Anal. Chem. 2004, 76, 4237-4240
Development of a Membrane-Based Immunofiltration Assay for the Detection of T-2 Toxin Arindam Pal, Debopam Acharya, Debjani Saha, and Tarun K. Dhar*
Department of Immunobiology, Indian Institute of Chemical Biology, 4, Raja S. C. Mullick Road, Jadavpur, Calcutta 700 032, India
An improved analytical device capable of performing simultaneous immunofiltration-based immunoassay on 30 samples in the presence of reference standards has been developed. The device consists of a rectangular membrane with 36 antibody spotted zones, one end of which was attached to a semirigid polyethylene card. A piece of wetted filter paper between the membrane and the polyethylene card absorbs the added reagent. The assay is a competitive one using T-2 toxin-horseradish peroxidase (T-2 toxin-HRP) as the labeled analyte and 4-chloro-1 naphthol (4CN) as the substrate. Signal amplification was done by the Super-CARD signal amplification method. Semiquantitative results were obtained by visual comparison of the color intensity of a sample spot with those of reference standards. Densitometric analysis was used for quantitation. The method allows rapid and easy determination of T-2 toxin in wheat and poultry feed with detection limits of 12.5 and 25 µg kg-1, respectively, with accuracy and precision. Matrix interference was eliminated by appropriate dilution of sample extracts with assay buffer. The detection sensitivity in ELISA was 10fold higher than that in the membrane-based method. Noninfected samples were spiked with T-2 toxin at several concentrations and analyzed by the present method and rapid ELISA. Mean recoveries by both methods were between 80 and 108%. The correlation between the two methods was excellent (R2 ) 0.99). The Fusarium is the most prevalent toxin-producing soil fungi, which contaminate food grains in the temperate regions of America, Europe, and Asia. They produce mycotoxins of the class trichothecene, which are highly toxic to animals and humans. The trichothecene of interest in this study is T-2 toxin, well-known for causing alimentary toxicity. Estimation of trace amounts of T-2 toxin present in food sample due to fungal contamination is a difficult task, as it requires laborious and extensive cleanup treatments to remove interfering substances from the sample matrix before analysis.1-3 Therefore, there is a demand for rapid * To whom correspondence should be addressed. Phone: +91-33-2473-3493/ 197. Fax: +91-33-2473-0284/5197. E-mail:
[email protected]. (1) Sukhadin, D. L. Toxicol. Lett. 2003, 97-107. (2) Langseth, W.; Rundberget, T. J. Chromatogr. A 1998, 815, 103-121. (3) Pascale, M.; Haidukowski, M.; Visconti, A. J. Chromatogr. A 2003, 989, 257-264. 10.1021/ac049631s CCC: $27.50 Published on Web 06/03/2004
© 2004 American Chemical Society
and reliable methods for the detection of T-2 toxin in food samples. Immunochemical methods, especially by microtiter plate ELISA, offer the required sensitivity and are widely used for simultaneous analysis of a large number of samples.4,5 However, these methods are time-consuming and require sophisticated equipment and complicated sample cleanup (generally, solid-phase extraction or immunoaffinity column). During the past few years, there has been an increasing demand for rapid, visual, quantitative assays that could be performed outside the laboratory, for example, on farms, in storehouse and in factories and several simple devices for performing immunological assays have been reported.6-9 But these methods are not suitable for analyzing a large number of samples at a time. We have recently reported a novel analytical device for performing an immunofiltration-based assay for detection of aflatoxin B1.10 The device consists of 4 membrane strips with 16 antibody-immobilized zones. The top edge of the membrane is fixed on a polyethylene card with adhesive. During the assay, a moist filter paper placed between the membrane strips and the polyethylene card acts as the absorbent body. The void volume of the wetted absorbent body is sufficient to absorb the additional fluid introduced during the assay. To avoid time-consuming sample cleanups, dilute samples were used and the detection sensitivity was increased by the recently described improved catalyzed reporter deposition method termed Super-CARD method of signal amplification involving biotinylated tyramine (B-T) and avidin-horseradish peroxidase conjugate.11,12 The method allows semiquantitative analysis within 12 min. However, only 10 samples can be analyzed at a time by using a single test card. (4) Barna-Vetro, I.; Gyongyosi, A.; Solti, L. Appl. Environ. Microbiol. 1994, 60, 729-731. (5) Laamanen, I.; Veijalainen, P. Food Addit. Contam. 1992, 9, 337-343. (6) Schneider, E.; Dietrich, R.; Martlbauer, E.; Usleber, E.; G. Terplan, G.; Food Agric. Immunol. 1991, 3, 185-193. (7) De Sagar, S.; Van Peteghem, C. Appl. Environ. Microbiol. 1996, 62, 18801884. (8) Sibanda, L.; De Sagar, S.; Van Peteghem, C.; Grabarkiewicz-Szczesna, J.; Tomczak, M. J. Agric. Food Chem. 2000, 48, 5864-5867. (9) Ho, J.-A.; Wauchope, R. D. Anal. Chem. 2002, 74, 1493-1496. (10) Pal, A.; Dhar, T. K. Anal. Chem. 2004, 76, 98-104. (11) Bhattacarya, R.; Bhattacharya, D.; Dhar, T. K. J. Immunol. Methods 1999, 227, 31-39. (12) Bhattacarya, D.; Bhattacharya, R.; Dhar, T. K. J. Immunol. Methods 1999, 230, 71-86.
Analytical Chemistry, Vol. 76, No. 14, July 15, 2004 4237
Figure 1. Schematic diagram of the device containing a membrane with 1-6 rows for on-site immunoassay. Each row contains six antibody-immobilized zones (a-f).
We have continued to explore ways to simplify the construction of the analytical device and increase the number of samples that can be simultaneously analyzed and succeeded in developing an improved version of the device (Figure 1), which differs from the previously reported one on the following aspects. It uses a single rectangular membrane containing 36 spotted zones, which simplifies the assembly of the device. Second, the membrane is fixed on one side instead of the middle position of the polyethylene card. Third, a large number of samples can be analyzed using a single test card. In this paper, we describe the use of this device for rapid monitoring of T-2 toxin in wheat and poultry feed and compare its performance with that of rapid ELISA. EXPERIMENTAL SECTION Materials and Chemicals. Nitrocellulose membrane, 0.45 µm (Catalog No. HAHY00010), was from Millipore Corp., Bedford, MA. Filter paper (Whatmann No. 3) was from Whatman International Ltd., Maidstone, England. Flat-bottomed polystyrene microtiter plates (Maxisorp) and the 8-channel microplate washer were from Nunc, Denmark. The automatic microtiter plate reader (Multiscan MS) was from Labsystems, Finland. Semirigid polyethylene card and adhesive tape were purchased from a local market. T-2 toxin, HT-2, T-2 triol, T-2 tetrol, verrucarol, nivalenol, 4 CN, O-phenylenediamine (OPD), and other chemicals were from Sigma, St. Louis, MO. Densitometric analysis was carried out as described previously.10 Reagents. T-2 toxin hemisuccinate (T-2 HS) was synthesized according to standard procedure13 with slight modification. The immunogen, T-2 HS-BSA, was prepared by the NHS-ester method and the degree of conjugation of T-2 toxin/protein was found to be 10 by the TNBS method.14 Polyclonal antibody against T-2 toxin was raised in rabbits following the standard method used in our laboratory.15 The antisera obtained after 6 months was purified by precipitation with ammonium sulfate (50% saturation) and passing it through a BSA-Sepharose immunosorbent column. T-2 toxin-HRP and T-2 toxin-casein conjugates were prepared by the NHS-ester method. Cross-reaction with HT-2, T-2 triol, T-2 tetrol, verrucarol, and nivalenol was 5.07, 0.55, 0.12, 0.1, and 0.1%, (13) Chu, F. S.; Grossman, S.; Wei, R. D.; Mirocha, C. J. Appl. Environ. Microbiol. 1979, 37, 104-108. (14) Habeeb, A. F. S. A. Anal. Biochem. 1966, 14, 328-336. (15) Das Sarma, J.; Duttagupta, C.; Ali, E.; Dhar, T. K. J. Immunol. Methods 1995, 184, 1-6.
4238
Analytical Chemistry, Vol. 76, No. 14, July 15, 2004
respectively. B-T, 3-(p-hydroxyphenyl) propionic acid-casein conjugate (p-OH-PPA-casein) and tyramine-casein conjugate (T-casein) were prepared according to the method described previously.11,12 The glasswares and T-2 waste were decontaminated with 5% sodium hypochlorite solution. Preparation of Membranes. A rectangular piece of nitrocellulose membrane (84 × 98 mm) was marked with a pencil to give forty-two 14-mm squares. The membrane was soaked in blotting buffer (Tris-HCl, 20 mM, pH 8.0 containing 0.9% NaCl) and semidried. Anti-T-2 antibody diluted 100-fold with Tris-buffered saline (50 mM, pH 8.0) containing 200 µg mL-1 of BSA was applied in a volume of 5 µL at the center of each 14-mm square and spread to ∼7-mm diameter with the tip of the dispenser. The top portion of the rectangular membrane was not spotted and kept free for attachment to the polyethylene base. The membrane was then dried and blocked by incubating with a solution of 0.2% p-OHPPA-casein conjugate in carbonate-bicarbonate buffer (50 mM, pH 9.6) for 1 h with gentle shaking at room temperature. It was then washed (Tris-HCl, 20 mM, pH 8, containing 2.9% NaCl and 0.05% Tween 20) and dried as described previously.10 Assay with the Analytical Device. A 5 × 84 mm tape having adhesive on both sides was fixed horizontally in such a way that it is about 1 cm away from two adjacent sides of a polyethylene card (190 × 200 mm). The unspotted top portion of the rectangular membrane was then fixed on the tape. A rectangular piece of filter paper (170 × 180 mm) was placed between the polyethylene card and the membrane and wetted with a limited amount of water (10 mL). The air entrapped between the membrane and the absorbent body was removed by repeated rolling of a glass tube under slight pressure over the membrane. Standards (0, 0.5, 1, 2.5, 5, and 10 pg/25 µL) and samples were applied (2 × 25 µL/ spot) at the center of the antibody spotted zones of the membrane. Excess water was extruded by rolling a glass tube over the membrane and gently sponging the absorbent body with a tissue paper. To each spotted area 25 µL of T-2 toxin-HRP conjugate in assay buffer was added. The absorbent body was then removed and the membrane was thoroughly washed with an even stream of wash buffer from a wash bottle. After removal of the adhering buffer from the surface, it was amplified by the Super-CARD method of signal amplification and visualized with 4 CN as described previously.10 The intensity of the colored spots was then visually compared with those of reference standards to give an idea about the amount of T-2 toxin present in the sample. The calibration curve for T-2 toxin was constructed by plotting the percentage of the B/B0 value (x-axis), obtained from densitometric analysis against toxin concentration (y-axis), where B0 is the density of the zero standards (no T-2 toxin added) and B is the density of standards. Rapid ELISA. The microtiter plate wells were incubated with T-2 toxin-casein conjugate (0.4 µg/mL, 100 µL/well) in coating buffer (sodium-potassium phosphate, 50 mM, pH 7.6) overnight at 4 °C and blocked with 0.2% T-casein conjugate in coating buffer for 3 h at 37 °C. The wells were washed with washing buffer (coating buffer containing 0.05% Tween 20). To each well was added 25 µL of standard (0, 0.5, 1, 2, 5, 10, and 20 pg/25 µL or sample and 50 µL of anti-T-2 toxin antibody (1:5000) in assay buffer (sodium-sodium phosphate buffer, 50 mM, pH 7.4 containing 0.9% of NaCl, 0.2% of BSA, and 0.01% of
thimerosal). The plate was incubated at room temperature for 10 min and washed. Then 50 µL of goat anti-rabbit IgG-HRP conjugate (diluted 1:2000) was added, incubated at room temperature for 3 min and washed. For amplification, 50 µL of B-T at a concentration of 100 µM in Tris-HCl buffer (50 mM, pH 8.0) containing 0.01% H2O2 was added to each well and incubated for 3 min at room temperature. After the plate was washed, 50 µL of avidin-HRP conjugate (diluted 1:2500) in assay buffer was added and the plate incubated for 3 min at room temperature. The plate was again washed and 50 µL of substrate solution containing OPD was added. After 2 min, the reaction was terminated by adding 50 µL of sulfuric acid (4 N) to each well and the absorbance at 492 nm was measured. The calibration curve was constructed from absorbance data as described above. Extraction of Samples. Samples were milled in a coffee grinder and 2-g portions of each sample were soaked in 5 mL of methanol-water (80: 20) for 15 min. It was then centrifuged and the clear supernatant was decanted. The wheat and poultry feed extracts were diluted in assay buffer and used directly in the assay without any further cleanup. Spiked Samples. For preparation of spiked samples the appropriate volume of a solution of T-2 toxin in methanol (10 ng/ µL) was added to 2 g of milled noninfected seeds/feed and incubated at 37 °C for 24 h. The dried samples were extracted by methanol:water as described above. RESULTS AND DISCUSSION Influence of Attachment Position of the Membrane on Reproducibility of Spot Intensity. In our previous communication, we have demonstrated the importance of the area of the absorbent body in the analytical device for maintaining assay reproducibility.10 A single strip having four spotted zones and absorbent body of 32 cm2 gave good reproducibility. On the basis of these results, we have tested the suitability of the absorbent body of 160 × 180 and 170 × 180 mm sizes for the present study with 36 spotted zones. In this case the membrane was fixed at the middle of one edge of the polyethylene card. The assay was carried out in absence of T-2 toxin and measurement of the time required for complete absorption of the applied fluid from the spotted zones (flow rate). The results showed that when an absorbent body of 160 × 180 or 170 × 180 mm size was used, the measured flow rate for the standards (2 × 25 µL) from each spotted zones in the row numbers 1-6 was about 4-5, 5-6, 6-7, 7-8, 9-11, and 10-12 s, respectively. The corresponding flow rate after removal of excess fluid for enzyme conjugate (25 µL) was about 4-5, 5-6, 6-7, 7-8, 8-9, and 10-12 s, respectively. There was a gradual decrease in spot intensity from row 1 to row 6 (100 to 82%). Further increase in the area of the absorbent body (180 × 180 or 180 × 200 mm) gave similar results. It may be pointed out that the addition of standards and enzyme conjugate to the spotted zones were in the order of 1, 2, 3, 4, 5, and 6 rows. Presumably, the addition of fluid in the first row decreases the void volume of the filter paper and thereby decreases the flow rate of the subsequent row leading to more spreading. For all subsequent rows the effect would be cumulative. We then investigated the effect on flow rate by changing the attachment position of membrane from the middle to the corner of the polyethylene card. With an absorbent body of 170 × 180 mm size, the time required for complete absorption of the zero
Figure 2. Dose response curves of T-2 toxin under optimized conditions: Rapid ELISA (-O-); membrane-based method (-b-). Each point represents the mean and standard deviation of four measurements in duplicate. Table 1. Comparison between Membrane-Based Method and Rapid ELISA analytical variables
rapid ELISA
membrane method
sensitivity (pg/well or spot) I50 valuea (pg/mL) standard curve range (pg/well or spot) number of washing steps total assay timeb (min)
0.25 5.8 0.5-20 4 40
0.5 7.1 1-20 3 22
a I , the midpoint of the curve. b Time required for analysis of 30 50 prediluted samples.
standard in the spotted zones of rows 1, 2, 3, 4, 5 and 6 were about 10-10.5, 10-10.5, 10-10.5, 10-10.5, 10-11, and 10-12 s respectively (2 × 25 µL). After removal of excess fluid from absorbent body, the corresponding flow rate for enzyme conjugate (25 µL) was about 9-10, 9-10, 9-10, 9-10, 10-11, and 10-12 s, respectively. The flow rate with 160 × 180 mm size absorbent body for standards or enzyme conjugate gave variation of about 4 s from row 1 to row 6. We therefore selected 170 × 180 mm size filter paper and attachment of the membrane at the corner of the polyethylene card for performing the assay. The reproducibility of the spot intensities was found to be excellent with a coefficient of variation of 2.4% (n ) 36). Comparison between Membrane-Based Method and Rapid ELISA. T-2 toxin was assayed by both methods under optimized conditions using the same reagents. The standard curves shown in Figure 2 for the two methods represent the mean of four independent standard curves, each obtained with duplicate determinations. The detection limit of T-2 toxin and the range and slope of the two curves were quite different. The lower limit of detection (distinguishable from blank by twice SD) using membrane-based and rapid ELISA was 0.5 and 0.25 pg, respectively, corresponding to B/B0 values of 95 ( 2.5% and 96 ( 2.5%, Analytical Chemistry, Vol. 76, No. 14, July 15, 2004
4239
Table 2. Recovery of T-2 Toxin Spiked to Samples T-2 toxin (µg/kg) founda
% recovery
added
rapid ELISA
membrane method
rapid ELISA
membrane method
wheat
50 125 250
42.5 135.0 220.0
45 127.5 227.0
85 108 88
90 102 91
poultry feed
50 125 250
50.0 135.0 200.0
42.5 132.5 220.0
100 108 80
85 106 88
matrix
a
Mean of duplicate determinations.
assayed by both the methods in duplicate (Table 3). The analytical recoveries ranged from 88 to 111% and 92.3 to 111% for the membrane-based method and ELISA, respectively. The intra- and inter-assay variability for both the methods were determined by analyzing three wheat samples containing a low, medium, and high concentration of T-2 toxin. Intra- and interassay CVs ranged between 1.2 and 14.3% and 3.4 and 19.7%, respectively. Correlation studies between the methods were performed on wheat and poultry feed samples spiked with T-2 toxin. The equation of correlation (using the statistical analysis module from Analyse-It Software Ltd., U.K.) obtained with 60 samples was
T-2 toxinMembrane-based method ) 0.33 µg kg-1 + (1.0044 × T-2 toxinRapid ELISA); R2 ) 0.99
Table 3. Analytical Recovery of Artificially Contaminated Wheat Samples T-2 toxin (µg/kg) founda
% recovery
rapid membrane rapid membrane endogenous added calcd ELISA method ELISA method 25 125 250
67.5 167.5 292.5
75 157 300
77.5 150.0 270.0
111.0 97.3 103.0
114.8 89.5 92.3
135
25 125 250
160 260 385
160 240 367.5
160.0 275.0 382.5
100.0 92.3 95.4
100.0 105.7 99.3
220
25 125 250
245 345 470
250 325 475
245.0 335.0 462.5
102.0 94.2 101.0
100.0 97.1 98.4
42.5
a
Mean of duplicate determination.
respectively. Rapid ELISA takes 40 min to be performed, with the possibility of analyzing dozens of samples simultaneously. The membrane-based method is also suitable for simultaneous analysis of 30 samples and takes about 22 min of assay time (Table 1). Assay of T-2 Toxin in Wheat and Poultry Feed. To assess the possible interference of sample constituents other than T-2 toxin, we assayed a noninfected sample extract diluted with assay buffer (10- to 1000-fold) by both the methods using the assay buffer as control. The results showed that, in the membrane-based assay, the matrix interference could be avoided if aqueous methanol extracts of wheat and poultry feed are diluted 1:250 and 1:500, respectively. Similarly, in rapid ELISA, wheat and poultry feed samples should be diluted 1:25 and 1:50, respectively, to prevent interference by matrixes. To study the extraction efficiency of T-2 toxin by aqueous methanol, noninfected samples of wheat and poultry feeds were spiked with a low, medium, and high level of T-2 toxin 1 day prior to extraction. The recoveries estimated by both the methods were excellent, being in the range of 80-108% (Table 2). To investigate the analytical recovery, three artificially contaminated sample extracts containing different levels endogenous of T-2 toxin were spiked with different levels of T-2 toxin and
4240
Analytical Chemistry, Vol. 76, No. 14, July 15, 2004
CONCLUSION An improved analytical device for performing immunofiltrationbased immunoassay of a large number of samples has been demonstrated. The method does not require a pump to create a vacuum to pull reagents through the membrane. The uniformity of the flow rate of the applied fluid in all the 36 spotted zones may be maintained not only by the size of the absorbent body but also by a change in the attachment position of the membrane over the polyethylene card. Results clearly prove that the method presented is able to analyze T-2 toxin in wheat and poultry feed with the detection limits of 12.5 and 25 µg kg-1, respectively, with accuracy and precision. This sensitivity is sufficient to analyze T-2 toxin at levels of regulatory relevance and comparable or much higher than other methods recently described.6-8 The quality of the analytical data and the ultrahigh sensitivity of the ELISA method based on Super-CARD amplification technology makes it suitable for confirmatory tests. Taking into account the sample dilution requirement and the operative working range, the limit of quantitation by the rapid ELISA method for wheat and poultry feed were 1.25 and 2.5 µg kg-1, respectively. This gave detection sensitivity in ELISA of about 10-fold higher compared to that of the membrane-based method. With precoated plates, the rapid ELISA takes 40 min to be performed, with the possibility of analyzing dozens of samples simultaneously. The membrane-based method described is suitable for simultaneous analysis of a large number of samples within 22 min. Thus, the method is well-suited for visual screening of food and feed in mycotoxin in regulatory programs. ACKNOWLEDGMENT This work has been supported in part by a grant from the Department of Biotechnology, Delhi. We thank the Council of Scientific and Industrial Research, Delhi, for a Research Fellowship to D.S. Received for review March 8, 2004. Accepted April 30, 2004. AC049631S