When using antibodies that are raised in the same host animal, multiple antigen labeling requires sequential labeling of each primary antibody. For double labeling, the order of the primary antibody application is very important as well as the different substrates used to achieve the different reaction product colors. Initial conditions such as the dilution must be determined in a single labeling protocol before performing double labeling. Once the individual primary antibody labeling conditions are optimized, we must determine which antibody will be applied first because one antibody may adversely affect the binding of the second antibody.
Unlike double-labeling using immunofluorescence, double-labeling using immunohistochemistry realistically can only be used when the antigenic sites are in different cells or different locations in the same cell type. In experiments localizing cytoplasmic antigens, typically the cell nuclear antigen is prioritized over cytoplasmic antigen, and cell membrane antigen is prioritized over cytoplasmic antigen. In cases where we do not know how one antibody will affect the other, test experiments are performed, reversing the order to find the best staining pattern intensity for both. Normally, the second primary antibody produces a lower intensity stain, so slightly increasing the concentration might be helpful.
We use a polymer-linked horseradish peroxidase (HRP) conjugated secondary antibody from Vector Labs as our detection because of its fast, one-step, highly sensitive, and ready-to-use factors; similar reagents are available from other companies. Polymer-linked peroxidase-conjugated secondary antibodies are far more expensive than conventional peroxidase conjugates, at least with respect to the reagent cost. However, because they save on primary antibody use, minimize
d missed observations, and decrease repeated experiments, we think they are well worth the price.
The combination of 3% PLP fixation, polyester wax embedding, and the polymer-linked peroxidase conjugates produce heightened sensitivity, such that we can often dilute the primary antibodies to a surprising extent. For example, with these methods, we routinely use the pendrin antibody in the featured JASN paper at 1:400,000 dilution. When we use the same antibody on frozen sections by immunofluorescence, the necessary dilution is 1:10,000, an order of magnitude more concentrated.
Finally, distinguishing between the location of the first and second primary antibodies bound to the tissue requires that the peroxidase reaction product for the sequential secondaries are distinguishable. There are multiple selections of enzyme substrates useful for double-labeling immunohistochemistry. Vector DAB (brown reaction product) and Vector SG (blue reaction product) are the two different substrates we typically use in our double-labeling experiments.