Dear Bruce, Thanks for your strong defense of what is clearly a methodology that is headed in the right direction, viz., the employing of fluorescent standard beads to convert the measured average cellular fluorescence intensity to some equivalent number of target receptor molecules per average cell. I have no argument with the use of beads; indeed, I was among the first persons in flow to attempt to elucidate the relationship between average cellular fluorescence intensity and the underlying number of target receptor/marker/antigen molecules which bound the fluorescent ligand to produce that fluorescence. It was clear to me even then, as Bohm and Sklar had both indicated before me, that a bead standard was the necessary vehicle to enable the conversion from numbers of fluorescent photons emitted by labelled cells to the numbers of bound fluorochromes underlying those fluorescences. In my original letter to this List, I stated the two main technical assumptions in employing beads then went on to elaborate some other factors which lead to what is the inhererent inaccuracy (not imprecision, thankyou Mario Roederer for your gentle distinction) in the numbers which are finally held aloft by Neptune [or was it Laertes or Medusa (like this issue)] emerging from the primeval soup: Spake Neptune, "Eureka, the number of CD71 receptors on a proliferating lymphocyte is 745,250 +/- 25,000!!" My concern is based on the observation that numbers have a way of becoming the gospel truth. I want to caution all players to the reality that although they may get very reproducible data (high precision), the numbers are still quite rubbery, because of the many uncertainties that occur through the chain of assumptions that is followed. A couple of examples: Consider an antigen whose expression can be up- and down- regulated. At low density, the Mab-FITC binding will be entirely monovalent. At high density, the Mab-FITC binding will be entirely bivalent once the targets are dense enough. This geometry needs to be addressed, as the numbers of bound FITC can be the same while the antigen target numbers changes dramatically. The second is pertinent to clinical states, where the number of marker molecules may change, but so also may the affinity be altered due to molecular defects in the marker itself. The fact is that the number of Mab-FITC ligands that bind is a function of both variables: the number of receptors expressed and their affinity for the binding ligand. Antibodies can vary considerably in their affinities for a particular receptor, indeed, affinity maturation is a feature of the immune system, so why could it not work the other way around? In summary, our task is not just to get a number, but to get the affinity as well. In this way the number data can be made unequivocal. Cheers, Bob
This archive was generated by hypermail 2b29 : Wed Apr 03 2002 - 11:50:04 EST