Flow cytometry has the potential for providing quantitative information about the relative expression of membrane proteins on different cells. For example, we may want to determine if cells, when stimulated, express more of a given protein than they do under resting conditions. In order to be able to draw quantitative conclusions from flow cytometric data, we have to be sure that the fluorescence intensity of a stained cell is directly and linearly proportional to the protein we are measuring. This involves using concentrations of reagents that are not limiting, so that the staining intensity of a cell will be related to the amount of the protein on a cell and not to the number of cells present nor to pipetting inaccuracies in the exact amount of reagent used. Determining the amount of stain to use involves titrating each antibody reagent to find a concentration that approaches saturation within the time of the reaction. With low affinity/avidity antibodies, strict saturation may not be possible in short staining times at reasonable antibody concentrations; therefore use of such antibodies makes conclusions about relative intensities difficult.
To determine a saturating concentration, run a titration curve by using halving dilutions of antibody, covering a range above and below that suggested by the manufacturer (if any). Make this 1:2 dilution series (using 10 halving dilutions) of both your antibody of interest and also of an irrelevant antibody matched by concentration and isotype. Then stain cells with the standard staining protocol, using both the irrelevant antibody and the antibody of interest from each of the 10 dilutions. As your final working concentration, pick a dilution of stain well on the intensity plateau for the stained cells, where small changes in stain concentration have little effect on cell fluorescence intensity. Very high antibody concentration is not, however, a panacea: cells often stain less well at very high antibody concentrations than they do at lower concentrations. Additionally, as can be seen from the cells stained with the control antibody, even negative controls often begin to stain non-specifically at high antibody levels. The important thing here is to maximize the signal-to-noise ratio: use a concentration on the plateau that gives maximal staining with the positive antibody but little staining with the control antibody.
With indirect staining, titration with halving dilutions is necessary for both the primary and the secondary antibody. Although it may seem as if there are too many unknowns in this equation, it is possible to guess at a reasonable concentration of one primary antibody, titrate halving dilutions of the secondary antibody with that concentration of the primary antibody, and then, using the concentration of secondary antibody determined to be optimal, titrate all primary antibodies in turn. Because non-specific staining will increase at high antibody concentrations, when titrating the secondary antibody it is particularly important to compare the staining of cells with and without primary antibody present. Cells stained with the secondary antibody, but without any primary antibody, are known as a "secondary control."As with titrating directly-conjugated antibodies, you want to use a level of secondary antibody that maximizes the signal-to-noise ratio, that is, gives a saturating signal when the primary antibody is present but little signal when the primary antibody is absent.
At staining concentrations that reach saturation, the amount of the stain present is hardly ever limiting: that is, there is usually enough antibody around to stain all the relevant cell proteins on all the cells without significantly lowering the concentration of free antibody (see Kantor and Roederer for examples of this calculation). The number of cells is not, therefore, critical to the staining intensity nor is the volume of the suspension. We simply need a high enough concentration of antibody so that effective equilibrium between bound and unbound antibody will be achieved in a reasonable staining time.