Structural comparison and classification of alpha-helical transmembrane domains based on helix interaction patterns

Area of Science:

• Structural biology
• Biochemistry
• Bioinformatics

Background:

• Membrane protein structural classification is limited by the scarcity of 3D structures.
• Technological advances are increasing the number of known membrane protein folds.

Purpose of the Study:

• To develop a membrane protein-specific classification system for alpha-helical proteins.
• To explore the structural universe of membrane proteins using helix architectures.

Main Methods:

• Representing each membrane protein by a helix interaction graph of transmembrane helices and residue contacts.
• Clustering proteins based on helix interaction graph similarity using the HISS score.
• Comparing the new classification with existing databases like SCOP and CATH.

Main Results:

• The HISS score effectively captures conserved transmembrane helix bundle architectures by focusing on helix interactions.
• The helix interaction-based classification aligns well with conventional structural databases (SCOP, CATH).
• Classification of known structures revealed 20 recurrent helix architectures and 15 singletons, showing variability and conserved patterns.

Conclusions:

• Helix interactions are key determinants of alpha-helical membrane protein folds.
• The new classification system effectively categorizes membrane proteins based on conserved helix architectures.
• Despite sequence diversity, common helix interaction patterns are conserved across membrane proteins.