I have recently been trying to stain human liver samples, fixed in NBFS and embedded in paraffin. We recently got no staining at all, with either the enzyme substrate DAB (for a number of different antibodies which should work on paraffin), or the counterstain hematoxylin, which is very odd. I can think of a number of potential issues that could've caused this outcome, including:
inadequate antigen retrieval (I used 1x Tris EDTA pH9, which should work for at least one of the antibodies used)
over-fixing of the tissue (which we can't do much about now)
inactivity of the reagents (both the DAB and hematoxylin are fairly old, so I think it worth replacing these)
What is very strange is that there was hardly any background staining of the tissue with the DAB substrate on any sections incubated in mouse monoclonal antibodies (which made me initially think the DAB substrate has "gone off") but much higher background with the rabbit polyclonal antibodies. This you would expect do to the difference in monoclonal and polyclonal antibodies, but surely this means the antibodies have bound and are being detected on some level, but perhaps aren't reaching their antigens successfully?
I have lots of experience staining mouse tissues but not human, and have not come across this before. Any help would be appreciated!