[Histonet] TUNEL_FFPE_very_long OT_Delete_if_uninterested

Tue Jul 14 09:34:48 CDT 2009

Hi Jerry, 

I would never try to persuade anyone.  I'm no smarter than the next lab worker trying to make sense of this and science biology in general. But this is how I see TUNEL and FFPE and after all the years I'm happy with things. 

I have not noticed false positives at all when looking at TUNEL FFPE versus frozen or whole cells and even when adapting the animal model to other verifying procedures (annexinV in flow) or caspace-3 when applicable.  Formalin can cause strand breaks, that is true but these are not amplifiable breaks.  For TdT to work, the break must be of particular kinds.  But that discussion is far beyond the realm of possibility in HistoNet.  Up at UW, if you haven't done so, there is a great Cell and Molecular Biology Grad level course that is required in grad school and I'd recommend it for anyone.  It is phenomenal to help in understanding molecular mechanisms. 

When I was down the hallway in grad school in Biological Structure, we had a PhD/MD student in adjacent lab who was working on apoptosis and macrophage scavenging in the developing limb buds (the web between "fingers") for a thesis.  Did TUNEL on FFPE and it was exquisite, beautiful and accepted as a (small part) of thesis.  Thousands of peer-reviewed articles in prestigous journals use TUNEL on FFPE.  Indeed, a review of specific articles where they are are trying to measure precisely strand breaks of DNA in UVA skin or germ-cell or other models, every one I see, the specimen is then placed into formalin or such for processing.  Why would they try to get a precise measurement of DNA strand breaks in their experimental model only to throw specimens into formalin if that is going to cause ubiquitous strand breaks and flood the assay?  All DNA strand breaks are not the same and the kind caused by formalin are simply not amplifiable by TdT. 

If they were, you could simply place a piece of normal, healthy mouse liver or some other tissue with non-apoptosing cells in formalin, process and every nuclei in liver, with broken strands from formalin would turn positive.  I've, never, ever seen such a thing in a well set up TUNEL assay.  And then if you do something to cause apoptosis, those nuclei are positive while others remain unaffected by formalin.  Since I worked in thymus a lot,I had a control block of 4 thymii, one no treatment and the other 3 with varying time treatments of hydrocortisone to induce steroidal T-cell apoptosis (widely accepted model).  The differences where easily recognizable in apoptosis and when run with other tissues, non-apoptosing cells were still clean as anything. 

(1) strand breaks by formalin not amplifiable in TUNEL, otherwise every nuclei formalin fixed would be pos.  Just not so. 

(2) microtomy could not possibly cause strandbreaks that are amplifiable. Cutting is working on a micron scale.  Pieces of DNA are at nanometer or Angstrom scale.  If it did, again, every nuclei would be possitive that the blade "sliced through",  And then why would frozen TUNEL be ok if a blade is slicing through DNA in those sections.  So I disagree that "every one of them (nuclei in section) is detectable by TUNEL".  Again as before a common piece of healthy mouse liver cut at 4 microns would show nuclei all over positive.  That just doesn't happen. 

What I've seen and done, with very specifically controlled experiments, the vast amount  of peer-reviewed literature and thinking through the processes at a molecular level, I just can't come to the conclusion that TUNEL on FFPE is a failed assay that cannot work.  No assay is perfect.  PCR has primer-dimer and other problems to deal with.  IHC has false pos and false neg to worry about.  Flow has Fc receptor problems to deal with.  ISH has stringency issues to create false pos and false neg.  I think there is beyond overwheming evidence that TUNEL on FFPE is an essential (if never perfect) tool in molecular science for apoptosis but as with every assay you have to be aware of limitations and problems.  I just don't believe at all that amplifiable formalin strand breaks and amplifiable microtome strand beaks are any problem at all and should not be a reason to turn from TUNEL on FFPE. 

But again, that is just my opinion that is no more valid than others who might differ. 

Ray 

Raymond Koelling 

PhenoPath Labs 

Seattle, WA  98155 

rocedure called microtomy. When a microtome bvlade passes through the nucleus of a cell it breaks a lot of DNA strands. And every one of them is detectable by TUNEL. 

I've heard of people getting rerasonable results with whole cells and frozen tissues, by for FFPE tissue, my current philosophy is: It is an assay that CANNOT work, even in principle. 

'Course, I've been wrong about other things, 

I'm open to persuasion. 

Jerry Ricks 
Research Scientist 
University of Washington 
Department of Pathology 

_________________________________________________________________ 
Lauren found her dream laptop. Find the PC that’s right for you. 
http://www.microsoft.com/windows/choosepc/?ocid=ftp_val_wl_290_______________________________________________ 
Histonet mailing list 
Histonet <@t> lists.utsouthwestern.edu 
http://lists.utsouthwestern.edu/mailman/listinfo/histonet 