Budding and vesiculation are prominent shape transformations of fluid lipid-bilayer vesicles. We discuss these transitions within the context of a curvature model which contains two types of bending energy. In addition to the usual local curvature elasticity κ, we include the effect of a relative areal stretching of the two monolayers. This area-difference elasticity leads to an effective nonlocal curvature energy characterized by another parameter κ¯. We argue that the two contributions to the curvature energy are typically comparable in magnitude. The model interpolates smoothly between the spontaneous-curvature model (κ¯=0) and the bilayer-couple model (κ¯→∞), discussed previously in the literature. Conceptually, this model is not new; however, neither its consequences nor its relation to experiment has previously been explored in detail. In particular, budding is discontinuous (first order) for small κ¯ but changes via a tricritical point to continuous (second order) for large κ¯. The order of the budding transition depends on both the ratio κ¯/κ (which is a material parameter) and the initial area difference between the inner and outer monolayers (which can be modified by appropriate treatment of the vesicle). Estimates suggest that, under typical laboratory conditions, the budding process should be discontinuous, in apparent disagreement with some recent experiments. Possible reasons for this discrepancy are discussed. We propose, in particular, that hysteretic effects are important and that the observed behavior may reflect a spinodal instability.

The plasma membrane of most animal cells conforms to the cytoskeleton and only occasionally separates to form blebs. Previous studies indicated that many weak interactions between cytoskeleton and the lipid bilayer kept the surfaces together to counteract the normal outward pressure of cytoplasm. Either the loss of adhesion strength or the formation of gaps in the cytoskeleton enables the pressure to form blebs. Membrane-associated cytoskeleton proteins, such as spectrin and filamin, can control the movement and aggregation of membrane proteins and lipids, e.g., phosphoinositol phospholipids (PIPs), as well as blebbing. At the same time, lipids (particularly PIPs) and membrane proteins affect cytoskeleton and signaling dynamics. We consider here the roles of the major phosphatidylinositol-4,5-diphosphate (PIP2) binding protein, MARCKS, and PIP2 levels in controlling cytoskeleton dynamics. Further understanding of dynamics will provide important clues about how membrane-cytoskeleton adhesion rapidly adjusts to cytoskeleton and membrane dynamics.