Large-Scale G Protein-Coupled Olfactory Receptor-Ligand Pairing

Area of Science:

• Biochemistry and Molecular Biology
• Computational Chemistry
• Neuroscience

Background:

• G protein-coupled receptors (GPCRs) exhibit diverse functions despite conserved structures.
• Olfactory receptors (ORs), the largest GPCR family, display varied responses to different ligands.
• Understanding how OR amino acid sequences dictate ligand specificity is crucial.

Purpose of the Study:

• To investigate the encoding of diversified ligand responses within OR amino acid sequences.
• To establish a predictive model for OR-odorant interactions.
• To accelerate the deorphanization of olfactory receptors.

Main Methods:

• Development of a proteochemometric (PCM) model integrating OR sequence data and ligand physicochemical properties.
• Utilizing supervised machine learning, specifically Random Forest (RF).
• Employing site-directed mutagenesis, in vitro functional assays, and molecular simulations for model construction and validation.

Main Results:

• Ligand selectivity in ORs is primarily determined by residues within 8 Å of the orthosteric pocket.
• The PCM-RF model achieved a 58% prediction hit rate for 111 ORs and 7 odorants.
• Discovered 64 novel OR-odorant pairs and deorphanized 25 ORs, with a 56% deorphanization rate for the best model.

Conclusions:

• The PCM-RF approach effectively predicts OR-odorant interactions.
• This method significantly advances OR-odorant mapping and the deorphanization process.
• The study provides insights into the sequence-function relationships governing olfactory receptor specificity.