Search

We have previously shown that an Escherichia coli heat-stable enterotoxin (STa)-biotin conjugate binds to polystyrene microtitre plates coated with avidin (Germani et al., 1992). In the present study the STa-biotin ELISA, based on inhibition of binding of anti-STa antibodies to avidin-bound STa-biotin conjugates, was compared with the conventional suckling mouse assay for the identification of STa from Biken agar extracts and from culture supernatants, using 150 E. coli isolates (50 STa-positive and 100 ST-negative). Pieces of Biken agar were a good source of toxin, 142 of 150 strains gave consistent results by both tests: 100 were negative and 42 were positive; seven of the remaining eight E. coli gave questionable but positive results in the STa-biotin ELISA and were positive by the suckling mouse test; the last E. coli gave negative result by both tests. The STa-biotin ELISA was 85.7% sensitive and 100% specific; the negative predictive value was 0.935 and the positive predictive value was 1. All the 150 strains tested for STa production from standard liquid cultures gave consistent results by both techniques. The STa-biotin ELISA detected 20 pg of partly purified STa compared to 15 pg in the suckling mouse assay.

Covalently prepared chimera antibodies were tested in a ganglioside GM1 erythro-immunoassay (CERIA) for E. coli heat-labile enterotoxin (LT) detection. The antibody specific for LT was conjugated with a polyclonal antibody specific for sheep erythrocytes. The assay is based on the specific binding of LT to polystyrene-adsorbed GM1 and subsequent erythro-adsorption via chimera antibody by which the bound toxin is visualized. Enterotoxin titers determined with this CERIA method were similar to those obtained with the Vero cell assay and with ELISA. 5 ng of cholera toxin/ml may be detected with the assay. The CERIA, as described, may be used either qualitatively or quantitatively and is well suited for routine laboratory diagnosis of LT in a culture supernatant of E. coli.