Abstract 224: Formation of Cholesterol-Enriched Nascent HDL Requires Helices 5 and 7

Heart disease claims nearly 400,000 individuals in the US per year. Large population studies have repeatedly demonstrated an inverse relationship between HDL concentration and risk. Despite the strong correlation, pharmacological trials that raise plasma HDL have shown little efficacy, suggesting that HDL function should be monitored. Studies were initiated into the mechanism of cholesterol efflux leading to nascent HDL (nHDL) formation. The most productive cholesterol removal occurs when lipid-poor apoA-I, the main structural protein in nHDL, forms particles after interacting with ABCA1. ApoA-I acquires cholesterol and phospholipid at the cell surface forming 10-12 nm diameter nHDL whose composition resembles lipid rafts. Because lipidated apoA-I has a significantly different 3-D conformation compared to its lipid-poor counterpart, the structural modifications leading to nHDL formation were investigated. First, we completed a 3-D solution structure of lipid-poor apoA-I using lysine specific chemical cross-linking in conjunction with specifically engineered double cysteine-containing apoA-I mutants. Strategically placed cysteines form disulfide bonds (locked) when residues are within ~0.5 nm. If no disulfide bond formed (unlocked) then the two cysteines were separated by more than ~0.5 nm. Both methods rely on mass spectrometry to identify amino acid sequence and distance constraints within each 10 repeating helices comprising 80% of full-length apoA-I. By comparing lipid-poor apoA-I conformation to our previously described 3-D structure of apoA-I on 10-12 nm nHDL, we formulated a series of hypothetical helical domain transitions that might drive protein-lipid interaction and particle formation. To test, 10 different double cysteine mutants were employed in either their “locked” (oxidized) or “unlocked” (reduced) state by evaluating their ability to form recombinant HDL using synthetic phospholipid or to form nHDL using HEK cells. After determining HDL particle yield, size and composition from all locked and unlocked double cysteine apoA-I mutants we conclude that both the N- and C-terminal ends of apoA-I are essential for the first steps in lipid acquisition, but that the central helices 5, 6 and 7 are essential for nHDL formation.