Paul,

I have done lots of fluorescence work in fish and retina; both are notorious 
for autofluorescence.  My observation has been that I can minimize the 
autofluorescence by keeping my tissue hydrated.  For instance, if I do IHC on 
fixed frozen sections, and then mount them in a 50:50 glycerol/buffer 
mixture, they look great for a day or two.  But with time, the glycerol, 
which is extremely hygroscopic, draws the water from the tissue, and the 
autofluorescence becomes screamingly bright.  The situation is extreme if I 
mount in something that hardens (like Mowiol), thereby drying the tissue 
completely.  At the time I was doing this research, I was working at 
Molecular Probes, and one of the fluorescence gurus at the company thought 
the reason this happens is that water preferentially quenches the endogenous 
fluorophores in the tissue, keeping the autofluorescence background low.  If 
his hypothesis is correct, I would suspect that there is residual paraffin in 
your tissue that's preventing complete tissue hydration.  To test this 
hypothses, you might try something like doing antigen-retrieval in a 
detergent-containing buffer to see if this diminishes the autofluorescence.

You are absolutely correct about the autofluorescence being worse in the 
fluorescein channels.  Apparently, there are more endogenous fluorophores 
that emit in the region of the spectrum.  One thing that is always 
recommended if autofluorescence is a problem is to use longer-wavelength dyes 
like Texas Red or Alexa 594 or some of the Cy dyes.  If you have a Texas Red 
filter set or a long-pass rhodamine filter set, you could try secondary 
reagents conjugated to one of these longer-wavelength dyes.  This should 
help.

There's also other tricks like incubating your tissue in 0.1% sodium 
borohydride in PBS for 30 minutes.  The sodium borohydride reacts across 
double-bonds, thereby reducing the pi-bond interactions that result in 
autofluorescence.

I hope this wildy speculative note on autofluorescence helps!

Karen in Oregon



Date:          Tue, 05 Oct 1999 11:01:37
From:          Paul Klosen <klosen@neurochem.u-strasbg.fr>
Subject:       Background fluorescence
To:            histonet@pathology.swmed.edu

Dear fellow Histonetters,

I recently started reusing immunofluorescence after years of
immunoperoxidase detection systems. I have encountered a problem with
"background" fluorescence on both paraffin and polyethylene glycol
sections. Under fluorescence illumination, these sections appear to
"fluoresce" all over the sections, even if I have not applied any
fluorochrome. This effect is more marked on the fluorescein channel, than
the rhodamine or the UV channel. I can see my immunostaining after
immunolabelling, but it has to stand out against a general red or green
background on the sections, which probably obscures faint signals.

I have checked sections of tissues fixed with different fixatives (aldehyde
and non-aldehyde), and its not linked to any particular fixative. I don't
observe this effect when doing immunofluorescence on both cell cultures or
fresh frozen tissue section fixed by methanol or formaldehyde. My sections
were mounted in PBS/glycerol with different antifading agents (no
correlation there) or unstained sections were also dehydrated and mounted
in Eukitt.

I am not sure whether this is really fluorescence or some kind of "light
scattering" on paraffin and PEG sections. So what is happening here ?? Has
anyone encountered similar problems and is there a way to reduce this
effect ??

Thanks

Paul
                                    -=-
                                   (o -) O
===============================oOo==(_)==OOo=================================
=
Paul Klosen, PhD
CNRS UMR 7518 Neurobiologie des Fonctions Rythmiques et Saisonnieres
Universite Louis Pasteur  12, rue de l'Universite
F-67000 Strasbourg, FRANCE
Tel. 03.88.35.85.04  Fax. 03.88.24.04.61
========================klosen@neurochem.u-strasbg.fr========================
=