Folks,

One point has been made is that power saturation will introduce errors in
measurement of fluor concentrations. This is true if reference solutions
are used, such as the fluorescein solutions I mentioned in my previous
response,  with different laser power settings. I find with fluorescein
using a krypton-argon ion laser on our BioRad 1024 that power saturation is
evident at even 1% power. That being said, to a first approximation, a
given image  "sees" the same power setting throughout so that the effective
quantum efficieny for any fluor in the image will be the same. Therefore,
it is reasonable to compare ratios within an image after correction for
black level offsets.

You can also estimate what the power saturation effect is for a given fluor
and power setting by calibration with reference solutions and use the
calibration to make an absolute estimate of the number of fluors in a
pixel. Again, in a real sample you can only get an estimate. In a given
image the actual excitation volume may vary somewhat from pixel to pixel,
and the power delivered may also vary because of differences in absorbance
and scattering from intervening material. In cases, where the labeling
density is extremely high one can also imagine that inner filter effects
could also result in underestimating the amount of fluor. This would mean,
though, that the fluor absorbance would have to be significant through a
depth of ~0.5 um when using a 1.4 NA objective. If you had fully packed
antibody averaging four fluors per molecule, I get an estimate of about
0.11 OD absorbance through half a micron assuming an extinction coefficient
of 80,000 for the dye. That means about a 25% drop in power through the
volume or an average illumination intensity of about 87%. That's not a big
effect by biological standards. Further, it needn't be said that fully
packed antibody to half a micron will never occur in a real sample. The
case where inner filter effects would come up is with small molecule
labeling, where you can get a very high local concentrations. One case
might be with DNA stains in the nucleus.

The interaction of dye molecules at high concentration has also been
mentioned as a contributor to non-linear effects. There are cases where
this can happen but it requires that fluors reach distances closer than say
about 100 angstroms so that energy transfer can occur between dyes or that
the dyes can interact directly forming dimers such as the case of
rhodamine, which gives rise to quenching. These conditions are not likely
to occur when using macromolecular probes such as antibodies or their like
since the dye concentrations never become significant in this context. One
mg/ml antibody will typically have a fluor concentration of 30 uM. Usually,
self quenching of fluors isn't seen until you get to concentrations greater
than 10 mM.

Regards, one and all

Mario M. Moronne, Ph.D.
Dept. of Subcellular Structure M/S 6-2100
Lawrence Berkeley National Laboratory
1 Cyclotron Rd.
Berkeley, CA
94720

ph (510) 486-4236
FAX (510) 486-5664
[log in to unmask]  or [log in to unmask]