Re: BETTER RED THAN VIOLET? the colur bias...

Howard Shapiro (hms@shapirolab.com)
Wed, 25 Jun 1997 19:40:54 -0400 (EDT)

The problems with UV/violet excitation of fluorescent antibodies have been:

1. UV/violet sources are not ideal - Argon and krypton UV/violet lasers are
expensive and power hungry, and burn out plasma tubes much faster than when
run at 488 (although this may have improved; He-Cd lasers are noisy and
temperamental, and arc lamps put out relatively low power in the UV and
violet, although they work quite well in systems designed for them, such as
Bio-Rad's, Partec's, and the old B-D FACS Analyzer and Cytomutt I. UV does,
of course, excite at least some phycobiliproteins and other relatively long
wavelength dyes, such as Texas red, but that only makes for compensation
problems when you use these with longer wavelength excitation...

2. Pyridine and flavin nucleotides, which are major sources of tissue and
cell autofluorescence, are, respectively, UV-excited/blue fluorescent and
blue-violet excited/yellow-green fluorescent. Autofluorescence of flavins
is even a problem with 488 excitation; that and the PE absorption maximum at
about 545 accounts for the fact that the old FACS Analyzer with 546 nm Hg
arc excitation gave ratios of fluorescence from PE-stained and unstained
cells about 10x higher than a FACS with any level of laser power at 488.

UV-excited tandems might fix problem 2, but not problem 1. However, there
is still problem 3...

3. There are a few dyes which are only useful with UV or violet excitation;
if you want to stain live cells for DNA, you need Hoechst 33342, and indo-1
is still the easiest calcium probe to use, and anybody who wants to combine
either of these with antibody measurements more or less has to use the
longer wavelengths to excite antibody labels. Chromomycin dyes and many GFP
variants also require violet excitation.

Within a few years, we'll probably have tractable solid state lasers running
from the violet to the infrared, and diodes from the red (or maybe the
yellow-green) to the infrared, leveling the playing field in that respect,
but I suspect that it will still make sense in most cases to excite antibody
labels at longer rather than shorter wavelengths.

-Howard