Referring to the question of Andreas Schroeder a few points should be
made. The paper in the Journal of Immunological Methods he cites,
states that DiOC(18)/PI canNOT be used for flow cytometric analysis of
cell mediated cytotoxicity; another paper in the same journal (volume
166, p. 45, 1993) demonstrates successful use of this approach.
Association of target and effector cells is a prequisite for cell
mediated cytoxicity to occur. So, one should not be surprised that
it actually does take place. In the flow cytometer one scores doubly
positive (i.e. green/red positive) vs singly positive (only green
positive cells). That takes care of the association of unlabeled
effector cells with labeled target cells.
If one has a lot of clumping in a sample (in particular higher order
clumps), filtering may help. Then, one has to check that there
filtering does not selectively exclude a cell population;
that may lead to a bias in your assay.
Another way of dealing with clumps of particles is by taking into
account that a clump of two spheres which passes by a norrow beam of
light will have a different ratio of pulse width vs pulse height. In
a plot of pulse width vs pulse height, clumps are off the diagonal
(which represents "single" events). Wolfgang Goehde (University of
Muenster, Germany) recognized this phenomenon; it is incorporated in
some flow cytometers, where it is known as the "pulse processor".
Hope this helps,
Martin Poot