The image illustrates predictions of our model for inter-leaflet coupling in a phase-separated lipid bilayer. There is intense debate about the role of “lipid rafts” (nanometer-sized domains) in cell membranes, and how closely they are related to bona fide thermodynamic phases (e.g., Wilson et al., Biophys. J. April 7, 2015). In any case, phase behavior in model systems contributes an understanding of generic interactions among lipids and/or proteins, which is essential to inferring their possible cellular roles. Furthermore, understanding this behavior could allow smarter design of artificial membranes. Our work is associated with the EPSRC (Engineering and Physical Sciences Research Council) CAPiTALS Program, which is interested in technological uses of membranes grounded in an understanding of the physics governing their curvature, asymmetry, and patterning.
Because bilayers contain two leaflets, it is essential to consider how phase-separated domains register (align) across the leaflets. By incorporating hydrophobic length mismatch in a lattice model, our theory shows how apparently conflicting observations of registration and anti-registration may be reconciled: anti-registration can be kinetically favored and can occur first in the phase-separation process, but is typically metastable. In analogy to a large body of work in the colloid and metallurgy literatures, the model predicts fascinatingly complex kinetics driven by metastable states.
Rafts are implicated in such diverse processes as signal transduction, trafficking, and virus entry. Whether and how they couple across leaflets is a crucial aspect of their function; for example, such coupling could influence co-localization of signaling proteins anchored in opposite leaflets. Do proteins and lipids of longer than average hydrophobic length in one leaflet align with similar species in the other, maintaining similarity of structure across the bilayer? Or, do they align with shorter than average species, to maintain a more uniform bilayer thickness profile? Our work suggests that kinetics in model membranes could be intimately connected to these simple, very important questions.
Our theoretical model began as an idealized lattice simulation. Detailed simulation studies will explore the theory’s predictions, and guide future molecular simulations or experiments on the thermodynamics and kinetics of domain registration both in vitro and in vivo. The image was created from simulation snapshots that exemplify the unusual kinetics in which registered domains nucleate from a metastable, anti-registered background pattern. The competing arrangements follow a color scheme inspired by a popular British candy, hence the title Lipid Allsorts. The final image was composed in Inkscape. We wish to especially thank Alexander Stukowski for the excellent OVITO package, which was used to render the simulation.
Further information on our research can be found at
–John Williamson and Peter Olmsted