New vistas in GPCR 3D structure prediction

Area of Science:

• Biochemistry
• Structural Biology
• Pharmacology

Background:

• Human G-protein coupled receptors (hGPCRs) are crucial drug targets, with over 50% of approved drugs acting on them.
• Accurate 3D structure prediction of hGPCRs is vital for drug discovery but hindered by limited resolved structures (only four X-ray structures available).
• Sequence identity between hGPCRs and potential templates is often below 30%, complicating traditional sequence alignment methods.

Purpose of the Study:

• To optimize the use of available crystal structures for molecular modeling of hGPCRs.
• To develop improved methods for predicting the 3D structures of hGPCRs, facilitating drug discovery.
• To identify suitable template regions within transmembrane domains for homology modeling of GPCRs.

Main Methods:

• Analysis of a large database of human G-protein coupled receptors (Family A).
• Identification and utilization of specific trans-membrane domain regions from reference receptor helices as modeling templates.
• Comparative modeling using crystal structures of bovine rhodopsin (1F88) and human A2A adenosine receptor (3EML) as templates.
• Refinement of modeled residues using alternative techniques.

Main Results:

• A novel approach is proposed: using specific trans-membrane helix segments from known GPCR structures as templates for modeling other GPCRs.
• The proposed method was tested by modeling the human beta2-adrenergic receptor.
• The generated models showed improved quality compared to those produced by standard Modeller software.

Conclusions:

• Specific parts of trans-membrane domains can serve as effective templates for GPCR molecular modeling.
• This strategy enhances the accuracy of predicted GPCR structures, aiding drug discovery and development.
• The findings offer a more efficient pathway for modeling GPCRs with limited structural data.