Intrinsic curvature hypothesis for biomembrane lipid composition: a role for nonbilayer lipids.

A rationale is presented for the mix of "bilayer" and "nonbilayer" lipids, which occurs in biomembranes. A theory for the L alpha-HII phase transition and experimental tests of the theory are reviewed. It is suggested that the phase behavior is largely the result of a competition between the tendency for certain lipid monolayers to curl and the hydrocarbon packing strains that result. The tendency to curl is quantitatively given by the intrinsic radius of curvature, Ro, which minimizes the bending energy of a lipid monolayer. When bilayer (large Ro) and nonbilayer (small Ro) lipids are properly mixed, the resulting layer has a value of Ro that is at the critical edge of bilayer stability. In this case, bilayers may be destabilized by the protein-mediated introduction of hydrophobic molecules, such as dolichol. An x-ray diffraction investigation of the effect of dolichol on such a lipid mixture is described. This leads to the hypothesis that biomembranes homeostatically adjust their intrinsic curvatures to fall into an optimum range. Experimental strategies for testing the hypothesis are outlined.