Eosinophils are easily identified in almost any laser source flow cytometer
with two or more detectors in positions orthogonal to the beam using
measurements of orthogonal scatter in planes of polarization parallel to
("polarized") and perpendicular to ("depolarized") the plane of polarization
of the illuminating laser beam; the method was first described by Bart de
Grooth, Leon Terstappen, and their colleagues. A patent on the procedure is
now assigned to Abbott, and the method is used in the Cell-Dyn 3000 and
later hematology counters from that manufacturer; as far as I know, they
haven't gone after individuals doing the measurement on other instruments,
but B-D, Coulter, Cytomation, Partec etc. won't tell you you can do this
because Abbott would probably sue them.
You can either buy a pair of polarizing filters in the size appropriate to
your instrument or cut them out of Polaroid and tape them in place. Instead
of a dichroic where the first beamsplitter goes in your instrument, use a
simple beamsplitter, or, better still, a cover slip or microscope slide,
which will reflect about 5% of the incident light (mostly orthogonal
scatter) to the detector normally used for orthogonal scatter. While
running beads, place a polarizing filter in front of the detector and rotate
it until the signal from the beads is maximized. This will be the
"polarized" scatter detector. Next, use the dichroic you normally use to
divert light to the orthogonal scatter PMT to reflect light at the laser
wavelength to a second PMT; put a polarizing filter on this and rotate it
until the signal from beads is minimized. This is the "depolarized"
detector. If you examine lysed whole blood or buffy coat from most
mammalian species, the majority of cells will cluster along a line in a
bivariate plot of polarized (x axis) vs. depolarized (y axis) scatter; the
eosinophils are shifted off in the depolarized direction.
If you'd rather not infringe patents, you can detect eosinophils by their
green autofluorescence, which, in unstained preparations, excited at 488 nm,
is equivalent to around 3,000 molecules of fluorescein (use a fluorescein
filter). You could probably get away with using PerCP or PE-Cy5 (or
red-excited) antibodies without interfering with eosinophil
autofluorescence; FITC is obviously out, and I'd be careful about PE.
Since the birefringence of eosinophils was described well before flow and
image cytometry were invented, the Abbott patent shouldn't prevent anybody
from identifying eosinophils by polarized extinction measurements; the
problem is that no current commercial flow cytometers measure extinction.
The old Ortho System 30 and 50 did; they are also among the few laser source
instruments which might not be able to do polarized/depolarized scatter
measurements because the fiber optics they use to relay signals to detectors
probably don't preserve polarization. If it is still impossible to change
filters in a FACScan, that won't work either. However, as was pointed out
by people from Abbott and DeGrooth's group at the ISAC Rimini meeting, you
can make a diode source flow cytometer to do this trick for a few hundred
dollars, so you shouldn't need to spend big money on a new instrument in any
case.
-Howard
This archive was generated by hypermail 2b29 : Wed Apr 03 2002 - 11:49:51 EST