Ideally you should have a known positive control tissue, fixed in the same way and of the same species as your test material. Different organs may be appropriate to use depending on the expected protein expression.
Although primary antibodies generally come with a suggested working dilution for each application, the best working dilution of an antibody for your lab needs to be determined in conjunction with the detection reagents to be used.
It is customary to do a doubling dilution of a primary antibody and run them in parallel on serial sections of a positive control block. This will indicate the dilution giving the best signal to noise ratio, and if your detection is sensitive enough to enable further dilution of the primary antibody.
Sometimes the pH and composition of the dilution buffer may affect the strength of primary antibody staining. The information provided where manufacturers have recommended specific dilution buffers may be useful to indicate which antibodies might benefit in a variation from your standard antibody diluent. BSA can be a common addition to antibody diluents, but may cause issues when working with primary antibodies of bovine, ovine or caprine origin.
Manufacturers will also likely recommend any unmasking protocols in the antibody information, particularly for antibodies recommended for use on FFPE. The heating method is usually fairly flexible, but some antibodies may require a particular pH of unmasking solution for best results. Be aware that over digestion or unmasking can change staining patterns, so it is important to check that the pattern still conforms to expectations.
Incubation times and temperatures
Due to reaction kinetics, the optimal dilution factor for an antibody is a function of the temperature and time of incubation. If it is intended to run the primary antibody overnight at 4ºC or for a shorter time at room temperature, these should be the conditions applied when optimising the primary. If following another laboratories' protocol, room temperature is often an overlooked variable.
It is recommended to run a positive control of an already optimised primary on a serial section from the tissue block alongside the optimisation run to check that the detection reagents are all working and that the tissue is appropriately fixed. The antibody should require the same unmasking as the test antibody. A lack of staining on only the new antibody is thus demonstrated to be due to the primary antibody.
Once the positive control material demonstrates the expected staining pattern, a control on tissue, fixed and unmasked in the same way as the test sample, but known to be negative for the antigen in question, (either a knock out model, or a tissue/cell type from a different organ) can be run along side the positive control. The primary could also be incubated with its immunogen to inhibit binding, although this does not guarantee the immunogen will not itself give rise to inappropriate staining. This control helps demonstrate the primary antibodies' specificity, but the staining pattern could be due to non-specific interaction of the sample with any immunoglobulin.
Use of a non immune Ig of the same species can be used to demonstrate that it is the specific interaction of the primary antibody with the tissue causing staining as opposed to a general interaction of the sample with the immunoglobulin isotype and species of the primary antibody, but does not demonstrate that the primary antibody is specific for its target.