A lipid-phase separation model of low-temperature damage to biological membranes.

An hypothesis is proposed to explain the damage caused to biological membranes exposed to low temperatures. The thesis rests on the general observation that the lipid components of most membranes are heterogeneous and undergo phase transitions from gel-phase lamellae to liquid-crystalline lamellae and some to a non-lamellar, hexagonal-II phase over a wide range of temperatures. As a consequence of these phase transitions the lateral distribution of the lipids characteristic of the growth temperature is disturbed and redistribution takes place on the basis of the temperature at which phase transitions occur. When membranes are cooled, first the non-lamellar forming lipids pass through a transition to a fluid lamellar phase and are miscible with bilayer-forming lipids into which they diffuse. On further cooling the high-melting-point lipids begin to crystallize and separate into a lamellar gel phase, in the process excluding the low-melting point lipids and intrinsic proteins. The lipids in these remaining regions form a gel phase at the lowest temperature. It is suggested that, because the non-lamellar lipids tend to undergo a liquid-crystalline to gel-phase transition at higher temperatures than lamellar-forming lipids, these will tend to phase separate into a gel phase domain rich in these lipids. Damage results when the membrane is reheated, whereupon the hexagonal-II-forming lipids give rise to non-lamellar structures. These probably take the form of inverted micelles sandwiched within the lipid bilayer and they completely destroy the permeability barrier properties of the membrane. The model is consistent with the phase behavior of membrane lipids and the action of cryoprotective agents in modifying lipid phase properties.