 |
 |
 |
 |
There are two issues involved:
1. optics (getting, installing, and aligning)
2. data analysis: computing polarization from detector pulse height
measurements.
Optics:
We've done fluorescence polarization using 488nm excitation. Optics
include a 520LP filter (to remove 488nm side scatter), a 50/50 beam
splitter (placed in the dichroic filter position between detectors
FL1 and FL2) to separate vertical and parallel components, and
polarizers in front of FL2 (vertically-oriented with respect to the
plane of the laser) and FL1 (horizontally oriented).
The optical system was balanced using a half wave retarder in a
rotation mount to adjust the vertical plane of the laser to 45
degrees, and adjusting PMT high voltages such that the response from
both detectors was equal. (The half wave plate was mounted on an
aluminum plate fixed to one of the laser cover mounting positions in
front of the 2W argon laser. Yes, the laser safety covers are
removed. The cover mount on the cast optical bench is non-precision
(that is, it is not parallel to the plane of the bench) so shims were
used to get the half wave plate vertical plane perpendicular.)
Optics were purchased from Newport Optical (Fountain Valley,
California) except for the 50/50 beam splitter mounted in an aluminum
ring (special order; the beam splitter is very thin and fragile, you
want it mounted in a ring) from Omega Optical (Brattleboro, Vermont).
The beam splitter in its mounting ring was mounted in a
custom-machined filter holder (working drawings available as a
device-independent PostScript file by anonymous ftp at
flowcyt.cyto.purdue.edu; see following article for instructions on
downloading.)
Computation of Polarization
Fluorescence polarization is the difference of the parallel and
perpendicular components divided by the sum of the two. Computing
this value for each cell needs to be done using the parallel and
perpendicular list mode values since there are no analog processing
electronics on the FACStarPlus.
There are several ways to compute the polarization values from list
mode data. We accomplished this using a simple package in Mathematica
which computes the polarization for each value (multiplied by a
suitable constant for proper scaling), and writes a new list mode
file with the additional derived parameter.
A brief summary was presented in a poster at the last ISAC meeting.
Dave Coder
Dept. of Immunology
Univ. of Washington
Internet: dcoder@u.washington.edu
fax: 206-543-3480
tel: 206-685-3014
|
|
|
|