Principles, problems, and strategies in the use of antigenic mixtures for the enzyme-linked immunosorbent assay.

Competition between proteins and other macromolecules for adsorption sites on plastic was studied with the enzyme-linked immunosorbent assay (ELISA) to determine effects of the use of antigenic mixtures or extracts of organisms on assays of antibodies and antigens by ELISA. A comparison of a number of different polystyrene microplates with bovine albumin and human immunoglobulin G (IgG) as antigens showed two major classes of plates; those which adsorbed albumin poorly and those which adsorbed albumin well. IgG adsorbed well on all plates, but plates which adsorbed albumin best also gave significant background levels of nonspecific binding of conjugate. When mixtures of IgG and bovine serum albumin were used as coating antigens, significant competition was observed; the component present at 1% or less in the mixture was essentially undetectable unless excessive amounts of conjugate were used. The important factor was the ratio of competitor to antigen, not the absolute amount. Other proteins (ovalbumin, rabbit albumin, human albumin, and gelatin) were equally effective competitors for adsorption sites on plastic. Nonionic detergents (Tween 20, and Triton X-100) were strong competitors even at 10:1 competitor-to-antigen ratios. In antigen capture assays, normal serum components blocked attachment of antigen-specific IgG, but this competition could be lessened to a degree by the use of strongly binding polystyrene plates. In indirect ELISA for measurement of serum antibody, the use of antigenic mixtures gave significantly lower antibody titers when the desired antigen was less than 1% of the total protein coated. Therefore, the use of mixed or crude antigens in ELISA presents significant problems concerning the sensitivity and specificity of tests.