Novel small molecules therapeutics for FTD/ALS reduce TDP43 related pathology

Abstract

AbstractBackgroundTDP43 aggregation and mis‐localization is a hallmark of sporadic Amyotrophic Lateral Sclerosis (ALS) and TDP‐related Frontotemporal dementia (FTD‐TDP). Hyperphosphorylated and ubiquitinated cytosolic TDP43 aggregates have been identified in post‐mortem tissue from disease patients. However, no therapies targeting TDP43 are in the clinic.MethodUsing our novel machine learning platform, we identified compounds that can disaggregate toxic oligomers of amyloid like proteins including TDP43.ResultWe show that a TDP43 peptide can spontaneously form trimers, tetramers, hexamers and octamers in vitro using high resolution ion‐mobility mass spectrometry. Two of our compounds reduce these presumed toxic oligomers and restore physiologically relevant monomers and dimers. Furthermore, our compounds dose‐dependently reduce cytoplasmic aggregates in a cellular TDP43 aggregation model. These compounds show low plasma and brain homogenate binding, high metabolic stability, are orally bioavailable and are highly blood‐brain barrier permeable. Finally, our compounds were tested in a doxycycline‐repressible iTDP43A315T mouse model of ALS. Daily oral dosing of the compounds for approximately two months did not cause any toxicity in the mice at two doses. Both drugs were detectable at high concentrations in the brains of the mice at the end of the experiment. Importantly, drug treatments showed improvement in a motor behavior test and alleviated a disinhibition phenotype in an elevated plus maze test in the diseased mice. Further analysis of target engagement and other downstream efficacy markers is ongoing.ConclusionAll together, these in vitro cell‐based and preliminary in vivo mouse model data show that the Acelot machine learning platform can identify small molecules that have therapeutic potential for ALS, FTD and other amyloid‐associated diseases.