A comparative differential scanning calorimetry study of the effects of cholesterol and various oxysterols on the thermotropic phase behavior of dipalmitoylphosphatidylcholine bilayer membranes

Cholesterol (C) is an abundant and essential lipid component particularly in the plasma membranes of the cells of higher animals (Finegold, 1993, Nes and McKean, 1977, Yeagle, 1988) and is known to have many effects on the thermotropic phase behavior and organization of lipid bilayers in both model and biological membranes (Demel and De Kruyff, 1976, Finegold, 1993, McMullen and McElhaney, 1996, Nes and McKean, 1977, Vist and Davis, 1990, Yeagle, 1988). These effects include a broadening and eventual elimination of the cooperative gel/liquid-crystalline phase transition and the concomitant progressive replacement of the L_β and L_α phases by a state with an intermediate degree of organization (the L_o phase). This L_o phase is characterized by a higher phospholipid hydrocarbon chain ordering, a restricted rate of lateral diffusion, and a reduced area per molecule compared to the liquid-crystalline state which would exist at physiological temperatures in the absence of C. As well, the presence of C increases the thickness and mechanical strength of the phospholipid bilayer and reduces its permeability. In addition, the simultaneous presence of the phospholipid-rich L_α and sphingolipid- and C-rich L_o phases in lipid bilayers formed in certain unsaturated phospholipid/sphingomyelin/C ternary lipid mixtures has prompted some investigators to postulate the existence of specialized detergent-insoluble “lipid rafts” in animal cell membranes (Brown and London, 2000, Silvius, 2003, Simons and Ikonen, 2000), although this hypothesis remains controversial (Edidin, 2003, McMullen et al., 2004, Munro, 2003). Nevertheless, there is a great deal of evidence that the presence of C does modulate a number of different membrane functions, either directly or through its general effects on the structure, physical properties and possibly on the lateral organization of phospholipid bilayers, in both model and biological membranes (Dahl and Dahl, 1988, McElhaney, 1992a, McElhaney, 1992b, Yeagle, 1988).

The oxysterols are a heterogeneous group of C-derived compounds with one or sometimes more additional oxygen substitutions, most commonly hydroxyl, carbonyl or epoxy groups, either on the steroid ring or the alkyl side chain (Schroepfer, 2000) (Fig. 1 and Supplemental Fig. 1). The oxysterols can be formed by the autoxidation of C or by a number of enzyme-catalyzed reactions, which typically take place in microsomes or mitochondria and usually involve cytochrome P-450. Although normally present in cells and plasma at low levels, various oxysterols have been shown to have a diverse array of important regulatory functions, most prominently in the control of C biosynthesis, but also in sphingolipid metabolism, platelet aggregation, and apoptosis, among many other roles. Although many oxysterols have been shown to exert their major biological effects through binding to various specific protein receptors, some evidence has been presented that certain of these compounds may act at least in part by inserting into and modifying the physical properties of the lipid bilayers of biological membranes (Olkkonen and Hynynen, 2009, Olsen et al., 2012).

One common property of oxysterols is that the additional oxygen-containing groups present in these molecules makes them considerably more hydrophilic than C itself. Moreover, the presence of additional oxygenated functions can alter the 3-dimensional conformation of the oxysterol and particularly the distribution of polarity over the molecular surface, and thus its packing in phospholipid bilayers. For these reasons, oxysterols such as 7β-HC, 7-KC and 25-HC transfer between lipid bilayer model membranes at rate orders of magnitude faster than C itself (Morel et al., 1996, Theunissen et al., 1986, Vila et al., 2001). It has also been reported that the rate of oxysterol transfer from erythrocytes to plasma acceptors correlates with the distance between the 3β-hydroxyl group and the second oxygen function, such that these rates increase as the oxygen function moves away from the C3 position (Meaney et al., 2002). Other studies have shown that various oxysterols have distinct effects on the thermotropic phase behavior and order of lipid bilayer model membranes, depending on the nature and location of the oxygen substitution, with the oxysterols typically increasing order but less effectively than C itself (Massey and Pownall, 2005, Rooney et al., 1986, Theunissen et al., 1986, Verhagen et al., 1996, Wang et al., 2004). These differing effects on lipid model membranes are thought to be due at least in part to differences in the orientation and degree of penetration into the hydrophobic core of the lipid bilayer of the various oxysterols as compared with C. At least under low lateral pressures in lipid monolayer films, oxysterols with oxygen-containing groups on the sterol ring adopt a tilted orientation relative to the film surface, so that both oxygen functions can interact with the aqueous phase (Kauffman et al., 2000, Li et al., 2003, Massey and Pownall, 2005), although this does not seem to be the case at the higher surface pressures characteristic of phospholipid bilayers (Theunissen et al., 1986). As well, again under low lateral surface pressures, oxysterols with a hydroxyl group in the alkyl side chain may even adopt a horizontal orientation at the monolayer film surface, so that both hydroxyl groups can interact with water (Kauffman et al., 2000), although again this does not seem to be the case at more biologically relevant lateral pressures (Theunissen et al., 1986). Moreover, oxysterols modified in the sterol ring system exhibit a weaker monolayer condensing ability compared to C itself (Kauffman et al., 2000, Phillips et al., 2001, Rooney et al., 1986, Theunissen et al., 1986). Oxysterols also have a weaker ability to reduce the permeability of model (Gale et al., 2009, Holmes and Yoss, 1984, Theunissen et al., 1986) and biological membranes (Boissonneault and Heiniger, 1985) as compared to C, and 25-HC in particular can actually markedly increase passive permeability in these systems. Finally, oxysterols may either be less effective (7β-HC and 7-KC) or as effective (25-HC) as C itself in inducing the formation of L_o domains (lipid rafts) in ternary lipid mixtures (Beattie et al., 2005, Massey and Pownall, 2005, Wang et al., 2004).

DSC is a powerful and non-perturbing thermodynamic technique that can provide unique and useful information about the thermotropic phase behavior and organization of model and biological membranes (Lewis and McElhaney, 2011, Mannock et al., 2010b, McElhaney, 1982). However, to our knowledge, only two DSC studies of the effect of oxysterols on the thermotropic phase behavior of phospholipid bilayer membranes have been published to date. In one study, the only calorimetric data presented was a plot of the enthalpy of the main phase transition of DPPC versus the concentration of 7-KC, 7β-HC, 7α-HC and 25-HC up to 30 mol%, where it was reported that the rate of decrease in the enthalpy was less than for C and decreased in this order (Theunissen et al., 1986). In the other study, only two concentrations of oxysterols (about 10 and 20 mol%) were investigated, such that the comparative effects of C and the four oxysterols studied on the pretransition of DPPC could not be studied, since it was already abolished at the lowest sterol concentration tested (Egli et al., 1984), as was also the case for the first study as well. In addition, the solubility limits of these sterols could not be determined in either study since only relatively low sterol concentrations were examined. Also, in the other study, 7-KC, 7α-HC and 7β-HC incorporation were reported to produce comparable decreases in the enthalpy of the main phase transition of DPPC, with only 25-HC producing smaller decreases in enthalpy (Egli et al., 1984). We thus present here the results of a comparative high-sensitivity DSC study of mixtures of C and the eight most physiologically relevant oxysterols (Fig. 1 and Supplemental Fig. 1) with the well-studied DPPC bilayer, using an experimental protocol that permits the broad, poorly energetic phase transitions which occur at high sterol levels to be accurately monitored (McMullen et al., 1993, McMullen and McElhaney, 1995) and covering the sterol concentration range from 1 to 50 mol%. We also deconvolve the sharp and broad components of the main phase transition of DPPC, which are present at lower sterol levels, and determine the thermodynamic parameters associated with each component. In this way we hope to resolve some of the apparently discrepant results in the previous literature and to provide insight into the possible bases for the observed differences in the effects of C and a number of its oxysterol analogs on phospholipid bilayer membranes.