53
would be preferable to work with uncompensated data. If compensation is required later for
data presentation, it could in theory be done with software. Compensation is not really
necessary except to permit correlation of each fluorescence channel on the flow cytometer
with a single fluorochrome. No information is added (in fact some is lost) by compensation.
When systematically applied, this procedure is not difficult. However, failure to adhere to the
rules about not changing gain settings during the procedure, or failure to run each
fluorochrome individually, can result in great confusion. When setting compensation for more
than two fluorochromes, the conflicting requirements and the possibility of multiple
"spillovers" can be very confusing. Usually compensation is set by running appropriately
stained biological samples, but bead kits for this procedure are commercially available. It must
be stressed that in order to use beads it is absolutely necessary that they have exactly the same
emission spectrum as the corresponding fluorochrome in the biological sample.
CONTROLS
Achieving stable beads with correct emission spectra is technically difficult. In order to insure
stability it is preferable to place the fluorochrome within the bead to protect it from
environmental degradation. However, the emission spectrum is very likely to be different
inside the plastic of the bead that it is on the biological sample. Conversely, placing
fluorochromes on the outside of beads makes them subject to instability. Most available
particles with correct emission spectra are not as stable as one would like.
It might be argued that beads need not have the same emission spectrum as the biological
particle if the relationship between the two spectra is known and is constant. This may be
true, but it assumes that filters in the instrument(s) have not changed. The possibility of high
intensity laser light damaging delicate interference filters in a flow cytometer is not remote,
and thus most laboratories have decided that they need to have particles with emission spectra
similar to the particles of interest.
The next topic that must be addressed is that of controls. Controls are designed to allow one
to detect errors within data or an experiment. Controls for flow cytometry are not different
than controls for any scientific endeavour and thus should not require discussion here.
However, there is frequently considerable confusion about controls in flow cytometery. This
usually stems either from a misconception about the need for controls or from fuzzy thinking
about what is being controlled for. There is obviously no magic in a flow cytometer that
relieves one of the necessity of running controls, but there are a number of levels at which
would be preferable to work with uncompensated data. If compensation is required later for
data presentation, it could in theory be done with software. Compensation is not really
necessary except to permit correlation of each fluorescence channel on the flow cytometer
with a single fluorochrome. No information is added (in fact some is lost) by compensation.
When systematically applied, this procedure is not difficult. However, failure to adhere to the
rules about not changing gain settings during the procedure, or failure to run each
fluorochrome individually, can result in great confusion. When setting compensation for more
than two fluorochromes, the conflicting requirements and the possibility of multiple
"spillovers" can be very confusing. Usually compensation is set by running appropriately
stained biological samples, but bead kits for this procedure are commercially available. It must
be stressed that in order to use beads it is absolutely necessary that they have exactly the same
emission spectrum as the corresponding fluorochrome in the biological sample.
CONTROLS
Achieving stable beads with correct emission spectra is technically difficult. In order to insure
stability it is preferable to place the fluorochrome within the bead to protect it from
environmental degradation. However, the emission spectrum is very likely to be different
inside the plastic of the bead that it is on the biological sample. Conversely, placing
fluorochromes on the outside of beads makes them subject to instability. Most available
particles with correct emission spectra are not as stable as one would like.
It might be argued that beads need not have the same emission spectrum as the biological
particle if the relationship between the two spectra is known and is constant. This may be
true, but it assumes that filters in the instrument(s) have not changed. The possibility of high
intensity laser light damaging delicate interference filters in a flow cytometer is not remote,
and thus most laboratories have decided that they need to have particles with emission spectra
similar to the particles of interest.
The next topic that must be addressed is that of controls. Controls are designed to allow one
to detect errors within data or an experiment. Controls for flow cytometry are not different
than controls for any scientific endeavour and thus should not require discussion here.
However, there is frequently considerable confusion about controls in flow cytometery. This
usually stems either from a misconception about the need for controls or from fuzzy thinking
about what is being controlled for. There is obviously no magic in a flow cytometer that
relieves one of the necessity of running controls, but there are a number of levels at which