Advertisement

Human apoA-I[Lys107del] mutation affects lipid surface behavior of apoA-I and its ability to form large nascent HDL

Open AccessPublished:December 13, 2022DOI:https://doi.org/10.1016/j.jlr.2022.100319

      Abstract

      Population studies have found that a natural human apoA-I variant, apoA-I[K107del], is strongly associated with low HDL-C but normal plasma apoA-I levels. We aimed to reveal properties of this variant that contribute to its unusual phenotype associated with atherosclerosis. Our oil-drop tensiometry studies revealed that compared to WT, recombinant apoA-I[K107del] adsorbed to surfaces of POPC-coated triolein drops at faster rates, remodeled the surfaces to a greater extent, and was ejected from the surfaces at higher surface pressures on compression of the lipid drops. These properties may drive increased binding of apoA-I[K107del] to and its better retention on large TG-rich lipoproteins, thereby increasing the variant’s content on these lipoproteins. While K107del did not affect apoA-I capacity to promote ABCA1-mediated cholesterol efflux from J774 cells, it impaired the biogenesis of large nascent HDL (nHDL) particles resulting in formation of predominantly smaller nHDL. Size exclusion chromatography of spontaneously reconstituted dimyristoylphosphatidylcholine (DMPC)-apoA-I complexes showed that apoA-I[K107del] had a hampered ability to form larger complexes but formed efficiently smaller-sized complexes. CD analysis revealed a reduced ability of apoA-I[K107del] to increase α-helical structure on binding to DMPC or in the presence of trifluoroethanol. This property may hinder formation of apoA-I[K107del]-containing large discoidal and spherical HDL but not smaller HDL. Both factors, the increased content of apoA-I[K107del] on TG-rich lipoproteins and the impaired ability of the variant to stabilize large HDL particles resulting in reduced lipid:protein ratios in HDL, may contribute to normal plasma apoA-I levels along with low HDL-C and increased risk for CVD.

      Keywords

      Abbreviations:

      DMPC,1 (2-dimyristoylphosphatidylcholine), nHDL (nascent HDL), POPC/TO/W (POPC/ triolein/ water), SEC (size exclusion chromatography), SUV (small unilamellar vesicles), TO (triolein), TO / W (triolein / water)

      Introduction

      ApoA-I is a major protein component of HDL and is also present in large triglyceride (TG)-rich lipoproteins, VLDL and chylomicrons (
      • Phillips M.C.
      New insights into the determination of HDL structure by apolipoproteins: Thematic Review Series: High Density Lipoprotein Structure, Function, and Metabolism.
      ,
      • Dominiczak M.H.
      • Caslake M.J.
      Apolipoproteins: metabolic role and clinical biochemistry applications.
      ). Plasma levels of HDL-C and apoA-I are inversely related to risk of CVD and atherosclerosis (
      • Sharrett A.R.
      • Ballantyne C.M.
      • Coady S.A.
      • Heiss G.
      • Sorlie P.D.
      • Catellier D.
      • Patsch W.
      Coronary heart disease prediction from lipoprotein cholesterol levels, triglycerides, lipoprotein(a), apolipoproteins A-I and B, and HDL density subfractions: The Atherosclerosis Risk in Communities (ARIC) Study.
      ). However, pharmacological treatments that increased HDL-C did not result in significant reduction of clinical cardiovascular events [(
      • Rader D.J.
      • Tall A.R.
      The not-so-simple HDL story: is it time to revise the HDL cholesterol hypothesis?.
      ,
      • Rye K.A.
      • Barter P.J.
      Cardioprotective functions of HDLs.
      ) and references cited therein]. This outcome supports the concept that correct properties of HDL are vital for their cardioprotective functions (
      • Rader D.J.
      • Tall A.R.
      The not-so-simple HDL story: is it time to revise the HDL cholesterol hypothesis?.
      ,
      • Rye K.A.
      • Barter P.J.
      Cardioprotective functions of HDLs.
      ,
      • Pownall H.J.
      • Rosales C.
      • Gillard B.K.
      • Gotto Antonio M.
      • M Jr., A.
      High-density lipoproteins, reverse cholesterol transport and atherogenesis.
      ). The cardioprotective functions of HDL mainly relate to its involvement in the pathways of reverse cholesterol transport, with apoA-I being involved in the critical steps of these pathways (
      • Rye K.A.
      • Barter P.J.
      Cardioprotective functions of HDLs.
      ,
      • Pownall H.J.
      • Rosales C.
      • Gillard B.K.
      • Gotto Antonio M.
      • M Jr., A.
      High-density lipoproteins, reverse cholesterol transport and atherogenesis.
      ,
      • Rosenson R.S.
      • Brewer Jr., H.B.
      • Davidson W.S.
      • Fayad Z.A.
      • Fuster V.
      • Goldstein J.
      • Hellerstein M.
      • Jiang X.C.
      • Phillips M.C.
      • Rader D.J.
      • Remaley A.T.
      • Rothblat G.H.
      • Tall A.R.
      • Yvan-Charvet L.
      Cholesterol efflux and atheroprotection: advancing the concept of reverse cholesterol transport.
      ). Interaction of apoA-I with ABCA1 promotes cellular cholesterol efflux and results in formation of discoidal nascent HDL (nHDL). Maturation of nHDL into spherical particles involves LCAT-mediated cholesterol esterification, with apoA-I being a principal activator of the enzyme. Interaction of HDL-bound apoA-I with hepatic scavenger receptor B1 promotes extraction of cholesteryl ester from the circulation. Mutations in apoA-I affect the protein stability and metabolism and, thereby, plasma concentrations of apoA-I and HDL (
      • Sorci-Thomas M.G.
      • Thomas M.J.
      The effects of altered apolipoprotein A-I structure on plasma HDL concentration.
      ). Many of the naturally occurring apoA-I mutations are associated with low levels of both plasma apoA-I and HDL-C and with increased risk for CVD and atherosclerosis (http://www.hgmd.cf.ac.uk/ac/gene.php?gene=APOA1). Population studies also demonstrated a strong association of low HDL-C levels with low plasma apoA-I levels (
      • Asztalos B.F.
      • Demissie S.
      • Cupples L.A.
      • Collins D.
      • Cox C.E.
      • Horvath K.V.
      • Bloomfield H.E.
      • Robins S.J.
      • Schaefer E.J.
      LpA-I, LpA-I: A-II HDL and CHD-risk: The Framingham Offspring Study and the Veterans Affairs HDL Intervention Trial.
      ).
      However, two large population studies, Copenhagen City Heart and Copenhagen General Population Studies (
      • Haase C.L.
      • Frikke-Schmidt R.
      • Nordestgaard B.G.
      • Tybjærg-Hansen A.
      Population-Based Resequencing of APOA1 in 10,330 Individuals: Spectrum of Genetic Variation, Phenotype, and Comparison with Extreme Phenotype Approach.
      ), found that Lys107del mutation of apoA-I (apoA-I[K107del]), a naturally occurring apoA-I heterozygous variant with a deletion of Lysine 107, was strongly associated with low plasma concentrations of HDL-C but normal levels of apoA-I. Reduced plasma HDL-C without a corresponding reduction in plasma apoA-I concentrations were also reported earlier in subjects with apoA-I[K107del in Finland and Germany (
      • Tilly-Kiesi M.
      • Lichtenstein A.H.
      • Ordovas J.M.
      • Dolnikowski G.
      • Malmström R.
      • Taskinen M.-R.
      • Schaefer E.J.
      Subjects With ApoA-I(Lys107→0) Exhibit Enhanced Fractional Catabolic Rate of ApoA-I in Lp(AI) and ApoA-II in Lp(AI With AII).
      ,
      • Nofer J.R.
      • von Eckardstein A.
      • Wiebusch H.
      • Weng W.
      • Funke H.
      • Schulte H.
      • Köhler E.
      • Assmann G.
      Screening for naturally occurring apolipoprotein A-I variants: apo A-I(delta K107) is associated with low HDL-cholesterol levels in men but not in women.
      ). Such dissociation between plasma concentrations of apoA-I and HDL-C seems especially surprising for the apoA-I[K107del] variant that is also associated with hereditary amyloidosis (
      • Amarzguioui M.
      • Mucchiano G.
      • Häggqvist B.
      • Westermark P.
      • Kavlie A.
      • Sletten K.
      • Prydz H.
      Extensive intimal apolipoprotein A1-derived amyloid deposits in a patient with an apolipoprotein A1 mutation.
      ,
      • Rowczenio D.
      • Dogan A.
      • Theis J.D.
      • Vrana J.A.
      • Lachmann H.J.
      • Wechalekar A.D.
      • Gilbertson J.A.
      • Hunt T.
      • Gibbs S.D.
      • Sattianayagam P.T.
      • Pinney J.H.
      • Hawkins P.N.
      • Gillmore J.,D.
      Amyloidogenicity and clinical phenotype associated with five novel mutations in apolipoprotein A-I.
      ), a disease characterized by enhanced proteolysis of mutated apoA-I and deposition of protein fragments as fibrils in vital organs that lead to removal of apoA-I from the circulation. Previous studies have shown that K107del leads to apoA-I destabilization (
      • Gorshkova I.N.
      • Mei X.
      • Atkinson D.
      Binding of human apoA-I[K107del] variant to TG-rich particles: implications for mechanisms underlying hypertriglyceridemia.
      ,
      • Ramella N.A.
      • Schinella G.R.
      • Ferreira S.T.
      • Prieto E.D.
      • Vela M.E.
      • Ríos J.L.
      • Tricerri M.A.
      • Rimoldi O.J.
      Human apolipoprotein A-I natural variants: molecular mechanisms underlying amyloidogenic propensity.
      ), enhanced fractional catabolism (
      • Tilly-Kiesi M.
      • Lichtenstein A.H.
      • Ordovas J.M.
      • Dolnikowski G.
      • Malmström R.
      • Taskinen M.-R.
      • Schaefer E.J.
      Subjects With ApoA-I(Lys107→0) Exhibit Enhanced Fractional Catabolic Rate of ApoA-I in Lp(AI) and ApoA-II in Lp(AI With AII).
      ), and reduced binding to HDL (
      • Huang W.
      • Matsunaga A.
      • Li W.
      • Han H.
      • Hoang A.
      • Kugi M.
      • Koga T.
      • Sviridov D.
      • Fidge N.
      • Sasaki J.
      Recombinant proapoA-I(Lys107del) shows impaired lipid binding associated with reduced binding to plasma high density lipoprotein.
      ). Each of these properties of apoA-I[K107del] should also lead to reduced plasma apoA-I concentrations. Nevertheless, the K107del mutation was not found to be associated with reduced plasma apoA-I concentrations (
      • Haase C.L.
      • Frikke-Schmidt R.
      • Nordestgaard B.G.
      • Tybjærg-Hansen A.
      Population-Based Resequencing of APOA1 in 10,330 Individuals: Spectrum of Genetic Variation, Phenotype, and Comparison with Extreme Phenotype Approach.
      ,
      • Tilly-Kiesi M.
      • Lichtenstein A.H.
      • Ordovas J.M.
      • Dolnikowski G.
      • Malmström R.
      • Taskinen M.-R.
      • Schaefer E.J.
      Subjects With ApoA-I(Lys107→0) Exhibit Enhanced Fractional Catabolic Rate of ApoA-I in Lp(AI) and ApoA-II in Lp(AI With AII).
      ), implying that some compensatory mechanisms may exist that maintain normal plasma apoA-I levels in subjects carrying the mutation. In this study, we investigated potential molecular mechanisms that may contribute to the reduced HDL-C coupled with normal apoA-I concentrations in plasma of subjects with the K107del mutation. Insights into the causes of this unusual phenotype are important for better understanding the function of the apoA-I[K107del] variant that is also linked to atherosclerosis (
      • Amarzguioui M.
      • Mucchiano G.
      • Häggqvist B.
      • Westermark P.
      • Kavlie A.
      • Sletten K.
      • Prydz H.
      Extensive intimal apolipoprotein A1-derived amyloid deposits in a patient with an apolipoprotein A1 mutation.
      ), premature coronary heart disease (
      • Tilly-Kiesi M.
      • Zhang Q.
      • Ehnholm S.
      • Kahri J.
      • Lahdenperä S.
      • Ehnholm C.
      • Taskinen M.-R.
      ApoA-IHelsinki (Lys107-->0) associated with reduced HDL cholesterol and LpA-I: A-II deficiency.
      ), and elevated plasma TG (
      • Nofer J.R.
      • von Eckardstein A.
      • Wiebusch H.
      • Weng W.
      • Funke H.
      • Schulte H.
      • Köhler E.
      • Assmann G.
      Screening for naturally occurring apolipoprotein A-I variants: apo A-I(delta K107) is associated with low HDL-cholesterol levels in men but not in women.
      ).
      One of the mechanisms contributing to the normal plasma apoA-I levels in carriers of the apoA-I[K107del] variant may relate to the increased apoA-I production into HDL containing apoA-I without apoA-II (but not in HDL containing both apoA-I and apoA-II) (
      • Tilly-Kiesi M.
      • Lichtenstein A.H.
      • Ordovas J.M.
      • Dolnikowski G.
      • Malmström R.
      • Taskinen M.-R.
      • Schaefer E.J.
      Subjects With ApoA-I(Lys107→0) Exhibit Enhanced Fractional Catabolic Rate of ApoA-I in Lp(AI) and ApoA-II in Lp(AI With AII).
      ). However, apoA-I has increased catabolic rate in both HDL subspecies in individuals carrying the K107del mutation (
      • Tilly-Kiesi M.
      • Lichtenstein A.H.
      • Ordovas J.M.
      • Dolnikowski G.
      • Malmström R.
      • Taskinen M.-R.
      • Schaefer E.J.
      Subjects With ApoA-I(Lys107→0) Exhibit Enhanced Fractional Catabolic Rate of ApoA-I in Lp(AI) and ApoA-II in Lp(AI With AII).
      ,
      • Tilly-Kiesi M.
      • Packard C.J.
      • Kahrij J.
      • Ehnholm C.
      • Shepherd J.
      • Taskinen M.R.
      In vivo metabolism of apoA-I and apo A-II in subjects with apo A-I(Lys107-->0) associated with reduced HDL cholesterol and Lp(AI w AII) deficiency.
      ). Another factor contributing to the reduced HDL-C coupled with normal plasma apoA-I levels may be an increased fraction of apoA-I on TG-rich lipoproteins along with reduced plasma HDL concentrations. ApoA-I can exist in plasma in lipid-free, lipid-poor, and lipid-bound states and exchange among lipoproteins (
      • Curtiss L.K.
      • Valenta D.T.
      • Hime N.J.
      • Rye K.A.
      What Is So Special About Apolipoprotein AI in Reverse Cholesterol Transport?.
      ). Our earlier in-vitro studies showed that incubation of WT or apoA-I[K107del] with TG-POPC synthetic emulsion particles that mimic VLDL-sized particles resulted in more molecules of the mutant bound to the particles compared to WT(15). Contrarily, when WT or pro-apoA-I[K107del] were incubated with the lipoprotein fraction of plasma, a smaller amount of the mutant, compared to WT, bound to HDL (
      • Huang W.
      • Matsunaga A.
      • Li W.
      • Han H.
      • Hoang A.
      • Kugi M.
      • Koga T.
      • Sviridov D.
      • Fidge N.
      • Sasaki J.
      Recombinant proapoA-I(Lys107del) shows impaired lipid binding associated with reduced binding to plasma high density lipoprotein.
      ). These observations suggest that compared to WT, apoA-I[K107del] may have a higher affinity to TG-rich lipoproteins, but a lower affinity to HDL, despite the fact that in plasma, the majority of apoA-I is bound to HDL. It is not clear why the K107 deletion may affect apoA-I binding to larger VLDL and to smaller HDL in opposite ways. To look into the factors that may affect binding of apoA-I[K107del] to lipoprotein surfaces, we studied the surface behavior of apoA-I[K107del] at triolein (TO)/water (W) and POPC /TO/W interfaces using oil-drop tensiometry (
      • Wang L.
      • Atkinson D.
      • Small D.M.
      The interfacial properties of ApoA-I and an amphipathic alpha-helix consensus peptide of exchangeable apolipoproteins at the triolein/water interface.
      ,
      • Meyers N.L.
      • Wang L.
      • Gursky O.
      • Small D.M.
      Changes in helical content or net charge of apolipoprotein C-I alter its affinity for lipid/water interfaces.
      ,
      • Wang L.
      • Mei X.
      • Atkinson D.
      • Small D.M.
      Surface behavior of apolipoprotein A-I and its deletion mutants at model lipoprotein interfaces.
      ,
      • Small D.M.
      • Wang L.
      • Mitsche M.A.
      The adsorption of biological peptides and proteins at the oil/water interface. A potentially important but largely unexplored field.
      ,
      • Meyers N.L.
      • Wang L.
      • Small D.M.
      Apolipoprotein C-I binds more strongly to phospholipid/triolein/water than triolein/water interfaces: a possible model for inhibiting cholesterol ester transfer protein activity and triacylglycerol-rich lipoprotein uptake.
      ,
      • Mitsche M.A.
      • Wang L.
      • Small D.M.
      Adsorption of egg phosphatidylcholine to an air/water and triolein/water bubble interface: use of the 2-dimensional phase rule to estimate the surface composition of a phospholipid/triolein/water surface as a function of surface pressure.
      ). TO is a common TG, and POPC is the most abundant phospholipid in lipoproteins. The TO/W interface provides a model for protein interactions with a hydrophobic TG core, a major component of TG-rich lipoproteins, while POPC/TO/W interfaces model the surface of TG-rich lipoproteins (
      • Small D.M.
      • Wang L.
      • Mitsche M.A.
      The adsorption of biological peptides and proteins at the oil/water interface. A potentially important but largely unexplored field.
      ,
      • Meyers N.L.
      • Wang L.
      • Small D.M.
      Apolipoprotein C-I binds more strongly to phospholipid/triolein/water than triolein/water interfaces: a possible model for inhibiting cholesterol ester transfer protein activity and triacylglycerol-rich lipoprotein uptake.
      ). Apolipoprotein adsorption to and desorption from the lipid surfaces in response to surface pressure changes is thought to mirror apolipoprotein binding to and dissociation from lipoproteins in-vivo (
      • Wang L.
      • Atkinson D.
      • Small D.M.
      The interfacial properties of ApoA-I and an amphipathic alpha-helix consensus peptide of exchangeable apolipoproteins at the triolein/water interface.
      ,
      • Meyers N.L.
      • Wang L.
      • Gursky O.
      • Small D.M.
      Changes in helical content or net charge of apolipoprotein C-I alter its affinity for lipid/water interfaces.
      ,
      • Wang L.
      • Mei X.
      • Atkinson D.
      • Small D.M.
      Surface behavior of apolipoprotein A-I and its deletion mutants at model lipoprotein interfaces.
      ,
      • Small D.M.
      • Wang L.
      • Mitsche M.A.
      The adsorption of biological peptides and proteins at the oil/water interface. A potentially important but largely unexplored field.
      ,
      • Meyers N.L.
      • Wang L.
      • Small D.M.
      Apolipoprotein C-I binds more strongly to phospholipid/triolein/water than triolein/water interfaces: a possible model for inhibiting cholesterol ester transfer protein activity and triacylglycerol-rich lipoprotein uptake.
      ). These surface pressure changes occur throughout the lipoprotein remodeling involving LCAT and lipase reactions, as well as transfer of TG, cholesteryl ester, and phospholipid molecules between lipoproteins mediated by plasma lipid transfers proteins (
      • Curtiss L.K.
      • Valenta D.T.
      • Hime N.J.
      • Rye K.A.
      What Is So Special About Apolipoprotein AI in Reverse Cholesterol Transport?.
      ,
      • Small D.M.
      • Wang L.
      • Mitsche M.A.
      The adsorption of biological peptides and proteins at the oil/water interface. A potentially important but largely unexplored field.
      ,
      • Meyers N.L.
      • Wang L.
      • Small D.M.
      Apolipoprotein C-I binds more strongly to phospholipid/triolein/water than triolein/water interfaces: a possible model for inhibiting cholesterol ester transfer protein activity and triacylglycerol-rich lipoprotein uptake.
      ).
      We also investigated the effects of the K107del mutation on the biogenesis of nHDL promoted by apoA-I during ABCA1-mediated cholesterol efflux from J774 cells. It has been shown earlier that deletion of K107 in apoA-I does not affect net ABCA-1 mediated cholesterol efflux from various cell lines (
      • Huang W.
      • Matsunaga A.
      • Li W.
      • Han H.
      • Hoang A.
      • Kugi M.
      • Koga T.
      • Sviridov D.
      • Fidge N.
      • Sasaki J.
      Recombinant proapoA-I(Lys107del) shows impaired lipid binding associated with reduced binding to plasma high density lipoprotein.
      ,
      • von Eckardstein A.
      • Castro G.
      • Wybranska I.
      • Theret N.
      • Duchateau P.
      • Duverger N.
      • Fruchart J.-C.
      • Ailhaud
      • Assmann G.
      Interaction of reconstituted high density lipoprotein discs containing human apolipoprotein A-I (apoA-I) variants with murine adipocytes and macrophages.
      ,
      • Gonzalez M.C.
      • Toledo J.D.
      • Tricerri M.A.
      • Garda H.A.
      The central type Y amphipathic a-helices of apolipoprotein AI are involved in the mobilization of intracellular cholesterol depots.
      ). In this study, we looked for the first time into the effect of the K107del mutation on the size distribution of nHDL, as changes in nHDL sizes and structural arrangement may contribute to the altered plasma ratio of apoA-I to HDL-C. As we found that deletion of K107 led to impaired formation of large and very large nHDL particles, we wanted to probe a structural basis of the altered size distribution of nHDL. To that end, we investigated the effects of the K107 deletion on the formation of dimyristoylphosphatidylcholine (DMPC)-apoA-I complexes at various lipid:protein ratios and the ability of apoA-I to form additional α-helical structure in the presence of a helical structure inducer, trifluoroethanol (TFE).

      Materials and methods

      Reagents

      POPC (25 mg/ml in chloroform), 1, 2-dimyristoyl phosphatidylcholine (DMPC), TO and BODIPY-cholesterol (TopFluor) were purchased from Avanti Polar Lipids (Alabaster, AL). TFE (extra pure) was purchased from Thermo Fisher Scientific. CodonPlus-RIL cells and the QuickChange mutagenesis kit were purchased from Stratagene (LaJolla, CA). Media (RPMI and MEM), fetal bovine serum, and penicillin streptomycin were from Invitrogen (Waltham, MA), Cpt-cAMP (cAMP), HRP-conjugated-cholera toxin subunit B and protease inhibitors were from Sigma. Polyclonal goat antibodies to apoA-I were purchased from Calbiochem and Invitrogen. Polyclonal antibodies against ABCA1 were from Novus Biologicals (Littelton, CO), and monoclonal antibodies against pan actin were from Cytoskeleton Inc. (Denver, CO).

      ApoA-I expression and purification

      WT apoA-I (WT) and apoA-I[K107del] were expressed, purified, and stored as described previously (
      • Gorshkova I.N.
      • Mei X.
      • Atkinson D.
      Binding of human apoA-I[K107del] variant to TG-rich particles: implications for mechanisms underlying hypertriglyceridemia.
      ). Before each experiment, a protein aliquot was thawed and freshly refolded by dialysis against 4M guanidine hydrochloride followed by extensive dialysis against an appropriate buffer: 2 mM sodium phosphate, pH 7.4, for oil-drop tensiometry experiments; TBS, pH 7.4, for formation of DMPC-apoA-I complexes, or 10 mM sodium phosphate, pH 7.6, for all other experiments.

      Drop-tensiometry

      Interfacial tension measurements

      An I.T. Concept Tracker oil-drop tensiometer (Longessaigne, France) was used to measure interfacial tension (γ) (
      • Wang L.
      • Atkinson D.
      • Small D.M.
      The interfacial properties of ApoA-I and an amphipathic alpha-helix consensus peptide of exchangeable apolipoproteins at the triolein/water interface.
      ,
      • Meyers N.L.
      • Wang L.
      • Gursky O.
      • Small D.M.
      Changes in helical content or net charge of apolipoprotein C-I alter its affinity for lipid/water interfaces.
      ,
      • Wang L.
      • Mei X.
      • Atkinson D.
      • Small D.M.
      Surface behavior of apolipoprotein A-I and its deletion mutants at model lipoprotein interfaces.
      ,
      • Meyers N.L.
      • Wang L.
      • Small D.M.
      Apolipoprotein C-I binds more strongly to phospholipid/triolein/water than triolein/water interfaces: a possible model for inhibiting cholesterol ester transfer protein activity and triacylglycerol-rich lipoprotein uptake.
      ). Experiments were carried out at 25 + 0.1oC in a thermostated system and repeated at least twice. To measure γ of the TO/W interface, TO drops of 16 μl were formed in a cuvette containing 6 ml of gently stirred standard buffer with a given amount of protein, and γ was monitored continuously until it reached equilibrium.
      For the experiments with POPC/TO/W interfaces, small unilamellar vesicles (SUV) were prepared from POPC solution in chloroform by drying the chloroform under nitrogen and resuspending the dried lipid film in standard buffer (2 mM sodium phosphate, pH 7.4) to a concentration of 2.5-5 mg/ml. The lipid suspension was then sonicated for 60 min with pulsed duty cycle of 30%. The resulting suspension was virtually clear and contained SUV particles with a diameter of 21-28 nm as estimated by negative staining electron microscopy. To measure γ of the POPC/TO/W interface, a 16 μl TO drop was formed in 6 ml gently stirred standard buffer containing 0 - 1 mg of POPC SUV stock and γ was recorded continuously. Adsorption of POPC molecules onto the TO drop lowered γ. After POPC was adsorbed, the excess POPC vesicles were removed from the aqueous phase by the buffer exchange procedure as described (
      • Mitsche M.A.
      • Wang L.
      • Small D.M.
      Adsorption of egg phosphatidylcholine to an air/water and triolein/water bubble interface: use of the 2-dimensional phase rule to estimate the surface composition of a phospholipid/triolein/water surface as a function of surface pressure.
      ), and γ stayed constant. After TO/W interfaces had stabilized at an initial γ (γi) of about 32 mN/m or POPC/TO/W interfaces reached γi of about 25 mN/m, varied amounts of WT or apoA-I[K107del] were added to the aqueous phase to achieve different protein concentrations in the range 1.3 - 5.3 x 10-7M. The γ of the interface was recorded continuously until it reached an equilibrium value (γeq).

      Penetration of the proteins into the POPC/TO/W interface and measurement of exclusion pressure

      To compare the ability of WT and apoA-I[K107del] to penetrate the POPC/TO/W interface, the adsorption isotherms were measured as described above after the initial surface pressure of the interface (Πi) was set at various target values. In short, after the TO drop (16 μl) was coated with POPC and γ reached ∼25 mN/m, the excess POPC vesicles were removed from the aqueous phase. Then the interface was either expanded or compressed as described (
      • Wang L.
      • Mei X.
      • Atkinson D.
      • Small D.M.
      Surface behavior of apolipoprotein A-I and its deletion mutants at model lipoprotein interfaces.
      ), resulting in a decrease or increase of the surface concentration of POPC (ᴦPOPC) and the corresponding changes in Πi. Values for Πi were obtained from γ of a TO/W interface (γTO = 32 mN) minus γ of the interface with POPC (Πi = γTO – γi). Larger values of ᴦPOPC lead to smaller γi and larger Πi. An aliquot of protein was then added to the aqueous phase (at final concentration of about 6 x 10-8M), γeq was measured, and equilibrium pressure Πeq was obtained (Πeq = γTO – γeq). The values of the increase in surface pressure (ΔΠ = Πeq - Πi) were plotted against Πi and the data were fit to a linear regression. A linear extrapolation of the plot gave values of exclusion pressure (ΠEX) at the x-intercept where ΔΠ = 0. At a surface pressure equal to or higher than ΠEX, protein cannot penetrate the POPC/TO/W interface.

      Slow expansion and compression of the POPC/TO/W interface and measurement of envelope (retention) pressure

      Slow expansion and compression of the POPC/TO/W interfaces were performed to estimate values of retention pressure at which protein begins to be ejected from the surface upon compression (
      • Mitsche M.A.
      • Wang L.
      • Small D.M.
      Adsorption of egg phosphatidylcholine to an air/water and triolein/water bubble interface: use of the 2-dimensional phase rule to estimate the surface composition of a phospholipid/triolein/water surface as a function of surface pressure.
      ). After POPC adsorption to the TO drop lowered γ to about 25 mN/m and buffer was exchanged with POPC-free buffer, Πi was set at various values by increasing or decreasing the drop volume. The values of Πi were used to calculate ᴦPOPC as a percentage of TO drop surface covered by POPC (
      • Mitsche M.A.
      • Wang L.
      • Small D.M.
      Adsorption of egg phosphatidylcholine to an air/water and triolein/water bubble interface: use of the 2-dimensional phase rule to estimate the surface composition of a phospholipid/triolein/water surface as a function of surface pressure.
      ). Protein was added to the buffer and adsorption lowered γ to γeq. The protein in the aqueous phase was removed by buffer exchange and the interface was slowly expanded to a volume of 25-28 μl. After γ equilibrated for 200 s or longer, the volume was decreased at the rate 0.02 μl/s with varying lower limit. Pressure-area (Π-A) curves (examples in Figure 3A ) were derived from the compression procedure as previously described (
      • Mitsche M.A.
      • Wang L.
      • Small D.M.
      Adsorption of egg phosphatidylcholine to an air/water and triolein/water bubble interface: use of the 2-dimensional phase rule to estimate the surface composition of a phospholipid/triolein/water surface as a function of surface pressure.
      ), with pressure Π calculated from γ when the surface area (A) changed. Each Π-A curve exhibits an envelope point, which marks a change in slope (marked by stars in Figure 3A). The envelope point was determined by plotting the values of the derivative dΠ/dA against area A. The envelope point is the surface area (AENV) and pressure (ΠENV) at which protein begins to be ejected from a POPC/TO/W interface. The values of ΠENV reflect the retention pressure.
      Figure thumbnail gr1
      Figure 1Apo A-I-induced remodeling of various lipid surfaces. A) and C) Interfacial tension (γ) changes for TO/W (A) and POPC/TO/W (C) interfaces after adding the proteins to the aqueous phase. The curves are example of the adsorption isotherms (γ versus time) for WT and apoA-I[K107del] that were added to the aqueous phase at (2.4 + 0.2) x 10-7 M. B) and D): Dependence of equilibrium γ on the protein concentration for the TO/W (B) and POPC/TO/W (D) interfaces. Each point represents the average + S.D. from at least three adsorption isotherms.
      Figure thumbnail gr2
      Figure 2Exclusion pressure ΠEX for WT and apoA-I[K107del] at POPC/TO/W interfaces. The initial surface pressure Πi and the corresponding ᴦPOPC of POPC/TO/W interfaces were set as described in Material and Methods. Adsorption of WT or apoA-I[K107del] to the interfaces increased the surface pressure to a different extent, ΔΠ, depending on the initial surface pressure Πi. For each protein, changes in surface pressure ΔΠ were plotted against corresponding values of Πi, and the data were fitted to linear regressions. X-intercepts of the linear regression at ΔΠ = 0 mN/M represent values of ΠEX, the exclusion pressure at which a given protein cannot bind POPC/TO/W interfaces.
      Figure thumbnail gr3
      Figure 3Pressure-area (Π-A) isotherms and retention pressure (ΠENV). A) Examples of pressure-area (Π-A) isotherms for WT and apoA-I[K107del] adsorbed to a POPC/TO/W interface of ᴦPOPC = 34.4+ 0.1%. WT or apoA-I[K107del], added at 2.7 x 10-7 M in the aqueous phase, adsorbed to POPC/TO/W interfaces. Shown here are compression isotherms derived from slow compression following slow expansion of the POPC/TO/W interface with one of the proteins adsorbed to it. The arrow shows the direction of compression. Asterisks mark envelope points (corresponding to AENV and ΠENV), where an abrupt change in slope shows the beginning of ejection of the given protein from the surface upon compression. B) Dependence of ΠENV on POPC surface concentration, ᴦPOPC, for WT and apoA-I[K107del]. Proteins were added at 2.7 x 10-7 M in the aqueous phase, adsorbed to the TO/W interface or POPC/TO/W interfaces with varied ᴦPOPC. Envelop pressures (ΠENV) were determined from Π-A isotherm for each protein and plotted against ᴦPOPC. ΠENV values are the mean + SD, n=2-3. ΠENV values at TO/W interface (ᴦPOPC = 0%) are outlined by a rectangle.

      Cholesterol efflux and nHDL biogenesis

      Mouse macrophage-derived J774 cells were maintained in media (RPMI) supplemented with 10% fetal bovine serum, penicillin, and streptomycin. For the experiments, cells were plated in 24-well plates and incubated with BODIPY-cholesterol complexed to methyl ß-cyclodextrin to label the cholesterol pool as described (
      • Sankaranarayanan S.
      • Kellner-Weibel G.
      • de la Llera-Moya M.
      • Phillips M.C.
      • Asztalos B.F.
      • Bittman R.
      • Rothblat G.H.
      A sensitive assay for ABCA1-mediated cholesterol efflux using BODIPY-cholesterol.
      ,
      • Mei X.
      • Liu M.
      • Herscovitz H.
      • Atkinson D.
      Probing the C-terminal domain of lipid-free apoA-I demonstrates the vital role of the H10B sequence repeat in HDL formation.
      ). ABCA1 was then upregulated by incubating the cells for 20-24 h in medium A (RPMI supplemented with 0.2% BSA and 2 μg/ml ACAT inhibitor) with or without 0.3 mM cpt-cAMP (cAMP). Cells were then washed and incubated with either WT or apoA-I[K107del] (10 μg/ml) in medium B (MEM-HEPES containing 0.01% BSA supplemented with 2 μg / ml ACAT inhibitor) containing 0.3 mM cAMP for 6 or 24 h. Control cells were incubated in media lacking acceptors (apoA-I) or cAMP. After the incubation, media were harvested and filtered through 0.22 μm membrane to remove cell debris. Cells were lysed by incubation in ice-cold 1% sodium cholate containing protease inhibitors for 2 h. Fluorescence intensity in media and cell lysates was measured using a plate reader (Tecan Infinite M1000) as described (
      • Liu M.
      • Mei M.
      • Herscovitz H.
      • Atkinson D.
      N-terminal mutation of apoA-I and interaction with ABCA1 reveal mechanisms of nascent HDL biogenesis.
      ). Fluorescence readings were adjusted by subtraction of a background fluorescence in media collected from cells incubated in media lacking acceptors. Efflux was calculated as a percentage of total fluorescence in media and cells. All experiments were carried out in duplicates or triplicates and were repeated 3-5 times.
      nHDL particles were detected by resolving the efflux media on 4-15% native PAGE followed by immunoblotting with antibodies to apoA-I. Bands corresponding to apoA-I-containing nHDL and lipid-free apoA-I were visualized using the ECL system and quantified using KwikQuant image analyzer. The membranes were then stripped with 0.25M glycine, pH 2.5, for 1 h. Blots were then subjected to ligand blotting using HRP-conjugated cholera toxin subunit B for 30 minutes to detect the ganglioside, monosialotetrahexosylganglioside (GM1) (
      • Duong P.T.
      • Collins H.L.
      • Nickel M.
      • Lund-Katz S.
      • Rothblat G.H.
      • Phillips M.C.
      Characterization of nascent HDL particles and microparticles formed by ABCA1-mediated efflux of cellular lipids to apoA-I.
      ). The membranes were imaged and quantified as described above. ABCA1 in cells was evaluated by resolving cell lysates on 4-15% SDS-PAGE follow ed by immunoblotting using antibodies to ABCA1. Actin, the loading control, was detected by probing the membranes using antibodies against pan actin. Bands were visualized and quantified as described above.

      Reconstitution and CD analysis of DMPC-apoA-I complexes

      DMPC-apoA-I complexes were obtained by spontaneous reconstitution as previously described (
      • Ludovico I.D.
      • Gisonno R.A.
      • Gonzalez M.C.
      • Garda H.A.
      • Ramella N.A.
      • Tricerri M.A.
      Understanding the role of apolipoproteinA-I in atherosclerosis. Post-translational modifications synergize dysfunction?.
      ), with some modifications. To prepare DMPC stock suspension (5 mg/ml), 5 mg DMPC in chloroform was dried under nitrogen in a borosilicate test tube and placed under vacuum overnight. One milliliter of TBS was added to the film of dry lipid and vortexed vigorously for 5 min at room temperature. During this process, the lipid dispersion was heated periodically, 2-3 times, for 30 s to the temperature above the phase transition of DMPC (24°C) to facilitate the multilamellar vesicle preparation. DMPC suspension was mixed with WT or apoA-I[K107del] solutions in TBS in the presence of 0.04% sodium azide to give lipid/protein ratios 2, 4, or 8 (w:w) and apoA-I concentration of 0.1 mg/ml. The mixtures were then incubated at 24oC with gentle shaking for 66 h. WT and apoA-I[K107del] in TBS without DMPC were incubated along. Following the incubation, the samples were eluted through a Superose 6 HP 10/30 column (Pharmacia) using Ӓkta FPLC system, equilibrated with TBS, pH 7.4, at a flow of 0.5 ml/min. The size exclusion chromatography (SEC) was used to separate DMPC-apoA-I complexes from the remaining lipid-free apoA-I and compare the elution profiles that reflect sizes of the resultant complexes. For each DMPC:apoA-I ratio, the experiment was repeated at least twice using a fresh preparation of DMPC multilamellar vesicles in each case. Fractions containing DMPC-apoA-I complexes were pooled and analyzed by immunoblotting with antibodies to apoA-I following separation of the particles on 4-15% native PAGE, as described above.
      The pooled fractions containing DMPC-apoA-I particles were dialyzed against 10 mM phosphate buffer, concentrated if necessary, and far-UV CD spectra of each sample were recorded at 25oC on Jasco J-1000 CD spectrometer (Jasco Inc., Easton, MD). The α-helical content was determined from the mean residue ellipticity at 222 nm as described (
      • Gorshkova I.N.
      • Liu T.
      • Kan H.Y.
      • Chroni A.
      • Zannis V.I.
      • Atkinson D.
      Structure and stability of apolipoprotein A-I in solution and in discoidal high-density lipoprotein probed by double charge ablation and deletion mutation.
      ). Protein concentration in the samples was determined by BCA protein assay (Thermo Scientific) in the presence of 1 mg/ml SDS in the working reagent.

      ApoA-I α-helical content in the presence of TFE

      The helical structure inducer, TFE, was used to assess the ability of WT and apoA-I[K107del] to form additional α-helical structure. Various amounts of TFE were added to the protein solutions in 10 mM phosphate buffer to have TFE concentrations 20, 40, 60, or 80 % (v/v) and protein concentration 0.04 mg/ml. After incubation of the samples for 1 hour at 4oC, CD spectra were recorded at 250C and the α-helical content was determined as described above. The experiment was repeated 3 times.

      Results

      Effect of K107 deletion on apoA-I affinity for lipid/water interfaces

      Protein adsorption to TO/W and POPC/TO/W interfaces

      We showed previously (
      • Wang L.
      • Mei X.
      • Atkinson D.
      • Small D.M.
      Surface behavior of apolipoprotein A-I and its deletion mutants at model lipoprotein interfaces.
      ) that WT apoA-I binds to TO/W and POPC/TO/W interfaces and decreases interfacial tension γ to equilibrium values, γeq. These decreases in γ correspond to the increases in surface pressure Π calculated as ΔΠ = Πeq – Πi. We compared ΔΠ values for WT and apoA-I[K107del] at apolar TO/W and more polar POPC/TO/W interfaces to assess how K107 deletion affects the protein binding to these surfaces. Adsorption isotherms (tension versus time) were obtained at various protein concentrations in the range 1.3-5.3 x10-7 M. In general, the equilibrium γ depends on protein concentration, with higher protein concentrations resulting in lower γeq and less time to reach equilibrium. Figures 1A and 1C show typical pairs of adsorption curves for the proteins at TO/W and POPC/TO/W interfaces, respectively, at protein concentration of (2.4 + 0.2) x 10-7 M. The initial γ of the TO/W interface was 32 + 0.5 mN/m and that of the POPC/TO/W interface was 25 + 0.5 mN/m, which corresponds to ᴦPOPC of ∼37% (
      • Mitsche M.A.
      • Wang L.
      • Small D.M.
      Adsorption of egg phosphatidylcholine to an air/water and triolein/water bubble interface: use of the 2-dimensional phase rule to estimate the surface composition of a phospholipid/triolein/water surface as a function of surface pressure.
      ). Similar to WT, apoA-I[K107del] lowered γ of both interfaces to reach equilibrium values, but the effect was greater for the mutant than for WT (Figures 1A and 1C).
      Based on at least three adsorption isotherms for each WT and apoA-I[107del] at (2.4 + 0.2) x 10-7 M at the TO/W interface, WT lowed γ to an equilibrium value of 16.8 + 0.5 mN/m, while apoA-I[K107del] lowered γ to an equilibrium value of 14.3 + 0.4 mN/m, The increases in the surface pressure due to adsorption of the proteins, ΔΠ, was higher for apoA-I[K107del] than for WT apoA-I (17.8 + 0.5 mN/m versus 15.7 + 0.5 mN/m, P < 0.05) indicating that at the same protein concentration, the mutant remodels the interface to a greater extent. The half time needed to reach equilibrium was calculated from at least three absorption isotherms for each WT and apoA-I[107del] at (2.4 + 0.2) x 10-7 M. The mean half time needed to reach equilibrium at the TO/W interface did not differ significantly between the proteins (85 + 21 s for apoA-I[K107del] and 99 + 16 s for WT, P>0.05). Figure 1B shows the tension, γ, of the TO/W surface at equilibrium after adding various concentrations of the proteins. Smaller concentrations of apoA-I[K107del] resulted in larger decreases of surface tension than the higher concentrations of WT apoA-I. For WT at the concentration of 5.3 x10-7 M, γeq was still higher than γeq for apoA-I[K107del] at three times lower concentration, 1.7 x10-7 M, indicating that the variant remodels the TO/W interface to greater extent than WT, which is consistent with higher affinity of apoA-I[K107del] to the TO/W interface.
      Based on at least three isotherms for each protein at (2.4 + 0.2) x 10-7 M at the POPC/TO/W interface, WT lowered γ to the equilibrium value of (14.5 + 0.4) mN/ m, while γeq with apoA-I[K107del] was 11.8 + 0.4 mN/ m. Accordingly, the increase in the surface pressure ΔΠ resulting from the protein adsorption was higher for apoA-I[K107del] than for WT (13.2 + 0.5 mN/ m versus 10.6 + 0.5 mN/ m P<0.025), indicating that at a similar protein concentration, the variant remodels the POPC/TO/W interface to a greater extent than WT. The half time needed to reach equilibrium calculated from the isotherms was significantly shorter with apoA-I[K107del] than with WT (65 + 12 s versus 186 + 31 s, P < 0.01). This difference indicates that in contrast to binding to the TO/W interface, binding to the POPC/TO/W interface was faster for apoA-I[K107del] than for WT. Comparison of γeq for the POPC/TO/W interface following adsorption of the proteins at various concentrations (Figure 1D) shows that lower concentrations of apoA-I[K107del] resulted in significantly lower values of γeq than the much higher concentrations of WT apoA-I. At the relatively high concentration of 5.3 x10-7 M, WT adsorption resulted in γeq that were still higher than γeq for apoA-I[K107del] at the three times lower concentration, 1.7 x10-7 M, or even at the four times lower concentration, 1.3 x10-7 M. Taken together, these data indicate that apoA-I[K107del] has a stronger ability than WT to bind to both the TO/W and POPC/TO/W interfaces, and the difference between the proteins is more pronounced at the POPC/TO/W interfaces.

      Protein penetration into POPC/TO/W interfaces and values of exclusion pressure (ΠEX)

      To compare the penetration behavior of WT and apoA-I[K107del] at POPC/TO/W interfaces, the adsorption curves were obtained for each protein at various initial surface pressure Πi, corresponding to various surface concentrations of POPC. The values of the increases in the surface pressure, ΔΠ, were plotted against Πi, and the data were fitted to linear regression (Figure 2). Linear regressions for both proteins were significant (R=0.99). Linear regression of the ΔΠ-Πi data for each protein gives ΠEX at the x-intercept where ΔΠ = 0 nm/m. ΠEX is the surface pressure at which a protein cannot penetrate and bind (i.e., is excluded from) the interface. The values for ΠEX were similar for WT and apoA-I[K107del] (25.0 - 25.5 nM/m). However, the plots show that at each initial surface pressure, binding of apoA-I[K107del] results in larger increases in the surface pressure, indicating that the mutant apoA-I remodels various surfaces of POPC-coated TO droplets to a greater extent and thus, interacts more strongly with the surfaces with various surface concentrations of POPC.

      Retention pressure (ΠENV) on POPC/TO/W interfaces

      Envelope pressure, ΠENV, for WT and apoA-I[K107del] at interfaces with varied POPC surface concentration, ᴦPOPC, was determined to compare differences in the protein retention over a range of lipid/water interfaces. To obtain the values of ΠENV, compression Π-A isotherms were recorded for POPC/TO/W surfaces with various POPC surface concentrations. Figure 3A shows examples of the Π-A isotherms for WT and apoA-I[K107del] adsorbed to POPC/TO/W interface with ᴦPOPC = 34.4 + 0.1 %. The data were generated from γ and surface area profiles when the POPC-coated TO drop with a protein adsorbed into the surface was slowly expanded and then compressed. The arrow marks the direction of compression. The asterisk (*) marks a point of change of the isotherm slope and corresponds to the envelope point, the surface area and pressure (ΠENV) at which the protein begins to be ejected from the surface on compression. The higher value of the ΠENV for the mutant (21.5 + 0.5 mN/m versus 19.6 + 0.5 mN/m for WT, P<0.05) indicates that compared to WT, apoA-I[K107del] begins to be ejected from the surface at higher surface pressure suggesting better retention of the variant on the POPC/TO surface. The Π-A isotherms for both proteins were obtained for various initial drop volumes corresponding to the various initial surface pressure and ᴦPOPC. Values of ΠENV determined from the isotherms for various ᴦPOPC are shown in Figure 3B. The values of ΠENV at ᴦPOPC = 0% (marked by a box) were determined from Π-A isotherms for the TO drop without POPC coating. Retention pressure for the two proteins did not differ significantly on the surface of the TO drop without POPC. In contrast, for POPC-coated TO drops, at each POPC surface concentration studied, apoA-I[K107del] was ejected from the surfaces at higher surface pressures than WT, indicating stronger retention of the variant on the POPC/TO/W interfaces.

      Effect of K107 deletion on cholesterol efflux and nHDL biogenesis

      Although the capacity of apoA-I[K107del] to promote ABCA1-mediated cholesterol efflux was reported by others using different cells (
      • Huang W.
      • Matsunaga A.
      • Li W.
      • Han H.
      • Hoang A.
      • Kugi M.
      • Koga T.
      • Sviridov D.
      • Fidge N.
      • Sasaki J.
      Recombinant proapoA-I(Lys107del) shows impaired lipid binding associated with reduced binding to plasma high density lipoprotein.
      ,
      • von Eckardstein A.
      • Castro G.
      • Wybranska I.
      • Theret N.
      • Duchateau P.
      • Duverger N.
      • Fruchart J.-C.
      • Ailhaud
      • Assmann G.
      Interaction of reconstituted high density lipoprotein discs containing human apolipoprotein A-I (apoA-I) variants with murine adipocytes and macrophages.
      ,
      • Gonzalez M.C.
      • Toledo J.D.
      • Tricerri M.A.
      • Garda H.A.
      The central type Y amphipathic a-helices of apolipoprotein AI are involved in the mobilization of intracellular cholesterol depots.
      ), we confirmed this process in J774 cells used in this study. We therefore, analyzed the capacity of the mutant apoA-I[K107del] to promote cholesterol efflux over a 6- and 24-h incubation period and compared the data with WT apoA-I. We chose these time points to investigate potential differences that may be occurring either during short and/or long incubation times (
      • Liu M.
      • Mei M.
      • Herscovitz H.
      • Atkinson D.
      N-terminal mutation of apoA-I and interaction with ABCA1 reveal mechanisms of nascent HDL biogenesis.
      ).
      Figure 4A shows that, as expected, following a 24-h incubation, the efflux promoted both by the WT and apoA-I[K107del] was relatively low from cells not incubated with cAMP and corresponded to 7.7 ± 4.2 and 7.3 ± 1.1%, respectively. Indeed, these cells have very low or undetectable level of ABCA1 (Figure 4B, lanes 2 and 6). However, both apoA-I[K107del] and WT promoted higher efflux (Figure 4A) from cells treated with cAMP that upregulated ABCA1 (Figure 4B, lanes 3-5 and 7-9). Percent cholesterol efflux promoted by apoA-I[K107del] was similar to that promoted by WT and corresponded to 27.9 ± 4.6 and 28.7 ± 2.8 %, respectively. Thus, the majority of the efflux was ABCA1-mediated. Furthermore, efflux promoted by apoA-I[K107del] after 6 h incubation was also similar to that of WT and corresponded to ∼ 60% of the efflux measured after 24-h incubation (data not shown).
      Figure thumbnail gr4
      Figure 4ABCA-1 mediated cholesterol efflux and nHDL formation. A) Cholesterol efflux. Cells pre-labeled with BODIPY-cholesterol were incubated with either WT or apoA-I[K107del] in media containing (gray bars) or not cAMP (black bars) for 24 h. Cholesterol efflux was determined by measuring net fluorescence in media and expressed as a percentage of total fluorescence in cells plus media. Bars represent means + SD (-cAMP, n=3; +cAMP, n=4). B) ABCA1 levels. Cells incubated with either WT or apoA-I[K107del] as described in Panel A were analyzed by Western blotting. Membranes were probed with antibodies to ABCA1 or to pan actin (loading control). Lane 1 is from cells incubated without protein acceptors. C) and D) Immunoblots showing apoA-I. Media harvested following a 6- (C) or 24-h (D) incubation with acceptors were analyzed by nondenaturing gradient (4-15%) PAGE followed by Immunoblotting. Immunoblots were probed with antibodies to apoA-I. Lanes 1-4 and 5-8 represent media derived from cells incubated with WT or lapoA-I[K107del], respectively. Lanes 1 and 5 represent media from cells incubated in the absence of cAMP. Lanes 2-4 and 6-8 represent media from cells incubated with cAMP. Bands marked by a rectangle represent nHDL > 8.5 nm-diameter.. E) Relative abundance of large nHDL (>8.5 nm-diameter)-apoA-I formed after 6- or 24-h incubation. Bands corresponding to nHDL-apoA-I depicted in immunoblots shown in panels C and D were quantified and large nHDL (marked by a rectangle) were expressed as a percentage of total nHDL. Values for apoA-I[K107del] (grey bars) were expressed as percentage of WT (set to 100%, black bars). Bars represent means ± SD, n = 5. ** P < 0.001.
      The level of ABCA1 in cells incubated with apoA-I[K107del] for 24 h was similar to that in cells incubated with WT (as determined in 2 experiments run in triplicates and normalized to actin level). In addition, cells treated with cAMP that were incubated in media not containing protein acceptors expressed a very low level of ABCA1 (Figure 4B, lane 1). Thus, similar to previous reports (
      • Huang W.
      • Matsunaga A.
      • Li W.
      • Han H.
      • Hoang A.
      • Kugi M.
      • Koga T.
      • Sviridov D.
      • Fidge N.
      • Sasaki J.
      Recombinant proapoA-I(Lys107del) shows impaired lipid binding associated with reduced binding to plasma high density lipoprotein.
      ,
      • Gonzalez M.C.
      • Toledo J.D.
      • Tricerri M.A.
      • Garda H.A.
      The central type Y amphipathic a-helices of apolipoprotein AI are involved in the mobilization of intracellular cholesterol depots.
      ,
      • Sankaranarayanan S.
      • Kellner-Weibel G.
      • de la Llera-Moya M.
      • Phillips M.C.
      • Asztalos B.F.
      • Bittman R.
      • Rothblat G.H.
      A sensitive assay for ABCA1-mediated cholesterol efflux using BODIPY-cholesterol.
      ), K107 deletion does not impair the ability of the mutant to promote cholesterol efflux as efficiently as the WT.
      Given that efflux promoted by the apoA-I[K107del] mutant was similar to that promoted by the WT, it was important to determine whether the mutant can bind the effluxed lipids as efficiently as the WT to form nHDL particles. To that end, we analyzed efflux media collected from cells after 6- and 24- h incubation by Western blotting. Figures 4C and 4D show representative immunoblots (probed with antibodies to apoA-I) of efflux media harvested after a 6- and 24-h incubation, respectively. As shown in these immunoblots, there was a major band of apoA-I in media derived from cells incubated without cAMP representing lipid-free apoA-I (lanes 1 and 5) that was not lipidated due to low, or undetectable level of ABCA1 (Figure 4B, lanes 2 and 6) resulting in minimal efflux (Figure 4A). When WT or apoA-I[K107del] were added to cells in the presence of cAMP to upregulate ABCA1 (Figure 4B, lanes 3-5 and lanes 7-9), they became lipidated and formed nHDL particles as early as 6 h as indicated by the presence of multiple bands representing a heterogenous population of presumably discoidal nHDL particles ranging from ∼6.8 to 12.2 nm-diameter (Figures 4C and 4D, lanes 2-4, and lanes 6-8, respectively). However, quantification of all bands representing nHDL showed that the overall capacity to form particles with apoA-I[K107del] was reduced to 92 ± 6 % (P < 0.05) and 93 ±7 % (P=0.07) of that for WT following 6- and 24-h incubation, respectively. The reduction in the ability of apoA-I[K107del] to bind the effluxed lipids was primarily due to its reduced capacity to form nHDL particles larger than 8.5 nm-diameter (compare lanes 6-8 to lanes 2-4 of Figures 4C and 4D) representing large and very large nHDL particle (
      • Mendivil C.O.
      • Furtado J.
      • Morton A.M.
      • Wang L.
      • Sacks F.M.
      Novel pathways of apolipoprotein A-I metabolism in HDL of different sizes in humans.
      ). After 6-and 24-h incubation, the relative abundance of large and very large nHDL particles formed by the mutant was significantly lower than those formed by the WT (Figure 4E and F). Large nHDL particles formed by the mutant after 6- and 24-h incubation accounted for 11.7 ± 2.3% (P < 0.001) and 23.4 ± 4.2 (P < 0.02), of total nHDL, respectively. Very large nHDL particles formed by apoA-I[K107del] after 6- and 24- h incubation accounted for 1.9 ± 0.6 (P < 0.05) and 6.4 ± 2 % (P < 0.005) of total nHDL, respectively. For comparison, the relative abundance of large nHDL particles formed by the WT after 6- and 24-h incubation was 34.1 ± 1.5 and 32.2 ± 3.1 %, respectively and the relative abundance of very large nHDL particles was 10.8 ± 0.8 and 14.8 ± 5.7%, respectively.
      It is notable that continued incubation from 6 to 24 hours led to a small increase in the relative abundance of very large nHDL particles formed by the WT from 10.8 to 14.8 % a and a significant (P < 0.05) increase by the mutant from 1.9 to 6.4 % of total nHDL, respectively. On the other hand, while the abundance of large nHDL particles formed by the WT did not change, this population of nHDL particles formed by the mutant increased significantly (P < 0.05) from 11.7 to 23.4 % of total nHDL (Figures 4E and 4F, respectively). Nevertheless, the overall abundance of combined large and very large nHDL particles formed by the mutant both after 6 and 24 hours was significantly lower than those formed by WT (P < 0.0001 and P < 0.001, respectively).
      Overall, the relative abundance of large + very large nHDL particles formed by apoA-I[K107del] after 6- and 24- h incubation was significantly reduced to 30.4 ± 13% (P < 0.0001) and 63.4 ± 3 % (P < 0.002) of the relative level of these particles formed by the WT, respectively (Supplemental Figure 1).
      These findings indicate that the mutant apoA-I[K107del] formed large and very large nHDL particles either more slowly or of lower stability than WT and thus was unable to attain WT levels even after a 24-h incubation.
      Since we previously demonstrated that nHDL particles formed by apoA-I following efflux from transfected HEK 293 cells contained the ganglioside GM1 (
      • Duong P.T.
      • Collins H.L.
      • Nickel M.
      • Lund-Katz S.
      • Rothblat G.H.
      • Phillips M.C.
      Characterization of nascent HDL particles and microparticles formed by ABCA1-mediated efflux of cellular lipids to apoA-I.
      ), it was of interest to determine, whether a) nHDL particles formed by apoA-I incubated with J774 cells also contain GM1 and whether the mutant has the capacity to bind GM1as efficiently as WT and b) to confirm that the lipid distribution (represented by GM1) is similar to the distribution of apoA-I. To determine the content of GM1 in nHDL particles, we probed the membranes with CTB which specifically binds to GM1 (
      • Duong P.T.
      • Collins H.L.
      • Nickel M.
      • Lund-Katz S.
      • Rothblat G.H.
      • Phillips M.C.
      Characterization of nascent HDL particles and microparticles formed by ABCA1-mediated efflux of cellular lipids to apoA-I.
      ). The representative ligand blots in Supplemental Figures 2A and 2B show that the distribution of GM1 in media derived from both WT and apoA-I[K107del] after 6- and 24-h incubation, parallels the distribution of lipidated apoA-I (Figures 4C and 4D, lanes 2-4 and 6-8). Importantly, all nHDL particles formed by WT and apoA-I[K107del] contained GM1 (lanes 2-4 and 6-8). No GM1 was detected in media derived from cells that were not incubated with cAMP (Supplemental Figures 2A and B, lanes 1 and 5). Large + very large (e.g. >8.5 nm-diameter) nHDL particles formed by the mutant contained 22 ± 6% and 25 ± 6 %, of total GM1 in nHDL following 6- and 24-h incubation, respectively, which were significantly lower than the relative level of these populations of nHDL particles formed by the WT and corresponded to 49 ± 15 % and 61± 9%, of WT, respectively (P < 0.01) (Supplemental Figure 2C). Overall, deletion of K107 did not impair the ability of apoA-I to bind GM1 but due to diminished biogenesis of large and very large nHDL particles, there was proportionally less GM1.
      Together these findings suggest that deletion of K107 impairs the ability of apoA-I to bind the effluxed lipids and form nHDL particles as efficiently as the WT and may impair its ability to bind enough lipid molecules/protein to form large and very large nHDL particles resulting in biogenesis of primarily smaller nHDL particles. Such changes are likely to have a profound impact in vivo.

      Effect of K107 deletion on sizes of DMPC-apoA-I complexes and α-helicity of apoA-I

      We aimed to investigate the properties of the apoA-I[K107del] that may contribute to the impaired ability of the protein to recruit lipids and form large and very large nHDL particles. To this end, we incubated apoA-I with DMPC at increasing lipid:protein ratios and studied the resultant DMPC-apoA-I complexes, which are commonly used as simple models of nascent HDL. Figure 5, A-C, shows SEC elution profiles of the resultant DMPC-apoA-I complexes. Both lipid-free WT and apoA-I[K107del] incubated without DMPC were eluted from the column between 17 and 19 ml, with a peak volume of 18 ml (not shown). Therefore, the fractions between 12 and 16.5 ml collected for analysis contained DMPC-apoA-I complexes with no lipid-free apoA-I. For each DMPC:apoA-I ratio, the elution profiles for apoA-I[K107del]-containing complexes were shifted to later elution volumes compared to those for WT-containing complexes, indicating overall smaller hydrodynamic diameters of the particles formed by the mutant. For each DMPC:apoA-I ratio, the elution profiles show two major populations of particle with different hydrodynamic diameters, with the apparent peak heights corresponding to the smaller particle being greater for the mutant than for WT.
      Figure thumbnail gr5
      Figure 5Size exclusion chromatography of DMPC-apoA-I complexes. A) – C) Gel filtration elution profile of DMPC-apoA-I complexes. WT and apoA-I[K107del] were incubated with DMPC suspension for 66 h at 24oC at DMPC:apoA-I weight ratio of 2 (A), 4 (B), or 8 (C). The mixtures were then eluted through Superose 6 HP 10/30 column equilibrated with TBS. Fractions containing DMPC-apoA-I complexes (grey-shaded areas) were pooled for analysis. Profiles for WT and apoA-I[K107del] shown by empty circles and filled circles, correspondingly. D) Representative immunoblot of DPMC-apoA-I complexes. Pooled fractions containing DMPC:apoA-I complexes were analyzed by nondenaturing gradient (4-15%) PAGE followed by immunoblotting with antibodies to apoA-I. Lanes 1, 3, and 5: WT-containing complexes separated by SEC for DMPC:apoA-I ratios of 2, 4, and 8, correspondingly. Lanes 2, 4, and 6: apoA-I[K107del]-containing complexes separated by SEC for DMPC:apoA-I ratios of 2, 4, and 8, correspondingly. Position of lipid-free apoA-I on the blot is indicated by the arrow.
      Accordingly, the representative immunoblot of the pooled fractions (Figure 5D) for each DMPC:apoA-I ratio shows a band corresponding to smaller particles (with Stokes diameters between 7.4 – 7.8 nm) and one band or two bands corresponding to larger particles (with Stokes diameters between 10.5 - 14 nm). Thus, for each DMPC:apoA-I ratio, apoA-I[K107del]-containing complexes show increased relative proportion of smaller particles and reduced abundance of larger particles. Furthermore, although increasing DMPC:apoA-I ratio from 2 to 8 led to increased relative abundance of larger particles for both WT and the mutant, WT formed more of the larger particles. Remarkably, increasing DMPC:apoA-I ratio up to 8 led to the formation by WT of a distinct population of particles with 14 nm diameter in addition to the particles with 10.5 nm diameter. The mutant, on the other hand, was unable to form this population of particles at this ratio (compare lane 5 to lane 6 in Figure 5D). Notably, the blot showed no presence of lipid-free apoA-I in any of the samples of WT- or apoA-I[K107del]-containing complexes separated by SEC and used for our analysis. Overall, our findings are consistent with the hampered ability of apoA-I[K107del] variant to form larger discoidal complexes with DMPC.
      Our findings agree with a significant shift of DMPC-apoA-I complexes toward smaller sizes as a result of the K107del mutation that was found by Huang et al. (
      • Huang W.
      • Matsunaga A.
      • Li W.
      • Han H.
      • Hoang A.
      • Kugi M.
      • Koga T.
      • Sviridov D.
      • Fidge N.
      • Sasaki J.
      Recombinant proapoA-I(Lys107del) shows impaired lipid binding associated with reduced binding to plasma high density lipoprotein.
      ) for the complexes formed by sodium cholate dialysis at DMPC:apoA-I weight ratio of 3.9 (molar ratio of 150:1). For a close DMPC:apoA-I weight ratio of 3.8 (molar ratio of 145:1), Ludovico et all (
      • Ludovico I.D.
      • Gisonno R.A.
      • Gonzalez M.C.
      • Garda H.A.
      • Ramella N.A.
      • Tricerri M.A.
      Understanding the role of apolipoproteinA-I in atherosclerosis. Post-translational modifications synergize dysfunction?.
      ) found no significant effect of the K107del mutation on sizes of DMPC-apoA-I complexes, based on the quantification of gradient non-denaturing gels. Interestingly that when DMPC-apoA-I complexes were formed by sodium cholate dialysis (
      • Huang W.
      • Matsunaga A.
      • Li W.
      • Han H.
      • Hoang A.
      • Kugi M.
      • Koga T.
      • Sviridov D.
      • Fidge N.
      • Sasaki J.
      Recombinant proapoA-I(Lys107del) shows impaired lipid binding associated with reduced binding to plasma high density lipoprotein.
      ), deletion of K107 resulted in an additional population of smaller particles that were not formed by WT and in a lack of a minor population of the largest particles were formed by WT. In our studies, both WT and apoA-I[K107del] were able to form a population of smaller particles. Thus, the effect of K107 deletion was not quite the same for DMPC-apoA-I particles prepared by sodium cholate dialysis and spontaneously reconstituted DMPC-apoA-I particles. This agrees with the recent findings by Bedi et al. (
      • Bedi S.
      • Morris J.
      • Shah A.
      • Hart R.C.
      • Jerome W.G.
      • Aller S.G.
      • Tang C.
      • Vaisar T.
      • Bornfeldt K.E.
      • Segrest J.P.
      • Heinecke J.W.
      • Davidson W.S.
      Conformational flexibility of apolipoprotein A-I amino- and carboxy-termini is necessary for lipid binding but not cholesterol efflux.
      ) suggesting that structural requirements for apoA-I to form complexes with phospholipids spontaneously may be different from those needed for the detergent-assisted formation of the complexes. Given that K107del mutation affects the protein conformation and stability (
      • Gorshkova I.N.
      • Mei X.
      • Atkinson D.
      Binding of human apoA-I[K107del] variant to TG-rich particles: implications for mechanisms underlying hypertriglyceridemia.
      ,
      • Ramella N.A.
      • Schinella G.R.
      • Ferreira S.T.
      • Prieto E.D.
      • Vela M.E.
      • Ríos J.L.
      • Tricerri M.A.
      • Rimoldi O.J.
      Human apolipoprotein A-I natural variants: molecular mechanisms underlying amyloidogenic propensity.
      ), the altered structural properties of apoA-I[K107del] apparently translated into slightly different effects for DMPC-apoA-I complexes when they were formed spontaneously or by cholate dialysis. Importantly, regardless of the method of the particle formation, the K107del mutation resulted in the shifts of DMPC-apoA-I particles to smaller sizes.
      It is well documented that the interaction of apoA-I with phospholipids to form discoidal HDL complexes is driven by an increase in amphipathic helical content in the protein on lipid interaction. To test the effect of the K107 deletion on the ability of apoA-I to increase α-helical structure on binding to DMPC, we determined the α-helical content of apoA-I in the DMPC-apoA-I complexes formed at increasing DMPC:apoA-I ratios. For the CD analysis, we used the same pooled fractions collected from SEC (Figure 5 A-C) that we used for the immunoblot analysis. The dependence of apoA-I α-helix on DMPC:apoA-I ratio is shown in Figure 6. Consistent with our earlier findings (
      • Gorshkova I.N.
      • Mei X.
      • Atkinson D.
      Binding of human apoA-I[K107del] variant to TG-rich particles: implications for mechanisms underlying hypertriglyceridemia.
      ), the K107del mutation did not change significantly the α-helix content of lipid-free apoA-I. However, the mutation resulted in the reduced α-helical content of apoA-I in DMPC-bound state at each DMPC:apoA-I ratio. Consistent with the SEC data, the largest effect of the K107 deletion on the protein α-helical content (10% decrease) was observed at DMPC:apoA-I ratio of 4. This reduction in the α-helical content, as a result of the K107del mutation, corresponds to unfolding of 24-residue segment, that is roughly one sequence repeat in apoA-I. At DMPC:apoA-I ratios of 2 and 8, the K107 deletion resulted in smaller but statistically significant reductions in the protein α-helical content, 6 - 7% and 5 - 6%, correspondingly. These reductions in the α-helical content resulted from the deletion of K107 correspond to unfolding of segments that are smaller than one sequence repeat. Thus, CD analysis of DMPC-apoA-I complexes indicate that deletion of K107 impedes the ability of apoA-I to increase α-helical structure on binding to the phospholipids.
      Figure thumbnail gr6
      Figure 6The α-helical content of apoA-I in DMPC-apoA-I complexes. DMPC-apoA-I complexes obtained by spontaneous reconstitution by incubating WT and apoA-I[K107del] with DMPC at various lipid:protein ratios and separated by SEC, as shown in , A – C, were analyzed by far-UV CD. The α-helical content of apoA-I in the DMPC-apoA-I complexes is plotted against DMPC:apoA-I ratios in the incubation mixtures. Data are means + SD, n >3. Letters above the data points indicate significance of the difference between the values for apoA-I[K107del] and WT by unpaired t-test, a: not significant, b: P < 0.05, c: P < 0.01.

      Effect of K107 deletion on the α-helical content of apoA-I in the presence of TFE

      To test if K107del affects the ability of apoA-I to fold additional α-helices in other environments that stimulate formation of α-helical structure, we determined the α-helical content of WT and apoA-I[K107del] in the presence of various concentrations of the helical structure inducer TFE (Figure 7). For both proteins, the α-helical content progressively increased when TFE concentration rose from 0 to 10-20%, and further increase in the TFE concentration did not result in significant changes of the protein α-helicity. However, while in the aqueous buffer (TFE = 0%) the α-helical content of the two proteins did not differ significantly, apoA-I[K107del] had a significantly lower α-helical content than WT in the presence of the various concentrations of TFE. The maximum α-helicity induced by TFE was on average 10% lower for apoA-I[K107del] compared to WT, again corresponding to the number of residues in one sequence repeat. Thus, K107del mutation impairs the ability of apoA-I to increase α-helical structure in the presence of TFE.
      Figure thumbnail gr7
      Figure 7The α-helical content of apoA-I in the presence of TFE. WT or apoA-I[K107del] were incubated in 10 mM phosphate buffer for 1 hour at 4OC with various concentrations of TFE, and the α-helical content was determined from far-UV CD spectra. Data are means + SD, n=3. Letters above the data points indicate significance of the difference between the values for apoA-I[K107del] and WT by unpaired t-test, a: not significant, b: P < 0.05, c: P<0.01.

      Discussion

      Despite many published studies on human apoA-I[K107del] variant that is associated with low HDL-C levels (
      • Haase C.L.
      • Frikke-Schmidt R.
      • Nordestgaard B.G.
      • Tybjærg-Hansen A.
      Population-Based Resequencing of APOA1 in 10,330 Individuals: Spectrum of Genetic Variation, Phenotype, and Comparison with Extreme Phenotype Approach.
      ,
      • Tilly-Kiesi M.
      • Lichtenstein A.H.
      • Ordovas J.M.
      • Dolnikowski G.
      • Malmström R.
      • Taskinen M.-R.
      • Schaefer E.J.
      Subjects With ApoA-I(Lys107→0) Exhibit Enhanced Fractional Catabolic Rate of ApoA-I in Lp(AI) and ApoA-II in Lp(AI With AII).
      ,
      • Nofer J.R.
      • von Eckardstein A.
      • Wiebusch H.
      • Weng W.
      • Funke H.
      • Schulte H.
      • Köhler E.
      • Assmann G.
      Screening for naturally occurring apolipoprotein A-I variants: apo A-I(delta K107) is associated with low HDL-cholesterol levels in men but not in women.
      ,
      • Tilly-Kiesi M.
      • Zhang Q.
      • Ehnholm S.
      • Kahri J.
      • Lahdenperä S.
      • Ehnholm C.
      • Taskinen M.-R.
      ApoA-IHelsinki (Lys107-->0) associated with reduced HDL cholesterol and LpA-I: A-II deficiency.
      ) and premature CVD (
      • Amarzguioui M.
      • Mucchiano G.
      • Häggqvist B.
      • Westermark P.
      • Kavlie A.
      • Sletten K.
      • Prydz H.
      Extensive intimal apolipoprotein A1-derived amyloid deposits in a patient with an apolipoprotein A1 mutation.
      ,
      • Tilly-Kiesi M.
      • Zhang Q.
      • Ehnholm S.
      • Kahri J.
      • Lahdenperä S.
      • Ehnholm C.
      • Taskinen M.-R.
      ApoA-IHelsinki (Lys107-->0) associated with reduced HDL cholesterol and LpA-I: A-II deficiency.
      ), its properties are not fully understood. Some conclusions regarding the effects of K107 deletion on apoA-I properties, such as LCAT activation (
      • Huang W.
      • Matsunaga A.
      • Li W.
      • Han H.
      • Hoang A.
      • Kugi M.
      • Koga T.
      • Sviridov D.
      • Fidge N.
      • Sasaki J.
      Recombinant proapoA-I(Lys107del) shows impaired lipid binding associated with reduced binding to plasma high density lipoprotein.
      ,
      • Tilly-Kiesi M.
      • Zhang Q.
      • Ehnholm S.
      • Kahri J.
      • Lahdenperä S.
      • Ehnholm C.
      • Taskinen M.-R.
      ApoA-IHelsinki (Lys107-->0) associated with reduced HDL cholesterol and LpA-I: A-II deficiency.
      ,
      • Rall Jr., S.C.
      • Weisgraber K.H.
      • Mahley R.W.
      • Ogawa Y.
      • Fielding C.J.
      • Utermann G.
      • Haas J.
      • Steinmetz A.
      • Menzel H.J.
      • Assmann G.
      Abnormal lecithin:cholesterol acyltransferase activation by a human apolipoprotein A-I variant in which a single lysine residue is deleted.
      ,
      • Jonas A.
      • von Eckardstein A.
      • Churgay L.
      • Mantulin W.W.
      • Assmann G.
      Structural and functional properties of natural and chemical variants of apolipoprotein A-I.
      ), DMPC solubilization rate (
      • Gorshkova I.N.
      • Mei X.
      • Atkinson D.
      Binding of human apoA-I[K107del] variant to TG-rich particles: implications for mechanisms underlying hypertriglyceridemia.
      ,
      • Huang W.
      • Matsunaga A.
      • Li W.
      • Han H.
      • Hoang A.
      • Kugi M.
      • Koga T.
      • Sviridov D.
      • Fidge N.
      • Sasaki J.
      Recombinant proapoA-I(Lys107del) shows impaired lipid binding associated with reduced binding to plasma high density lipoprotein.
      ,
      • Ludovico I.D.
      • Gisonno R.A.
      • Gonzalez M.C.
      • Garda H.A.
      • Ramella N.A.
      • Tricerri M.A.
      Understanding the role of apolipoproteinA-I in atherosclerosis. Post-translational modifications synergize dysfunction?.
      ), or aggregation in solution (
      • Ramella N.A.
      • Schinella G.R.
      • Ferreira S.T.
      • Prieto E.D.
      • Vela M.E.
      • Ríos J.L.
      • Tricerri M.A.
      • Rimoldi O.J.
      Human apolipoprotein A-I natural variants: molecular mechanisms underlying amyloidogenic propensity.
      ,
      • Ludovico I.D.
      • Bedi S.
      • Melchior J.T.
      • Gonzalez M.
      • Garda H.
      • Davidson S.
      Deletion of Lys 107 Modify the Ability of ApoA-I to Self-associate in Solution.
      ), are inconsistent. One of the unusual findings about apoA-I[K107del] is that it is not associated with low plasma apoA-I concentrations, despite being associated with low HDL-C (
      • Haase C.L.
      • Frikke-Schmidt R.
      • Nordestgaard B.G.
      • Tybjærg-Hansen A.
      Population-Based Resequencing of APOA1 in 10,330 Individuals: Spectrum of Genetic Variation, Phenotype, and Comparison with Extreme Phenotype Approach.
      ,
      • Tilly-Kiesi M.
      • Lichtenstein A.H.
      • Ordovas J.M.
      • Dolnikowski G.
      • Malmström R.
      • Taskinen M.-R.
      • Schaefer E.J.
      Subjects With ApoA-I(Lys107→0) Exhibit Enhanced Fractional Catabolic Rate of ApoA-I in Lp(AI) and ApoA-II in Lp(AI With AII).
      ). This phenotype contrasts with that of most human apoA-I mutations that are associated with either both reduced plasma HDL-C and reduced apoA-I levels (
      • Sorci-Thomas M.G.
      • Thomas M.J.
      The effects of altered apolipoprotein A-I structure on plasma HDL concentration.
      ,
      • Haase C.L.
      • Frikke-Schmidt R.
      • Nordestgaard B.G.
      • Tybjærg-Hansen A.
      Population-Based Resequencing of APOA1 in 10,330 Individuals: Spectrum of Genetic Variation, Phenotype, and Comparison with Extreme Phenotype Approach.
      ,
      • Haase C.L.
      • Tybjærg-Hansen A.
      • Grande P.
      • Frikke-Schmidt R.
      Genetically elevated apolipoprotein A-I, high-density lipoprotein cholesterol levels, and risk of ischemic heart disease.
      ,
      • Weisgraber K.H.
      • Rall S.C.
      • Bersot T.P.
      • Mahley R.W.
      • Franceschini G.,
      • Sirtori C.R.
      Apolipoprotein A-I milano - detection of normal A-I in affected subjects and evidence for a cysteine for arginine substitution in the variant A-I.
      ,
      • Anthanont P.
      • Polisecki E.
      • Asztalos B.F.
      • Diffenderfer M.R.
      • Barrett P.H.
      • Millar J.S.
      • Billheimer J.
      • Cuchel M.
      • Rader D.J.
      • Schaefer E.J.
      A novel ApoA-I truncation (ApoA-I Mytilene) associated with decreased ApoA-I production.
      ), or elevated HDL-C along with elevated plasma apoA-I levels (
      • Haase C.L.
      • Tybjærg-Hansen A.
      • Grande P.
      • Frikke-Schmidt R.
      Genetically elevated apolipoprotein A-I, high-density lipoprotein cholesterol levels, and risk of ischemic heart disease.
      ). In this work, we revealed important properties of apoA-I[K107del] that may contribute to its unusual phenotype.
      Our earlier studies on binding of apoA-I to synthetic TG-rich emulsion particles (
      • Gorshkova I.N.
      • Mei X.
      • Atkinson D.
      Binding of human apoA-I[K107del] variant to TG-rich particles: implications for mechanisms underlying hypertriglyceridemia.
      ) suggested that the K107del mutation may enhance binding of apoA-I to large TG-rich lipoproteins in vivo. However, no other published experimental results on this variant or data on patients carrying this mutation that would directly support this hypothesis were known. In the current study, we studied for the first time lipid surface behavior of apoA-I[K107del] and revealed properties of apoA-I[K107del] that can drive increase affinity of this variant to large TG-rich lipoproteins. We found that the K107 deletion led to increases in both the extent and the rate of adsorption of apoA-I to the surfaces of POPC-coated TO drops and increased retention of apoA-I[K107del] on the surfaces. The differences between apoA-I[K107del] and WT in adsorption (Figure 1) and especially, desorption (Figure 3B) behaviors were more pronounced for the more polar POPC-TO surfaces than for the surface of a TO drop without POPC. The higher ΠENV values for apoA-I[K107del] on the surface of POPC-coated TO drops (Figure 3B) imply that in vivo, during lipoprotein remodeling and inferred lipoprotein surface pressure changes, apoA-I[K107del] remains on the surfaces of TG-rich particles up to higher pressures than WT. The larger surface pressure changes that resulted from adsorption of apoA-I[K107del] to the surfaces of POPC-coated TO drop (Figure 2) indicate that apoA-I[K107del] remodels the lipid surfaces to a greater extent than WT, reflective of stronger interactions of the variant with the surface lipids of TG-rich lipoproteins. Taken together, these observations imply that enhanced binding of apoA-I[K107del] to the phospholipid surfaces of TG-rich lipoproteins and its better retention on these surfaces may lead to an increased content of apoA-I[K107del] on TG-rich lipoproteins in vivo.
      It has been shown that compared to WT, apoA-I[K107del] has a less stable and more loosely folded lipid-free conformation with greater exposure of hydrophobic surfaces (
      • Gorshkova I.N.
      • Mei X.
      • Atkinson D.
      Binding of human apoA-I[K107del] variant to TG-rich particles: implications for mechanisms underlying hypertriglyceridemia.
      ,
      • Ramella N.A.
      • Schinella G.R.
      • Ferreira S.T.
      • Prieto E.D.
      • Vela M.E.
      • Ríos J.L.
      • Tricerri M.A.
      • Rimoldi O.J.
      Human apolipoprotein A-I natural variants: molecular mechanisms underlying amyloidogenic propensity.
      ). It has been proposed that lipid-free apoA-I forms a four-segment bundle that opens when the protein binds to lipids or lipoprotein lipid surfaces [(
      • Phillips M.C.
      New insights into the determination of HDL structure by apolipoproteins: Thematic Review Series: High Density Lipoprotein Structure, Function, and Metabolism.
      ) and references sited therein]. In our earlier studies, we suggested that the K107 deletion results in helical registry shift and ensuing disruption of salt bridges, both intra-helical (K107-E110 and K107-E111) and inter-helical (E111-H155 and E111-R151) that leads to destabilization of the N-terminal helical bundle structure and greater exposure of hydrophobic surfaces in lipid-free apoA-I (
      • Gorshkova I.N.
      • Mei X.
      • Atkinson D.
      Binding of human apoA-I[K107del] variant to TG-rich particles: implications for mechanisms underlying hypertriglyceridemia.
      ). These characteristics are consistent with the position and interactions of K107 in the crystal structure of the [1-184]apoA-I dimer, as well as the proposed “domain swapped” monomer (
      • Mei X.
      • Atkinson D.
      Crystal structure of C-terminal truncated apolipoprotein A-I reveals the assembly of high density lipoprotein (HDL) by dimerization.
      ) and the consensus structure of lipid-free apoA-I (
      • Melchior J.T.
      • Walker R.G.
      • Cooke A.L.
      • Morris J.
      • Castleberry M.
      • Thompson T.B.
      • Jones M.K.
      • Song H.D.
      • Rye K.A.
      • Oda M.N.
      • Sorci-Thomas M.G.
      • Thomas M.J.
      • Heinecke J.W.
      • Mei X.
      • Atkinson D.
      • Segrest J.P.
      • Lund-Katz S.
      • Phillips M.C.
      • Davidson W.S.
      A consensus model of human apolipoprotein A-I in its monomeric and lipid-free state.
      ). Specifically, in all three structural models, K107 forms a strong intra-helical salt bridge with E110 and E111 located in helix 4 that is a part of the N-terminal helix bundle. In addition, as a consequence of the loss of the K107/E111 interaction inter-helical salt bridges (E111-H155 and E111-R151) that stabilize the interaction between helix 4 and helix 6 in the bundle are also likely to be perturbed. Thus, deletion of K107 is expected shift the helical registry, disrupts these salt bridge interactions, and thereby, leads to destabilization of the helix bundle and exposure of the hydrophobic core of the bundle. The lower stability and less stable folded conformation leading to the exposed hydrophobic surface would provide flexibility and adaptability for conformational changes that are required for protein binding to the surface of large TG-rich lipoprotein particles (
      • Gorshkova I.N.
      • Mei X.
      • Atkinson D.
      Binding of human apoA-I[K107del] variant to TG-rich particles: implications for mechanisms underlying hypertriglyceridemia.
      ). These processes represent the structural basis of the enhanced binding of apoA-I[K107del] to the surfaces of TG-rich lipoproteins (or lipid drops in the drop tensiometry experiments). It should be noted that the general structural organization of apoA-I on TG rich particles must be different from that on discoidal or spherical HDL particles. Given that sizes of TG-rich particles are much larger than HDL (50-100 nm for VLDL versus 7-12 nm for HDL), and the relative content of apoA-I on TG-rich particles is expected to be immensely lower than on HDL. ApoA-I is not able to form cage-like structures on TG-rich particles. Furthermore, the lipid drops used in our drop-tensiometry experiments are much larger than VLDL.
      Remarkably, animal studies showed that when human WT or variant forms of apoA-I were expressed in apoA-I -/- mice, apoA-I mutations that led to similar conformational changes (less stable and more loose folding with greater exposure of hydrophobic surfaces) resulted in a noticeable portion of apoA-I detected in plasma TG-rich lipoproteins of the animals (
      • Chroni A.
      • Kan H.Y.
      • Kypreos K.E.
      • Gorshkova I.N.
      • Shkodrani A.
      • Zannis V.I.
      Substitutions of Glu110 and Glu111 in the middle helix 4 of human ApoA-I by Alanine affect the structure and in vitro functions of apoA-I and induce severe hypertriglyceridemia in ApoA-I-deficient mice.
      ,
      • Kateifides A.K.
      • Gorshkova I.N.
      • Duka A.
      • Chroni A.
      • Kardassis D.
      • Zannis V.I.
      Alteration of negatively charged residues in the 89 and 96 domain of apoA-I affects lipid homeostasis and maturation of HDL..
      , ref.
      • Gorshkova I.N.
      • Atkinson D.
      Enhanced Binding of Apolipoproteins A-I Variants Associated with Hypertriglyceridemia to Triglyceride-Rich Particles.
      and references therein). In one study (
      • Kateifides A.K.
      • Gorshkova I.N.
      • Duka A.
      • Chroni A.
      • Kardassis D.
      • Zannis V.I.
      Alteration of negatively charged residues in the 89 and 96 domain of apoA-I affects lipid homeostasis and maturation of HDL..
      ), up to 40% of an apoA-I variant with the similar conformational characteristics were detected in plasma TG-rich lipoproteins.
      Another factor contributing to low HDL-C along with normal apoA-I levels in plasma of individuals with the K107 deletion mutation may relate to the smaller sizes and reduced cholesterol content of their HDL particles, first reported by Tilly-Kiesi et al (
      • Tilly-Kiesi M.
      • Zhang Q.
      • Ehnholm S.
      • Kahri J.
      • Lahdenperä S.
      • Ehnholm C.
      • Taskinen M.-R.
      ApoA-IHelsinki (Lys107-->0) associated with reduced HDL cholesterol and LpA-I: A-II deficiency.
      ). As apoA-I is an activator of LCAT that promotes enlargement of HDL in the circulation, several studies investigated if the K107 deletion affected the ability of apoA-I to activate LCAT. The LCAT activation ability of apoA-I[K107del] isolated from heterozygous individuals was found to be either lower than (
      • Rall Jr., S.C.
      • Weisgraber K.H.
      • Mahley R.W.
      • Ogawa Y.
      • Fielding C.J.
      • Utermann G.
      • Haas J.
      • Steinmetz A.
      • Menzel H.J.
      • Assmann G.
      Abnormal lecithin:cholesterol acyltransferase activation by a human apolipoprotein A-I variant in which a single lysine residue is deleted.
      ,
      • Jonas A.
      • von Eckardstein A.
      • Churgay L.
      • Mantulin W.W.
      • Assmann G.
      Structural and functional properties of natural and chemical variants of apolipoprotein A-I.
      ) or similar to (
      • Tilly-Kiesi M.
      • Zhang Q.
      • Ehnholm S.
      • Kahri J.
      • Lahdenperä S.
      • Ehnholm C.
      • Taskinen M.-R.
      ApoA-IHelsinki (Lys107-->0) associated with reduced HDL cholesterol and LpA-I: A-II deficiency.
      ) that of WT apoA-I isolated from the same individuals. Studies of recombinant pro-apoA-I also did not find any effect of K107 deletion on the protein’s LCAT activation ability (
      • Huang W.
      • Matsunaga A.
      • Li W.
      • Han H.
      • Hoang A.
      • Kugi M.
      • Koga T.
      • Sviridov D.
      • Fidge N.
      • Sasaki J.
      Recombinant proapoA-I(Lys107del) shows impaired lipid binding associated with reduced binding to plasma high density lipoprotein.
      ). Thus, it seems unclear if the reduced HDL size associated with apoA-I[K107del] may relate to impaired cholesterol esterification by LCAT. We posited that structural changes in apoA-I[K107del] may modify the initial step of HDL biogenesis. To test this hypothesis, we investigated the effect of K107del on ABCA1-mediated cholesterol efflux and sizes of nHDL.
      We found no effect of K107 deletion on apoA-I ability to promote ABCA1-mediated cholesterol efflux from J774 cells, in agreement with an earlier reports (
      • Huang W.
      • Matsunaga A.
      • Li W.
      • Han H.
      • Hoang A.
      • Kugi M.
      • Koga T.
      • Sviridov D.
      • Fidge N.
      • Sasaki J.
      Recombinant proapoA-I(Lys107del) shows impaired lipid binding associated with reduced binding to plasma high density lipoprotein.
      ,
      • von Eckardstein A.
      • Castro G.
      • Wybranska I.
      • Theret N.
      • Duchateau P.
      • Duverger N.
      • Fruchart J.-C.
      • Ailhaud
      • Assmann G.
      Interaction of reconstituted high density lipoprotein discs containing human apolipoprotein A-I (apoA-I) variants with murine adipocytes and macrophages.
      ,
      • Gonzalez M.C.
      • Toledo J.D.
      • Tricerri M.A.
      • Garda H.A.
      The central type Y amphipathic a-helices of apolipoprotein AI are involved in the mobilization of intracellular cholesterol depots.
      ) showing that this mutation did not affect net cholesterol efflux from fibroblasts, CHO cells and murine adipocytes. As apoA-I is believed to directly interact with ABCA1 on cell membranes (
      • Liu M.
      • Mei M.
      • Herscovitz H.
      • Atkinson D.
      N-terminal mutation of apoA-I and interaction with ABCA1 reveal mechanisms of nascent HDL biogenesis.
      ,
      • Chroni A.
      • Liu T.
      • Gorshkova I.
      • Kan H.Y.
      • Uehara Y.
      • von Eskardstein A.
      • Zannis V.I.
      The central helices of apoA-I can promote ATP-binding cassette transporter A1 (ABCA1)-mediated lipid efflux. Amino acid residues 220-231 of the wild-type apoA-I are required for lipid-efflux in vitro and high density lipoprotein formation in vivo.
      ), these data suggest that deletion of K107 does not affect binding of apoA-I to ABCA1. Similar levels of ABCA1 expression in the presence of WT or apoA-I[K107del] (Figure 4B) are also consistent with similar interactions of both proteins with ABCA1. It is known that apolipoproteins stabilize ABCA1 in cells by protecting it from proteases (
      • Arakawa R.
      • Yokoyama S.
      Helical apolipoproteins stabilize ATP-binding cassette transporter A1 by protecting it from thiol protease-mediated degradation.
      ). Interestingly, Vedhachalam et al. (
      • Vedhachalam C.
      • Chetty P.S.
      • Nickel M.
      • Dhanasekaran P.
      • Lund-Katz S.
      • Rothblat G.H.
      • Phillips M.C.
      Influence of apolipoprotein (Apo) A-I structure on nascent high density lipoprotein (HDL) particle size distribution.
      ) proposed that hydrophobicity of the C-terminal domain of apoA-I is critical for effective ABCA1-mediated cholesterol efflux in agreement with the concept that the C-terminal repeat is the first lipid interacting domain (
      • Liu M.
      • Mei M.
      • Herscovitz H.
      • Atkinson D.
      N-terminal mutation of apoA-I and interaction with ABCA1 reveal mechanisms of nascent HDL biogenesis.
      ), while destabilization of the N-terminal bundle may increase the effectiveness, and a cooperation of the N-terminal and C-terminal domains enhance cholesterol efflux to apoA-I via ABCA1. It is possible that the intact C-terminal domain and cooperation between the N- and C-terminal domains preserved in apoA-I[K107del] override the potentially enhanced ability of the destabilized N-terminal bundle to promote cholesterol efflux. These observations are similar to those for the L38G/K40G mutation of apoA-I that we studied previously (
      • Liu M.
      • Mei M.
      • Herscovitz H.
      • Atkinson D.
      N-terminal mutation of apoA-I and interaction with ABCA1 reveal mechanisms of nascent HDL biogenesis.
      ). In both cases, the mutation did not result in changes in ABCA1-mediated cholesterol efflux from cells but had a marked effect on nHDL biogenesis, supporting our suggestion that efflux and nHDL formation are uncoupled processes (
      • Liu M.
      • Mei M.
      • Herscovitz H.
      • Atkinson D.
      N-terminal mutation of apoA-I and interaction with ABCA1 reveal mechanisms of nascent HDL biogenesis.
      ). The L38G/K40G mutation, designed to destabilize a hinge region but have little effect on the helical backbone, resulted in a significantly enhanced ability to form nHDL, which suggests that a destabilized N-terminal bundle facilitates nHDL formation. In contrast, despite no effect of on ABCA1-mediated cholesterol efflux, the K107 deletion mutation modulated the size distribution of nHDL particles, resulting in a reduced abundance of larger particles and an increased proportion of smaller particles. While the relative level of large + very large nHDL particles formed by the mutant is higher after a 24-h incubation compared to 6-h incubation (e.g. 30.4 and 63.4% of control level, respectively), this population is still significantly lower than with the WT (P<0.002). Furthermore, representative blots (Figures 4C and 4D) demonstrate that, in particular, the largest nHDL (with diameter >10 nm) were formed by WT, but not by apoA-I[K107del]. The consequences of the delayed formation of large and very large nHDL particles in vivo may be profound as discussed below.
      In our efflux experiments after a 6-h incubation, 38.5% of mutant apoA-I remained lipid-free. As a consequence of its tendency to bind to TG-rich lipoproteins combined with its reduced ability to form large nHDL, it is likely that under physiological conditions, a fraction of the lipid-free mutant binds to VLDL thus reducing the available lipid-free pool to continue to form large and very large nHDL particles leading to a further reduction in this population.
      Importantly, it was demonstrated that the relative amount of cholesterol varies with the size of HDL particles. Mendivil et al showed that the ratio of cholesterol to apoA-I increases with HDL size in humans (
      • Mendivil C.O.
      • Furtado J.
      • Morton A.M.
      • Wang L.
      • Sacks F.M.
      Novel pathways of apolipoprotein A-I metabolism in HDL of different sizes in humans.
      ). Similarly, Liu et al. reported that the cholesterol:phospholipid ratio in the larger nascent HDL particles formed by apoA-I following incubation with J774 cells is 3 times higher compared to the smaller particles (
      • Liu L.
      • Bortnick A.E.
      • Nickel M.
      • Dhanasekaran P.
      • Subbaiah P.V.
      • Lund-Katz S.
      • Rothblat G.H.
      • Phillips M.C.
      Effects of apolipoprotein A-I on ATP-binding cassette transporter A1-mediated efflux of macrophage phospholipid and cholesterol: formation of nascent high density lipoprotein particles.
      ). Therefore, in our system, even after the prolonged incubation of 24 h, the 36% reduction in the relative abundance of large + very large nHDL translates into a much larger reduction in total HDL-cholesterol. Since very small and medium nHDL particles are enriched in phospholipids, their relative increase (Figure 4 E and F) cannot compensate for the reduction in total cholesterol, thus providing an explanation for the low HDL-C observed in patients’ plasma (
      • Tilly-Kiesi M.
      • Zhang Q.
      • Ehnholm S.
      • Kahri J.
      • Lahdenperä S.
      • Ehnholm C.
      • Taskinen M.-R.
      ApoA-IHelsinki (Lys107-->0) associated with reduced HDL cholesterol and LpA-I: A-II deficiency.
      ).
      Not only are the larger HDL particles formed more slowly by the mutant but they are also probably less stable than WT-HDL presumably due to the inability to form the salt bridge stabilizing interactions between apoA-I molecules. These properties are consistent with the faster catabolic rate of HDL observed in patients carrying the mutation (
      • Tilly-Kiesi M.
      • Lichtenstein A.H.
      • Ordovas J.M.
      • Dolnikowski G.
      • Malmström R.
      • Taskinen M.-R.
      • Schaefer E.J.
      Subjects With ApoA-I(Lys107→0) Exhibit Enhanced Fractional Catabolic Rate of ApoA-I in Lp(AI) and ApoA-II in Lp(AI With AII).
      ).
      Finally, since the mutant promoted the same efflux as WT but was unable to bind the same amount of effluxed lipids, it is expected that RCT will be impaired and that the unbound cholesterol may be deposited in the arterial wall. In contrast, patients that carry apoA-I variants that cause amyloidosis appear to be protected from CVD despite lower affinity for lipids and biogenesis of small HDL particles (
      • Nilsson O.
      • Lindvall M.
      • Obici L.
      • Ekström S.
      • Lagerstedt J.O.
      • Del Giudice R.J.
      Structure dynamics of ApoA-I amyloidogenic variants in small HDL increase their ability to mediate cholesterol efflux.
      ). It appears that compared to WT, these variants have both increased capacity to promote efflux and increased biogenesis of small HDL thus, allowing for more efficient RCT. apoA-I[K107del] on the other hand, has similar capacity to WT to promote efflux coupled with less efficient binding of the effluxed lipids and therefore, is unable to compensate for the lack of larger HDL particles biogenesis leading to less efficient RCT.
      Together these data provide an explanation for low HDL-C and a shift in the size of HDL from large (HDL2) to small (HDL3) particles (
      • Tilly-Kiesi M.
      • Zhang Q.
      • Ehnholm S.
      • Kahri J.
      • Lahdenperä S.
      • Ehnholm C.
      • Taskinen M.-R.
      ApoA-IHelsinki (Lys107-->0) associated with reduced HDL cholesterol and LpA-I: A-II deficiency.
      ). Both changes increase the risk of atherosclerosis and cardiovascular disease.
      Thus, apoA-I[K107del] exerts the similar effect on the size of HDL particles, whether they are spherical plasma HDL or discoidal particles created by ABCA1 reaction or by solubilization of DMPC vesicles.
      .In contrast to apoAI binding to large TG rich particles, an important structural self-association of apoA-I molecules is required when apoA-I binds to HDL, or solubilizes lipids on the surface of ABCA1-expressing cells to form discoidal nHDL or solubilizes phospholipid vesicles to form discoidal phospholipid-apoA-I complexes,. ApoA-I must dimerize to form “double-belt” structures around the lipid bilayer of the nHDL discs. The general structural organization of apoA-I is thought to be similar overall on discoidal and spherical HDL particles, with the “double-belt” being the fundamental organizational motif for apoA-I (
      • Silva R.A.
      • Huang R.
      • Morris J.
      • Fang J.
      • Gracheva E.O.
      • Ren G.
      • Kontush A.
      • Jerome W.G.
      • Rye K.A.
      • Davidson W.S.
      Structure of apolipoprotein A-I in spherical high density lipoproteins of different sizes.
      ,
      • Huang R.
      • Silva R.A.
      • Jerome W.G.
      • Kontush A.
      • Chapman M.J.
      • Curtiss L.K.
      • Hodges T.J.
      • Davidson W.S.
      Apolipoprotein A-I structural organization in high-density lipoproteins isolated from human plasma.
      ). In discoidal particles, the “double-belt”, comprised of two antiparallel apoA-I molecules, wraps around the edge of the discs (
      • Segrest J.P.
      • Jones M.K.
      • Klon A.E.
      • Sheldahl C.J.
      • Hellinger M.
      • De Loof H.
      • Harvey S.C.
      A detailed molecular belt model for apolipoprotein A-I in discoidal high density lipoprotein.
      ), and in spherical particles, 3 to 5 “double-belted” apoA-I molecules form a symmetrical cage-like structure around the spheres (
      • Silva R.A.
      • Huang R.
      • Morris J.
      • Fang J.
      • Gracheva E.O.
      • Ren G.
      • Kontush A.
      • Jerome W.G.
      • Rye K.A.
      • Davidson W.S.
      Structure of apolipoprotein A-I in spherical high density lipoproteins of different sizes.
      ,
      • Huang R.
      • Silva R.A.
      • Jerome W.G.
      • Kontush A.
      • Chapman M.J.
      • Curtiss L.K.
      • Hodges T.J.
      • Davidson W.S.
      Apolipoprotein A-I structural organization in high-density lipoproteins isolated from human plasma.
      ). Some data suggest that the presence of a third apoA-I molecule in a ‘hairpin” conformation cannot be ruled out for discs with diameters exceeding 10 nm (
      • Phillips M.C.
      New insights into the determination of HDL structure by apolipoproteins: Thematic Review Series: High Density Lipoprotein Structure, Function, and Metabolism.
      ,
      • Li L.
      • Chen J.
      • Mishra V.K.
      • Kurtz J.A.
      • Cao D.
      • Klon A.E.
      • Harvey S.C.
      • Anantharamaiah G.M.
      • Segrest J.P.
      Double belt structure of discoidal high density lipoproteins: molecular basis for size heterogeneity.
      ,
      • Davidson W.S.
      • Silva R.A.
      Apolipoprotein structural organization in high density lipoproteins: belts, bundles, hinges and hairpins.
      ).
      It is well documented that the sequence of apoA-I contains ten, potentially amphipathic helical repeats in the exon 4 encoded region (residues 44-243) that form this double belt together with the N-terminal (residues 1-43) domain. In the “double-belt” arrangement, the polar surface of apoA-I amphipathic helices faces the aqueous environment, while the non-polar face of the amphipathic helices interacts with the hydrophobic regions of phospholipids, shielding them from exposure to water and thereby stabilizing the HDL particle (
      • Segrest J.P.
      • Jones M.K.
      • Klon A.E.
      • Sheldahl C.J.
      • Hellinger M.
      • De Loof H.
      • Harvey S.C.
      A detailed molecular belt model for apolipoprotein A-I in discoidal high density lipoprotein.
      ). Segments of the apoA-I molecules forming these “double belts” may have loop or unstructured character in some of the repeat segments of the apoA-I molecules that are displaced from the particle surface or twisted to modulate particle diameter in response to various amount of lipid cargo (
      • Silva R.A.
      • Huang R.
      • Morris J.
      • Fang J.
      • Gracheva E.O.
      • Ren G.
      • Kontush A.
      • Jerome W.G.
      • Rye K.A.
      • Davidson W.S.
      Structure of apolipoprotein A-I in spherical high density lipoproteins of different sizes.
      ,
      • Li L.
      • Chen J.
      • Mishra V.K.
      • Kurtz J.A.
      • Cao D.
      • Klon A.E.
      • Harvey S.C.
      • Anantharamaiah G.M.
      • Segrest J.P.
      Double belt structure of discoidal high density lipoproteins: molecular basis for size heterogeneity.
      ,
      • Davidson W.S.
      • Silva R.A.
      Apolipoprotein structural organization in high density lipoproteins: belts, bundles, hinges and hairpins.
      ,
      • Chetty P.S.
      • Mayne L.
      • Kan Z.-Y.
      • Lund-Katz S.
      • Englander S.W.
      • Phillips M.C.
      Apolipoprotein A-I helical structure and stability in discoidal high-density lipoprotein (HDL) particles by hydrogen exchange and mass spectrometry.
      ). Thus, the formation of the double belt to stabilize smaller sized discoidal HDL particles does not require all ten repeat segments to form helical conformation. As particles increase in size, unstructured segments of the sequence repeats are driven to form in-register amphipathic helical structure to expand the anti-parallel double-belt and stabilize the increased perimeter of the particle. The α-helical content of apoA-I on both discoidal and spherical synthetic HDL is known to increase when the lipid:apoA-I ratio and diameter of the particles increases (
      • Sparks D.L.
      • Lund-Katz S.
      • Phillips M.C.
      The charge and structural stability of apolipoprotein A-I in discoidal and spherical recombinant high density lipoprotein particles.
      ). This observation is supported by hydrogen exchange and mass spectroscopy analysis of discoidal particles (
      • Chetty P.S.
      • Mayne L.
      • Kan Z.-Y.
      • Lund-Katz S.
      • Englander S.W.
      • Phillips M.C.
      Apolipoprotein A-I helical structure and stability in discoidal high-density lipoprotein (HDL) particles by hydrogen exchange and mass spectrometry.
      ). Our analysis of the α-helical content of WT and apoA-I[K107del] with increasing concentrations of DMPC or TFE reveals that in both environments, apoA-I[K107del] was not able to increase α-helical structure as much as WT (Figures 6 and 7))
      Most importantly, stabilizing interactions between the two apoA-I molecules forming the double belt are achieved through extensive salt bridges between the antiparallel helices. As discussed earlier (
      • Gorshkova I.N.
      • Mei X.
      • Atkinson D.
      Binding of human apoA-I[K107del] variant to TG-rich particles: implications for mechanisms underlying hypertriglyceridemia.
      ), in addition to residue K107 forming an important stabilizing intra-helical salt-bridge (K107- E111) in the structure of WT apoA-I, K107 together with several other residues is involved in critical salt-bridge networks that stabilize the helix 4 - helix 6 interaction of the apposed molecules in the double belt. Thus, overall, the deletion of K107, together with the resulting disruption of the registry of the polar and apolar helical faces, disrupts and destabilizes this region with consequent disruption of important inter-helical stabilizing interactions that form the double belt. The hampered ability of apoA-I[K107del] to form correctly structured α-helix at higher lipid:protein ratios impairs the variant’s ability to form an appropriately structured and stabilized “double-belt” around larger HDL, both discoidal and perhaps spherical HDL. In the small nHDL particles, the H4/H6 region may adopt a loop/unstructured conformation instead of helical structure due to lower lipid load. Thus, the apoA-I[K107del] mutant maintains its ability to form small particles but has impaired ability to form larger particles. Our efflux experiments, at both 6 and 24 hours showed that a decreased amount of larger sized nHDL were formed by the K107 mutant. After 24 hours, apoA-I[K107del] demonstrated some ability to form some larger sized nHDL particles predominantly less than 10.4 nm in size possibly driven by continued ABCA1 action but still significantly less compared to WT apoA-I. The detailed structural organization and stability of these particles, and how apoA-I is organized on the particle is unknown but must be different to that of WT.
      Additionally, of important note is that the apposition of helix 4 and helix 6 in the antiparallel double belt forming the nHDL particle is thought to be the interaction site for LCAT. This apposition creates a cluster of charged residues that possibly form the LCAT binding motif. Disruption of helix 4 by the K107 deletion together with changes in the organization of the charged cluster formed between helices 4 and 6 is likely to impact the LCAT interaction resulting in particles that are not only hindered in their formation but are dysfunctional in the important step of cholesterol esterification critical to RCT.
      A suggested possible “hairpin-belt” model of apoA-I organization on larger discoidal particles assumes that the stabilizing salt-bridges in this model are identical to those in the “double-belt” model, except they are intra-helical versus inter-helical for the “hairpin-belt” versus the double belt, respectively (
      • Davidson W.S.
      • Silva R.A.
      Apolipoprotein structural organization in high density lipoproteins: belts, bundles, hinges and hairpins.
      ). Thus, the disruption of the intra- and inter-helical salt bridges in apoA-I resulting from deletion of Lys107 (
      • Gorshkova I.N.
      • Mei X.
      • Atkinson D.
      Binding of human apoA-I[K107del] variant to TG-rich particles: implications for mechanisms underlying hypertriglyceridemia.
      ) would hinder the ability of apoA-I[K107del] to form the “hairpin-belt” structure as well on larger discoidal particles. Since K107del disrupts the intra-helical salt bridges in the vicinity of the deleted residue and the inter-helical salt bridges between registered helices 4 and 6 (residues 99-120 and 143-164, respectively) of the two anti-parallel apoA-I molecules in the “double belt” (or in the same apoA-I molecule in the “hairpin” conformation), but likely does not disrupt salt bridges in the other helical regions of the apoA-I (
      • Gorshkova I.N.
      • Mei X.
      • Atkinson D.
      Binding of human apoA-I[K107del] variant to TG-rich particles: implications for mechanisms underlying hypertriglyceridemia.
      ), apoA-I[K107del] can efficiently form smaller nHDL. It is likely that reduced binding of pro-apoA-I[K107del] to plasma HDL observed by Huang et al. (
      • Huang W.
      • Matsunaga A.
      • Li W.
      • Han H.
      • Hoang A.
      • Kugi M.
      • Koga T.
      • Sviridov D.
      • Fidge N.
      • Sasaki J.
      Recombinant proapoA-I(Lys107del) shows impaired lipid binding associated with reduced binding to plasma high density lipoprotein.
      ) may relate to the impaired ability of the variant to bind to a population of larger HDL particles due the reasons described above.
      In conclusion, the structural changes in apoA-I[K107del] may lead to 1) enhanced binding of the variant to and its better retention on plasma TG-rich lipoproteins and 2) impaired ability of apoA-I to form larger nHDL and stabilize larger spherical HDL. The former may result in an additional pool of plasma apoA-I on TG-rich lipoproteins that can contribute to normal plasma apoA-I levels, even with reduced HDL-bound apoA-I levels. The latter results in predominantly smaller plasma HDL particles with primarily reduced protein ratio and therefore, may contribute to the reduced plasma HDL-C levels along with normal apoA-I levels. In addition, the shift in the HDL size distribution resulting in a lower abundance of large HDL particles may lead to compromised antiatherogenic properties of HDL that are apparently influenced by particle size (reviewed in ref. 5).

      Conflict of interest

      The authors state no conflict of interest

      Acknowledgements

      The authors are grateful to Dr. Anne Tybjærg-Hansen of the Copenhagen University Hospital and the University of Copenhagen, Denmark, for helpful communications on epidemiological findings. We thank Dr. Olga Gursky for allowing us to use the instrumentation in her lab and Dr. Shobini Jayaraman for help with chromatography and data analysis. We are indebted to Dr. Donald M. Small (1931-2019) who was one of the inventors of oil-drop tensiometry, advisor to N.L. Meyers, and our teacher and colleague.

      Supplementary Data

      References

        • Phillips M.C.
        New insights into the determination of HDL structure by apolipoproteins: Thematic Review Series: High Density Lipoprotein Structure, Function, and Metabolism.
        J. Lipid Res. 2013; 54: 2034-2048
        • Dominiczak M.H.
        • Caslake M.J.
        Apolipoproteins: metabolic role and clinical biochemistry applications.
        Ann Clin Biochem. 2011; 48: 498-515
        • Sharrett A.R.
        • Ballantyne C.M.
        • Coady S.A.
        • Heiss G.
        • Sorlie P.D.
        • Catellier D.
        • Patsch W.
        Coronary heart disease prediction from lipoprotein cholesterol levels, triglycerides, lipoprotein(a), apolipoproteins A-I and B, and HDL density subfractions: The Atherosclerosis Risk in Communities (ARIC) Study.
        Circulation. 2001; 104: 1108-1113
        • Rader D.J.
        • Tall A.R.
        The not-so-simple HDL story: is it time to revise the HDL cholesterol hypothesis?.
        Nat. Med. 2012; 18: 1344-1346
        • Rye K.A.
        • Barter P.J.
        Cardioprotective functions of HDLs.
        J. Lipid Res. 2014; 55: 168-179
        • Pownall H.J.
        • Rosales C.
        • Gillard B.K.
        • Gotto Antonio M.
        • M Jr., A.
        High-density lipoproteins, reverse cholesterol transport and atherogenesis.
        Nat. Rev. Cardiol. 2021; 18: 712-723
        • Rosenson R.S.
        • Brewer Jr., H.B.
        • Davidson W.S.
        • Fayad Z.A.
        • Fuster V.
        • Goldstein J.
        • Hellerstein M.
        • Jiang X.C.
        • Phillips M.C.
        • Rader D.J.
        • Remaley A.T.
        • Rothblat G.H.
        • Tall A.R.
        • Yvan-Charvet L.
        Cholesterol efflux and atheroprotection: advancing the concept of reverse cholesterol transport.
        Circulation. 2012; 125: 1905-1919
        • Sorci-Thomas M.G.
        • Thomas M.J.
        The effects of altered apolipoprotein A-I structure on plasma HDL concentration.
        Trends Cardiovasc. Med. 2002; 12: 121-128
        • Asztalos B.F.
        • Demissie S.
        • Cupples L.A.
        • Collins D.
        • Cox C.E.
        • Horvath K.V.
        • Bloomfield H.E.
        • Robins S.J.
        • Schaefer E.J.
        LpA-I, LpA-I: A-II HDL and CHD-risk: The Framingham Offspring Study and the Veterans Affairs HDL Intervention Trial.
        Atherosclerosis. 2006; 188: 59-67
        • Haase C.L.
        • Frikke-Schmidt R.
        • Nordestgaard B.G.
        • Tybjærg-Hansen A.
        Population-Based Resequencing of APOA1 in 10,330 Individuals: Spectrum of Genetic Variation, Phenotype, and Comparison with Extreme Phenotype Approach.
        PLoS Genet. 2012; 8: e1003063
        • Tilly-Kiesi M.
        • Lichtenstein A.H.
        • Ordovas J.M.
        • Dolnikowski G.
        • Malmström R.
        • Taskinen M.-R.
        • Schaefer E.J.
        Subjects With ApoA-I(Lys107→0) Exhibit Enhanced Fractional Catabolic Rate of ApoA-I in Lp(AI) and ApoA-II in Lp(AI With AII).
        Atherosclerosis Thrombosis and Vascular. Biology. 1997; 17: 873-880
        • Nofer J.R.
        • von Eckardstein A.
        • Wiebusch H.
        • Weng W.
        • Funke H.
        • Schulte H.
        • Köhler E.
        • Assmann G.
        Screening for naturally occurring apolipoprotein A-I variants: apo A-I(delta K107) is associated with low HDL-cholesterol levels in men but not in women.
        Hum. Genet. 1995; 96: 177-182
        • Amarzguioui M.
        • Mucchiano G.
        • Häggqvist B.
        • Westermark P.
        • Kavlie A.
        • Sletten K.
        • Prydz H.
        Extensive intimal apolipoprotein A1-derived amyloid deposits in a patient with an apolipoprotein A1 mutation.
        Biochem. Biophys. Res. Commun. 1998; 242: 534-539
        • Rowczenio D.
        • Dogan A.
        • Theis J.D.
        • Vrana J.A.
        • Lachmann H.J.
        • Wechalekar A.D.
        • Gilbertson J.A.
        • Hunt T.
        • Gibbs S.D.
        • Sattianayagam P.T.
        • Pinney J.H.
        • Hawkins P.N.
        • Gillmore J.,D.
        Amyloidogenicity and clinical phenotype associated with five novel mutations in apolipoprotein A-I.
        Am J Pathol. 2011; 179: 1978-1987
        • Gorshkova I.N.
        • Mei X.
        • Atkinson D.
        Binding of human apoA-I[K107del] variant to TG-rich particles: implications for mechanisms underlying hypertriglyceridemia.
        J. Lipid Res. 2014; 55: 1876-1885
        • Ramella N.A.
        • Schinella G.R.
        • Ferreira S.T.
        • Prieto E.D.
        • Vela M.E.
        • Ríos J.L.
        • Tricerri M.A.
        • Rimoldi O.J.
        Human apolipoprotein A-I natural variants: molecular mechanisms underlying amyloidogenic propensity.
        PLoS One. 2012; 7: e43755
        • Huang W.
        • Matsunaga A.
        • Li W.
        • Han H.
        • Hoang A.
        • Kugi M.
        • Koga T.
        • Sviridov D.
        • Fidge N.
        • Sasaki J.
        Recombinant proapoA-I(Lys107del) shows impaired lipid binding associated with reduced binding to plasma high density lipoprotein.
        Atherosclerosis. 2001; 159: 85-91
        • Tilly-Kiesi M.
        • Zhang Q.
        • Ehnholm S.
        • Kahri J.
        • Lahdenperä S.
        • Ehnholm C.
        • Taskinen M.-R.
        ApoA-IHelsinki (Lys107-->0) associated with reduced HDL cholesterol and LpA-I: A-II deficiency.
        Arterioscler. Thromb. Vasc. Biol. 1995; 15: 294-306
        • Tilly-Kiesi M.
        • Packard C.J.
        • Kahrij J.
        • Ehnholm C.
        • Shepherd J.
        • Taskinen M.R.
        In vivo metabolism of apoA-I and apo A-II in subjects with apo A-I(Lys107-->0) associated with reduced HDL cholesterol and Lp(AI w AII) deficiency.
        Atherosclerosis. 1997; 128: 213-222
        • Curtiss L.K.
        • Valenta D.T.
        • Hime N.J.
        • Rye K.A.
        What Is So Special About Apolipoprotein AI in Reverse Cholesterol Transport?.
        Arterioscler. Thromb. Vasc. Biol. 2006; 26: 12-19
        • Wang L.
        • Atkinson D.
        • Small D.M.
        The interfacial properties of ApoA-I and an amphipathic alpha-helix consensus peptide of exchangeable apolipoproteins at the triolein/water interface.
        J. Biol. Chem. 2005; 280: 4154-4165
        • Meyers N.L.
        • Wang L.
        • Gursky O.
        • Small D.M.
        Changes in helical content or net charge of apolipoprotein C-I alter its affinity for lipid/water interfaces.
        J. Lipid Res. 2013; 54: 1927-1938
        • Wang L.
        • Mei X.
        • Atkinson D.
        • Small D.M.
        Surface behavior of apolipoprotein A-I and its deletion mutants at model lipoprotein interfaces.
        J. Lipid Res. 2014; 55: 478-492
        • Small D.M.
        • Wang L.
        • Mitsche M.A.
        The adsorption of biological peptides and proteins at the oil/water interface. A potentially important but largely unexplored field.
        J. Lipid Res. 2009; 50: S329-S334
        • Meyers N.L.
        • Wang L.
        • Small D.M.
        Apolipoprotein C-I binds more strongly to phospholipid/triolein/water than triolein/water interfaces: a possible model for inhibiting cholesterol ester transfer protein activity and triacylglycerol-rich lipoprotein uptake.
        Biochemistry. 2012; 5: 1238-1248
        • Mitsche M.A.
        • Wang L.
        • Small D.M.
        Adsorption of egg phosphatidylcholine to an air/water and triolein/water bubble interface: use of the 2-dimensional phase rule to estimate the surface composition of a phospholipid/triolein/water surface as a function of surface pressure.
        J. Phys. Chem. B. 2010; 114: 3276-3284
        • von Eckardstein A.
        • Castro G.
        • Wybranska I.
        • Theret N.
        • Duchateau P.
        • Duverger N.
        • Fruchart J.-C.
        • Ailhaud
        • Assmann G.
        Interaction of reconstituted high density lipoprotein discs containing human apolipoprotein A-I (apoA-I) variants with murine adipocytes and macrophages.
        J.Biol. Chem. 1993; 268: 2616-2622
        • Gonzalez M.C.
        • Toledo J.D.
        • Tricerri M.A.
        • Garda H.A.
        The central type Y amphipathic a-helices of apolipoprotein AI are involved in the mobilization of intracellular cholesterol depots.
        Arch Biochem and Biophys. 2008; 473: 34-41
        • Sankaranarayanan S.
        • Kellner-Weibel G.
        • de la Llera-Moya M.
        • Phillips M.C.
        • Asztalos B.F.
        • Bittman R.
        • Rothblat G.H.
        A sensitive assay for ABCA1-mediated cholesterol efflux using BODIPY-cholesterol.
        J. Lipid Res. 2011; 52: 2332-2340
        • Mei X.
        • Liu M.
        • Herscovitz H.
        • Atkinson D.
        Probing the C-terminal domain of lipid-free apoA-I demonstrates the vital role of the H10B sequence repeat in HDL formation.
        J. Lipid Res. 2016; 57: 1507-1517
        • Liu M.
        • Mei M.
        • Herscovitz H.
        • Atkinson D.
        N-terminal mutation of apoA-I and interaction with ABCA1 reveal mechanisms of nascent HDL biogenesis.
        J Lipid Res. 2019; 60: 44-57
        • Duong P.T.
        • Collins H.L.
        • Nickel M.
        • Lund-Katz S.
        • Rothblat G.H.
        • Phillips M.C.
        Characterization of nascent HDL particles and microparticles formed by ABCA1-mediated efflux of cellular lipids to apoA-I.
        J. Lipid Res. 2006; 47: 832-843
        • Ludovico I.D.
        • Gisonno R.A.
        • Gonzalez M.C.
        • Garda H.A.
        • Ramella N.A.
        • Tricerri M.A.
        Understanding the role of apolipoproteinA-I in atherosclerosis. Post-translational modifications synergize dysfunction?.
        Biochim. Biophys. Acta. Gen Subj. 2021; 1865: 129732
        • Gorshkova I.N.
        • Liu T.
        • Kan H.Y.
        • Chroni A.
        • Zannis V.I.
        • Atkinson D.
        Structure and stability of apolipoprotein A-I in solution and in discoidal high-density lipoprotein probed by double charge ablation and deletion mutation.
        Biochemistry. 2006; 45: 1242-1254
        • Mendivil C.O.
        • Furtado J.
        • Morton A.M.
        • Wang L.
        • Sacks F.M.
        Novel pathways of apolipoprotein A-I metabolism in HDL of different sizes in humans.
        Arterioscler. Thromb. Vasc. Biol. 2016; 36: 156-165
        • Bedi S.
        • Morris J.
        • Shah A.
        • Hart R.C.
        • Jerome W.G.
        • Aller S.G.
        • Tang C.
        • Vaisar T.
        • Bornfeldt K.E.
        • Segrest J.P.
        • Heinecke J.W.
        • Davidson W.S.
        Conformational flexibility of apolipoprotein A-I amino- and carboxy-termini is necessary for lipid binding but not cholesterol efflux.
        J Lipid Res. 2022; 63: 100168
        • Rall Jr., S.C.
        • Weisgraber K.H.
        • Mahley R.W.
        • Ogawa Y.
        • Fielding C.J.
        • Utermann G.
        • Haas J.
        • Steinmetz A.
        • Menzel H.J.
        • Assmann G.
        Abnormal lecithin:cholesterol acyltransferase activation by a human apolipoprotein A-I variant in which a single lysine residue is deleted.
        J. Biol. Chem. 1984; 259: 10063-10070
        • Jonas A.
        • von Eckardstein A.
        • Churgay L.
        • Mantulin W.W.
        • Assmann G.
        Structural and functional properties of natural and chemical variants of apolipoprotein A-I.
        Biochim. Biophys. Acta. 1993; 1166: 202-210
        • Ludovico I.D.
        • Bedi S.
        • Melchior J.T.
        • Gonzalez M.
        • Garda H.
        • Davidson S.
        Deletion of Lys 107 Modify the Ability of ApoA-I to Self-associate in Solution.
        Arterioscler. Thromb. Vasc. Biol. 2020; 40: A207
        • Haase C.L.
        • Tybjærg-Hansen A.
        • Grande P.
        • Frikke-Schmidt R.
        Genetically elevated apolipoprotein A-I, high-density lipoprotein cholesterol levels, and risk of ischemic heart disease.
        J Clin. Endocrinol. Metab. 2010; 95: E500-E510
        • Weisgraber K.H.
        • Rall S.C.
        • Bersot T.P.
        • Mahley R.W.
        • Franceschini G.,
        • Sirtori C.R.
        Apolipoprotein A-I milano - detection of normal A-I in affected subjects and evidence for a cysteine for arginine substitution in the variant A-I.
        J. Biol. Chem. 1983; 258: 2508-2513
        • Anthanont P.
        • Polisecki E.
        • Asztalos B.F.
        • Diffenderfer M.R.
        • Barrett P.H.
        • Millar J.S.
        • Billheimer J.
        • Cuchel M.
        • Rader D.J.
        • Schaefer E.J.
        A novel ApoA-I truncation (ApoA-I Mytilene) associated with decreased ApoA-I production.
        Atherosclerosis. 2014; 235: 470-476
        • Mei X.
        • Atkinson D.
        Crystal structure of C-terminal truncated apolipoprotein A-I reveals the assembly of high density lipoprotein (HDL) by dimerization.
        J Biol Chem. 2011; 286: 38570-38582
        • Melchior J.T.
        • Walker R.G.
        • Cooke A.L.
        • Morris J.
        • Castleberry M.
        • Thompson T.B.
        • Jones M.K.
        • Song H.D.
        • Rye K.A.
        • Oda M.N.
        • Sorci-Thomas M.G.
        • Thomas M.J.
        • Heinecke J.W.
        • Mei X.
        • Atkinson D.
        • Segrest J.P.
        • Lund-Katz S.
        • Phillips M.C.
        • Davidson W.S.
        A consensus model of human apolipoprotein A-I in its monomeric and lipid-free state.
        Nat Struct Mol Biol. 2017; 24: 1093-1099
        • Chroni A.
        • Kan H.Y.
        • Kypreos K.E.
        • Gorshkova I.N.
        • Shkodrani A.
        • Zannis V.I.
        Substitutions of Glu110 and Glu111 in the middle helix 4 of human ApoA-I by Alanine affect the structure and in vitro functions of apoA-I and induce severe hypertriglyceridemia in ApoA-I-deficient mice.
        Biochemistry. 2004; 43: 10442-10457
        • Kateifides A.K.
        • Gorshkova I.N.
        • Duka A.
        • Chroni A.
        • Kardassis D.
        • Zannis V.I.
        Alteration of negatively charged residues in the 89 and 96 domain of apoA-I affects lipid homeostasis and maturation of HDL..
        J. Lipid Res. 2011; 52: 1363-1372
        • Gorshkova I.N.
        • Atkinson D.
        Enhanced Binding of Apolipoproteins A-I Variants Associated with Hypertriglyceridemia to Triglyceride-Rich Particles.
        Biochemistry. 2011; 50: 2040-2047
        • Chroni A.
        • Liu T.
        • Gorshkova I.
        • Kan H.Y.
        • Uehara Y.
        • von Eskardstein A.
        • Zannis V.I.
        The central helices of apoA-I can promote ATP-binding cassette transporter A1 (ABCA1)-mediated lipid efflux. Amino acid residues 220-231 of the wild-type apoA-I are required for lipid-efflux in vitro and high density lipoprotein formation in vivo.
        J. Biol.Chem. 2003; 278: 6719-6730
        • Arakawa R.
        • Yokoyama S.
        Helical apolipoproteins stabilize ATP-binding cassette transporter A1 by protecting it from thiol protease-mediated degradation.
        J. Biol. Chem. 2002; 277: 22426-22429
        • Vedhachalam C.
        • Chetty P.S.
        • Nickel M.
        • Dhanasekaran P.
        • Lund-Katz S.
        • Rothblat G.H.
        • Phillips M.C.
        Influence of apolipoprotein (Apo) A-I structure on nascent high density lipoprotein (HDL) particle size distribution.
        J. Biol. Chem. 2010; 285: 31965-31973
        • Liu L.
        • Bortnick A.E.
        • Nickel M.
        • Dhanasekaran P.
        • Subbaiah P.V.
        • Lund-Katz S.
        • Rothblat G.H.
        • Phillips M.C.
        Effects of apolipoprotein A-I on ATP-binding cassette transporter A1-mediated efflux of macrophage phospholipid and cholesterol: formation of nascent high density lipoprotein particles.
        J Biol Chem. 2003; 278: 42976-42984
        • Nilsson O.
        • Lindvall M.
        • Obici L.
        • Ekström S.
        • Lagerstedt J.O.
        • Del Giudice R.J.
        Structure dynamics of ApoA-I amyloidogenic variants in small HDL increase their ability to mediate cholesterol efflux.
        J. Lipid Res. 2021; 62: 1-13
        • Silva R.A.
        • Huang R.
        • Morris J.
        • Fang J.
        • Gracheva E.O.
        • Ren G.
        • Kontush A.
        • Jerome W.G.
        • Rye K.A.
        • Davidson W.S.
        Structure of apolipoprotein A-I in spherical high density lipoproteins of different sizes.
        Proc. Natl. Acad. Sci. U S A. 2008; 105: 12176-12181
        • Huang R.
        • Silva R.A.
        • Jerome W.G.
        • Kontush A.
        • Chapman M.J.
        • Curtiss L.K.
        • Hodges T.J.
        • Davidson W.S.
        Apolipoprotein A-I structural organization in high-density lipoproteins isolated from human plasma.
        Nat. Struct. Mol. Biol. 2011; 18: 416-422
        • Segrest J.P.
        • Jones M.K.
        • Klon A.E.
        • Sheldahl C.J.
        • Hellinger M.
        • De Loof H.
        • Harvey S.C.
        A detailed molecular belt model for apolipoprotein A-I in discoidal high density lipoprotein.
        J. Biol. Chem. 1999; 274: 31755-31758
        • Li L.
        • Chen J.
        • Mishra V.K.
        • Kurtz J.A.
        • Cao D.
        • Klon A.E.
        • Harvey S.C.
        • Anantharamaiah G.M.
        • Segrest J.P.
        Double belt structure of discoidal high density lipoproteins: molecular basis for size heterogeneity.
        J. Mol. Biol. 2004; 343: 1293-1311
        • Davidson W.S.
        • Silva R.A.
        Apolipoprotein structural organization in high density lipoproteins: belts, bundles, hinges and hairpins.
        Curr Opin Lipidol. 2005; 16: 295-300
        • Chetty P.S.
        • Mayne L.
        • Kan Z.-Y.
        • Lund-Katz S.
        • Englander S.W.
        • Phillips M.C.
        Apolipoprotein A-I helical structure and stability in discoidal high-density lipoprotein (HDL) particles by hydrogen exchange and mass spectrometry.
        Proc. Natl. Acad. Sci. U S A. 2012; 109: 11687-11692
        • Sparks D.L.
        • Lund-Katz S.
        • Phillips M.C.
        The charge and structural stability of apolipoprotein A-I in discoidal and spherical recombinant high density lipoprotein particles.
        J Biol Chem. 1992; 267: 25839-25847