I have been working on a proliferation study on developing lungs of rodents and am having a problem getting specific staining of proliferating lung epithelial cells. The animals are injected with BrdU 2hrs prior to harvesting the lungs. They are fixed in 1%PFM and embedded in paraffin and sectioned. I am getting labelled epithelial cells but I am also getting nonspecific labeling of something in the parenchyma. It initially appears to be some sort of peroxidase but the negative controls (+ primary antibody/ - secondary antibody or vice-a-versa) are clean. Moreover, the labeling in the parenchyma is non-nuclear. The only idea I have is that there is some amount of free BrdU in the surrounding tissue which is being picked up by the antibodies. Any ideas?

Well, I realy have no experience in these sort of things, but maybe you should try to change the fixation protocol and see if it cleans it up.
(e.g. use formaldehyde instead of PFM).