How does the DYT1 dystonia torsin A mutation affect neuronal development?

Dystonia is the third most common movement disorder with a prevalence of 48.5 cases per 100000. Despite improvements in the treatment of dystonic symptoms, the disease remains incurable and its aetiology is poorly defined. The first dystonia gene was identified in 1997 when heterozygosity for a three base-pair mutation in TOR1A/torsinA was uncovered as the cause of juvenile-onset DYT1/DYT-TOR1A dystonia. Recently, TOR1A mutations were also shown to cause recessive neurological disease. This is more severe than dominant DYT-TOR1A dystonia, and infants that are born with homozygous TOR1A mutations suffer from broad neurological dysfunction and symptoms that originate in utero. The fact that mutations in TOR1A/torsinA lead to a spectrum of neurodevelopmental disease highlights the importance of understanding TOR1A function during neural development. This information is essential to design therapies against TOR1A/torsinA disease.

TorsinA is a AAA+ ATPase family member. These proteins use the energy of ATP hydrolysis to drive structural changes in substrates. Several lines of evidence strongly suggest that torsinA operates in the nuclear envelope (NE) during neural development. However, the specific nature of this activity remains elusive. This is due to the fact that the role(s) of torsinA have been mostly studied in vitro and in non-neuronal cell types. Consequently, this research uncovered a variety of cellular functions for torsinA, but it is unclear which are active in neural development and/or whether there are specific cell types or developmental stages that require torsinA activity. Notably, several studies have reported a relationship between torsinA and the LINC complex that couples the nucleoskeleton with the cytoskeleton. The LINC complex is formed by Sun proteins and Nesprins, which are transmembrane proteins in the inner and the outer nuclear membranes, respectively. The nature of the relationship between torsinA and the LINC complex seems to be cell-context dependent. Some cell types appear to use torsinA to remove NE LINC complexes, while in other cases torsinA appears to add LINC complexes to the NE. Moreover, it is unclear whether torsinA interacts with Sun proteins and/or with Nesprins. Resolving which torsinA ‘functions’ are important in neurodevelopment has clear importance for explaining the cascade between torsinA mutation and disease, and it is the main focus of this thesis.

In this work, we utilized a Tor1a/torsinA knock-out mouse to investigate the role(s) of torsinA in neural development. We identified a previously unrecognized defect whereby Tor1a deletion leads to semi-penetrant defects in brain morphogenesis. This suggests that the full spectrum of human TOR1A disease also remains undiscovered and may extend to include patients with grossly abnormal brain development. We also identified excess levels of NE LINC complex levels in the neural progenitor cells of these Tor1a-/- mice. In turn, we strongly associated this abundance of LINC complexes with increased radial glial cell progenitor proliferation and breakdown of the cytoarchitecture of the developing brain - ultimately resulting in grossly abnormal brain development. This includes that we genetically demonstrated - via Sun2 deletion - a causal relationship between excess LINC complexes and abnormal brain morphogenesis in Tor1a-/- embryos. Furthermore, the concept that torsinA negatively regulates LINC complexes in neural development was further supported by characterization of Sun1/Sun2 double knock-out mice. This identified that Sun1/Sun2 double knock-out developing neurons display membrane defects that resemble those we previously associated with excess torsinA activity in vitro. This identified for the first time a role of Sun proteins in nuclear membrane biogenesis and indicates that torsinA regulates cellular lipid metabolism during CNS development. We also show that, as opposed to radial glial cells, developing neurons exchange LINC complex components in a torsinA-independent manner. We also investigated whether LINC complex excess underlie the development of NE membrane abnormalities in developing neurons. However, we found that torsinA NE membrane abnormalities form in absence of LINC complexes. Thus, these data show that torsinA does not regulate LINC complex levels in the NE of developing neurons.

Overall, this study significantly extends in vitro work by defining the neurodevelopmentally-relevant nature of the torsinA-LINC complex interaction. We demonstrate that torsinA is necessary in early neural development in radial glial cells to downregulate the levels of LINC complex in the NE. Failure of this exchange leads to excess LINC complexes that in turn impairs radial glial biology, ultimately leading to brain developmental defects. Moreover, we demonstrate that the exchange of LINC complexes occurs independently of torsinA in developing neurons. Considered together, these data implicate the LINC complex and the dysfunction of radial glial cells in TOR1A-pathology. Moreover, it predicts that the full spectrum of TOR1A recessive disease will include a semi-penetrant pathology where structural brain defects arise downstream of radial glial cell dysfunction.