A sandwich enzyme-linked immunosorbent assay was developed for measuring Staphylococcus aureus alpha-toxin. This assay was 500 to 1,000 times more sensitive than the commonly used hemolytic titration assay and was less variable. The binding of alpha-toxin to the adsorbed antibody was most effective after an overnight incubation at 27°C. The toxin was detectable even at a log2 17 dilution of an S. aureus culture supernatant. Staphylococcal alpha-toxin is generally measured by the hemolytic titration assay (8, 16, 22, 24, 39, 40). However, this assay has several disadvantages, such as variability (3, 7, 8, 14) and lack of standardization (25, 27), which limit its use as a quantitative method. The use of enzyme-labeled antibodies in the form of the enzyme-linked immunosorbent assay (ELISA) was pioneered by Engvall and Perlmann (18, 19) and Van Weemen and Schuurs (33, 34). Since its introduction, ELISA techniques have gained wide acceptance (1, 2, 6, 13, 15, 28, 30, 32, 37, 41) over the equally sensitive radioimmunoassay, for a number of reasons as reviewed by Wisdom (38) and O'Sullivan et al. (26). This paper describes the development of a double-antibody sandwich (35, 36) ELISA for the detection of the alphatoxin of Staphylococcus aureus Wood 46; this form of ELISA overcomes the limitations of the hemolytic titration assay. MATERIALS AND METHODS Production of alpha-toxin. The production of alpha-toxin by S. aureus Wood 46 has been described (24). The bacteria were grown under 10% carbon dioxide at 37°C in Dolman-Wilson medium (17) for 36 to 48 h. The culture was then centrifuged at 10,000 x g for 20 min, and the supernatant with alpha-toxin was stored at-20°C. Purification was achieved by adsorption chromatography on controlled pore glass (9). Toxoid. Purified alpha-toxin was rendered nonhemolytic by heating it at 60°C for 30 min. This material is referred to as toxoid in accordance with the terminology of Burnet (11). Production and purification of antisera. Antitoxin was produced by the method of Lo and Fackrell (24). New Zealand white rabbits were hyperimmunized by intravenous and subcutaneous injections of purified toxoid at a concentration of 2 mg/ml in phosphate-buffered saline (PBS). Booster doses were given every 3 days for 3 weeks, and the antisera were obtained 1 month after the last injection. Immunoglobulin G (IgG) was purified by ammonium sulfate fractionation followed by ion-exchange chromatography (12) and then dialyzed against coating buffer. Purity was established by the appearance of a single line when purified antitoxin was diffused against crude toxin. Protein determination. Protein concentrations were determined by the method of Bradford (10). Hemolytic titration. The 50%-endpoint method (40) was modified (24) for use with Microtiter plates (Cooke Labora-* Corresponding author. tory Products), by proportional reduction of the reagent volumes. From an S. aureus culture supernatant, 50 Il was serially diluted in 50 ,u of PBS in the Microtiter plate. A 50