Discerning intersecting fusion-activation pathways in the Nipah virus using machine learning.

Area of Science:

• Virology
• Structural Biology
• Biophysics

Background:

• Nipah virus fusion with host cells is mediated by the viral glycoproteins G and F.
• Ephrin binding to the G protein's head domain triggers a conformational change, activating the F protein for fusion.

Purpose of the Study:

• To quantitatively analyze the conformational changes in the Nipah virus G protein's head domain upon ephrin binding.
• To understand the signal transduction pathway from the ephrin-binding site to the stalk domain of the G protein.

Main Methods:

• Molecular dynamics simulations were used to generate conformational ensembles of the G protein head domain in unbound and ephrin-bound states.
• A novel machine learning method was employed to quantify the differences between these conformational ensembles.
• Human ephrins B2 and B3, and a double mutant of B2, were used as ephrin ligands.

Main Results:

• A significant portion (approximately 25%) of the G protein head domain residues exhibit altered conformational density upon ephrin binding.
• This affected subspace includes residues critical for F protein activation, indicating a conserved signaling pathway.
• The spatial distribution of these residues supports a model where signal transduction occurs across the G protein head dimer interface.

Conclusions:

• Ephrin binding induces substantial conformational changes in the Nipah virus G protein head domain, crucial for initiating virus-host fusion.
• Signal transduction involves a large subspace of residues and likely occurs via the dimer interface.
• Protein dynamics and conformational changes, not solely backbone deviations, play a key role in viral glycoprotein function.