The physical and functional behavior of capture antibodies adsorbed on polystyrene.

Six monoclonal and two polyclonal antibodies to fluorescein (FLU) were affinity purified and immobilized on Immulon 2 polystyrene as capture antibodies (CAbs): (a) by passive adsorption at pH 9.6, (b) via a streptavidin bridge to a biotinylated carrier molecule, and (c) via an antiglobulin which had been previously adsorbed passively to the polystyrene. Data show that less than 3.0% of the binding sites of monoclonal CAbs and approximately 5-10% of those of polyclonal CAbs were capable of capturing antigen (FLU4.2-BSA) after passive adsorption. Immobilization of CAbs via an antiglobulin or a streptavidin bridge, resulted in the preservation of antibody binding sites to greater than 70% for some monoclonals although immobilization via the streptavidin bridge resulted in the highest number of functional sites/well. The data presented are consistent with studies on other adsorbed proteins which demonstrate that passive adsorption on polystyrene results in the loss of protein function. Furthermore, these data show that generally less than half of the binding sites of antibodies available in solution are available after solid-phase immobilization even when non-adsorptive methods are employed. Some polyclonal anti-FLU also have lower average avidity following passive adsorption compared with CAbs immobilization via a streptavidin bridge. Immunochemical studies revealed that adsorbed polyclonal-CAbs performed like monoclonals when tested with multivalent antigens (FLU10-IgA) but in an expected heterogeneous manner in Scatchard plots when tested using univalent FLU-insulin. This observation implied cross-linking of immobilized CAbs by the multivalent antigen. Because only 5-10% of adsorbed polyclonal CAbs are active, the survivors must be non-randomly distributed in clusters to explain the cross-linking. This was confirmed by scanning electron microscopy which gave rise to the hypothesis that antibodies which retain activity after adsorption, are those present in clusters, i.e., the functional adsorbed CAb is an antibody cluster. Data presented in this report on the behavior of adsorbed CAbs, and reviewed from the work of others for various adsorbed proteins, indicate that the method of passive adsorption at pH 9.6, which is widely used in popular microtiter ELISAs, and which has in many ways revolutionized immunoassay, is a method of protein denaturation. Assayists that utilize passive adsorption of proteins on hydrophobic supports as part of their research need to be cognizant of this phenomenon, while inventors of immunoassay should develop alternative methods of immobilization which do not destroy 90% of the functional activity of solid-phase reactant.

Application of the amplified enzyme-linked immunosorbent assay: comparative quantitation of bovine serum IgG1, IgG2, IgA, and IgM antibodies.

The comparative quantitation of serum antibodies to a defined antigen, using the amplified enzyme-linked immunosorbent assay (a-ELISA), has been demonstrated in a model system in which bovine immunoglobulins (Ig) IgG1, IgG2, IgA, and IgM antibodies to human serum albumin (HSA) were measured. Comparative measurements are facilitated because the same enzyme-antibody complex is used for measuring all 4 isotypes. Serum dilutions from 1:100 to 1:50,000,000 were titrated and, when graphed logarithmically, yielded dose-response plots that contained a linear segment for all but IgA anti-HSA. Data obtained with whole serum and fractions enriched in IgA- and IgM-anti-HSA indicated that dimeric IgA antibodies may compete poorly with those of the IgM and IgG classes and that IgM antibodies are more avid than those of the IgG1 and IgG2 subclasses. Plateauing of complete a-ELISA titrations at optical density (OD)400 nm values lower than those observed for their standard-curve counterparts was interpreted to result from saturation of the antigen. Quantitation was accomplished through the use of standard curves prepared by adsorbing purified, radiolabeled Ig of the 4 isotypes directly to polystyrene. These plots were also valuable in evaluating antiglobulin specificity and potency and for ascertaining linearity in the absence of the primary antibody to be measured, as well as standard curves for determining antibody content in absolute terms. The absolute amounts of IgG1 and IgG2 antibodies to HSA were simliar to those determined by quantitative precipitation and constituted about 25% of the total IgG1 and IgG2 in a hyperimmune bovine serum. Only 5% of the serum IgM was specific antibody to HSA. The specificity of various anti-bovine globulin reagents was further shown by demonstrating the charactristic distribution of IgG1, IgG2, IgM, and IgA anti-HSA in serum fractionated on Sephadex QAE-50 and in sucrose density-gradients. Finally, data on the influence of enzyme-complex concentrations, complex-step incubation times, and the reaction kinetics of soluble antibody-enzyme complexes on a ELISA results are presented.