Computational screening of natural products as tryptophan 2,3-dioxygenase inhibitors: Insights from CNN-based QSAR, molecular docking, ADMET, and molecular dynamics simulations.
Journal:
Computers in biology and medicine
PMID:
40233673
Abstract
Parkinson's disease (PD) is characterised by a complex array of motor, psychiatric, and gastrointestinal symptoms, many of which are linked to disruptions in neuroactive metabolites. Dysregulated activity of tryptophan 2,3-dioxygenase (TDO), a key enzyme in the kynurenine pathway (KP), has been implicated in these disturbances. TDO's regulation of tryptophan metabolism outside the central nervous system (CNS) plays a critical role in maintaining the balance between serotonin and kynurenine-derived metabolites, with its dysfunction contributing to the worsening of PD symptoms. Recent studies suggest that targeting TDO may help alleviate non-motor symptoms of PD, providing an alternative approach to conventional dopamine replacement therapies. In this study, a data-driven computational pipeline was employed to identify natural products as potential TDO inhibitors. Machine learning and convolutional neural network-based QSAR models were developed to predict TDO inhibitory activity. Molecular docking revealed strong binding affinities for several compounds, with docking scores ranging from -9.6 to -10.71 kcal/mol, surpassing that of tryptophan (-6.86 kcal/mol), and indicating favourable interactions. ADMET profiling assessed pharmacokinetic properties, confirming that the selected compounds could cross the blood-brain barrier (BBB), suggesting potential CNS activity. Molecular dynamics (MD) simulations provided further insight into the binding stability and dynamic behaviour of the top candidates within the TDO active site under physiological conditions. Notably, Peniciherquamide C maintained stronger and more stable interactions than the native substrate tryptophan throughout the simulation. MM/PBSA decomposition analysis highlighted the energetic contributions of van der Waals, electrostatic, and solvation forces, supporting the binding stability of key compounds. This integrated computational approach highlights the potential of natural products as TDO inhibitors, identifying promising leads that address PD symptoms beyond traditional dopamine-centric therapies. Nonetheless, experimental validation is necessary to confirm these findings.