The Role of TorsinA in Brain Lipid Metabolism

Mutations in TOR1A/TorsinA cause poorly explained and symptomatically complex neurological diseases. Dominant inheritance of a three-base pair deletion (Δgag) that removes a single amino acid (+/ΔE) underlies a partially-penetrant isolated dystonia that occurs despite a structurally normal brain. In contrast, recessive TorsinA disease caused by several biallelic combinations of mutations is characterized by a wide variety of severe neurological defects, can be lethal, and is associated with neurodegeneration. The aim of this PhD project was to examine whether common or distinct molecular mechanisms account for such different diseases. TorsinA has been linked to phosphatidic acid (PA) lipid metabolism in Drosophila melanogaster. We started by evaluating whether the lipid regulating function of Torsin is conserved in mammals. For that we evaluated the role of PA phosphatase (PAP) enzymes in TOR1A diseases using induced pluripotent stem cells (iPSC)-derived neurons from patients, and mouse models of recessive Tor1a disease. We found that Lipin PAP enzyme activity was abnormally elevated in human DYT-TOR1A dystonia patient cells and in the brains of three Tor1a mouse models (-/- ; ΔE/ΔE; cKO/ΔE). We assessed the pathogenic role of excess Lipin activity in the neurological dysfunction of Tor1a disease mouse models by interbreeding these with Lpin1 knock-out mice. Genetic reduction of Lpin1 improved the survival of recessive Tor1a disease-model mice, alongside suppressing neurodegeneration, motor dysfunction, and nuclear membrane pathology. From these data we conclude that recessive TOR1A disease mutations cause abnormal PA metabolism. Next, we studied whether excess Lipin activity also occurred when there is a mild loss of TorsinA function, as is the case in the +/ΔE situation. While animal models of the severe recessive disease displayed many behavioural and cellular phenotypes that phenocopy the human disease, models of the dominant DYT-TOR1A dystonia have failed to do so, at least in a consistent and replicable manner. Moreover, the lack of phenotypes makes it extremely difficult to model and study the reduced penetrance of the disease. We showed that +/ΔE acts with reduced penetrance in inbred mice. Lipid metabolism was abnormal in the brains of ~30% of +/ΔE embryos, and this defined which animals developed neurological symptoms later in life. The variability was not genetically encoded. Instead, abnormal lipid metabolism primarily emerged in female +/ΔE embryos and was influenced by the maternal diet. The metabolic defect resolved during post-natal development but was still sufficient to cause ~30% of female juvenile +/ΔE mice to develop dystonia-like symptoms. Furthermore, symptoms were prevented by genetically suppressing the abnormal PA metabolism. We conclude that +/ΔE embryos poorly buffer metabolic stress, and in turn some experience a period of abnormal lipid metabolism that hardwires the brain for dystonia in later life. Additionally, female sex strongly predisposes +/ΔE embryos towards metabolic defects, illustrating the profound effect that sex can have on the disease phenotypes. Therefore, we conclude that, despite being discrete pathologies, TorsinA diseases are caused by varying degrees of TorsinA loss and subsequent elevation of Lipin activity. Our data also demonstrate that hyperactivity of a single lipid metabolic reaction is sufficient to cause complex neurological dysfunction and redefine TorsinA diseases as inborn errors of metabolism for which Lipin inhibition is a feasible druggable target.status: publishe