 |
 |
 |
 |
ANSWER: Actually, you would NOT want a saturating concentration of mAb when
you are initially doing the quality control on your antibody to define the
curve of fluorescence intensity vs. molar amount of antibody bound. As I
mentioned in the note I posted previously, you would initially want to use
amounts of antibody which will ALL be used up by the cells, so that you can
calculate how many moles of antibody are bound to the cells and thus how
much fluorescence per mole. Now when you are running an unknown you want to
be in the same range as your curve, since linear extrapolation of the
fluorescence vs molesofantibody curve will probably not be valid. Because
of this, it would be prudent ot run your curve on a cell type that has LOTS
of the antigen in question on the surface. That way your unknown cells will
be in the range of your control curve and you can back calculate moles of
antibody (and thus moles of antigen) from the fluorescence. You do not
necessarily know for an arbitrary antibody that it is univalent (ie has
only one binding site per molecule of antigen). Therefore you must have
some independent information about this from the maker of the antibody or
from the literature, but this is often available. When you stain your
unknown cells, of course the antibody must saturate the antigen.
>
>2. Do you or anybody else on this mailing list have any ideas as to the
>degree of error which one could expect from this method (providing an
>essentially saturated condition can be achieved) due to such things as:
> -non-specific staining via Fc receptors?
> -compromised membrane integrity (allowing cells to stain
> non-specifically) due to cells just in the process of dying but that
> are not yet easily differentiated from the cells of interest (i.e.
> cannot be gated out)?
> -mAb polyspecificity?
ANSWER: Don't know, but would be interested in other peoples' results on this.
>
> Your method would work I believe, as long as one can determine when
>cells are as close to being saturated with the mAb as possible and still
>know the amount of Ab used. As well, I believe that it would be very
>useful to have a general idea of the degree of error one could expect. I
>would be interested in people's opinions of the potential accuracy of
>this method. i.e. would the fact that you can never, theoretically,
>achieve saturation (and therefore, get less staining), counterbalance the
>fact that you would get higher staining due to non-specific interactions,
>or would one outweigh the other?
COMMENT: I don't believe that it is impossible to get saturating conditions
with a monoclonal antibody. Many IgG monoclonals have very high affinities
for their ligand, and should be able to saturate well. The apparent
"nonsaturation" that people observe (eg continuous increase in fluorescence
intensity with time with MoAb staining) in my opinion is largely due to
nonspecific staining. It is possible to address at least some of this
nonspecific staining in the same way that the radioiodinators do and adding
cold antibody to block the SPECIFIC component of the staining. To do all of
this right does take some doing, and I cannot pretend to have done it right
myself.
>
> Thanks for your ideas Dr. Milford,
>
> Byram W. Bridle
Edgar L. Milford, M.D.
Brigham and Women's Hospital
Tissue Typing Laboratory
75 Francis Street, Boston, MA, 02115
Tel: 617 - 732-5872
Fax: 617 - 566-6176
|
|
|
|