[Histonet] RE: PCR and ISH in fixed tissues

Histo Jock histojock <@t> hotmail.com
Tue Jul 5 20:05:58 CDT 2005

Hi Caroline,

Here's what I know:

The issue isn't whether or not you can "do" RNA work on fixed tissues, it's 
whether or not you can trust the result.

At my former place of employment we ran a study comparing ISH and IHC 
results on FFPE and matched snap frozen specimens. The frozen results were 
always rock steady and matched the IHC dead on. The ISH on FFPE was all over 
the map - when you actually got staining that was. Results I've seen from 
other places follow the same pattern. Absolutely no consistancy with the 
fixed material and much, much tougher to optimize the signal.

I know that fixation procedure is a least partly to blame for this. If the 
initial specimen was a big piece you'll often see less signal in the center 
as it took longer for fixative to infiltrate there. Where perfusion isn't an 
option, a lot of researchers will insist on cutting their animal specimens 
into tiny little pieces and dropping them in fixative as quickly as possible 
to prevent this. Real pain to work with those!

You also might as well forget about anything that was embedded before the 
age of buffered formalin. Formic acid is death to nucleic acids. Old, 
unbuffered formalin that's gone acidic is a sure way to kill an ISH signal. 
I've never seen even a hint of a signal out of anything embedded pre-1980's

Finally, keep in mind that different species of RNA are subject to different 
rates of degradation. I've seen FFPE specimens that showed great results for 
total RNA, but were useless when probed for a specific species. This is 
particularly true if you try standardizing against an rRNA which is almoist 
necer a good idea.

All of the above do not apply if you use snap frozen material. Never had any 
problems at all with it.

I'm not doubting that some places do ISH on fixed slides and get a signal 
from it. The literature is full a this and some times that's all you've got 
so you go with it. Given the choice, though, I'd stick with the frozen stuff 
every time despite the extra hassle involved. If you do use fixed stuff only 
trust it for localization, not quantitation.


>Hey guys,
>My boss is submitting a grant and he does not believe that you can do
>PCR (RT) or ISH with fixed tissues.  I thought you could but I don't
>know any details.  Could someone elaborate?  Is there a particular way
>the tissue should be fixed?  In other words do I have to use FFPE
>sections or can I use 4% paraformaldehyde perfusion?
>Any advice would be helpful.
>Caroline Bass

Express yourself instantly with MSN Messenger! Download today - it's FREE! 

More information about the Histonet mailing list