High performance thin layer chromatography ELISAGRAM

177

Journal of Immunological Methods, 136 (1991) 177-183 © 1991 Elsevier Science Publishers B.V. 0022-1759/91/$03.50 ADONIS 0022175991000744

JIM 05819

High performance thin layer chromatography ELISAGRAM Application of a multi-hapten immunoassay to analysis of the zearalenone and aflatoxin mycotoxin families J a m e s J. Pestka Department of Food Science and Human Nutrition, Department of Microbiology and Public Health, Center for Environmental Toxicology, Michigan State University, East Lansing, MI, U.S.A. (Received 25 July 1990; revised received 10 October 1990, accepted 10 October 1990)

A new immunoblot approach called ELISAGRAM was devised that combines the sensitivity and selectivity of competitive ELISA with the capacity of high performance thin layer chromatography (HPTLC) to separate structurally related haptens. The procedure involved (1) separation of halr e n s by HPTLC, (2) blotting of the HtrI'LC plate with nitrocellulose (NC) that was coated with hapten ~pecific monoclonal antibody, (3) incubation of NC with hapten-enzyme conjugate to identify unreacted antibody binding sites, (4) detection of bound enzyme conjugate with a precipitating substrate and (5) visual or densitometric assessment of inhibition bands indicative of a cross-reacting hapten. The technique was applied to two major mycotoxin families, the zearalenones and aflatoxins (AFs), which are to toxicological significance. Detection limit for zearalenone and a-zearalenol in the method was 300 pg/assay. AFB 1, AFB 2, and AFG1 were detectable at 380 pg/assay and A F G 2 was detectable at 1500 pg/assay. Multiple standard curves for the zearalenones and AFs could be constructed using scanning densitometry. Cross-reactivity in E L I S A G R A M curves was analogous to that found in competitive ELISA. This procedure should be widely applicable to the simultaneous quantitation and confirmation of multiple haptens with a single cross-reactive antibody. Key words." lmmunoblot; ELISA; Enzyme immunoassay; Zearalenone; Aflatoxin; Mycotoxin; Chromatograpy, high performance thin layer

Introduction

Competitive enzyme-linked-immunosorbent assays (ELISAs) have been developed for a wide range of biologically significant haptens including drugs, hormones, pesticides, environmental contaminants and natural toxicants. These have found extensive application in human and animal health

Correspondence to." J.J. Pestka, 236 Food Science, Michigan State University, East Lansing, MI 48824-1224, U.S.A.

for the analysis of clinical samples and, more recently, in agriculture for food safety assessment (Newsome, 1986; Pestka, 1988). There are some major difficulties encountered with the application of ELISAs for haptens related to antibody specificity and confirmation. Antibodies used in these assays may react in differing degrees with metabolites and analogues that are structurally related to the parent hapten. Often these compounds may be concurrently present in a sample. Thus the quantitative capacity of a competitive ELISA is highly dependent on the antibody

178

specificity, selection of hapten standard(s) and presence of structurally related compounds in the sample. Also, because there is the potential for presence of cross-reacting haptens or nonspecific interference in a sample, it is frequently necessary to confirm the presence of the analyte by an equally or sometimes less sensitive method such as high performance liquid chromatography (HPLC), gas chromatography (GC) or mass spectrometry (MS). To overcome the above problems, a new immunoblot approach called E L I S A G R A M was devised that combines the sensitivity and selectivity of competitive ELISA with the capacity of high performance thin layer chromatography (HPTLC) to separate structurally related haptens. The procedure involves (1) separation of haptens by HPTLC, (2) blotting of the H P T L C plate with antibody-coated nitrocellulose, (3) incubation with hapten-enzyme conjugate to identify unreacted antibody sites, (4) detection of bound conjugate with a precipitating substrate and (5) visual or densitometric assessment of inhibition zones indicative of a cross-reacting hapten. This report describes the application of E L I S A G R A M to the zearalenone (ZEA) and aflatoxin (AF) mycotoxin families. The results suggest that this assay will be widely applicable to the simultaneous quantitation and confirmation of multiple haptens using a single cross reactive antibody.

Materials and methods

Reagents All inorganic chemicals and organic solvents were reagent grade or better. Bovine serum albumin (fraction V); dioctylsodium sulfosuccinate (DONS); 3,5,3',5'-tetramethylbenzidine (TMB), hydrogen peroxide, horseradish peroxidase (type VI, HRP), and Tween 20 were purchased from Sigma Chemical Co. (St. Louis, MO). Ascites fluid containing ZEA monoclonal antibody 2G3-6E3 and AFB~ monoclonal antibody 5 C l l were prepared, subjected to a m m o n i u m sulfate precipitation and brought up in 0.01 M phosphate buffered saline (pH 7.5, PBS) to a concentration of 5 m g / m l as described previously (Dixon et al., 1987; Dixon-Holland et al., 1988). ZEA and AFB 1 were first

converted to O-carboxymethylamine derivatives by the method of Chu et al. (1977) and then conjugated to H R P by the N-hydroxysuccinimide procedure of Kitagawa et al. (1981) for use as marker ligands in the E L I S A G R A M .

High performance (HPTLC)

thin-layer

chromatography

Zearalenone (ZEN) and a-zearalenol (ZEL) standards were kindly supplied by Pittman-Moore, (Terre Haute, IN). AFBI, AFB2, AFG~ and A F G 2 standards were obtained from Sigma. Mycotoxin standards were dissolved in chloroform in a range from 20 to 2000 n g / m l and 5 ffl aliquots spotted onto Whatman L H P - K (Clifton, N J) channeled linear high performance T L C plates (10 x 10 cm). Zearalenones and AF were separated on the T L C plates to a 5 cm height from the origin using chloroform : methanol (97 + 3) (Richardson et al., 1984) and chloroform : acetone (9 + 1) (Spillman, 1985), respectively. Plates were dried under a stream of warm air. Relative Rrs of zearalenones and AFs were determined by viewing under short and long wave UV illumination, respectively.

EL1SA GRAM Steps for the E L I S A G R A M are summarized in Fig. 1. They were specifically as follows. (1) Nitrocellulose (NC) membranes (0.45 ~tm; Schleicher and Schuell, Keene, N H ) were cut into 6 x 5 cm sheets and placed into 6 x 7 cm freezer bags (Dazey Micro-Seal, Industrial Airport, KS) sealed on three sides. AFB 1 or zearalenone monoclonal antibodies were diluted between 1/100 to 1/500 in PBS and 3 ml added per bag. The open portion of the bag was sealed with a Dazey freezer bag heat sealer. Bags were incubated overnight at 4°C with gentle agitation and could be stored additionally for up to 2 weeks at 4°C. (2) Just prior to assay, antibody-coated N C was blocked for 15 rain in PBS containing 0.5% ( w / v ) bovine serum albumin (PBS-BSA) and then placed in PBS. H P T L C plates containing the separated mycotoxins were sprayed with PBS till saturated. After shaking off excess PBS, N C was layered onto the H P T L C plate with the lower 5 cm margin at the origin and covering 4 - 5 of the 1 cm wide channels. A

179 1. SEPARATE HAPTENS ON CHANNELED HPTLC SILICA GELPLATE(R) 2. IMMUNOSPECIFIC TRANSFER: FILTER PAPER (E:::a) ANTIBODY-COATED NC ('~) HPTLC ( 1 )

1 1

88

3. REACT NC WITH HAPTEN-HRP CONJUGATE ( ~ ) 4. INCUBATE NC WITH PRECIPITATING HRP SUBSTRATE. IDENTIFY HAPTENS AS INHIBITORY BANDS

l tttt

5. PERFORM SCANNING DENSITOMETRY

Fig. 1. Summary of ELISAGRAM procedure.

second N C sheet was added to cover remaining channels. Two layers of Whatman no. 1 filter paper were placed over the NC and the entire complex was pressed lightly for 30 s with a weighted block to achieve close contact between the N C and T L C plate. (3) Filter paper was removed and NC lanes marked by pencil. The N C was removed, placed in a new 6 x 7 cm bag and then 3 ml of either zearalenone- or AFBl-horseradish peroxidase conjugate (0.5 m g / m l ) diluted (1/200 to 1/500) in PBS-BSA was added to the bag. The bag was sealed and gently agitated at 25°C for 10 rain. (4) N C sheets were removed, placed in petri dishes and subjected to three 5 min washes in PBS containing 0.2% Tween 20. N C was rinsed briefly in distilled water and placed in a petri dish. (5) N C sheets were incubated for 10-20 min at 25°C with 10 ml of D O N S / T M B substrate prepared as described by Koch et al. (1985). The reaction was stopped by washing the N C in distilled water and then incubating for 10 min with a solution consisting of ethanol (13

ml), D O N S (100 mg) made up to 50 ml with distilled water. N C was then dried between filter paper and stored in the dark.

Scanning densitometry E L I S A G R A M S were analyzed on a Shimadzu CS-930 Dual Wavelength Scanner with DR-2 Data Recorder (Giangarlo Scientific, Pittsburgh, PA) using a linear scan in the fluorescence photo mode with both excitation and emission wavelengths set at 370 nm. In this manner, the intense blue background of the E L I S A G R A M effectively quenched any reflected illumination and yielded a baseline of 'zero' absorbance units whereas inhibition zones yielded reflected light that were recorded as ' p e a k ' absorbance units.

Results

The ZEA and AF mycotoxin families were chosen as models for evaluating the feasibility of specific transfer of haptens from T L C to NC. These compounds fluoresce under UV illumination and this enabled rapid visualization and optimization of NC transfer protocols. Based on preliminary studies, immunospecific blotting could be achieved after 15-30 s of contact between the antibody impregnated NC and T L C plates. Following blotting, haptens that were previously separated by TLC could be effectively identified in a modified ELISA that involved incubating the NC sheets with mycotoxin-HRP conjugates and then developing with a precipitating substrate. The resultant image from this entire procedure (Fig. 1) was a distinct white band against a dark blue background. Fig. 2 is a photograph of E L I S A G R A M that was prepared from an H P T L C separation of ZEA and ZEL and that employed zearalenone 2G3-6E3 monoclonal antibody and Z E A - H R P conjugate. The limit for visual detection of these compounds by the E L I S A G R A M procedure was 300 p g / a s say. Relative areas of the E L I S A G R A M bands were readily measured by scanning densitometry (Fig. 3). Both ZEA and Z E L concentrations could be plotted against densitometer response based on an E L I S A G R A M (Fig. 4). Cross-reactivity of this monoclonal antibody for the two compounds in

180

500

C) •

&-. 400

ZEA

0 ZEA • ZEL

0

/

/0

//

E

~.E300

ZEL

/

LJ

< ,v, 2 0 0 .<

///

/

// 0,,' /

0

/0

,' /

kd D-

0

lO0

0

.63

ZEA

'

•

J

Fig. 4. Simultaneous quantitation of zearalenones by ELISAG R A M densitometry.

Fig. 2. E L I S A G R A M of ZEA and ZEL. ZEA and ZEL were separated by HPTLC and E L I S A G R A M prepared with ZEA specific monoclonal antibody and ZEA-HRP.

/

i

2 3 4 5 LOG CONCENTRATION ( p g / a s s a y )

1.3 2.5 5.0 10 TOXlN(ng)

/

i

/

the E L I S A G R A M was analogous to that found in competitive EL1SA (Dixon et al., 1987). The E L I S A G R A M approach was similarly applied to the detection of AFB], AFB 2, AFG] and A F G 2 using AFB] monoclonal antibody 5 C l l and A F B ] - H R P conjugate (Fig. 5). AFB 1, AFB 2, and AFG~ were detectable at 380 p g / a s s a y and for

ZEL

/

'~

25

0

f

\

l

\,

fl

/\

a 1

B2

G1 G2

0

"o

o2

I 0.8

] 0.6

I 0.4

Rf Fig. 3. Densitometry scan of EL1SAGRAM for zearalenones. E L I S A G R A M prepared as indicated in Fig. 2. ZEA and ZEL amounts were 5.0 ng (a), 2.5 (b), 1.5 ng (c) and 0.75 ng (d).

.75 1.5 3.0 6.0 TOXIN (ng) Fig. 5. E L I S A G R A M of AFB], AFB 2, A F G l and A F G 2. Aflatoxins were separated by HPTLC and E L I S A G R A M prepared with AFByspecific monclonal antibody and A F B F H R P .

181 BI

50-

A

_

_

G,

was observed for this monoclonal antibody when these aflatoxins were analyzed by competitive ELISA (Dixon-Holland et al., 1988).

J

Discussion

o o Q. 25In

o

"o

0

i

i

0.8

I

0.6

OA

Rf Fig. 6. Densitometry scan of ELISAGRAM for aflatoxins. ELISAGRAM prepared as indicated in Fig. 5. AFB1, AFB2, AFG] and AFG2 amounts were 5.0 ng (a), 2.5 ng (b) and 1.2 ng (c). A F G z at 1500 p g / a s s a y . Bands for all four AFs were resolvable by scanning densitometry (Fig. 6). When standard curves were developed simultaneously on a single E L I S A G R A M for all four aflatoxins, the following order of cross reactivity was noted: AFB 1 > A F G ] > AFB 2 and A F G 2 (Fig. 7). A similar rank order of cross reactivity 700

O •

600 500 E E ~-" 400

O AFB 1 • AFB 2

A

~ AFG 1

•

•

AFG 2

/

<

lad

O/°

O

he"

< 300 v <

,,, 200 D..

~A/O /

lO0

__.

2

3

4

LOG CONCENTRATION ( p g / a s s a y ) Fig. 7. Simultaneous quantitation of aflatoxins by E L I S A G R A M densitometry.

This report describes a hybrid procedure ( E L I S A G R A M ) that combines HPTLC, immunoblotting and competitive ELISA in order to facilitate quantitation and confirmation of multiple haptens. To illustrate this technique, it was applied to two major mycotoxin families, the zearalenones and aflatoxins, which are of toxicological significance to human and animal health (Pestka and Casale, 1990). The E L I S A G R A M represents a unique approach to the detection of haptens because it yields a 'negative', image of an H P T L C chromatogram that is achieved by the use of a modified competitive ELISA. The ELISAG R A M concept is based on previously described immunoblot techniques in which proteins from the gel phase of electrophoresis are transferred to a solid phase for immunoreaction thereby permitting the optimal combination of high resolution gel electrophoresis with the sensitivity and simplicity of solid phase assays (Towbin and Gordon, 1984). The use of antibody-coated NC to detect water-soluble polymeric glycoconjugates which do not bind directly to N C has been described ( H a n d m a n and Jarvis, 1985) but differs from the E L I S A G R A M because in this approach N C is used as the capture site and the polymeric antigen is detected with identical antibody that is radiolabeled. Several applications of immunoblotting to the analysis of compounds separated by T L C have been also described. Again these differ from the E L I S A G R A M because they utilize a labeled antibody to (1) detect a high molecular weight antigen directly on a T L C plate (Mattsby-Baltzer and Alving, 1984), (2) detect antigen that is chemically fixed to a T L C plate (Magnani, 1985), or (3) detect antigen passively transferred from T L C to N C (Towbin et al., 1984). The E L I S A G R A M exhibited a high degree of sensitivity when compared to conventional thin layer chromatography. Using the zearalenones and aflatoxins as examples, the E L I S A G R A M was approximately 60 and ten times more sensitive,

182

respectively, than existing T L C detection methods for these toxins using fluorescence under UV illumination (Gimeno, 1983; Stoloff and Scott, 1984). The approach should improve even further the sensitivity and selectivity of TLC detection of non-fluorescing haptens which currently require secondary chemical reactions to identify general classes of haptenic compounds. Although the E L I S A G R A M S were approximately ten times less sensitive than ELISA (Dixon et al., 1987; Dixon-Holland et al., 1988), an advantage of E L I S A G R A M over ELISA is the ability to selectively quantitate members of a hapten family in a single assay using a cross-reactive antibody with the provision that there is adequate separation of the haptens by HPTLC. As an illustrative example, the aflatoxins are regulated worldwide because of their potent carcinogenicity (Pestka and Casale, 1990). In the competitive ELISA, the 5 C l l monoclonal antibody that was raised against AFB 1 reacts to differing degrees with the four major AF including AFB1, AFB2, A F G 1 and A F G 2 (Dixon-Holland et al., 1988). Based on biologic potency as assessed by mutagenicity and in vivo carcinogenicity, these toxins exhibit the following approximate rank order: AFB 1 > AFG1 > AFB 2 > A F G 2 (Pestka and Casale, 1990). However, competitive ELISA of a food sample for aflatoxins does not discriminate among this structurally related group and yields a single absorbance value from which total AF content must be estimated. The final estimate will thus vary depending on which AF or group of AFs is selected as a standard. In contrast, the E L I S A G R A M exploits the cross-reactivity of an antibody for a hapten family and actually facilitates the simultaneous analysis of multiple haptens such as the AFs simultaneously. Another important use of the E L I S A G R A M would be in the confirmation of positive ELISA tests. Where regulatory and legal actions related to aflatoxin contamination or drug testing are of concern, confirmation could be readily achieved by high resolution of H P T L C coupled with the selectivity of antibodies. Finally, a further advantage of the ELISAG R A M is the relative speed and ease in which the assay can be carried out. Using a channeled H P T L C plate, nine sample extracts can be sep-

arated, blotted and assayed in 60 min. Although this is a longer assay time than that required for current ELISA kits, it should be noted that multiple analytes can be separated simultaneously in the E L I S A G R A M . In conclusion, the E L I S A G R A M combines the sensitivity and selectivity of competitive ELISA with the high resolution of HPTLC. It should be readily applicable to the simultaneous analysis and confirmation of members of biologically important hapten families including drugs, hormones, natural toxicants, pesticides or environmental contaminants. From a research standpoint, the E L I S A G R A M will be particularly valuable in detection of unknown in vivo metabolites or naturally-occurring analogues of toxins, drugs or hormones.

Acknowledgements This work was also supported by the Research Excellence Fund from the State of Michigan and by the Michigan State University Agricultural Experiment Station. We thank D. Klein and J. Azcona for assistance with manuscript preparation.

References Chu, F.S., Hsia, M.-T.S. and Sun, P.S. (1977) Preparation and characterization of aflatoxin B l - l - ( O - c a r b o x y m e t h y l ) oxime. J. Assoc. Off. Anal. Chem. 60, 791-794. Dixon, D.E., Warner, R.L., Hart, L.P. and Pestka, J.J. (1987) Hybridoma cell line production of a specific monoclonal antibody to the mycotoxins zearalenone and alphazearalenol. J. Agric. Food Chem. 35, 122-126. Dixon-Holland, D.E., Pestka, J.J., Bidigare, B.A., Casale, W.L., Warner, R.L., Ram, B.P. and Hart, L.P. (1988) Production of sensitive monoclonal antibodies to aflatoxin B~ and aflatoxin M 1 and their application to ELISA of naturally contaminated foods. J. Food Prot. 51, 201-204. Gimeno, A. (1983) Rapid thin layer chromatographic determination of zearalenone in corn, sorghum and wheat. J. Assoc. Off. Anal. Chem. 66, 565-569. H a n d m a n , E. and Jarvis, H.M. (1985) Nitrocellulose-based assays for the detection of glycolipids and other antigens: mechanism of binding to nitrocellulose. J. Immunol. Methods 83, 113-123. Kitagawa, T., Shimozono, T., Kawa, T.A., Yoshida, T. and Nishimura, H. (1981) Preparation and characterization of

183 heterobifunctional cross-linking reagents for protein modifications. Chem. Pharm. Bull. 29, 1130-1135. Koch, C., Skjodt, K. and Laursen, I. (1985) A simple immunoblotting method after separation of proteins in agarose gel. J. Immunol. Methods 84, 271-278. Magnani, J.L. (1985) lmmunostaining free oligosaccharides directly on thin-layer chromatograms. Anal. Biochem. 150, 13-17. Mattsby-Baltzer, I. and Alving, C.R. (1984) Lipid A fractions analyzed by a technique involving thin-layer chromatography and enzyme-linked immunosorbent assay. Eur. J. Biochem. 138, 333-337. Newsome, W.H. (1986) Potential advantages of immunochemical methods for analysis of food. J. Assoc. Off. Anal. Chem. 69, 919-923. Pestka, J.J. (1988) Enhanced surveillance of foodborne mycotoxins by immunochemical assay. J. Assoc. Off. Anal. Chem. 71, 1075-1081. Pestka, J.J. and Casale, W. (1990) Naturally occurring fungal toxins. In: J.O. Nriagu and M.S. Simmons (Eds.), Food

Contamination from Environmental Sources. J. Wiley and Sons, New York, pp. 613-638. Richardson, K.E., Hagler, Jr., W.M. and Hamilton, P.B. (1984) Method for detecting production of zearalenone, zearalenol, T-2 toxin and deoxynivalenol by Fusarium isolates. Appl. Environ. Microbiol. 47, 643-646. Spillman, J.R. (1985) Modification of the rapid screening method for aflatoxin in corn for quantitative use. J. Assoc. Off. Anal. Chem. 68, 453-456. Stoloff, L. and Scott, P.M. (1984) Natural poisons. In: S. Williams (Ed.), Official Methods of Analysis, 14th edn. Association of Official Analytical Chemists, Arlington, VA, Secs. 26.049-26.060, pp. 484-486. Towbin, H. and Gordon, J. (1984) Immunoblotting and dot immunobinding. Current status and outlook. J. Immunol. Methods 72, 313-340. Towbin, H., Schoenenberger, C., Ball, R., Braun, D.G. and Rosenfelder, G. (1984) Glycosphingolipid-blotting: an immunological detection procedure after separation by thin layer chromatography. J. Immunol. Methods 72, 471-479.