Structure-Kinetics Relationships of Opioids from Metadynamics and Machine Learning.

Area of Science:

• Pharmacology
• Computational Chemistry
• Drug Discovery

Background:

• Opioid overdose deaths, primarily from synthetic opioids like fentanyl, reached record highs in 2021.
• Naloxone, an FDA-approved opioid reversal agent, acts by competitive binding at the mu-opioid receptor (mOR).
• Understanding opioid residence time at the mOR is crucial for evaluating naloxone's efficacy.

Approach:

• Residence times of 15 fentanyl and 4 morphine analogs were estimated using molecular dynamics simulations (metadynamics).
• Simulation results were compared with experimental kinetic and binding data for opioid analogs.
• A machine learning (ML) model was developed to analyze the influence of fentanyl substituents on mOR interactions and dissociation kinetics.

Key Points:

• Microscopic simulations revealed common binding mechanisms and molecular determinants of dissociation kinetics for fentanyl analogs.
• The study provides insights into how structural modifications affect opioid residence times at the mOR.
• A novel ML approach was established to predict the kinetic impact of fentanyl's substituents.

Conclusions:

• This research enhances understanding of opioid-mOR interactions, critical for assessing naloxone effectiveness.
• The developed ML approach offers a generalizable method for predicting and tuning ligand residence times in drug discovery.
• These findings support the development of novel strategies to combat the opioid crisis and advance computer-aided drug design.