Foxg1: intracellular localization and possible pathogenetic mechanisms in Rett syndrome

The Forkhead box G1 (FoxG1) is a transcription factor essential for the forebrain development and involved in pathogenesis of Rett syndrome. Notwithstanding the importance of this protein, little is known about the modalities by which it exerts its cellular functions. In this thesis I investigated, in cell culture and in animal model, the molecular mechanisms of Foxg1 action and the pathophysiological consequences of Foxg1 haploinsufficiency. Using fluorescence recovery after photobleaching strategy, I investigated the chromatin binding dynamics of Foxg1 in its wild-type form or carrying some Rett syndrome-causing mutations. The experiments show that GFP-Foxg1 mobile molecules have a chromatin affinity that decreases in truncated mutants with the extension of the protein deletion. Conversely, the immobile fraction does not decrease gradually with the extension of protein deletion but is significantly higher for a truncation associated with a severe phenotype, suggesting a possible dominant negative function of some RTT mutants. As already present in literature, GFP-Foxg1 localizes almost exclusively in the nucleus. However, no one reported data on Foxg1-GFP C-terminal fusion protein. Surprisingly, Foxg1 wt-GFP, in addition to the preserved nuclear localization, shows a clear mitochondrial localization. Further investigation led me to discover that Foxg1 undergoes a proteolytic cleavage and the resulting C-terminal peptide localizes to mitochondria, as evidenced by CFP-Foxg1 wt-YFP fusion protein as well as by other biochemical and immunohistochemical assays. Tripan Blue and proteinase K protection experiments in living cells show that nearly 80% of the protein reside in the mitochondrial matrix. Considering that Foxg1 does not have a classical N terminal mitochondrial targeting signal, I looked for Foxg1 amino acids able to drive a fused fluorescent tag into mitochondria and identified amino acids 277-302 as critical. A pull down experiment showed that a "mitochondrial" Foxg1 peptide interacts with some proteins involved in mitochondrial trafficking and function such as voltage dependent anion channel 1 and 2. Functionally, a C-terminal "mitochondrial" peptide of Foxg1 (aa 272-481) is able to enhance ATP production and mitochondrial membrane potential. Full length Foxg1 strikingly enhances mitochondrial potential and promotes cellular proliferation and mitochondrial fission; on the contrary, "mitochondrial" Foxg1 promotes mitochondrial fusion and an initial stage of cellular differentiation. In an animal model, I investigated the functional result of Foxg1 haploinsufficiency in mouse visual system. We chose visual system since it is known that Foxg1 is expressed in the developing retina and chiasm and the knockout mouse embryos show defects in retinal ganglion cells axonal navigation. It is also known that Foxg1 haploinsufficiency in mice results in subtle cortical defects. My data show that Foxg1+/Cre heterozygous mice have a profound impairment in visual acuity, despite a preserved retinal organization. Defects in visual cortical circuitry suggest that acuity loss may be due to altered cortical mechanisms. Taken together, these results suggest the following hypothesis: Foxg1 is a nuclear and mitochondrial protein that upon its internal processing localizes in the mitochondrial compartment and becomes a key coordination part of the machinery that brings about cell differentiation, replication and bioenergetics