C. GENIGEORGIS and W. W. SADLER Department of Public Health, School of Veterilzary Medicine University of California, Davis, California 95616 lmmunofluorescent Detection of Staphylococcal Enterotoxin B SUMMARY Application of the fluorescent antibody tech- nique (FAT) to detection of staphylococcal en- terotoxin B and of the cells that produce it in culture media was investigated. Enterotoxin B was detected by FAT in culture media both in the presence and in the absence of bacterial cells. Two methods of detection were developed. The first involved staining of fixed smears with fluorescein-isothiocyanate-conjugated anti- enterotoxin B serum. The second technique in- volved the precipitation of enterotoxin by mix- ing one drop of fluid containing either living cells and toxin or only toxin with two drops of serially diluted conjugate. After incubation, the formed precipitates were caught on Milli- pore filter membranes and washed, and impres- sion smears were made on slides. Enterotoxin B was demonstrated with the first technique only when there was a minimum of 15 fig/ml. The second technique detected as little as 1 fig toxin/ml. INTRODUCTION Much progress has been made lately in methodology for detecting staphylococcal enterotoxins. Biological tests (feeding mon- keys or injecting cats with toxic filtrates) have been replaced more and more by specific serologic procedures. The antigenicity of two of the known enterotoxins (A and B) has been demonstrated. A has been purified above 70% (Bergdoll, 1962), and B above 99% (Schantz et al., 1965). Specific anti- sera against these two types of toxins are now available, and both toxins are demon- strable in cultures or food extracts by single and double gel diffusion methods. Quantities of 1 pg of toxin per 1 ml of culture can be detected by the slide gel diffusion test, while tube double gel diffusion can increase the sensitivity to 0.05 pg/ml (Hall et al., 1963, 1965). Type B strains usually produce large amounts of toxin in broth cultures (over 100 pg/ml) whereas type A strains produce only minute amounts. Casman and Bennett (1965) reported that with IS-20 billion microorgan- isms per ml the production of type A toxin I. Detection in Culture Media was about 2-4 pg per ml of aerated brain- heart infusion (BHI) broth. Since foods implicated in food poisoning rarely have such a great number of staphylococci, much less toxin would be expected to be present. The demonstration of enterotoxins in foods involves two basic problems : 1) extraction ; and 2) concentration to detectable quanti- ties. Experiments have been conducted by adding known amounts of toxin in foods and then recovering them by extraction. Recov- eries have been variable: Casman and Ben- nett (1965) reported recovering 687% of en- terotoxin A and 48 to 72% of enterotoxin B ; and Hall et al. (1965) reported recovering 4So/o of A and 75% of type B. Ion-exchange chromatography and gel filtration are used for selective separation of enterotoxins from broth cultures or food extracts, which are further concentrated until the concentration is high enough to be within the sensitivity limit of the gel diffusion test. At least 24 hr are necessary for the detection of toxins when their concentration is in the area of 1 pg/ml broth or food extract. An indirect hemagglutinatioI-inhibition procedure has been developed (Robinson and Thatcher, 1965 ; Brown and Brown, 1965) ; and a flotation system is reported which can detect enterotoxin B of 1 pg/ml in 2-3 hr (Hopper, 1963, 1965). Both of these meth- ods still need more refined techniques of ex- traction of enterotoxins from foods. It is obvious from the foregoing that no suitable method is available which can dem- onstrate enterotoxins or enterotoxic cells in foods in a short period without resorting to extractions, concentrations, and gel diffusion. Since enterotoxins react specifically with their antibodies, the fluorescent antibody technique (FAT) might be applicable for a direct detection of toxins and toxic strains in foods in a short period by the use of conju- gated antienterotoxin sera. To study this possibility, experiments were first conducted with pure broth and agar cultures of toxic r 4411