Letters in Applied Microbiology 1994, 18, 337-339

Effect of heterologous paralytic shellfish poisoning toxin-enzyme conjugates on the cross-reactivity of a saxitoxin enzyme immunoassay

E. Usleber, R. Dietrich, E. Martlbauer and G. Terplan Institute for Hygiene and Technology of Food of Animal Origin, Veterinary Faculty, University of Munich, Munich, Germany

MS/322: accepted 10 January 1994

E. USLEBER, R . DIETRICH, E. MARTLBAUER AND G. TERPLAN. 1994. A polyclonal antiserum against saxitoxin (STX) was used in a competitive enzyme immunoassay for the detection of paralytic shellfish poisoning toxins. The extent of cross-reactions was determined from the amounts of neoSTX, decarbamoylSTX and gonyautoxin 2/3 (GTX2/3) that gave 50% inhibition in the assay. Horseradish peroxidase (HRP) conjugates of the toxins and a bovine serum albumin conjugate of STX (STX-BSA) were used. When compared with STX-BSA and STX as standard, the extent varied to which heterologous conjugates affected the binding values of the other toxins to the antibodies. The antibodies did not bind GTX2/3-HRP. By use of neoSTX-HRP or decarbamoylSTX-HRP as the labelled antigen instead of STX-HRP, the detection limit for neoSTX was improved to 100 pg ml-'.

INTRODUCTION

Accumulation of paralytic shellfish poisoning (PSP) toxins, such as saxitoxin (STX), in shellfish is a human health risk which requires quick and reliable analytical methods. Because of the variable composition of structurally different toxins, which may be present in a single sample, assay specificity is a major problem in immunochemical approaches to the detection of PSP toxins. Enzyme immu- noassay (EIA) methods have been described for saxitoxin (Chu and Fan 1985) and neosaxitoxin (Chu et al. 1992). These assays, and those we have described (Renz and Terplan 1988; Usleber et al. 1991), have high sensitivity for the parent toxin but cross-react extensively with other PSP toxins. The assay sensitivity of these tests should be suffi- cient to avoid false-negative results at the EEC tolerance level of 80 pg PSP toxins per 100 g shellfish (European Communities 1991). I t is, however, possible that significant over- or underestimation of the total toxin content of a sample is a disadvantage for a screening method. We describe an approach to reducing the differences in toxin recognition with heterologous enzyme immunoassays for saxitoxin.

Cowespondence to : Dr E. Usleber, Lchrstuhl fur Hygiene und Technologie der Milch, Universitat Munchen, Schellingstrasse IOIIII, 80799 Munchen, Germany.

MATERIALS AND METHODS

Saxitoxin (STX), neoSTX, decarbamoylSTX (dcSTX) and gonyautoxin 213 (GTX2/3) were generous gifts from M.V. Laycock, Institute for Marine Biosciences, National Research Council, Halifax, Canada. STX was also pur- chased from Calbiochem, Bad Soden, Germany. A poly- clonal rabbit antiserum against STX was used as described elsewhere (Renz and Terplan 1988).

Preparatlon of toxin conjugates

An STX-bovine serum albumin conjugate (STX-BSA) was prepared by formaldehyde treatment (Johnson et al. 1964) according to the method of Chu and Fan (1985) and Renz and Terplan (1988). To prepare toxin-HRP conjugates, HRP (4 mg) was activated with periodate according to Wilson and Nakane (1978). To portions of each 1 mg of activated HRP (in ca 250 pl of 0.1 mol 1-' acetate buffer, pH 4-4), 20 pg of toxin (neoSTX, dcSTX, GTX2/3; in 200 pl of 0.1 mol I - ' acetic acid) was added, and the pH was adjusted to 7.5 with sodium carbonate buffer (pH 9.6; approx. 450 pl). After 60 min at 20°C, 30 pl of NaBH, (4 mg ml-') was added, the mixtures were incubated for 30 min at P C , and finally dialysed separately against 0.15 mol 1-' phosphate buffered saline for 2 d (pH 6.0,3 x 5 1).

338 E. USLEBER ET AL

Table 1 Influence of the labelled antigen on the relative cross-reactivities (RCR) of enzyme immunoassay for saxitoxin with related paralytic shellfish poisoning toxins

Direct EIA Indirect EIA

STX-HRP NeoSTX-HRP dcSTX-HRP STX-BSA

Toxin (mol. wt)

50% dose pmol mI- ' (Pg ml-

50% dose pmoI mI-'

% RCR (pg ml-')

50% dose pmol mI-'

% RCR (pg m1-l)

Saxitoxin (STX) (299)

Neosaxitoxin (neoSTX) (315)

Decarbamoylaaxitoxin (dcSTX) (256)

Gonyautoxin 2/3 (GTX2/3) (395)

0.05 (15.0)

1.61 (510.0)

0-18 (47.5)

0.41 (163.5)

100 0.09 (27.0)

3.1 1 a03 (326.5)

27.8 0.29 (75.5)

12.1 0.88 (348.0)

100 0.05 (1 5.0)

8.7 0.72 (272.0)

31.0 0.31 (79.5)

10-2 0.82 (324.9)

50% dose pmoI mI- '

% RCR (pgml-') % RCR

100 0.46 100 (138)

6-9 22.09 2.1

16.1 1.75 29.3 (404.0)

6.1 4.0 11.5

(6982)

(1 582)

HRP, Horseradish peroxidase; BSA, bovine serum albumin ; EIA, enzyme immunoassay.

Enzyme Immunoarray dcSTX and GTX2/3 were established for each test system. Toxin concentrations that gave 50% competitive inhibition of labelled-antigen binding to the antibodies in each test system were used to calculate relative cross-reactivities.

Indirect EIA with STX-BSA as the solid phase antigen and goat anti-rabbit antibody-HRP conjugate was per- formed as described by Renz and Terplan (1988). Toxin- HRP conjugates were used in direct EIAs as previously described for a STX-HRP conjugate (Usleber et al. 1991). Standard curves for the determination of STX, neoSTX,

Toxin concentration (pg rn1-I)

Fig. 1 Standard curves for the detection of saxitoxin (STX; a), neosaxitoxin (neoSTX; A), decarbamoylsaxitoxin (dcSTX; m) and gonyautoxin 2/3 (GTX2/3 mixture; V) in the enzyme immunoassay with polyclonal antibodies against STX and a dcSTX-horseradish peroxidase conjugate as the labelled antigen. The x-axis (log scale) indicates the toxin concentration, the y-axis indicates the percent absorbance relative to the toxin-free control ( B , ; absolute B , values were 1*25-1.35 absorbance units). Each point represents the mean of four replicates, the coefficients of variation were between 1.5% and 8.5%

RESULTS AND DISCUSSION

The alkaline conditions for the preparation of antibody- enzyme conjugates, described by Wilson and Nakane (1978) for the periodate method, are unsuitable for alkali-labile PSP toxins. This method was, however, successfully used to prepare HRP conjugates of STX, neoSTX and dcSTX; if the pH of the reaction mixtures was kept near neutral and the incubation times were reduced, only GTX2/3-HRP failed to react with the STX-antibodies. The conjugation ratios (moles toxin bound per mole HRP) could not be determined because PSP toxins do not suffkiently absorb U.V. Thus, it is unclear whether the conjugation procedure failed for this toxin, or whether the conjugated GTX2/3 toxins were not recognized by the antibodies. However, the STX-HRP (Usleber et al. 1991) and two novel toxin conju- gates (dcSTX-HRP ; neoSTX-HRP) prepared by the modified periodate method showed that this technique is a promising approach for the conjugation of PSP toxins to HRP or other glycoproteins, particularly if, as in this study, only small quantities of toxins are available. These toxins, which are commercially available from National Research Council, Canada, have been well characterized (Pleasance et al. 1992) and are of higher purity than those used in an earlier study (Usleber et al. 1991).

A comparison of the individual sensitivity levels and spe- cificities of the four test systems, as determined in competi- tive direct (STX-HRP ; neoSTX-HRP ; dcSTX-HRP) and indirect (STX-BSA) EIA is given in Table 1. The indirect

C R 0 S S - R E A C T I V I T Y 0 F S A X I TO X I N I M M U N 0 A S S A Y S 339

assay had the least overall sensitivity, particularly for neoSTX. Similar results were also found by Chu and Fan (1985). By contrast, all three direct assays were very sensi- tive, with 50% inhibition binding values ranging from 15-20 pg ml-' for STX to 270-510 pg ml-' for neoSTX. With these characteristics, the direct assays were more sen- sitive for neoSTX than a previous indirect EIA for neoSTX (Chu et al. 1992). Compared with the homologous STX-HRP, heterologous dcSTX-HRP and neoSTX-HRP increased the relative cross-reactivity with neoSTX about two and three times, respectively.

In the direct EIA with dcSTX-HRP, the detection limits, calculated from the toxin concentrations giving 25% inhibition binding, of the standard curves for STX, neoSTX, dcSTX and GTX2/3 (Fig. 1) were 6.2, 68.0, 24.3 and 96-1 pg ml-', respectively. Compared with the homologous assays, the differences between toxins were smaller, thus reducing the risk of a marked over- or under- estimate of the total toxin content in an unknown mixture. Assuming a minimum dilution factor of 400 for shellfish sample extracts for EIA, as found earlier for the STX- HRP-based assay, the corresponding detection limits would be about 2.5, 27.2, 9.7 and 38.4 ng g- ' shellfish for STX, neoSTX, dcSTX and GTX2/3, respectively. Thus, all the toxins could be detected at concentrations much lower than the EEC tolerance level for PSP toxins (800 ng g - ').

ACKNOWLEDGEMENT

The authors thank M. Straka for excellent technical assist- ance.

REFERENCES

Chu, F.S. and Fan, T.S.L. (1985) Indirect enzyme-linked immu- nosorbent assay for saxitoxin in shellfish. Journal of the Assocz- ation of OflciaI Analytical Chemists 68, 13-16.

Chu, F.S., Huang, X. and Hall, S. (1992) Production and charac- terization of antibodies against neosaxitoxin. Journal of the A O A C International 75,341-345.

European Communities (1991) Richtlinie des Rates vom 15. Juli 1991 zur Festlegung von Hygienevorschriften fur die Eneu- gung und Vermarktung lebender Muscheln (91/492/EWG). Amtsblatt der Europaischen Gemeinschaften L 286, 1-1 4.

Johnson, H.M., Frey, P.A., Angelotti, R., Campbell, R.E. and Lewis, H.K. (1964) Haptenic properties of paralytic shellfish poison conjugated to proteins by formaldehyde treatment. Pro- ceedings of the Society for Experimental Biology and Medicine 117,425-430.

Pleasance, S., Ayer, S.W., Laycock, M.V. and Thibault, P. (1992) Ionspray mass spectrometry marine toxins. 111. Analysis of paralytic shellfish poisoning toxins by flow-injection analysis, liquid chromatography/mass spectrometry and capillary electrophoresis/mass spectrometry. Rapid Communications in Mass Spectrometry 6, 1424.

Renz, V. and Terplan, G. (1988) Ein enzymimmunologischer Nachweis von Saxitoxin. Archiv f i r Lebensmittelhygiene 39, 30-33.

Usleber, E., Schneider, E. and Terplan, G. (1991) Direct enzyme immunoassay in microtitration plate and test strip format for the detection of saxitoxin in shellfish. Letters in Applied Micro- biology 13, 275-277.

Wilson, B. and Nakane, P.K. (1978) Recent development in the periodate method of conjugating HRP to antibodies. In Immu- nojuorescence and Related Staining Techniques ed. Knapp, W . pp. 215-224. Amsterdam: Elsevier.