Neuroprotection is a fundamental preventive alternative for Parkinson’s disease (PD) therapeutics. Dopaminergic replacement has been used for PD treatment with positive effects over motor symptomatology but without effect over disease progression and early-treatment perspective. Epidemiological studies have shown that nicotine consumption decreases PD prevalence through the activation of neuroprotective mechanisms over dopaminergic neurons. Nicotine neuroprotection has been associated with the overstimulation of PI3K/Akt signaling pathway (SP), induced by nicotinic acetylcholine receptors (e.g a7 nAChR), leading to Bcl-2 over-expression. Due to nicotine toxicity and dependency, analogs with less secondary effects and similar neuroprotective activity are relevant for PD pharmacology. In this work a computational strategy integrating structural bioinformatics, SP manual reconstruction, and machine learning was performed to predict the neuroprotective activity of a series of 8 novel nicotine analogs over PI3K/Akt behavior. We made a molecular docking between analogs and a7 nAChR receptor using a conformational analysis and performed a physicochemical characterization of the analogs. In order to elucidate the relationship between a7 nAChR receptor and Bcl-2 activation a Markov Chain Monte-Carlo transition matrix approach was constructed across PI3K/Akt SP. To develop a machine learning model, random training datasets were generated from a manually curated neuroprotective database of physicochemical properties using dimensional reduction. Contrasting 4 methods of prediction with each dimensional reduction, we developed an artificial neural network model using the synthetic datasets as training. Our model predicted the neuroprotective capacity of seven nicotine analogs suggesting them as putative PD neuroprotective molecules. Hereby we present a new computational strategy in which a predictive machine learning model is used to assess the neuroprotective activity of ligands based on SP. Our new method is applicable to a variety of diseases based in SP and can be coupled with drug discovery methodologies to improve pharmacological research in neurodegenerative diseases.

Novel computational strategy for neuroprotection prediction: Identification of novel nicotine-analogs as potential Parkinson therapeutic agents

2019 Fall National Meeting

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.