Proteochemometric Modeling Identifies Chemically Diverse Norepinephrine Transporter Inhibitors.

Area of Science:

• Computational chemistry and drug discovery
• Molecular pharmacology and transporter biology

Background:

• Solute carriers (SLCs) are crucial in disease but underexplored, except for the norepinephrine transporter (NET/SLC6A2).
• Existing NET ligands lack chemical diversity, hindering the discovery of novel inhibitors.

Purpose of the Study:

• To develop a computational screening pipeline for identifying novel, chemically diverse NET inhibitors.
• To expand the chemical space for modeling NET ligands by incorporating data from related SLC proteins.

Main Methods:

• A data-driven approach selected related proteins to enhance proteochemometric models for NET.
• Optimized proteochemometric models were built using stepwise feature selection and stacked machine learning.
• The model screened 600 million virtual compounds, clustering results to yield 46 diverse candidates for synthesis and testing.

Main Results:

• Thirty-two candidate compounds were synthesized and tested for NET inhibition.
• Five compounds demonstrated sub-micromolar inhibitory potency against NET, achieving a 16% hit rate.
• The identified hits represent promising starting points for further experimental research.

Conclusions:

• A novel computational strategy successfully diversified chemical space to identify new NET inhibitors.
• This study represents the first instance of selecting related targets based on sequence similarity for ligand discovery.
• The findings highlight the potential of data-driven approaches in uncovering novel therapeutic agents.