Co-evolution analysis to predict protein-protein interactions within influenza virus envelope

Area of Science:

• Virology
• Molecular Biology
• Bioinformatics

Background:

• Integral membrane proteins hemagglutinin (HA), neuraminidase (NA), M2, and matrix protein M1 are essential for influenza A virus assembly.
• Protein-protein interactions at the interfaces of these viral proteins are critical for forming functional virions.

Purpose of the Study:

• To investigate the hypothesis that amino acid residues at protein interfaces co-evolve due to physical constraints.
• To predict co-evolving residue pairs within influenza A virus proteins HA, NA, M2, and M1.

Main Methods:

• Applied the EvFold program to analyze large sequence samplings from the Influenza Research Database.
• Focused analysis on the cytoplasmic tails of HA, NA, and M2, and the C-terminal domain of M1.
• Utilized Direct Information scores to identify correlated positions.

Main Results:

• Identified six pairs of correlated positions among HA, M2, and M1 proteins.
• Found one, two, and three correlated position pairs for HA-M2, HA-M1, and M2-M1 interactions, respectively.
• Observed no co-varying positions for NA-HA, NA-M1, and NA-M2 pairs, likely due to the high conservation of the NA cytoplasmic tail.
• Determined high evolutionary sustainability for observed amino acid patterns in correlated positions (sum of frequencies up to 0.99).

Conclusions:

• The study provides evidence for co-evolutionary constraints on interacting influenza A virus proteins.
• Predicted co-evolving residue pairs can inform the development of hypothetical models for viral protein interactions within the envelope.
• Findings contribute to understanding the molecular mechanisms underlying influenza virus assembly and evolution.