Precision Disease Modeling Using Patient Derived Neurons
Induced pluripotent stem cells (iPSC) technology enables the derivation of human neurons with the exact genetic background from any patient without the need to identify the mutations that cause the symptoms. This is of extreme importance when dealing with psychiatric and neurodegenerative disease where the precise genetic origins are often unknown. I will demonstrate how this method, along with electrophysiology (whole cell patch clamp) and computational modeling, allowed for an in-depth study of mechanisms underlying neuropathology in bipolar disorder and Parkinson’s disease. I will show that bipolar disorder hippocampal neurons exhibit a hyperexcitability phenotype along with a physiological instability that are caused by an altered basal state of reduced sodium currents and increased fast potassium currents, both in experiment and in a numerical simulation. Further, machine learning algorithms that utilize different biomarkers that I measured, allowed a prediction of the patient’s responsiveness to lithium with a low error rate and within days, saving valuable time of finding a correct treatment. I will show that in Parkinson’s disease dopaminergic neurons from patients with six different Parkinson’s causing mutations as well as from idiopathic Parkinson’s patients (where the cause is unknown) all exhibited a common phenotype of reduced synaptic activity, suggesting that reduced synaptic activity may be the earliest phenotype that is common in Parkinson’s disease, and this can now be used as an input for prediction algorithms of disease onset in individuals at risk.