Localization of immunoreactive, dystonia-associated protein torsinA near the Golgi apparatus of cultured rodent astrocytes

An in-frame deletion of a single glutamic acid codon in the TOR1A gene causes the neurological disorder DYT1 dystonia, but the cellular pathophysiology of this disorder remains elusive. A current model postulates that the wild-type (WT) torsinA protein is mainly localized to the endoplasmic reticulum (ER), but that the mutant form ({Delta}E-torsinA) is diverted to the nuclear envelope and cytoplasmic inclusion bodies. This mis-localization has been observed by overexpressing the proteins in neuronal and non-neuronal cells. However, it is not clear whether this model is valid for the astrocytic glial cells that support and modify neuronal functions. Here we report, using rodent astrocytes in primary culture, that the overexpressed torsinA proteins were distributed as predicted by the mis-localization model. We also found by immunocytochemistry that the cultured astrocytes express torsinA endogenously. Most of the signals from endogenous protein, whether the WT or {Delta}E form, were localized near a cis-Golgi marker GM130. Such localization of endogenous proteins was found in glial cells from several sources: the hippocampus of WT rats, the hippocampus and striatum of WT mice, and the hippocampus and striatum of {Delta}E-torsinA knock-in mice, a model of DYT1 dystonia. These results show that the mis-localization model is applicable to overexpressed torsinA proteins, but not applicable to those expressed at endogenous levels, at least in cultured rodent astrocytes. These discrepancies in the distribution of overexpressed versus endogenous torsinA proteins highlight the potential need for caution in interpreting the results of overexpression studies.