond to 3.12, 6.25, 12.5, and 25.0 ng PLTX/ml extract. The toxin recovery from algae was around 100% (range: 95.6−105.5%; coefficient of variability: 0.9− 12.3%)(Table 2). 2039



Article

range between 95.7% and 104.6% (mussels), 95.6−105.5% (microalgae) and 57.3−85.3% (seawater), with a good precision. These results further support the ELISA suitability for preliminary PLTXs screening in shellfish during monitoring programs. A just developed LC-MS/MS method for PLTXs detection20 has a LOQ of 10 μg/kg shellfish meat, but it requires sample processing steps as well as instrumental and personal skills that can be problematic in monitoring programs. This assay could also be applied to algal net samples to identify PLTXs-producing algal species prior to blooming: in fact, similarly to other dinoflagellates, Ostreopsis do not always produce the same amount of toxins, if any.37 Although the assay cannot distinguish between PLTX and its analogs, it can be used as an early warning method for preliminary screening in monitoring programs, before confirmatory chemical analyses on ELISA positive samples. Furthermore, in one multiwall plate, more than 10 samples can be simultaneously analyzed with low cost equipment in about 6 h. This ELISA can act as an inexpensive early warning system to prevent human illness without the unnecessary economic losses for shellfish producers that can follow premature closing of fisheries due to nontoxic dinoflagellate strains. Similarly, economic losses for the tourism industry can be avoided, such as those in Provence (France) after hotel cancellations at the first report on Ostreopsis in seawater.14 In conclusion, a sensitive, specific, accurate and precise indirect sandwich ELISA has been developed to quantify PLTX in mussels (LOQ = 11 ng/mL), microalgae (LOQ = 9.6 ng/ mL) and seawater (LOQ = 2.4 ng/mL), and a single laboratory characterization was performed. It provides a useful detection method that does not require expensive instruments or specialized operators, and could be suitable for PLTX detection in monitoring programs or as self-control directly by the producers, when large numbers of samples have to be quickly processed.

The ELISA was set up with the mAb as capture reagent and the rabbit pAb as detection reagent. The working range for PLTX detection is 1.25−80 ng/mL, with LOD and LOQ of 1.1 and 2.2 ng/mL, respectively. The assay demonstrates very good accuracy (mean Bias = 2.1%) and very good intra-assay and interassay repeatability (RSDr = 6% and 9%, respectively). Within its working range, it detects also the natural analogue 42-OH-PLTX and PLTX-biotin. Due to lack of availability, at present we cannot evaluate in our assay the reactivity of ostreocin-D (42-hydroxy-3,26-didemethyl-19,44-dideoxy-palytoxin) and ovatoxin-a, the structure of which was very recently elucidated as 42-hydroxy-17,44,64-trideoxy PLTX.41,42 However, indirect evidence of ovatoxin-a binding by both the mAb and pAb was previously obtained by immunocytochemistry: the antibodies directly detected ovatoxin-a in single Ostreopsis cf. ovata cells, in which this toxin was the most abundant and no PLTX was found by high resolution LC-MS.37 Thus, we can predict that this assay should detect also ovatoxin-a, the major PLTX analog in Mediterranean Sea.41 We cannot address the question of cross-reactivity of ostreocin-D. However, this toxin has to date not been detected in Mediterranean seafood, despite the occurrence of Ostreopsis siamensis.42 The ELISA is also specific for PLTXs, as shown by the lack of reactivity toward other chemically unrelated algal toxins implicated in seafood contamination (okadaic acid, yessotoxin, domoic acid, brevetoxin-3, saxitoxin), even at concentrations 4 orders of magnitude higher than PLTX EC50. Thus, the specificity of this method will avoid false positives by these toxins during monitoring programs. The ELISA sensitivity (about 1 ng/mL) is of the same order of magnitude of other immunoassays,22,26,28 but lower than that of an immunoassay (0.5 pg/mL) recently developed using single-chain antibodies isolated by phage display technology.28 Despite the excellent sensitivity of the latter, the variable toxin recovery from mussels (64−113%) and clams (84−181%) will likely impair its suitability for monitoring purposes. On the contrary, the matrix interferences by mussel, microalgae and seawater extracts in our assay were minimal. Just 10-fold dilution of mussels, microalgae, and seawater extracts eliminated matrix effects. At this dilution, the accuracy appears very good and the LOQ in mussels (11 μg/kg) is far below that (228 μg/kg) recently determined by LC-MS/MS using the same extraction procedure,19 as well as below the safety limit suggested by the European Food Safety Authority in shellfish (30 μg/kg).29 Unfortunately, we are unable to compare the ELISA LOQ with that of all of the other methods published so far, due to the lack of such details in most reports.15,18,25,26 Yakes et al.28 reported a LOD of 1.4 ng/mL for PLTX in clam and 2.8 ng/ mL in grouper extracts, but the latter contained only 0.025 g meat/mL and the assay was not evaluated at higher matrix concentrations. The recent assay, based on PLTX binding to Na+/K+-ATPase and fluorescence polarization technique detection,21 has an instrumental LOD (2 nM, corresponding to 5.36 ng/mL) about 5-folds higher than that of this ELISA. Furthermore, the toxin recovery from mussels (86.6%) was lower than that we determined (95.7−104.6%). Moreover, it requires equipment often unavailable in monitoring situations. In net microalgae (without Ostreopsis cf. ovata) and seawater the LOQs for PLTX were 9.6 and 2.4 ng/mL, respectively, lower than those of LC-MS (127 and 131 ng/mL, respectively).38 Recoveries of PLTX from the three matrices, an index of both extraction efficiency and ELISA accuracy,

■

ASSOCIATED CONTENT

S Supporting Information *

Figure S1 shows SPR sensorgrams of mAb binding to PLTX or PLTX-biotin. Figure S2 shows SPR sensorgrams of PLTX and analogs binding to mAb. Figure S3 shows repeatability of the assay. Figure S4 shows mussels matrix effect on PLTX detection within the working range of the ELISA after different dilutions. Table S1 reports Bias values for PLTX analysis by ELISA. Table S2 reports intra-assay and interassay ELISA repeatability. Table S3 reports absorbance of PLTXs-negative matrices extracts after different dilutions. Table S4 reports Bias values for spiked extracts of mussels, microalgae and seawater diluted 1:10. This material is available free of charge via the Internet at http:// pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*Phone: +390405588835; fax: +390405583215; e-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by Regione Autonoma Friuli-Venezia Giulia, Direzione Risorse Rurali, Agroalimentari e Forestali. 2040



Article

(16) Yasumoto, T.; Oshima, Y.; Yamaguchi, M. Occurrence of a new type of toxic shellfish poisoning in the Tohoku district. Bull. Jpn. Soc. Sci. Fish. 1978, 44, 1249−1255. (17) EC (European Commission). Commission decision of 15 March 2002 laying down detailed rules for the implementation of Council Directive 91/492/ECC as regards the maximum levels and the methods of analysis of certain marine biotoxins in bivalve molluscs, echinoderms, tunicates and marine gastropods (2002/225/EC). Off. J. Eur. Commun. 2002, 175, 62−64. (18) Ciminiello, P.; Dell’Aversano, C.; Dello Iacovo, E.; Fattorusso, E.; Forino, M.; Tartaglione, L. LC-MS of palytoxin and its analogues: State of the art and future perspectives. Toxicon 2011, 57, 376−389. (19) Ciminiello, P.; Dell’Aversano, C.; Dello Iacovo, E.; Fattorusso, E.; Forino, M.; Tartaglione, L.; Rossi, R.; Soprano, V.; Capozzo, D.; Serpe, L. Palytoxin in seafood by liquid chromatography tandem mass spectrometry: Investigation of extraction efficiency and matrix effect. Anal. Bioanal. Chem. 2011, 401, 1043−1050. (20) Selwood, A. I.; van Ginkel, R.; Harwood, D. T.; McNabb, P. S.; Rhodes, L. R.; Holland, P. T. A sensitive assay for palytoxins, ovatoxins and ostreocins using LC-MS/MS analysis of cleavage fragments from micro-scale oxidation. Toxicon 2012, 60, 810−820. (21) Alfonso, A.; Fernández-Araujo, A.; Alfonso, C.; Caramés, B.; Tobio, A.; Louzao, M. C.; Vieytes, M. R.; Botana, L. M. Palytoxin detection and quantification using the fluorescence polarization technique. Anal. Biochem. 2012, 424, 64−70. (22) Bignami, G. S.; Raybould, T. J.; Sachinvala, N. D.; Grothaus, P. G.; Simpson, S. B.; Lazo, C. B.; Byrnes, J. B.; Moore, R. E.; Vann, D. C. Monoclonal antibody-based enzyme-linked immunoassays for the measurement of palytoxin in biological samples. Toxicon 1992, 30, 687−700. (23) Riobó, P.; Paz, B.; Franco, J. M.; Vázquez, J.; Murado, M. A. Proposal for a simple and sensitive haemolytic assay for palytoxin. Toxicological dynamics, kinetics, ouabain inhibition and thermal stability. Harmful Algae 2008, 7, 415−429. (24) Seemann, P.; Gernert, C.; Schmitt, S.; Mebs, D.; Hentschel, U. Detection of hemolytic bacteria from Palythoa caribaeorum (Cnidaria, Zoantharia) using a novel palytoxin-screening assay. Antonie Van Leeuwenhoek 2009, 96, 405−411. (25) Levine, L.; Fujiki, H.; Gjika, H. B.; Vanvunakis, H. A radioimmunoassay for palytoxin. Toxicon 1988, 26, 1115−1121. (26) Frolova, G. M.; Kuznetsova, T. A.; Mikhailov, V. V.; Eliakov, G. B. Enzyme linked immunosorbent assay for detecting palytoxinproducing bacteria. Russian J. Bioorg. Chem. 2000, 26, 285−289. (27) Garet, E.; Cabado, A. G.; Vieites, J. M.; Gonzalez-Fernandez, A. Rapid isolation of single-chain antibodies by phage display technology directed against one of the most potent marine toxins: Palytoxin. Toxicon 2010, 55, 1519−1526. (28) Yakes, B. J.; Degrasse, S. L.; Poli, M.; Deeds, J. R. Antibody characterization and immunoassays for palytoxin using an SPR biosensor. Anal. Bioanal. Chem. 2011, 400, 2865−2869. (29) EFSA (European Food Safety Authority). Scientific opinion on marine biotoxins in shellfish − palytoxin group. EFSA J. 2009, 7 (12) (1−38), 1393. (30) Ciminiello, P.; Dell’Aversano, C.; Dello Iacovo, E.; Fattorusso, E.; Forino, M.; Grauso, L.; Tartaglione, L.; Florio, C.; Lorenzon, P.; De Bortoli, M.; Tubaro, A.; Poli, M.; Bignami, G. S. Stereostructure and biological activity of 42-hydroxy-palytoxin: A new palytoxin analogue from Hawaiian Palythoa subspecies. Chem. Res. Toxicol. 2009, 22, 1851−1859. (31) Yoshitake, S.; Imagawa, M.; Ishikawa, E.; Niitsu, Y.; Urushizaki, I.; Nishiura, M.; Kanazawa, R.; Kurosaki, H.; Tachibana, S.; Nakazawa, N.; Ogawa, H. Mild and efficient conjugation of rabbit Fab’ and horseradish peroxidase using a maleimide compound and its use for enzyme immunoassay. J. Biochem. 1982, 92, 1413−1424. (32) Ishikawa, E.; Imagawa, M.; Hashida, S.; Yoshitake, S.; Hamaguchi, Y.; Ueno, T. Enzyme-labeling of antibodies and their fragments for enzyme immunoassay and immunohistochemical staining. J. Immunoassay 1983, 4, 209−327.

M.P. is grateful to the Italian Ministry of Education (cofin Prot. 20085M27SS). A.B and O.C. are grateful to the Centre National de la Recherche Scientifique (France). WE are grateful to Prof. G. Honsell (University of Udine, Italy) for providing and characterization of algal samples and to Dr. E. D’Orlando (University of Trieste, Italy) for her assistance.

■

REFERENCES

(1) Kerbrat, A. S.; Amzil, Z.; Pawlowiez, R.; Golubic, S.; Sibat, M.; Darius, H. T.; Chinain, M.; Laurent, D. First evidence of palytoxin and 42-hydroxy-palytoxin in the marine cyanobacterium Trichodesmium. Mar. Drugs 2011, 9, 543−560. (2) Tubaro, A.; Durando, P.; Del Favero, G.; Ansaldi, F.; Icardi, G.; Deeds, J. R.; Sosa, S. Case definitions for human poisonings postulated to palytoxins exposure. Toxicon 2011, 57, 478−495. (3) Ciminiello, P.; Dell’Aversano, C.; Dello Iacovo, E.; Fattorusso, E.; Forino, M.; Grauso, L.; Tartaglione, L.; Guerrini, F.; Pistocchi, R. Complex palytoxin-like profile of Ostreopsis ovata. Identification of four new ovatoxins by high-resolution liquid chromatography/mass spectrometry. Rapid Commun. Mass Spectrom. 2010, 24, 2735−2744. (4) Aligizaki, K.; Katikou, P.; Milandri, A.; Diogène, J. Occurrence of palytoxin-group toxins in seafood and future strategies to complement the present state of the art. Toxicon 2011, 57, 390−399. (5) Durando, P.; Ansaldi, F.; Oreste, P.; Moscatelli, P.; Marensi, L.; Grillo, C.; Gasparini, R.; Icardi, G. Ostreopsis ovata and human health: Epidemiological and clinical features of respiratory sindrome outbreaks from a two-year syndromic surveillance, 2005−2006, in north-west Italy. EuroSurveillance 2007, 12(6); http://www.eurosurveillance.org/ ew/2007/070607.asp#1. (6) Tichadou, L.; Glaizal, M.; Armengaud, A.; Grossel, H.; Lemée, R.; Kantin, R.; Lasalle, J.-R.; Drouet, G.; Rambaud, L.; Malfait, P.; De Haro, L. Health impact of unicellular algae of the Ostreopsis genus blooms in the Mediterranean Sea: Experience of the French Mediterranean coast surveillance network from 2006 to 2009. Clin. Toxicol. 2010, 48, 839−844. (7) Moshirfar, M.; Khalifa, Y. M.; Espandar, L.; Miffin, M. D. Aquarium coral keratoconjunctivitis. Arch. Ophtalmol. 2010, 128, 1360−1362. (8) Deeds, J. R.; Handy, S. M.; White, K. D.; Reimer, J. D. Palytoxin found in Palythoa sp. zoanthids (anthozoa, hexacorallia) sold in the home aquarium trade. PlosOne 2011, 6, e18235. (9) Deeds, J. R.; Schwartz, M. D. Human risk associated with palytoxin exposure. Toxicon 2010, 56, 150−162. (10) Alcala, C. C.; Alcala, L. C.; Garth, J. S.; Yasumura, D.; Yasumoto, T. Human fatality due to the ingestion of the crab Demania reynaudii that contained a palytoxin-like toxin. Toxicon 1988, 26, 105−107. (11) Onuma, Y.; Satake, M.; Ukena, T.; Roux, J.; Chanteau, S.; Rasolofonirina, N.; Ratsimaloto, M.; Naoki, H.; Yasumoto, T. Identification of putative palytoxin as the cause of clupeotoxism. Toxicon 1999, 37, 55−65. (12) Aligizaki, K.; Katikou, P.; Nikolaidis, G.; Panou, A. First episode of shellfish contamination by palytoxin-like compounds from Ostreopsis species (Aegean Sea, Greece). Toxicon 2008, 51, 418−427. (13) ARPAC (Agenzia Regionale Protezione Ambientale Campania). Il monitoraggio dell’Ostreopsis ovata lungo il litorale della Campania (giugno-agosto 2007). 2008. http://www.arpacampania.it/ pubblicazioni.asp (accessed September 14, 2012). (14) Lemée, R.; Mangialiajo, L.; Cohu, S.; Amzil, Z.; Blanfuné, A.; Chomerat, N.; Ganzin, N.; Gasparini, S.; Grossel, H.; Guidi-Guilvard, L.; Hoareau, L.; le Duff, F.; Marro, S.; Simon, N.; Nezan, E.; Pedrotti, M. L.; Sechet, V.; Soliveres, O.; Thibaut, T. Interactions between scientists, managers and policy makers in the framework of the French MediOs project on Ostreopsis (2008−10). Cryptogam.: Algol. 2012, 33, 137−142. (15) Riobό, P.; Franco, J. M. Palytoxins: Biological and chemical determination. Toxicon 2011, 57, 368−375. 2041



Article

(33) Partis, M. D.; Griffiths, D. G.; Roberts, G. C.; Beechey, R. B. Crosslinking of proteins by omega-maleimido alkanoyl N-hydroxysuccinimide esters. J. Protein. Chem. 1983, 2, 263−277. (34) Hashida, S.; Imagawa, M.; Inoue, S.; Ruan, K. H.; Ishikawa, E. More useful maleimide compounds for the conjugation of Fab’ to horseradish peroxidase through thiol groups in the hinge. J. Appl. Biochem. 1984, 6, 56−63. (35) Brinkley, M. A. A survey of methods for preparing protein conjugates with dyes, haptens and crosslinking reagents. Bioconjugate Chem. 1992, 3, 2−13. (36) Mattson, G.; Conklin, E.; Desai, S.; Nielander, G.; Savage, M. D.; Morgensen, S. A practical approach to crosslinking. Mol. Biol. Rep. 1993, 17, 167−183. (37) Honsell, G.; De Bortoli, M.; Boscolo, S.; Dell’Aversano, C.; Battocchi, C.; Fontanive, G.; Penna, A.; Berti, F.; Sosa, S.; Yasumoto, T.; Ciminiello, P.; Poli, M.; Tubaro, A. Environ. Sci. Technol. 2011, 45, 7051−7059. (38) Ciminiello, P.; Dell’Aversano, C.; Fattorusso, E.; Forino, M.; Magno, G. S.; Tartaglione, L.; Grillo, C.; Melchiorre, N. The Genoa 2005 outbreak. Determination of putative palytoxin in Mediterranean Ostreopsis ovata by a new liquid chromatography tandem mass spectrometry method. Anal. Chem. 2006, 78, 6153−6159. (39) Eurachem Guide. The fitness for purpose of analytical methods. A laboratory guide to method validation and related topics. 1988. http://www.eurachem.ul.pt (accessed September 14, 2012). (40) Zamolo, V. A.; Valenti, G.; Venturelli, E.; Chaloin, O.; Marcaccio, M.; Boscolo, S.; Castagnola, V.; Sosa, S.; Berti, F.; Fontanive, G.; Poli, M.; Tubaro, A.; Bianco, A.; Paolucci, F.; Prato, M. Highly sensitive electrochemiluminescent nanobiosensor for the detection of palytoxin. ACS Nano 2012, 25, 7989−7997. (41) Ciminiello, P.; Dell’Aversano, C.; Dello Iacovo, E.; Fattorusso, E.; Forino, M.; Grauso, L.; Tartaglione, L.; Guerrini, F.; Pezzolesi, L.; Pistocchi, R.; Vanucci, S. Isolation and structure elucidation of ovatoxin-a, the major toxin produced by Ostreopsis ovata. J. Am. Chem. Soc. 2012, 134, 1869−1875. (42) Parsons, M. L.; Aligizaki, K.; Dechraoui Bottein, M. Y.; Fraga, S.; Morton, S. L.; Penna, A.; Rhodes, L. Gambierdiscus and Ostreopsis: Reassessment of the state of knowledge of their taxonomy, geography, ecophysiology, and toxicology. Harmful Algae 2012, 14, 107−129.

2042


Sandwich ELISA Assay for the Quantitation of Palytoxin and Its

Recommend Documents