Donepezil as a new therapeutic potential in KCNQ2- and KCNQ3-related autism.

Introduction: The genes encode the voltage-gated K channel underlying the neuronal M-current, regulating neuronal excitability. Loss-of-function (LoF) variants cause neonatal epilepsy, treatable with the M-current-opener retigabine, which is no longer marketed due to side effects. Gain-of-function (GoF) variants cause developmental encephalopathy and autism that could be amenable to M-current, but such therapies are not clinically available. In this translational project, we investigated whether donepezil, a cholinergic drug used in Alzheimer's, suppresses M currents and improves cognitive symptoms in patients with GoF variants.

Methods: (1) The effect of 1 μM donepezil on the amplitude of the M-current was measured in excitatory and inhibitory neurons of mouse primary cultured hippocampal cells. M-current was measured using the standard deactivation protocol (holding at 0 mV and deactivation at -60 mV) in the voltage-clamp configuration of the whole-cell patch clamp technique. The impact of donepezil was also examined on the spontaneous firing activity of hippocampal neurons in the current-clamp configuration. (2) Four children with autism, aged 2.5-8 years, with the following GoF variants were enrolled: and . Patients were treated off-label with donepezil 2.5-5 mg/d for 12 months and assessed with: clinical Global Impression of Change (CGI-c), Childhood Autism Rating Scale 2 (CARS-2), Adaptive Behavior Assessment System-II (ABAS-II), and Child Development Inventory (CDI).

Results: (1) Application of donepezil for at least 6 min produced a significant inhibition of the M-current with an IC50 of 0.4 μM. At 1 μM, donepezil reduced by 67% the M-current density of excitatory neurons (2.4 ± 0.46 vs. 0.89 ± 0.15 pA/pF,  < 0.05). In inhibitory neurons, application of 1 μM donepezil produced a lesser inhibition of 59% of the M-current density (1.39 ± 0.43 vs. 0.57 ± 0.21,  > 0.05). Donepezil (1 μM) potently increased by 2.6-fold the spontaneous firing frequency, which was prevented by the muscarinic receptor antagonist atropine (10 μM). (2) The CARS-2 decreased by 3.8 ± 4.9 points ( > 0.05), but in two patients with variants, the improvement was over the 4.5 clinically relevant threshold. The global clinical change was also clinically significant in these patients (CGI-c = 1). The CDI increased by 65% ( < 0.05), while the ABAS-II remained unchanged.

Discussion: Donepezil should be repurposed as a novel alternative treatment for GoF variants in encephalopathy.