Anal. Chem. 1999, 71, 4198-4202
Development of an Ultrasensitive Immunoassay for Rapid Measurement of Okadaic Acid and Its Isomers Mark P. Kreuzer, Ciara K. O’Sullivan, and George G. Guilbault*
Laboratory of Sensor Development, Department of Analytical Chemistry, NUIC, Cork, Ireland
This report highlights the characteristics of an okadaic acid immunoassay with limits of detection in the subfemtomole range. Two different immunoassay formats were investigated and their characteristics compared in relation to linear ranges, limits of detection, and cross-reactivity with other seafood toxins present in water and/or mussel samples. The developed ELISA system can be manipulated to quantitatively measure total diarrhetic shellfish poisoning (DSP) content or for okadaic acid and dinophysistoxin-1 individual concentrations by variation of the format of the immunoassay. Real mussel samples were validated in percentage recovery test. Calibration curves were established, and aliquots of real samples were tested. Very good recoveries were attained, highlighting the validity of the ELISA system to accurately determine the DSP concentration in mussel samples. Each year, a large number of people fall victims to some sort of poisoning induced after ingestion of low-molecular-weight marine toxins. These toxins accumulate or are sometimes metabolized by the shellfish, which digest the planktonic algae (dinoflagellates). There are a number of families responsible for the poisonings, the symptoms and severity of which are different for each family. Paralytic shellfish poisoning (PSP) is associated with all derivatives of saxitoxin. Diarrhetic shellfish poisoning (DSP) is caused by a group of polyether toxins including okadaic acid, the dinophysistoxins, pectenotoxin, and yessotoxin. Neurotoxic shellfish poisoning (NSP) is the result of exposure to a group of polyethers called brevetoxins, and amnesic shellfish poisoning (ASP) is caused by the unusual amino acid, domoic acid. The symptoms after ingestion of contaminated shellfish are varied and largely dependent upon the concentration in the shellfish and the consumed amount. A summary can be seen in Table 1. During a seven-year period, between 1976 and 1982, 1300 cases of DSP were recorded in Japan, not to mention those which were unreported due to the similarity in symptoms to gastrointestinal infection.1 In 1981, 5000 cases were identified in Spain2 in that one year alone. (1) U.S. Food and Drug Administration Centre for Food Safety & Applied Nutrition. Foodborne pathogenic microorganisms and natural toxins (Bad Bug Book); U.S. FDA: Rockville, MD, 1992. (2) Campos, M. J.; Fraga, S., Marino, J.; Sanchez, F. J. International Council for the Exploration of the Sea Report; U.S. Dept. of State, U.S> Government Printing Office: Washington, DC, 1982; Vols. 1-8, pp 1977-1981. (3) Kat, M. In Toxic Dinoflagellate Blooms; Taylor, D. L., Seliger, H. H., Eds.; Elsevier, North-Holland: New York, 1979; pp 215-220.
4198 Analytical Chemistry, Vol. 71, No. 19, October 1, 1999
Sporadic occurrences have also been recorded in The Netherlands,3 Chile,4 and Norway.5 This exhibits a worldwide, geographical distribution, and its frequency and effects have brought about global concern. Along with the poisonings to humans, the financial scale of such a toxin bloom on the fishing industry can be overwhelming. To deal with the aforementioned problems, it has become necessary to devise practical and sensitive detection methods for these toxins. The toxicity in Europe is mainly due to the polyether toxin, okadaic acid (OA), which accumulates in the hepatopancreas of bivalves.5,6 Previous assays in the past have relied on experimental animals, such as the mouse bioassay.7 Problems were encountered with these tests not to mention the inherent dislike in using live animals, which has fallen out of favor in recent years. Nonspecific toxin detection and the risk of false positives caused by fatty acids have led to the search for more simplistic and reliable assays.8 Current HPLC techniques, based on the methods of Lee,9 will detect OA and DTXs by the spectrophotometric detection of 9-antharyldiazomethane derivatives. The level of sensitivity lies at ∼15 µg of OA/100 g of shellfish tissue (37.5 ng/mL or 4.6 × 10-8 M OA). However, instability of the derivatization agent and problems with sample clean up have imposed limitations on this approach. To counteract this problem, Aase and Rogstad10 have optimized a sample cleanup procedure. They suggested the use of solid-phase extraction using a silica column of 100 mg and washing solvents composed of dichloromethane instead of chloroform. Methods combining HPLC and mass spectroscopy have proved sensitive (2 ng of OA detectable).11 Nevertheless, it has been shown that up to 50% of the toxin can be lost in the derivatization and extraction procedure.12 (4) Guzman, L.; Campodonico, I. Publ. Inst. Patagonia Ser. Mon. 1975, 9, 6. (5) Kumagai, M.; Yanagi, T.; Murata, M.; Yasumoto, T.; Kat, M.; Lassus, R.; Rodriguez-Vazquez, J. A. Agric. Biol. Chem. 1986, 50, 2853-2857. (6) Carmody, E. P.; James, K. J.; Kelly, S. S. J. AOAC Int. 1995, 7B (6), 14031408. (7) Kimura, L. H.; Hokama, Y.; Abad, M. A.; Oyama, M.; Miyahara, J. T. Toxicon 1982, 20, 907. (8) Sajiki, J.; Takahashi, K. Lipids 1992, 27, 988-992. (9) Lee, J.; Yanagi, T.; Kanna, R.; Yasumoto, T. Agric. Biol. Chem. 1987, 51, 877-881. (10) Aase, B.; Rogstad, A. J. Chromatogr., A 1997, 764, 223-231. (11) Pleasance, S.; Quilliam, M.; deFreitas, A.; Mar, J.; Cembella, A. Rapid Commun. Mass Spectrom. 1990, 4, 206-213. (12) Gianna, R.; Bruno, M.; Volterra, L. Inquinamento, in press. (13) Gucci, P. M. B.; Serse, A. P.; Coccia, A. M.; Tubaro, A.; Della Loggia, R.; Gianna, R.; Bruno, M.; Volterra, L. Toxicol. Lett. 1994, 74, 91-97. 10.1021/ac9901642 CCC: $18.00
© 1999 American Chemical Society Published on Web 08/20/1999
Table 1. Syndromes and Mode of Action of Marine Algal Toxins disease
marine algae
PSP (paralytic)
Alexandrium spp., Pyrodinium spp., Gonyaulax spp., Gymnodinium spp.
DSP (diarrhetic)
Dinophysis spp., Prorocentrum spp.
NSP (neurotoxic)
Ptychodiscus brevis
ASP (amnesic)
Nitzschia spp., Pseudonitzschia australis, Nitzshcia pseudodelicatissima
symptoms tingling/numbness in mouth, ataxia, dizziness, headache, respiratory distress, and muscular paralysis (death) diarrhoea, nausea, vomiting, and abdominal pain oral and digital paraesthesia, ataxia, vomiting, diarrhea and abdominal pain nausea, vomiting, headache, abdominal pain, diarrhea, memory loss, disorientation, paralysis (death)
biochemical site of action site 1 on voltage-dependent sodium channel catalytic subunit of phosphorylase phosphatases site 5 on voltage-dependent sodium channel calcium channels Kainate receptor in CNS
Figure 1. Structure of okadaic acid (OA) and its isomers, the dinophysistoxins (DTXs).
Undoubtedly an inconsistency exists between the various detection methods. Gucci et al.13 confirmed this when their group tested extracts by four different assays. No clear quantitative agreement was found. More recent advances have led to a relatively rapid radioactive protein phosphatase (PP) assay, developed by Honkanen et al.14 It was used to detect OA in oyster extracts, and samples containing OA at g0.2 ng/g were considered positive. Results correlated well with LC determination. Tubaro et al.15 developed a colorimetric assay that could detect concentrations of OA as low as 2 ng/g of digestive glands using PP2A. In 1997, a patent was granted for a fluorometric PP assay that utilizes 4-methylumbelliferyl phosphate as substrate with a 20-fold improved sensitivity.16 This work highlights the characteristics of an okadaic acid immunoassay with very low limits of detection and develops an ELISA system to quantitate accurately the concentration of this toxin in water and mussel samples. This developed ELISA can be manipulated to quantitatively measure total DSP content (OA + DTX-1) or for okadaic acid and dinophysistoxin-1 individual concentrations by varying the format of the assay (Figure 1). EXPERIMENTAL SECTION Materials and Methods. Mouse anti-okadaic acid monoclonal antibodies, okadaic acid (Prorocentrum concavum), dinophysistoxin-1 (Halicondria okadai), and brevetoxin-3 (Ptychodiscus brevis) were purchased from Calbiochem (Nottingham, UK). Alkaline phosphatase (AP) (EC 3.1.3.1), 1-ethyl-3-(3-dimethylaminopropyl)(14) Honkanen, R. E.; Stapleton, J. D.; Bryan, D. E.; Abercrombie, J. Toxicon 1996, 34, 1385-1392. (15) Tubaro, A.; Florio, C.; Luxich, E.; Sosa, S.; Della Loggia, R.; Yasumoto, T. Toxicon 1996, 34, 746-752. (16) Vieytes, M. R.; Fontal, O. I.; Leira, F.; Babtista de Sousa, J. M. V.; Botana, L. M. Anal. Biochem. 1997, 248, 258-264.
carbodiimide (EDC), N-hydroxysuccinimide (NHS), p-nitrophenyl phosphate, bovine serum albumin (BSA), and Tween 20 were obtained from Sigma (Dublin, Ireland). Dimethyl sulfoxide (DMSO) was purchased from Merck (Schuchardt, Germany). All other chemicals were of reagent grade or better. Doubly distilled water was used throughout. Apparatus. All prepared immunoconjugates were analyzed for protein content using a Hewlett-Packard UV-visible spectrophotometer. All ELISAs were carried out on 96-well, Falcon Probind microtiter plates. Bio-Tek Instruments (Winooski, VT) supplied the microplate washer (model ELP-40) and reader (model EL311). Incubations performed at elevated temperatures were carried out in an oven supplied by Heraeus Instruments. Sephadex PD-10 columns were purchased from Supelco (Bellefonte, PA). A FreeZyme conjugation purification kit was purchased from Pierce (Rockford, IL). Optimization Studies. Enzyme Conjugation. 1. Conjugation of Okadaic Acid Antigen to Alkaline Phosphatase. All solutions were made up in ethanol, as OA is soluble in this solvent. To 10 µL of EDC and 20 µL NHS (20 mg/mL each) a 1 mg/mL solution of okadaic acid was added. The modification of the carboxylic acid group was maintained for 15 min at 37 °C. At this stage, 2.2 mg of AP was prepared and the activated antigen was added dropwise to the AP solution and left for 3 h at 37 °C. Once the reaction was complete, the mixture was purified by passing the solution through a Sephadex PD-10 column (Supelco). The fractions were analyzed with an UV-visible spectrometer and the fractions containing the desired conjugate stored at 4 °C in 0.1 M carbonate buffer containing 0.02% sodium azide. 2. Conjugation of r-Okadaic Acid Antibodies to Alkaline Phosphatase. A 0.5 mg aliquot of R-okadaic acid antibodies in 500 µL of Dulbecco’s phosphate-buffered saline (DPBS) was added Analytical Chemistry, Vol. 71, No. 19, October 1, 1999
4199
to 0.5 mg of AP. A 50 µL aliquot of a 1% glutaraldehyde (GA) solution was then added and the mixture gently swirled at room temperature for 2 h. At this point, 100 µL of a 0.1 M glycine solution was added to stop the reaction. The antibody-enzyme conjugate was further purified using a FreeZyme conjugation purification kit, and the fractions were analyzed by UV adsorption and stored at 4 °C in 0.1 M carbonate buffer containing 0.02% sodium azide. ELISA Procedure 1. Direct Competitive Assay Using Labeled Antigen (Ag*). Primary capture antibodies (20 µg/mL) in 0.1 M sodium carbonate buffer, pH 9.5, were precoated on the surface of a microtiter plate for 60 min at 37 °C. All washing steps involved automatic dispensing of a 0.1 M DPBS solution, pH 7.4, into all wells (3 times). Wells were subsequently blocked with 100 µL of modified DPBS solution containing 1% BSA (DBT) for 60 min at 37 °C. Dilutions of free and labeled antigen (13 µg/ mL) were plated out in duplicate in DBT and incubated for 60 min at 37 °C. Color development was facilitated by addition of p-nitrophenyl phosphate (1 mg/mL in carbonate buffer, pH 9.8). Aliquots of 50 µL each were used for incubations, with the exception of blocking and color development, where 100 µL was used. DPBS consisted of 137 mM NaCl, 2.7 mM KCl, 8 mM Na2HPO4, and 1.5 mM NaH2PO4 adjusted to pH 7.4. Primary antibody immobilization and color development were performed in 0.1 M sodium carbonate buffer, pH 9.5 and 9.8, respectively. The color development buffer contained 1 mM MgCl2‚6H2O and 0.1 mM ZnCl2, both of which are alkaline phosphatase activators. Two formats of each ELISA procedure were utilized in order to compare sensitivity, linear ranges, and limits of detection. These formats involved the basic procedure as described in ELISA procedure with the exception of the addition of free and labeled proteins. Competitive Format. In this system, a dilution of the free OA analyte and the labeled OA were added simultaneously and as the name suggests were allowed to compete for 60 min at 37 °C. This forms the basis of a heterogeneous, multistep immunoassay. Displacement Format. Using this system, the labeled OA was incubated for 60 min at 37 °C prior to the addition of the dilution range of analyte. This would make this format more favorable for a single-step, homogeneous immunoassay. ELISA Procedure 2. Direct Competitive Assay Using Labeled Antibody (Ab*). Okadaic acid-bovine serum albumin conjugate (25 µg/mL) was coated to the surface of a microwell plate under the same conditions as primary antibody in procedure 1 above. The plates were subsequently washed and blocked for 60 min in DBT buffer at 37 °C or overnight at 4 °C when viable. Approximately 4 µg/mL R-okadaic acid-alkaline phosphatase antibodies were then incubated for another 60 min in DBT at 37 °C with or without standard solutions of analyte, depending on the format used. This is as described in the competitive and displacement format sections. Color was developed as mentioned previously. Extraction Procedure for Mussel Samples. The extraction procedure was performed according to the method of Kelly et al.17 Briefly, the extraction procedure involved the following: 5 g (17) Kelly, S. S.; Bishop, A.; G.; Carmody, E. P.; James, K. J. J. Chromatog., A 1996, 749, 33-40.
4200 Analytical Chemistry, Vol. 71, No. 19, October 1, 1999
Figure 2. Schematic for ELISA systems utilized. (1) Antigen capture; (2) antibody capture. Table 2. List of Optimized Parameters for Both ELISA Systems format
capture Ab
antigen capture (1) 20 µg/mL antibody capture (2)
Ag* 13 µg/mL
OA-BSA
Ab*
25 µg/mL 4 µg/mL
of hepatopancreas was homogenized in 10 mL of 80:20 methanol/ water. This mixture was centrifuged at 3000 rpm for 10 min. A 2.5 mL aliquot of the supernatant was washed twice with 2.5 mL of hexane by vortex mixing for 1 min. The upper layer was discarded each time and 5 mL of 20:80 water/chloroform was added to the residual mixture, which was vortexed for 2 min. After centrifugation for 5 min to separate the layers, the lower chloroform layer was transferred to a 10 mL volumetric flask. The chloroform extraction was repeated and made up to 10 mL. An aliquot of this (0.5 mL) can be reconstituted in methanol after evaporation under nitrogen. RESULTS AND DISCUSSION Displacement and competition ELISAs were carried out using both labeled antigen (Ag*) and labeled antibodies (Ab*) in two different formats as shown in Figure 2. Primarily, optimization studies were performed using the method of checkerboard titrations. The optimized parameters of each ELISA system can be seen in Table 2. All subsequent ELISAs were performed using these parameters. These 2 systems were fully characterized in terms of linear ranges, limits of detection and cross-reactivity. Okadaic acid antigen dissolves only in a limited number of organic solvents. This can lead to a number of problems, particularly involving denaturing of the protein conjugates. For this system, it was determined that the solvent concentration must remain at
