Something in your protocol is auto-fluorescent. Could be residual parafin; could be PF. Do you quench after fixation?

Did you try Target Retrieval from Dako (soduim citrate buffer ph 6 or Tris-EDTA ph 9 for 30 or 20 mins)? This permeabilises bettter, Tween and Triton disturb the  Dapi staining. Use PBS and BSA for antibody diluent, and only PBS for Dapi. I use as mounting media the one from Dako.

Try to remove more parffin from you sections

@Raquel,  I use xylene to remove wax. you mean to say I should try to increase the no. of incubations in xylene or apart from xylene there is some other solvent.

It might be your fixatives, because the choice of fixatives depends with target proteins. Fixatives may surround the protein of interests, thus blocking it from binding with antibodies. Antibodies cannot pass through certain fixatives, thus important factor to consider when planning to do immunohistochemistry. Thus you will need to do antigen retrieval protocol (refer to paper linked here): "Antigen retrieval was achieved by boiling in 10 mm citric acid for 15 min".

Article Evidence of Adrenal Failure in Aging Dax1-Deficient Mice

We also got similar problem (but only occasionally). We switched to using Hoechst 33342 and it has worked great.

We use DAPI all the time on paraffin sections of human, mouse and rat brains  with no problems whatsoever. I would need to see your protocol to be able to give an answer to your problem.

• DAPI nuclear couterstaining of FFPE tissue requires that it  has been antigen retrieved with heated citrate pH 6 or tris pH 9 buffers.  The problem you describe sounds like an antigen retrieval issue.

you can try to slight dry your section before mounting.