Human Antibody Titration Analysis

Overview:

This protocol is to be used specifically in conjunction with FlowJo^TMflow cytometric data analysis software and Kaleidegraph data anlaysis and graphing software. Refer to the Antibody Titration Protocol for instructions on performing the actual titration. The analysis procedure consists of two steps:

1. Creation of a series of plots of othogonal scatter versus the fluorescence channel being tested for the full titration series.
2. Creation of a table consisting of median positive and negative fluorescences versus antibody titre which can then be exported into Kaleidegraph or other data analysis software.

A typical antibody titration is done as a set of serial (50%) dilutions of the antibody in question staining a constant volume and number of cells. The stains are then analyzed by flow cytometry and the data is imported into FlowJo.

Typically, the first step in the analysis is to create a lymphocyte gate (this assumes that the antigen being stained are found on lymphocytes). The lymphocyte population is then plotted as orthogonal scatter versus the fluorescence for each of the titration points (typically 10 to 12):
 

As the cells are stained with increasingly dilute concentrations of anibody, the intensity of fluorescence decreases. A rectangular gate is then placed upon a distinctly positive staining population (usually in the first point). In FlowJo, a frequency of parentstatistic is applied to this gated population. This  frequency of parent statistic is then used to determine the median positive fluorescence for each of the titration points.

The assumption being made is that the frequency of positive cells does not change from sample to sample. If this is true, then the median percentile of the positive cells can be calculated using the following equation:

median positive percentile = (100 - frequency of parent) + frequency of parent/ 2

A similar calculation applies for the median negative percentile which is used to follow the effects of non-specific staining:

median negative percentile = (100 - frequency of parent) / 2

The median postive percentile and median negative percentile are then applied to the parent population (lymphocytes) for each titration point:
 

In the above example, the frequency of parent is 22%. The calculated median positive percentileis 89, and the median negative percentileis 39. FlowJo then proceeds to calculate the median fluorescence intensity at each of these percentiles. An examination of the first titration point confirms that the calculations are indeed correct:
 

Once the statistics have been completed, the median positive and median negative percentiles can be exported from FlowJo into a data analysis program such as Kaleidegraph. In Kaleidegraph, these values are plotted against the volume or titre of antibody used. In this example, the titres ranged from 2.5 µl to 0.001 µl:
 

The equation used to determined the curve in Kaleidegraph looks like the following:

m1 + m2/(1 + m3/m0); m1=0.7; m2=123; m3=0.3

where m1 = y intercept, m2 = y asymptote, m3 = x value at 1/2 max.

The titration curve, in conjunction with the fluorescence plots, is used to determine the saturating and separating titres of the antibody. In this example, the saturating titre, occurs at ~2.5 µl. This is the point at which fluorescence does not increase significantly with increasing titre.The separating titre, the lowest titre that gives a good distinction between positively stained and unstained cells, occurs at ~0.3 µl, as determined by an inspection of the plots.