More information would be helpful. If you are labeling a neuronal culture, why would you expect labeling for astrocytes? Did you grow the culture on a bed of astrocytes? Are you controlling for astrocytes in the culture? What host animal is each antibody in? If you used the same host animal for your astrocyte and neuronal label and then used conjugated secondaries for that animal then both astrocytes and neurons would be marked/colocalizing with the same fluorescent marker. Hope that helps on the troubleshooting front!

Jennifer Fessler thank you for your response ! The cell line is progenitors that can differenciate in neurons and glial cell that's why i'm trying to label astrocytes. I simply induce differneciation by letting the cells without growth factors. GFAP antibodies are from rabbit and beta-3-tubulin from mouse so it is not the problem. I've seen an article that also has astrocytes labelled with B-III-tubulin and GFAP but I don't understand why neurons are GFAP+. At fisrt I thought that it was because of the myelin but even the cell body is marked.

Differentiating cell lines into either neurons or glia is not my specialty. However, cells that are not fully differentiated into either class may express unexpected proteins. On the IHC front, you may want to perform a control experiment on known cells with your conjugated secondary. Sometimes secondaries can have non-specific binding at concentrations that are too high or when they are not extremely clean. Nonspecific binding can create false labeling in the channel. For example, I have a rabbit 647 secondary that is nonspecifically binding to my 568 chicken. So, then I have false labeling in the 647 channel. Best of wishes!