Structural organization of lipid phase and protein-lipid interface in apolipoprotein-phospholipid recombinants:...

The complexes of individual human plasma apolipoproteins (apo) A-I, E and A-II with dipalmitoylphosphatidylcholine (DPPC) in the absence or in the presence of cholesterol (Chol) were prepared with initial DPPC/Chol/protein weight ratio as 3:0.15:1. ApoA-I/DPPC/Chol complexes with different protein content (initial DPPC/apoA-I weight ratios were changed from 10.5:1 to 2.6:1) but with a fixed initial DPPC/Chol weight ratio of 20:1 were also prepared. The complexes were isolated by gel-filtration and characterized by size and composition. ApoA-I- and apoA-II-complexes had the same size (80-84 A) and the complexes became more heterogeneous upon Chol inclusion; apoE-complexes were larger (97-100 A) and more homogeneous and Chol addition had no effect on their hydrodynamic properties. Chol seems to be excluded partially in the following manner for isolated complexes with different apo's: A-II > E > A-I. The possible existence of two lipid regions in the complexes differing in lipid dynamics - the lipid shell in the vicinity of apolipoprotein (boundary lipid) opposite to the remaining part of the lipid bilayer - has been studied by absorbance and fluorescence spectroscopy with cis-parinaric acid (cis-PA) and trans-parinaric acid (trans-PA) embedded into the complexes. Their application is based on a strong preference of trans-PA for solid lipid while cis-PA distributes more equally between co-existing fluid and solid lipid regions (Sklar et al. (1979) Biochemistry 18, 1707-1716). (1) For apoA-I-complexes, the partition of cis-PA between water and lipid phase at temperatures below and above the transition temperature of DPPC (T(t)) was insensitive to Chol and temperature, while partition of trans-PA into the lipid phase of Chol-containing complex was increased at high temperature and decreased at low temperature. These results seem to be related to trans-PA redistribution between Chol-rich and protein-rich lipid domains, the latter being more disordered at T < T(t) and more immobilized at T > T(t) compared to the bulk bilayer; cis-PA localizes preferentially in boundary lipid. This hypothesis was directly confirmed by measurements of energy transfer between apoA-I tryptophanyls and probe molecules. (2) The relative response of trans-PA fluorescence intensity to temperature-induced phase transition of DPPC in apoA-I/DPPC/Chol complexes was decreased as a function of apolipoprotein content in a non-monotonic fashion with a transition midpoint at a mol ratio DPPC/A-I of 250:1, probably indicating two different modes of apolipoprotein/DPPC interaction in different sized complexes. (3) The comparative study of lipid dynamics in apoA-I-, apoE- and apoA-II-containing complexes with temperature response to phospholipid phase transition with fluorescence parameters such as intensity and anisotropy of cis-PA and trans-PA revealed the presence of boundary lipid in all three complexes without Chol. In contrast to apoA-I-containing complexes, in apoA-II/DPPC/Chol complexes, trans-PA seems to move preferentially into boundary lipid and cis-PA to distribute between two different regions probably as a result of more ordering action induced by apoA-II compared to apoA-I on the nearest phospholipid molecules in Chol-containing complexes; the apoE action on trans-PA and cis-PA distribution could be intermediate. Based on these results, the degree of Chol exclusion from the boundary lipid region for complexes with different apo's increasing in the order A-II > E > A-I can be suggested. Different Chol distributions between two lipid regions in the complexes seems not to be a function of complex size, but rather is an inherent property of the particular apolipoprotein molecule.