Typically we would use frozen sections for confocal, as the sections are thicker than paraffin embedded sections >35um vs 7um respectively. Also, florescence or DAB staining changes protocols regarding IHC. What are you staining for? It is better to work backwards, i.e. find the antibody compatible and the manufacturers recommended substrate fixation procedures, e.g. works with FF-IHC (fresh frozen - immunohistochemistry), or PF-IHC (paraffin fixed -IHC) etc. Then design your post-fixation protocol to that endpoint.
Typically we would use frozen sections for confocal, as the sections are thicker than paraffin embedded sections >35um vs 7um respectively. Also, florescence or DAB staining changes protocols regarding IHC. What are you staining for? It is better to work backwards, i.e. find the antibody compatible and the manufacturers recommended substrate fixation procedures, e.g. works with FF-IHC (fresh frozen - immunohistochemistry), or PF-IHC (paraffin fixed -IHC) etc. Then design your post-fixation protocol to that endpoint.
It depends a lot on your antigen/antibody/tissue age too. Some antibodies work with any protocol, some are really sensitive to fix (type, pH, length), permabilisation, antigen retrieval, tissue type and age etc. Your best bet is almost always finding someone who has used the same antibody that you have and getting their protocol, and have a look at the manuafacturers data sheet as they often have starting protocols/links to papers that use their antibodies.
I think this is a useful overview of things you can tweak:
Generally I can say immunfluorescence on paraffin embedded sections is possible. But it depends on your question. Thin paraffin sections (around 5 µm) have better resolution. This is important to know when you are interested in small structures like receptors, synapses or organells. Even if your antibody is not working quite well you have the possibility of antgene retrieval. If you like to make fiber analayses or cell and fiber reconstructions thicker frozen or vibratom sections have more advantages. Most comercially available antibodies are suitable for formalin(paraformaldehyd) fixed, paraffin embedded (FFPE) tissue, but it depends on your protein of interest. You are working with mice brain, I recommand if it is possible to look for antibodies which are not generated in mice. The mice on mice antibody reaction leads to unspecific background staining and this makes trouble during fluorescence and confocal microscopy.