Abstract
The nitrocellulose model and microphotometry were used to investigate whether in immunoperoxidase cytochemical methods the amount of final reaction product reflects the amount of cell surface antigen. The results obtained with four cytochemical peroxidase methods, i.e., those using diaminobenzidine/H2O2 (DAB/H2O)2, DAB/H2O2/COCl2, DAB/H2O2/imidazole, and silver intensification of the DAB end product, were compared first. The quantitative DAB/H2O2/imidazole method proved to be the most sensitive and was selected for further studies. Cell surface antigens prepared by solubilization of peritoneal macrophages with octyl-beta-D-glucopyranoside were immobilized on nitrocellulose. Monoclonal antibody binding to these cell antigens was detected by peroxidase immunocytochemistry. Comparison of the sensitivity of the indirect immunoperoxidase and the biotin-(strept)avidin immunoperoxidase methods on the basis of the highest detectable dilution of a cell lysate showed that these methods were equally sensitive. A linear relationship between the absorbance of the peroxidase reaction product and the amount of cell lysate immobilized on nitrocellulose was found for all three indirect immunoperoxidase methods. This proves that the amount of final immunocytochemical peroxidase reaction product is proportional to the amount of antigen in cell lysates. However, the relative expression of antigens in intact cells differs from that in cell lysates. Therefore, the present method to solubilize cells and immobilize cell antigens cannot be used to quantitate the antigen content of cells.