Effect of Membrane Lipid Packing on Stable Binding of the ALPS Peptide

• Published in: Journal of chemical theory and computation 2019-02, Vol.15 (2), p.1418-1429
• Main Authors: Wildermuth, Kyle D, Monje-Galvan, Viviana, Warburton, Linnea M, Klauda, Jeffery B
• Format: Article
• Language: English
• Subjects: Computer simulation; Curvature; Defects; Image processing; Lipids; Molecular dynamics; Peptides; Physical properties; Proteins; Yeast
• ISSN: 1549-9618; 1549-9626
• DOI: 10.1021/acs.jctc.8b00945

The amphipathic lipid packing sensor (ALPS) motif, originally discovered on the ArfGAP1 membrane-binding protein, binds to pre-existing large packing defects in a membrane (spontaneous or due to membrane curvature), though a more precise relationship between the ALPS peptide and packing defect characteristics of a membrane remains unclear. We developed an image processing technique for identifying packing defects to quantify the relationship between the ALPS peptide of the Osh4 protein in yeast and packing defects on a membrane model using molecular dynamics simulations. We used the highly mobile membrane mimetic (HMMM) model to create very large packing defects and expedite the binding time scale. Most prominently, we show that the probability of the ALPS peptide moving toward the membrane increases when it is near a large packing defect. Deviations from this trend exist for very large packing defects (≳115 Å2), which we propose is due to an overwhelming hydrophobic effect and a reduced electrostatic effect when large portions of the nonpolar core are exposed and the peptide is oriented unfavorably. Furthermore, we compared our HMMM results to similar simulations using all-atom lipid membranes. The binding time scales of the ALPS peptide can be reduced by roughly 1 order of magnitude when HMMM is used, while still maintaining many of the important physical characteristics of the binding process observed when using an all-atom lipid membrane.