Differences in humoral insulin-antibody response among inbred Lou/M rats and epitope presentation differences in ELISA and radioimmune titration.

Insulin antibodies were not detected in an insulin-capture enzyme-linked immunosorbent assay (ELISA), but they were easily detected in an insulin-copolymer-capture ELISA. Thus, there is a high degree of steric hindrance because of the proximity of the epitopes on the insulin monomer. This is circumvented by substituting an insulin copolymer for insulin in the capture ELISA. A regression analysis comparing the titers of 28 Lou/M rat insulin antiserums measured by liquid-phase radioimmune titration (RIT) with titers obtained in the direct insulin ELISA was not significant (P greater than .05). Thus, epitopes on insulin available and/or masked for antibody binding in the RIT differ from those available and/or masked in the direct insulin ELISA. As more of the epitopes become available when an insulin copolymer is substituted for monomeric insulin in the ELISAs, a significant positive correlation (P less than .05) with the RIT was observed with these 28 insulin antiserums. Twenty-five percent (7 of 28) of these antiserums contained more antibodies that bound to epitopes available in the ELISAs that were masked in the RIT. Conversely, two antiserums contained more antibodies that bound to epitopes that were available in the RIT but were masked in the ELISAs. Thus, the amount of insulin antibodies measured in a given antiserum can vary substantially, depending on which epitopes are made available or are masked in the particular antibody-titration method used. These results demonstrate that the humoral immune response to insulin among inbred Lou/M rats can vary in insulin-antibody levels as well as the epitopes on insulin to which the antibodies bind.