Wilson, K. J.; Nguyen, H. Q.; Gervay-Hague, J.; Keller, S. L.

Despite longstanding excitement and progress toward understanding liquid-liquid phase separation in natural and artificial membranes, fundamental questions have persisted about which molecules are required for this phenomenon. Except in extraordinary circumstances, the smallest number of components that has produced large-scale, liquid-liquid phase separation in bilayers has stubbornly remained at three: a sterol, a phospholipid with ordered chains, and a phospholipid with disordered chains. This requirement of three components is puzzling for two reasons: (1) the Gibbs Phase Rule states that only two components are necessary, and (2) only two components are required for liquid-liquid phase separation in lipid monolayers, which resemble half of a bilayer. Inspired by reports that sterols interact closely with lipids with ordered chains, we tested whether phase separation would occur in bilayers in which a sterol and lipid were replaced by a single, joined sterol-lipid. By evaluating a panel of sterol-lipids, we discovered a minimal bilayer of only two components (PChemsPC and diPhyPC) that demixes into micron-scale, liquid phases. In this system, the sterol-lipid behaves as a 3:1 ratio of cholesterol to phospholipid. Our system gives the computation and theory community a two-component membrane that maps directly onto simplified theories and that can be used to validate simulation force fields. It suggests a new role for sterol-lipids in nature, and it gives experimental communities a membrane in which tie-lines (and, therefore, the lipid composition of each phase) are easily determined and will be consistent across multiple laboratories.