Sandwich ELISA formats designed to detect 17 kDa IL-1 beta significantly underestimate 35 kDa IL-1 beta.

The IL-1 beta precursor (proIL-1 beta) represents a significant component of total IL-1 beta production in certain cell types such as keratinocytes, fibroblasts and alveolar macrophages. It has been presumed that immunodetection systems for the mature 17 kDa IL-1 beta can be used interchangeably for the 35 kDa intracellular proIL-1 beta. However, during attempts to purify alveolar macrophage proIL-1 beta, we found that conventional enzyme-linked immunoassays (ELISAs) (using antibodies directed against the 17 kDa mature IL-1 beta) underestimated the amounts of 35 kDa proIL-1 beta by at least ten-fold compared to detection by Western blot techniques. This difference was due to the fact that ELISAs, with an antigen capture format (i.e., that use more than one epitope), can more readily see these distinct epitopes on mature or partially processed IL-1 beta than on the proIL-1 beta molecule. This problem does not occur with the Western blot technique, either because only one antibody is needed and hence there is no stearic blockade of a second epitope or because it denatures 35 kDa proIL-1 beta during the immobilization step, presumably better exposing epitopes as expressed on mature 17 kDa IL-1 beta. The problem with the ELISA can be partially corrected by proteolytic removal of the aminoterminus of 35 kDa proIL-1 beta with neutrophil elastase. More accurate determinations of proIL-1 beta by ELISA can be made by using 35 kDa proIL-1 beta as the reference standard (when the 35 kDa proIL-1 beta is free of molecular weight IL-1 beta). These data suggest that there are conformational differences between the carboxyterminus of 35 kDa proIL-1 beta and mature 17 kDa IL-1 beta which may affect immunodetection when using antibodies directed against mature 17 kDa IL-1 beta.