De novo identification of potent ingredients for proteasome activation in MT101-5 using an AI-driven approach.

Parkinson's disease (PD) is a neurodegenerative disorder characterized by the intracellular accumulation of α-synuclein (α-syn) protein within Lewy bodies, which are hallmark neuropathological lesions of α-synucleinopathies. Network analysis, incorporating prior knowledge such as biological pathways and gene expression data from patients with PD, was used to identify the pharmaceutically active ingredients of MT101-5 and elucidate their mechanisms of action related to proteasome activation in PD. To validate the artificial intelligence (AI) model's prediction that API-mediated α-syn aggregate clearance occurs through enhanced proteasome activity, we employed a reporter-based screening to assess proteasome function within the ubiquitin-proteasome system. Our findings demonstrated that diterpenes derived from Genkwae Flos in MT101-5 inhibited α-syn fibril formation by restoring proteasome activation through Nurr1 activity. Two diterpenoids mitigated behavioural deficits and dopaminergic neuron loss in mice subjected to subacute MPTP administration (30 mg/kg/day for 5 days). These results provide preliminary evidence supporting the therapeutic potential of the active ingredients in MT101-5 for the prevention or treatment of PD.