Recently attended a seminar on GFP labelled virus infecting axons
(cultured) and examined with confocal and some of the problems he pointed out:

Cells were unfixed, not even with formalin, to have optimal viewing of GFP
with confocal

Formalin or paraformaldehyde fixed cells will maintain GFP, but still not
100%, he abandoned fixed cells to have better results with GFP fluorescence. 

Solvents, temperatures (paraffin?), type of GFP (some are more stable than
others), types of cells transfected (hope I said that correctly) and
expressing GFP may work better than other cells, pH, all affect the protein
and its ability to fluoresce. Clontech has a manual called Living Colours,
probably available on their website to help you work with GFP (mounting
media, autofluorescence problems, etc). 

If you have to detect the GFP with antibodies, this defeats the purpose of
using this unique molecule that fluoresces.  Possibiltiy is to use another
system that you CAN visulize, like beta Galactosidase/XGAL protocols, at
least you CAN develop a blue color seen by light microscopy, plus stability
of chromogen.  

Frozen sections fixed with PFA or NBF, coverslipped with PBS (DO NOT SEAL
WITH NAIL POLISH, contains isopropyl alcohol that leaches into the aq PBS,
quenches GFP) have worked for us.  We thinned toluene based mounting media
to consistency of nail polish using toluene, nonmiscible with water, and
viewed immediately.  Considered ourselves lucky with this technic although
the GFP tended to fade after a good dose of UV light.  

The "nonspecific staining" of GFP in muscle is probably because the GFP is
EXPRESSED by cells in that tissue, and not truly staining, since one does
not STAIN with GFP unless the nonspecific staining is because of background
staining from trying to detect the GFP with antibodies, then it becomes a
matter of blocking per antiGFP antibody IHC protocols.  I would assume here
that GFP expressed by cells and then found in tissues not expected to see
GFP is an experimental question, trafficking of those cells to that tissue,
or some other reason. 

Go back in the histonet archives for more discussion of GFP, not entirely
the magic bullet for those of us in histopathology laboratories, but
certainly a wonderful tool.     


 

 
At 01:57 PM 1/9/01 +0000, you wrote:
>Dear Angeline,
>
>You may well find that if you paraffin embed you lose the fluorescence,
>although you might be able to use antibodies to detect it (which of
>course, opens up more problems!). we have used both fluorescence and
>antibodies to detect this protein, and have managed to accomodate most
>problems, (although if anyone can suggest ways to prevent non specific
>staining of GFP in muscle, like other antibdies, i would be grateful!).
>
>You should be able to see it with UV in the frozens, we certainly have
>with GFP in muscle and in tumours, though we havent really looked so
>much in brains...
>
>Let me know if i can help any more,
>
>Emma Carter
>Oxford BioMedica (UK) Ltd
>
>> -----Original Message-----
>> From:	Angeline Martin-Studdard [SMTP:AMARTIN@mail.mcg.edu]
>> Sent:	Tuesday, January 09, 2001 1:20 PM
>> To:	histonet@pathology.swmed.edu
>> Subject:	GFP
>> 
>> Hello All,
>> 
>> I will soon be delving into the visualization of GFP positive cells
>> that have been transplanted into normal mice.  I have seen a few blips
>> on the histonet regarding problems with this protein.  I will be
>> working with brain tissue sections and will try both frozen and
>> paraffin embedded.  I'd like to know if anyone else has experience
>> with this.  Can the fluorenscence from the protein still be seen after
>> processing of the tissue or does it require detection and/or
>> amplification? 
>> 
>> Any advice or assistance you can offer will be very helpful.
>> 
>> thank you,
>> angeline
>> 
>> angeline martin
>> cell biology and anatomy 
>> medical college of georgia
>> augusta, ga
>> amartin@mail.mcg.edu
>
>
>
Gayle Callis
Veterinary Molecular Biology
Montana State University
Bozeman MT 59717-3610
406 994-4705
406 994-4303