I am working with stroke animal models and have to perform the immunohistochemistry time and again. To label the astrocyte I generally used the GFAP marker with 1:500 dilution of the antibody. The problem was GFAP labels the astrocytes only in the edges of cortex and sometime in the layers in between cortex and striatum. It labels the brain slices from the sham operated brain slices as well (where actually it should not). So how can I eliminate this problem of overall staining in brain slices and not in sham operated group?