Studies of the lipid binding characteristics of the apolipoproteins from human high density lipoprotein. II. Calorimetry of the binding of apo AI and apo AII with phospholipids

The biotin-capture lipid affinity assay: a rapid method for determining lipid binding parameters for apolipoproteins

The lipid affinity of plasma apolipoproteins is an important modulator of lipoprotein metabolism. Mutagenesis techniques have been widely used to modulate apolipoprotein lipid affinity for studying biological function, but the approach requires rapid and reliable lipid affinity assays to compare the mutants. Here, we describe a novel method that measures apolipoprotein binding to a standardized preparation of small unilamellar vesicles (SUVs) containing trace biotinylated and fluorescent phospholipids. After a 30 min incubation at various apolipoprotein concentrations, vesicle-bound protein is rapidly separated from free protein on columns of immobilized streptavidin in a 96-well microplate format. Vesicle-bound protein and lipid are eluted and measured in a fluorescence microplate reader for calculation of a dissociation constant and the maximum number of potential binding sites on the SUVs. Using human apolipoprotein A-I (apoA-I), apoA-IV, and mutants of each, we show that the assay generates binding constants that are comparable to other methods and is reproducible across time and apolipoprotein preparations. The assay is easy to perform and can measure triplicate binding parameters for up to 10 separate apolipoproteins in 3.5 h, consuming only 120 microg of apolipoprotein in total. The benefits and potential drawbacks of the assay are discussed.

Formation of HDL-like complexes from apolipoprotein A-I(M) and DMPC

Conditions for the preparation of reconstituted high density lipoproteins (HDLs) by incubation of the synthetic lipid dimyristoylphosphatidylcholine (DMPC) and recombinant apolipoprotein A-I(M) have been investigated as a function of ratio of incubation lipid to protein, incubation temperature and the lipid form (multilamellar (MLV) or small unilamellar (SUV) vesicles). The size distributions of the resultant lipid-protein complex particles from various incubations have been evaluated by native gel electrophoresis. Structural changes of the protein after incorporation into these complex particles have been estimated by CD. Thermal characteristics of the particles has been examined by DSC and correlated with CD results. Titration calorimetry has been used to obtain interaction parameters based on a simplified binding model. It is hypothesized that the major enthalpic step in the production of rHDLs is the primary association step between protein and lipid vesicles. It has been shown that by raising the temperature and incubation ratio, the formation of rHDL particles can be directed towards smaller size and a narrower size distribution. The results have been described on the basis of a model where formation of discoidal particles requires prior saturation of vesicle surface area by adsorbed protein, thus explaining differences between particles formed from MLVs and SUVs.

The mechanism of human plasma phospholipid transfer protein-induced enlargement of high-density lipoprotein particles: evidence for particle fusion

1. Phospholipid transfer protein (PLTP) mediates conversion of high-density lipoprotein (HDL3) to large particles, with concomitant release of apolipoprotein A-I (apoA-I). To study the mechanisms involved in this conversion, reconstituted HDL (rHDL) particles containing either fluorescent pyrenylacyl cholesterol ester (PyrCE) in their core (PyrCE-rHDL) or pyrenylacyl phosphatidylcholine (PysPC) in their surface lipid layer (PyrPC-rHDL) were prepared. Upon incubation with PLTP they behaved as native HDL3, in that their size increased considerably. 2. When PyrPC-rHDL was incubated with HDL3 in the presence of PLTP, a rapid decline of the pyrene excimer/monomer fluorescence ratio (E/M) occurred, demonstrating that PLTP induced mixing of the surface lipids of PyrPC-rHDL and HDL3. As this mixing was almost complete before any significant increase in HDL particle size was observed, it represents PLTP-mediated phospholipid transfer or exchange that is not directly coupled to the formation of large HDL particles. 3. When core-labelled PyrCE-rHDL was incubated in the presence of PLTP, a much slower, time-dependent decrease of E/M was observed, demonstrating that PLTP also promotes mixing of the core lipids. The rate and extent of mixing of core lipids correlated with the amount of PLTP added and with the increase in particle size. The enlarged particles formed could be visualized as discrete, non-aggregated particles by electron microscopy. Concomitantly with the appearance of enlarged particles, lipid-poor apoA-I molecules were released. These data, together with the fact that PLTP has been shown not to mediate transfer of cholesterol esters, strongly suggest that particle fusion rather than (net) lipid transfer or particle aggregation is responsible for the enlargement of HDL particles observed upon incubation with PLTP.4.ApoA-I rHDL, but not apoA-II rHDL, were converted into large particles, suggesting that the presence of apoA-I is required for PLTP-mediated HDL fusion. A model for PLTP-mediated enlargement of HDL particles is presented.

Transformations of reconstituted high-density lipoprotein subclasses as a function of temperature or LDL concentration

The objectives of this study were to determine the structural changes in defined, reconstituted high density lipoproteins (rHDL) resulting from spontaneous phospholipid depletion in the presence or absence of low-density lipoproteins (LDL), to establish the precursor-product relationships among the rHDL particles and to assess the differences in behavior of rHDL particles containing apo A-I or apo A-II. The rHDL particles were prepared by the sodium cholate dialysis method, and were incubated in buffer at 50 degrees C, or in buffer containing different concentrations of LDL at 37 degrees C, for up to 24 h. The changes in the rHDL particle distributions with time were followed by non-denaturing gradient gel electrophoresis, and the rHDL were isolated at various time points for chemical analysis. We found that rHDL particles containing apo A-I or apo A-II lose phospholipid and gain cholesterol when incubated with LDL. Increasing LDL concentrations remove increasingly larger amounts of phospholipid. With phospholipid loss the apo A-I containing particles undergo major structural rearrangements that give rise to 78 A and 106 A particles from 86 A and 94 A precursors. The 78 A products appear to be the most stable, lipid-poor species. Reconstituted HDL particles prepared with apo A-II (94 and 101 A in diameter) are more resistant to structural rearrangements than the apo A-I counterparts under similar reaction conditions.

Secondary structure and orientation of the amphipathic peptide GALA in lipid structures. An infrared-spectroscopic approach

GALA, a synthetic, amphipathic 30-amino-acid peptide, based upon a Glu-Ala-Leu-Ala motive, was designed to mimic the behavior of viral fusion proteins. GALA is a water-soluble peptide with an aperiodic conformation at neutral pH, and becomes an amphipathic alpha helix as the pH is lowered to 5, where it interacts with phospholipid bilayers. Attenuated total-reflection infrared spectroscopy, using polarized light, provides information on the structure and orientation of the peptide and the lipids, which is not subject to artifacts due to light scattering with large particles. H/2H-exchange rate of the amide N-H group and analysis of the shape of the amide I' by Fourier self-deconvolution and curve fitting indicate that the alpha-helical content increases from 19% to 69%, on lowering the pH. A further increase to 100% alpha helix is observed after interaction with palmitoyloleoylglycerophosphocholine (PamOleGroPCho) vesicles. Dichroism data obtained with oriented bilayers of the PamOleGroPCho-GALA complex demonstrate that PamOleGroPCho hydrocarbon chains and the peptide alpha helical axis are essentially perpendicular (+/- 15) to the membrane plane. At neutral pH, in the presence of dimyristoylglycerophosphocholine (Myr2GroPCho), GALA is known to form discoidal structures similar to those formed under the same conditions by apolipoproteins AI and AII. In these discoidal complexes, the alpha-helical content was estimated to be 65%, with the rest of the structure being essentially unordered. No significant modification of the all-trans conformation of the hydrocarbon chain of Myr2GroPCho was detected upon disc formation. Dichroism measurements show that the alpha-helical axis is essentially parallel to the hydrocarbon chains. These data support a model in which a discoidal patch of the bilayer is surrounded by amphipathic helices which shield the hydrophobic region of the bilayer from the aqueous environment. The infrared spectrum of GALA in this complex was found to be very similar to those of apolipoproteins AI and AII which form discoidal complexes with Myr2GroPCho, but the spectrum is quite different from that of apolipoprotein B100 in low-density lipoproteins, which does not form discoidal complexes.

In vitro and in vivo interactions of Triton 1339 with plasma lipoproteins of normolipidemic rhesus monkeys. Preferential effects on high density lipoproteins

Triton WR-1339 was incubated in vitro in various proportions with plasma from normolipidemic rhesus monkeys or with ultracentrifugally purified lipoproteins, and the products were examined by isopycnic density gradient ultracentrifugation, agarose column chromatography, electrophoretic and immunochemical techniques, and electron microscopy. Some experiments used apo A-I, apo A-II, or Triton labeled with either 125I or 131I. At concentrations of less than 10 mg/ml plasma, Triton interacted preferentially with HDL, changing lipoprotein size and density; Triton was progressively incorporated into the HDL particles, displacing apo E, apo A-I, and apo A-II. At concentrations above 10 mg/ml plasma, Triton displaced all apo A-I from the particle, and much lipid was dissolved into the Triton micelles. When Triton-treated HDL particles were used as a substrate for the enzyme LCAT, enzyme activity decreased in parallel to the displacement of apo A-I. There was no displacement of apo B from LDL nor any loss of lipids; but the particles became deformed and formed rouleaux. A single intravenous dose of Triton WR-1339 administered to a normolipidemic monkey (N) and to a hypercholesterolemic monkey (H) resulted in concentration-dependent HDL changes similar to those observed in vitro. LDL was less affected by Triton, with changes occurring only at high doses. After these structural changes, intravenously injected 131I apo A-I disappeared rapidly from the circulation; 125I apo A-II disappeared less rapidly. These increased clearances were accompanied by a drop in apo A-I plasma levels and the disappearance of HDL particles from plasma. The lipoprotein and apolipoprotein patterns returned to normal 14 days after Triton. We conclude that Triton WR-1339, when exposed to rhesus plasma in vitro or in vivo, interacts preferentially with HDL in a dose-dependent manner. At low concentrations, Triton acts on surface components of the HDL particle; at higher concentrations, Triton penetrates the particle, causing structural disruption. Because of its high affinity for HDL, Triton WR-1339 is a useful reagent for study of HDL structure-function relationships.

Thermodynamics and kinetics of protein incorporation into membranes

The free energy and enthalpy of protein incorporation into membranes are calculated with special emphasis on the hitherto neglected effects of immobilization of protein and perturbation of lipid order in the membrane. The free energy change is found to be determined by the hydrophobic effect as the driving force for incorporation and the protein immobilization effect which leads to a considerable reduction of the free energy gained from the hydrophobic effect. For incorporation of a hydrophobic, bilayer-spanning alpha-helix, the free energy change obtained is of the order of -15 kcal/mol (1 cal = 4.184 J) in agreement with experimental results. The lipid perturbation effect yields only a small contribution to the free energy change due to an energy/entropy compensation inherent in lipid order. This effect dominates the enthalpy change, giving rise to values on the order of 100 kcal/mol with a pronounced temperature dependence around the lipid phase transition as observed experimentally. The kinetics of protein incorporation are even more strongly affected by the lipid perturbation effect, leading to an abrupt decrease of the rate of incorporation below the lipid phase transition.

Displacement of the human apoprotein A-I by the human apoprotein A-II from complexes of (apoprotein A-I)-phosphatidylcholine-cholesterol

Reassembly experiments, involving isolated human apoproteins A-I and A-II and (dimyristoylglycerophosphocholine)-cholesterol vesicles were performed with apoprotein mixtures at apoprotein A-I/A-II molar ratios varying between 0 and 3. The apoproteins were incubated at 24 degrees C. 28 degrees C and 32 degrees C with either pure dimyristoyl-glycerophosphocholine vesicles or with dimyristoylglycerophosphocholine cholesterol vesicles containing 2, 5, 10, 15 mol/100 mol cholesterol. The kinetics of association were followed by measuring the increase of the fluorescence polarization ratio after labeling the lipids with diphenyl hexatriene. The complexes were separated from the free protein by gradient ultracentrifugation. Total protein was assayed and the apoproteins A-I and A-II were quantified separately by immunonephelometry. The content of apoprotein A-I was also monitored by measuring the intrinsic tryptophan fluorescence. The results suggest that apoprotein A-II has a greater affinity than apoprotein A-I for the phospholipid-cholesterol vesicles and that apoprotein A-II is able to quantitatively displace apoprotein A-I from the lipid-protein complexes. The content of apoprotein A-II in the complexes increases proportionally to the concentration of apoprotein A-II in the incubation mixture until saturation is reached. At saturation the dimyristoylglycerophosphocholine/apoprotein A-II ratio in the complex is dependent upon the cholesterol content of the original vesicles and increases from 60 to 275 mol/mol between 0 and 15 mol/100 mol cholesterol. From these experiments one can calculate that 1 mol human apoprotein A-I is displaced by 2 mol human apoprotein A-II.

Reassembly of human apoproteins A-I and A-II with unilamellar phosphatidylcholine-cholesterol liposomes. Association kinetics and characterization of the complexes

The kinetics of association between the human apoprotein A-I and apoprotein A-II and cholesterol dimyristoyl phosphatidylcholine (DMPC) vesicles are compared in this study and the lipid-apoprotein complexes are characterized. The association kinetics are followed by turbidity measurements monitoring the decrease of the vesicular size and by fluorescence polarization measurements monitoring the decrease in the mobility of the phospholipid acyl chains during complex formation. The influence of the incubation temperature and of the cholesterol/DMPC ratio has been studied by both techniques. Under all incubation conditions the apoprotein A-II associates more readily with cholesterol-DMPC vesicles than apoprotein A-I, as the kinetics are faster and the complex yield larger. With both apoproteins optimal complex formation takes place around the phospholipid transition temperature and around 10 mol% cholesterol. The apoprotein A-I/lipid association seems restricted to this narrow range for the temperature and the cholesterol/DMPC ratio, while the apoprotein A-II still associates with vesicles containing 20 mol% cholesterol and at temperatures up to 32 degrees C. The lipid-apoprotein complexes were isolated by gradient ultracentrifugation and by gel chromatography. According to these data the apoprotein A-II associates more readily than apoprotein A-I with cholesterol-DMPC vesicles to form protein-rich complexes, whilst the optimal apoprotein A-I-lipid association requires a more disordered lipid structure.

[Presence and isolation of 2 lipoproteins immunologically related to apolipoprotein A I in human serum]

Very high density lipoproteins d : 1.23--1.25 g/ml (VHDL2) have been isolated from human serum by preparative ultracentrifugation. They contain 80 per cent proteins and 20 per cent lipids. Lipids are mainly phospholipids (80 per cent). The proportion of lysolecithin (50 per cent) is higher than that of lecithin (40 per cent). The quantity of cholesterol is low, the free cholesterol: total cholesterol ratio is 0.35. VHDL2 consisted principally in lipoprotein D and two lipoproteins immunologically apparented to apolipoprotein A I, called LP A I1 and LP A I2. The LP A I1 has a molecular weight slightly higher and a hydrated density lower than that of LP AI2. Our experiments suggest that LP A I1 exists in the serum before ultracentrifugation while LP A I 2 comes from HDL degradation during ultracentrifugation. The immunological heterogeneity of apo A I forming different protein-lipid complexes is discussed.

Covalent binding of photosensitive 1-(12-azido-[9,10(-3)H2]oleoyl)glycero-3-phosphocholine (lysolecithin) to human serum high density apolipoproteins

Human serum high density apoproteins were complexed with increasing concentrations of 1-(12-azido-[9,10(-3)H2]oleoyl)glycero-3-phosphocholine up to the saturation concentration (72 mol lysolecithin per mol apo HDL). Ultraviolet irradiation generated the nitrene which led to crosslinking with the two main apolipoproteins AI and AII. Methods are described for the removal of excess, unbound lipid and the column chromatographic separation of the lipopolypeptides AI and AII.

Structure and thermodynamic properties of the complexes between phospholipase A2 and lipid micelles

The interaction between porcine pancreatic phospholipase A2 and a homogeneous population of micelles of the subtrate analogue n-hexadecylphosphorylcholine containing 155 lipid monomers was studied by light scattering, equilibrium gel filtration, and isothermal calorimetry. From the detergent/protein molar ratio and the equivalent "molecular weight" of the resulting complex it is concluded that insertion of the enzyme into the detergent micelle results in a protein--detergent complex containing two phospholipase A2 molecules and 80 lipid monomers at 25 degrees C. The affinity constants and complex composition have been determined at different temperatures, allowing calculation of the thermodynamic parameters of the binding process. It is concluded that the interaction of phospholipase A2 with micellar lipids is predominantly hydrophobic.

Ionization behaviour of native apolipoproteins and of their complexes with lecithin. 1. Calorimetric and potentiometric titration of the native apoA-I protein and of the apoA-I protein-dimyristoyl lecithin complex

The ionization behaviour of native apoA-I protein is compare to that of its complex with synthetic dimyristoyl lecithin in studies using calorimetric, potentiometric and spectrophotometric titration. In the presence of phospholipids, 10 out of 21 lysines together with 22 acidic residues are masked in the complex. All tyrosines remain accessible to titration below pH 13. The apparent ionization enthalpy of the 11 lysine residues is not affected by the presence of phospholipids. These data are consistent with discrete binding sites located in the apoprotein helical segments as suggested by the model of Segrest et al. [FEBS Lett. 38, 247-253 (1974)]. A tentative localisation of lysine, arginine, aspartic acid and glutamic acid residues directly involved in phospholipid binding is suggested, assuming that such helical regions are involved in apoprotein-phospholipid association.

Ionization behaviour of native apolipoproteins and of their complexes with lecithin. 2. Potentiometric titration of the native apo-A-II, apoC-I, apoC-III proteins and of their complexes with dimyristoyl lecithin

A comparison of the ionization behaviour of the human apoA-II, apoC-I, apoC-III proteins and of their complexes with dimyristoyl lecithin is based on potentiometric titration of the basic and acidic residues and spectrophotometric titration of the phenolic groups. Experimental data suggest that a number of lysine, arginine, aspartic acid and glutamic acid residues are masked in the complexes. For each of these amino acids and in all three proteins the number of masked residues is consistent with the content of those regions predicted to be involved in lipid binding by the model of Segrest et al. [FEBS Lett. 38, 247-253 (1974)]. These data taken together with the results of calorimetric and titration experiments with the apoA-I protein reported in the accompanying article [Rosseneu et al. (1977) Eur. J. Biochem. 79, 251-257] strongly support the general nature of the proposed model and further suggest that ionic interactions have some role in the formation of the dimyristoyl lecithin/apolipoprotein complexes.

Thermodynamics of lipid protein associations. Thermodynamics of helix formation in the association of high density apolipoprotein A-I (apoA-I) to dimyristoyl phosphatidylcholine

The structure and phospholipid-binding properties of human plasma high density apolipoprotein A-I (apoA-I) has been studied at pH 7.4 and 3.1 by microcalorimetry, circular dichroism and density gradient ultracentrifugation. At pH values of 7.4 and 3.1, apoA-I binds to dimyristoyl phosphatidylcholine (DMPC) to form complexes of similar composition (molar ratio of DMPC/apoA-I of 100) and helical content (67%). At pH 7.4, the lipid-protein association is accompanied by an increase in helical content from 58 to 67% and an exothermic enthalpy of binding (deltaHB) of -90 kcal/mol apoA-I. At pH 3.1, the helical content of apoA-I is increased from 48 to 67% on binding to DMPC and the enthalpy of binding was -170 kcal/mol. We suggest that the difference in the enthalpies of binding (-80 kcal/mol) at pH 3.1 compared to 7.4 is due to the greater coil leads to helix transition at the lower pH.

Lipid-protein interactions in the plasma lipoproteins

The purpose of this review has been to discuss new information about the mechanism of lipid and apoprotein interaction in the plasma lipoproteins. A special form of the amphipathic helix has been identified as a major structural element of the apolipoproteins sequenced to date. Evidence is reviewed concerning the role of the amphipathic helix in the binding to phospholipids. Several different models for the organization of the components of HDL, LDL and LP-X have evolved from extensive structural studies. Resolution of the differences among these models will require additional experimental testing. Verification of models based on the study of reconstituted HDL will require rigorous proof of native structure in these particles. A detailed description of the molecular organization of the lipid and protein constituents of the plasma lipoproteins is still lacking. Further structural and sequence studies with apoB and the "arginine-rich" protein are needed. Crystallization of an apoprotein or lipoprotein and determination of the three-dimensional structure would be a major achievement. With such further detailed structural information, it may then be possible to correlate changes in structure with determinants of metabolism.

Phospholipid binding and self-association of the major apoprotein of human and baboon high-density lipoproteins

The purpose of this study was to establish a relationship between self-association and phospholipid binding of the human and the baboon apoA-I protein. The enthalpy changes on binding dimyristoyl lecithin and lysolecithin to either the human or the baboon native apoA-I protein were measured in a microcalorimeter. An endothermal process, most pronounced for the human apoprotein, was observed at low phospholipid levels. At higher phospholipid to protein ratios the binding was exothermal. Gel filtration experiments on Sephadex G-200 showed that the native apoprotein of both species consists of dimers and tetramers. The baboon native apoA-I protein contained a higher amount of dimers. After preincubation of the apoA-I protein with lysolecithin, the enthalpy changes measured on subsequent binding of dimyristoyl lecithin were shifted towards more exothermal values compared to the curve for the native apoprotein. The amplitude of this shift corresponds to that of the endothermal process observed on binding dimyristoyl lecithin to the native apoprotein. This process was attributed to a phospholipid-induced disaggregation of the apoA-I protein. Gel filtration data showed a decreased extent of aggregation in the apoA-I protein preincubated with lysolecithin. This sample consisted exclusively of dimers. Ultracentrifugal flotation of the complexes formed between the apoA-I protein, and respectively dimyristoyl lecithin and sphingomyelin indicated that preincubation with lysolecithin increased the extent of complex formation. These results suggest that the dimeric form of the apoA-I protein possesses the highest affinity for phospholipids. Any dissociation of higher polymers enhances the phospholipid-binding capacity of the human and the baboon apoA-I protein.

[Lipid-protein-interactions of human apolipoproteins-structural aspects and models of lipoproteins (author's transl)]

The plasma lipoproteins are complex macromolecular structures which play an essential role in fat transport and in energy and membrane metabolism of higher organized organisms. Much has been learned in the last decade about the structural and functional interrelationships of the different lipoprotein classes. Their protein moieties, the so-called apolipoproteins, have been purified and characterized, the primary structure of four of them is known. Initial recombination experiments showed a considerable potential of the (unfractionated) lipoprotein protein to bind to lipids and to form particles similar to native lipoproteins. Further binding experiments performed in several laboratories with the purified A- and C-apolipoproteins and different physico-chemically well defined lipids have lead to the identification of lipid binding sites within the protein molecules and the formation of amphipathic helices upon and during lipid binding. This possible common mechanism of lipid-protein fractions forms the basis of a recently proposed model of one lipoprotein class, namely the high density lipoproteins (HDL). The significance of protein-protein-interactions in the formation and maintenance of these lipoprotein particles is still unknown. Whether disturbed lipid protein interactions lead to structural and/or functional alterations of the corresponding lipoproteins is a topic of discussion. The pertinent literature is listed in this paper and the physiological relevance of these studies and their clinical aspects will be presented.

Interaction of the apoproteins of very low density and high density lipoproteins with synthetic phospholipids

The interaction of synthetic dimyristoyl phosphatidylcholine (lecithin) liposomes with isolated apoC-I and apoC-III proteins from very low density lipoproteins has been studied by microcalorimetry. Complex formation is a highly exothermal process characterized by a maximal enthalpy of -130 kcal/mol (-544 kJ) apoC-III-1 and -65 kcal/mol apoC-I proteins (-272 kJ). The complex composition determined after its isolation by ultracentrifugal flotation agrees with the value derived from the enthalpy binding curves. The binding of a constant amount of dimyristoyl lecithin to apoprotein mixtures containing various proportions of apoA-I and apoC-III failed to demonstrate the existence of any preferential association between the two apoproteins, in contrast with results obtained previously with apoA-I/apoA-II protein mixtures. Finally the various contributions to the enthalpy of binding such as that arising from an increase in apoprotein helicity have been evaluated. A classification of the apolipoproteins according to their lipid-binding affinity is proposed as: apoA-II congruent to apoC-III greater than apoC-I greater than apoA-I proteins.