The structural role of cholesterol in cell membranes: from condensed bilayers to lipid rafts.

CONSPECTUS: Defining the two-dimensional structure of cell membranes represents one of the most daunting challenges currently facing chemists, biochemists, and biophysicists. In particular, the time-averaged lateral organization of the lipids and proteins that make up these natural enclosures has yet to be established. As the classic Singer-Nicolson model of cell membranes has evolved over the past 40 years, special attention has focused on the structural role played by cholesterol, a key component that represents ca. 30% of the total lipids that are present. Despite extensive studies with model membranes, two fundamental issues have remained a mystery: (i) the mechanism by which cholesterol condenses low-melting lipids by uncoiling their acyl chains and (ii) the thermodynamics of the interaction between cholesterol and high- and low-melting lipids. The latter bears directly on one of the most popular notions in modern cell biology, that is, the lipid raft hypothesis, whereby cholesterol is thought to combine with high-melting lipids to form "lipid rafts" that float in a "sea" of low-melting lipids. In this Account, we first describe a chemical approach that we have developed in our laboratories that has allowed us to quantify the interactions between exchangeable mimics of cholesterol and low- and high-melting lipids in model membranes. In essence, this "nearest-neighbor recognition" (NNR) method involves the synthesis of dimeric forms of these lipids that contain a disulfide moiety as a linker. By means of thiolate-disulfide interchange reactions, equilibrium mixtures of dimers are then formed. These exchange reactions are initiated either by adding dithiothreitol to a liposomal dispersion to generate a small amount of thiol monomer or by including a small amount of thiol monomer in the liposomes at pH 5.0 and then raising the pH to 7.4. We then show how such NNR measurements have allowed us to distinguish between two very different mechanisms that have been proposed for cholesterol's condensing effect: (i) an umbrella mechanism in which the acyl chains and cholesterol become more tightly packed as cholesterol content increases because they share limited space under phospholipid headgroups and (ii) a template mechanism whereby cholesterol functions as a planar hydrophobic template at the membrane surface, thereby maximizing hydrophobic interactions and the hydrophobic effect. Specifically, our NNR experiments rule out the umbrella mechanism and provide strong support for the template mechanism. Similar NNR measurements have also allowed us to address the question of whether the interactions between low-melting kinked phospholipids and cholesterol can play a significant role in the formation of lipid rafts. Specifically, these NNR measurements have led to our discovery of a new physical principle in the lipids and membranes area that must be operating in biological membranes, that is, a "push-pull" mechanism, whereby cholesterol is pushed away from low-melting phospholipids and pulled toward high-melting lipids. Thus, to the extent that lipid rafts play a role in the functioning of cell membranes, low-melting phospholipids must be active participants.