[Histonet] RE: Staining of nucleus - real?

[C.M. van der Loos]

Dear Min-han,
The problem as you described whether or not the staining of an antibody is real is a bit like the chicken-versus-egg problem: difficult to say something sensible about it. I would like to share some thoughts from the practical point of view:
1. As you already said, consider your negative controls first. Tissue pretreatments, fixation, detection systems and chromogens all may have their influence on background staining. Of course this should be excluded first.
2. Try to find a "trustful" positive control. Tissues mentioned in literature, cultured cells, rare tumors, etc. may be very helpful. 
3. Consider the point of leaking away of antigens during fixation, and perhaps precipitating elsewhere at the tissue section. Especially when working with cyro's one should be careful with the interpretation of IHC using anti-cytokines, chemokines. 
4. The IgG concentration of the diluted antibody. Once above 2-5 ug/ml you should be suspicious about your staining result. Non-specific staining above 5 ug/ml easily occurs (although there are also some positive examples in this respect).
5. When working with FFPE sections test different tissue pretreatment procedures (nothing - pepsin or pronase - HIER/citrate6.0 - HIER/EDTA9.0 - "exotic" pretreatments). Again this may yield different staining patterns, but there should be at least some similar staining features.
6. Compare the staining patterns of different antibodies to the same antigen/epitope. Certainly not the most economic way, but sometimes very informative (not to say "revealing").
7. Finally, you have go from the testing phase to the target tissue. At this point you have to select an antibody, fixation, tissue pretreatment, detection system, etc. that all together yield a staining result you feel comfortable with. This remains a scarry moment. I do agree with that!

Personally I think that the IHC evaluation of all those new research-type of antibodies on the marked, is perhaps one of the biggest problems we have to struggle with in the next few years. 
I hope my thoughts at least gives you some moral support.

Chris van der Loos, PhD
Dept. of Pathology
Academical Medical Center
Amsterdam - The Netherlands


----- Original Message ----- 
>From  "Tan, MinHan" <MinHan.Tan <@t> vai.org> 
Date  Tue, 6 Jul 2004 21:28:06 -0400 
To  <histonet <@t> lists.utsouthwestern.edu> 
Subject  [Histonet] Staining of nucleus - real? 
Good evening,

I'd like to ask a question - we have recently purchased a polyclonal
cadherin antibody from Santa Cruz that does stain the inter-cellular
compartment, but unexpectedly, also stains the nucleus in some cells in
kidney tissue. (no other tissue reflects this kind of staining), as far
as we can assess. One pathologist thought it was a 'background' nuclear
staining; another felt it was likely to be a true stain. Although we are
using the ABC system, it is not a biotin-derived staining, since our
controls omitting primary antibody are negative and the staining is not
cytoplasmic.

How can we evaluate whether this staining is due to the antibody binding
to the relevant antigen, or whether this reflects cross-reactivity, or
whether this is 'background nuclear staining'?

Any advice would be very much appreciated!

Regards,
Min-han