 |
 |
 |
 |
BTW, we find that for most antibodies, the cell number being stained is
relatively unimportant (i.e., the amount of antibody is in vast excess
over antigen). However, this may not be true of all antibodies. You
should use consistent cell numbers (1 million per test is generally
used), and do a test to make sure that a sample with extra cells (3-5
million) has the same staining profile.
We also fit standard Michaelis binding curves to our data; you can use
Kaleidagraph or JMP [=SAS for Macs] (or for those with Consort-40 or -VAX,
you can use COTFIT) to do nonlinear least squares fitting of any function.
For those of you who will test manufacturer's reagents, you will find
that most of the phycobilli-conjugates (PE, Cy5-PE) are packaged well
below saturating concentrations. Most FITC and Biotin reagents are
packaged at saturating concentrations. For some conjugates, this is done
because saturating concentrations would make cells off-scale in
fluorescence (especially for Cy5-PE). Our approach is to "cut" these
reagents with unconjugated (i.e., cold competitor) such that the final
concentration of mAb is saturating and the final fluorescence is
on-scale.
It is important to understand the consequences of being below saturating
concentrations of these reagents: the final fluorescence that will be
obtained will be proportional to (1) the amount of mAb added, and (2)
the time of incubation with the antibody. Again, the cell number
usually isn't a factor, because even at subsaturating concentrations
that mAb can still be in vast excess over antigen. If you are doing
antigen density measurements, it becomes essential to have saturating
concentrations of reagent!
mr
|
|
|
|