Design of a New Class of Amphipathic Helical Peptides for the Plasma Apolipoproteins That Promote Cellular Cholesterol Efflux But Do Not Activate LCAT
Abstract Amphipathic helical peptides represent the lipid-binding units of the soluble plasma apolipoproteins. Several synthetic peptide analogues have been designed to mimic such structures and have been used to unravel some of the mechanisms involved in the physiological function of the apolipoproteins, including lipid binding, LCAT activation, and enhancement of cholesterol efflux from lipid-laden cells. A series of novel synthetic peptides, named ID peptides, was modeled on the basis of the structural properties common to the amphipathic helices of apolipoprotein (apo) A-I. In these new peptides, however, the segregation between hydrophobic and hydrophilic faces of the helices is more pronounced than in apoA-I, so that the surface of the hydrophobic and hydrophilic faces of the amphipathic helices is equal. Moreover, there are fewer negatively charged residues in the center of the hydrophilic face of the helical peptides. Most charged amino acids are located along the edge of the helix and are susceptible to forming salt bridges with residues of an antiparallel helix, such as around a discoidal phospholipid/peptide complex. The physicochemical characteristics of these peptides and their complexes with phospholipids were compared with those of the 18A peptide and its lipid/peptide complex. All ID peptides bind dimyristoylphosphatidylcholine vesicles more rapidly than the 18A peptide to yield discoidal peptide/phospholipid complexes of comparable size. The α-helical content of the lipid-free ID peptides is close to that of the 18A peptide and increases slightly on lipid binding. The stability of the ID and 18A peptides and of the phospholipid/peptide complexes against guanidinium hydrochloride denaturation is higher than that of lipid-free and lipid-bound apoA-I. LCAT activation by the 18A/phospholipid/cholesterol complexes equals that of apoA-I/phospholipid/cholestrol complexes, whereas none of the ID peptides tested is able to activate LCAT to a significant extent. Incubation of the peptide/phospholipid complexes with lipid-laden macrophages induces cellular cholesterol efflux and incorporation of cholesterol into the complexes. The cholesterol efflux capacity of the peptide/phospholipid complexes is comparable among the peptides and higher than that of apoprotein/phospholipid complexes. In conclusion, although the amphipathicity of the new peptides is higher than that of the 18A model peptide, the lack of LCAT activation by the ID peptides suggests that an enhanced segregation of the hydrophobic and hydrophilic residues, equal magnitude of hydrophobic and hydrophilic faces of the helix, and the absence of negatively charged residues in the central part of the hydrophilic face might account for the lack of LCAT activity of these peptides. These parameters do not affect the capacity of the peptide/phospholipid complexes to promote cellular cholesterol efflux.
- Received March 7, 1996.
- Accepted July 4, 1996.
A classification of the amphipathic helices of the water-soluble plasma apolipoproteins according to their molecular hydrophobicity potential was proposed by Brasseur et al.1 2 According to this concept, apolipoproteins can be subdivided into two classes; the first class consists of apoA-I, apoA-IV, and apoE, whose helices are less hydrophobic than those of apoA-II, apoC-II, and apoC-III, which belong to the second class of apolipoproteins.1 2 Isolated apolipoproteins can be reassembled with phospholipids to form discoidal particles of ≈10 nm diameter.3 4 5 6 The physicochemical analysis of these complexes, together with the theoretical assembly of their protein and lipid constituents by computer modeling, has provided a model for these discoidal particles in which the apolipoprotein helices are oriented parallel to the phospholipid acyl chains around the edge of the discs.7 8
Synthetic peptides, mainly 18 to 22 residues long, have been extensively used as models for the study of helix/lipid interactions.9 10 11 12 The sequences of these peptides either match those of the native apolipoproteins13 or represent a consensus sequence for the various helical repeats identified in apoA-I.14 In a previous study we compared the structure, composition, and physicochemical properties of the phospholipid/peptide complexes generated with the 18A peptide described by Segrest et al15 and with a number of 18A variants.16 17 These data showed that the structural properties, the mode of lipid association, and the stability of the complexes resemble those of the native apolipoproteins.17 Energy minimization calculations, together with the physicochemical analysis of these complexes, suggested that ionic interactions between residues belonging to two antiparallel peptides strongly contribute to the stability of a pair of peptides at a lipid/water interface even when these peptides are not linked through a β strand as in the apolipoproteins.17 Besides their lipid-binding properties, native apolipoproteins are able to activate the LCAT enzyme and promote cholesterol efflux from lipid-laden cells.18 19 20 These properties are shared by the synthetic model peptides, as LCAT activation by phospholipid/18A (blocked form of 18A) complexes is comparable to that of apoA-I,9 while lipid-free or lipid-bound 18A peptide can promote cholesterol efflux from lipid-laden macrophages.21 To assess the contribution of interhelix association and cooperativity as well as the role of hydrophilic residues in the function of amphipathic model peptides, novel amphipathic helical peptides were synthesized and their properties studied. A comparison of their properties to those of apoA-I and of previous model peptides should contribute to a better definition of the structural and compositional parameters modulating the function of the amphipathic helical segments.
Molecular Modeling of the Peptides by Energy Minimization
The method used for these calculations is that previously applied to the study of polypeptide conformation.22 The total conformational energy was first calculated as the sum of three components: the London–van der Waals energy of interaction between all pairs of non–mutually bonded atoms, the Coulomb electrostatic interaction between atomic charges, and the potential energy of rotation of torsional angles. The lowest conformational energy of the helical peptide corresponding to the most stable structure was obtained by subsequent simplex minimization at a resolution of <5° for each torsional angle.
For the calculation of the energy of interaction between a pair of peptides, the conformation of the pair of helices was kept in the lowest energy state after peptide translation and rotation. The energies of interaction between all atoms of this assembled system were calculated after energy minimization of the pair of peptides at the water/lipid interface. These interaction energies consist of the sum of the van der Waals, the hydrophobic, and the electrostatic energy of interaction.22 All calculations were performed using the PC-PROT+ and PC-TAMMO+ (Theoretical Analysis of Molecular Membrane Organization) programs.
The peptides were synthesized by solid-phase peptide synthesis by coupling on a TentaGel S-RAM resin (Rapp Polymere) as described previously.16 17 The amino termini of the completed peptide chains were acetylated using acetic anhydride, while the carboxyl termini were blocked by amidation. All syntheses were carried out on a Milligen 9050 Pepsynthetisizer using continuous flow procedures. After cleavage with trifluoroacetic acid in the presence of scavengers and extraction with t-butyl methyl ether, the purity of all peptides was checked by reverse-phase chromatography on a Waters 625 high-pressure chromatography system equipped with a Waters 490E detector and a Hitachi D-2500 integrator. Separations were carried out on a Pharmacia Pep/S C2/C18, 4×250-mm reverse-phase column. The purity of the peptides was further verified by amino acid analysis.
Preparation of Phospholipid/Peptide Complexes
Complexes were prepared by incubation of the peptides with DMPC (Sigma Chemical Co) vesicles, at a DMPC/peptide ratio of 3:1 (wt/wt) at 25°C for 16 hours. The DMPC vesicles were obtained by sonicating the phospholipid three times for 7 minutes each at 37°C under nitrogen. Complexes with DPPC (Sigma) used for the cell-culture experiments were prepared at a DPPC/peptide ratio of 3:1 (wt/wt), using the cholate dialysis procedure.23 The mixture was incubated at 42°C for 16 hours and the cholate removed by extensive dialysis. Complexes used for the LCAT activation assays were prepared by the same method but contained cholesterol (Sigma) at a cholesterol/phospholipid wt/wt ratio of 0.05.
The formation of the DMPC/peptide complexes was followed by monitoring the optical density decrease at 325 nm of multilamellar vesicles of DMPC with the peptides (phospholipid/peptide ratio 2:1 [wt/wt]) as a function of temperature.16 Complexes were isolated by gel filtration on a Superose 6HR column in a 5 mmol/L Tris-HCl buffer, pH 8.1, 0.15 mol/L NaCl, 0.2 g/L NaN3, in a fast protein liquid chromatography system (Waters). Complexes were detected by measuring the optical density at 280 nm and the Trp fluorescence emission at 330 nm. The composition and size of the complexes were determined on the fraction with maximal UV absorption in the elution profile.
The chemical composition of the isolated complexes was assayed as follows: phospholipids and cholesterol were measured enzymatically by using commercial kits (Biomérieux and Boehringer GmbH) and the peptides were assayed by Phe quantitation by reverse-phase HPLC.16
Electron Microscopy of the Phospholipid/Peptide Complexes
Phospholipid/peptide complexes at a peptide concentration of 150 μg/mL were negatively stained with a 20 g/L solution of potassium phosphotungstate (pH 7.4). Seven microliters of the samples was applied to Formvar carbon-coated grids and examined in a Zeiss EM 10C transmission electron microscope operating at 60 kV.4 The mean diameter and size of distribution of 150 particles were calculated for each sample.
Fluorescence measurements were performed on an Aminco SPF-500 spectrofluorimeter equipped with a special adapter (Aminoco-J4-9501) for fluorescence polarization measurements.4 Temperature scans were performed between 30°C and 55°C for the DPPC/peptide complexes. Denaturation experiments were carried out by following the maximal emission wavelength of the Trp residues after exposure to GdnHCl concentrations between 0 and 6 mol/L.
Circular Dichroism Measurements
Circular dichroic spectra of the peptides and isolated complexes with phospholipids were measured at room temperature in a Jasco 710 spectropolarimeter.16 Measurements were carried out at a protein concentration of 0.5 mg/mL in a 0.01 mol/L sodium phosphate buffer, pH 7.4. Nine spectra were collected and averaged for each sample. The secondary structure was estimated according to the generalized inverse method of Compton and Johnson.24 For the denaturation experiments, the native and lipid-associated peptides, at a peptide concentration of 100 μg/mL, were incubated for 48 hours in the presence of increasing GdnHCl concentrations before the circular dichroism measurements.
LCAT-Activating Properties of the Complexes
LCAT activation by DPPC/cholesterol/apolipoprotein complexes was determined by HPLC measurement of the amount of CE formed during the enzymatic reaction.4 The assay mixture, consisting of variable amounts of complexes (between 2 and 20 μmol/L cholesterol), 6 mmol/L β-mercaptoethanol, and 90 μmol/L defatted BSA (Sigma), was preincubated for 20 minutes at 37°C. The reaction was initiated by adding 5 μL semipurified LCAT enzyme (50 μg LCAT per milliliter) to 0.2 mL of the reaction mixture. The reaction was carried out at 37°C and was stopped by extraction of the incubation mixture with hexane/isopropanol (3:2, vol/vol), containing cholesteryl heptadecanoate as an internal standard. The samples were injected on a (5 μm) Zorbax ODS reverse-phase column, at 50°C. The CE was eluted isocratically at a flow rate of 1.0 mL/min with acetonitrile/isopropanol (1:1, vol/vol).
The course of the LCAT reaction was followed between 1 minute and 24 hours, and the initial velocities were determined in the linear portion of the curves, ie, between 0% and 15% CE formed, corresponding to incubation times between 0 and 30 minutes. The initial reaction rates (Vo) were analyzed using Lineweaver-Burk plots of 1/Vo versus 1/C, according to Michaelis-Menten kinetics. A linear regression analysis yielded the apparent kinetic parameters Vmax, Km, and Vmax/Km.
Cellular Cholesterol Efflux Experiments From J774 Macrophages
J774 murine macrophages were grown for 18 hours in DMEM (GIBCO) with the addition of 10% fetal calf serum, in a 5% CO2 atmosphere at 37°C and incubated with acetylated LDL at a concentration of 100 μg apoB per milliliter for 24 hours. The excess LDL was removed by washing the cells in medium containing lipoprotein-deficient serum. The apolipoprotein/DPPC or peptide/DPPC complexes (1:3 protein or peptide/phospholipid, wt/wt ratio) were then added to the medium (n=3) at concentrations varying between 10 and 200 μg apolipoprotein or peptide per milliliter and incubated for another 18 hours. Blanks containing no protein/phospholipid complexes but equivalent amounts of buffer were run in each experiment.
Cells were separated from the medium by centrifugation, and lipids were extracted from the isolated cells with 5 mL hexane/isopropanol (3:2, vol/vol). Cholesterol and CE were quantified by HPLC.4 18 The amount of free cholesterol in the medium was measured enzymatically (Boehringer) and expressed as micrograms cholesterol, representing the net efflux.
Molecular Modeling of the ID Peptides
A new amphipathic helical peptide (ID3) was designed to have a hydrophobic and hydrophilic face of roughly equal magnitude and a sequence such that it would enable formation of two salt bridges within a pair of peptides. The ID3 sequence was derived from that of the apoA-I 102-118 helical repeat and of the 18A and 18AM4 peptides (Fig 1⇓), since it was shown for both that salt bridges are essential for the stability of the complexes with lipids.17 Phe and Trp residues were introduced in the sequence to enable peptide quantification by Phe assay by HPLC and peptide detection by Trp fluorescence measurement.
Two mutants of the ID3 peptide were designed to both increase the hydrophobicity index of the hydrophobic face of the helix and decrease the number of charged residues in the hydrophilic face. This was achieved through an Ala11-Leu substitution in the hydrophobic face of the ID7 and ID9 peptides, whose mean hydrophobicity (calculated according to Eisenberg et al25 ) increased to a value close to that of the 18A peptide (Table 1⇓). The number of negatively charged residues was decreased through an Asp-Asn substitution at position 2 and 3 in the ID7 and ID9 peptides, respectively. The sequences and the Edmundson wheel representation of the ID peptides and the 18A peptide are shown in Fig 1⇑, demonstrating the segregation of hydrophilic and hydrophobic residues on opposite faces of the helix, typical for amphipathic helices.
Table 1⇑ summarizes the amphipathicity and hydrophobicity of the ID peptides, which have an equal number of hydrophobic and hydrophilic residues and a hydrophobic angle of 180°. The mean hydrophobicity is higher for the ID7 and ID9 peptides than for the ID3, due to the Ala11-Leu substitution, and is closer to that of the 18A peptide. An important feature of these peptides, compared with analogues, is the absence of negatively charged Asp or Glu residues in the center of the hydrophilic face of the helix. While the 18A peptide contains four Glu residues at positions 1, 8, 12, and 16, which form a central band of negative charges in the hydrophilic face, the ID peptides have only one Glu residue in position 10 in the center of the hydrophilic face. The ID peptides have, moreover, two Lys residues at positions 7 and 18, at the boundary between hydrophobic and hydrophilic sides.
Besides the characterization and modelization of the single peptides, we further calculated the most stable conformation corresponding to the minimal energy for a pair of peptides. The electrostatic interactions between Asp2-Lys17 in a pair of ID3 peptides or a pair of ID9 peptides and the Asp3-Lys18 interactions in a pair of ID7 peptides stabilize the pair of helical peptides. The distance between charged residues of the two helices lies between 0.30 and 0.48 nm, yielding a stable pair of antiparallel helices (Fig 2⇓).
Lipid-Binding Properties of the Peptides
Association of the peptides with lipids was followed by monitoring the turbidity decrease of DMPC vesicles as a function of the temperature after peptide addition (Fig 3⇓). The optical density decrease at 325 nm shows that all ID peptides associate with lipids even at a temperature of 15°C. The midpoint temperature of the turbidity decrease is 19°C, 18.9°C, 16.9°C, and 16.9°C for the 18A, ID3, ID7, and ID9 peptides, respectively. All peptides bind rapidly to DMPC, the complexes remain stable at temperatures between 15°C and 30°C, and the ID peptides seem to associate more readily with DMPC than the 18A peptide. Complex formation was further monitored by measuring the Trp fluorescence emission spectra (Table 2⇓). The maximal wavelength for Trp fluorescence emission is lower for the ID7 and ID9 peptides than for the 18A peptide. This might be due to the location of the Trp residue in the middle of the hydrophobic face of the ID peptides, thereby reducing its exposure to the solvent. Complex formation was further monitored by measuring the degree of fluorescence polarization of the phospholipid/peptide mixture after labeling with DPH. The fluorescence polarization decreased around 23°C, and the DMPC transition was shifted toward higher temperatures due to peptide/lipid interactions. Compared with that of pure DMPC, the transition temperature was shifted by 3.0°C, 3.0°C, 4.5°C, and 2.5°C in the complexes generated with the 18A, ID3, ID7, and ID9 peptides, respectively.
Isolation and Composition of the Peptide/DPPC Complexes
Complexes generated between the synthetic peptides and phospholipid were isolated by gel chromatography on a Superose 6 HR fast protein liquid chromatography column. The complexes were identified by measuring the phospholipid content and by monitoring the Trp fluorescence intensity of the eluant. The elution profiles of the DPPC/peptide mixtures are represented in Fig 4⇓, showing that homogeneous complexes eluting in a symmetrical peak are formed between DPPC and the model peptides. The efficiency of complex formation is also demonstrated by the absence of free peptide or phospholipid in the elution profile. The elution volumes of the peaks corresponding to the complexes are comparable for all peptides. A negative-staining electron micrograph of the 18A/DPPC complexes is shown as an inset in Fig 4⇓, showing the typical “rouleau” pattern of stacked discoidal complexes, as previously observed for apolipoprotein/phospholipid and peptide/phospholipid complexes.14 16
The size of the complexes was calculated as the Stokes radius from the elution volume of the Superose 6 HR column and estimated from the electron micrographs (Table 2⇑). According to the electron microscopy measurements, the discoidal complexes formed with the ID peptides are smaller than the 18A/DPPC complexes, whereas the gel filtration runs could not resolve the small differences in size.
The secondary structure of the peptides and of the peptide/phospholipid complexes was measured by circular dichroism, showing that the blocked ID peptides have an α-helical content of ≈40%. Phospholipid binding increased the α-helical content of the 18A peptide by ≈20%, while an α-helicity increase of only 10% was measured for the ID peptides (Table 2⇑).
Denaturation experiments were performed by addition of increasing quantities of GdnHCl and measurement of the helical content of the lipid-free (Fig 5A⇓) and lipid-bound peptide (Fig 5B⇓) by circular dichroism. The midpoint of the denaturation occurs at GdnHCl concentrations of 1.5, 2.6, 2.8, and 2.2 mol/L GdnHCl for the ID3, ID7, ID9, and 18A peptides, respectively, compared with 0.7 mol/L for apoA-I.17 All peptide/lipid complexes denature at higher GdnHCl concentrations, as the midpoint of the denaturation curves is shifted to 3.5, 5.7, 5.5, and 5.5 mol/L for the ID3, ID7, ID9, and 18A peptides in the DPPC/peptide complexes. The results are similar to those reported for apoA-I/lipid complexes, in which a shift of about 2 mol/L GdnHCl is observed between the denaturation of the lipid-free and lipid-bound apolipoprotein.26 The secondary structure of the short peptides is therefore stabilized by the association with lipids, as for the native apolipoproteins.
LCAT-Activation Properties of the Complexes
The time kinetics of the LCAT reaction with peptide/DPPC/cholesterol discoidal complexes as substrate were followed between 0 and 24 hours and compared with those of the apoA-I/DPPC/cholesterol complexes, showing that among the peptide/DPPC/cholesterol complexes the fastest kinetics are observed with the 18A peptide. Esterification rates were determined as a function of the substrate concentrations, and the apparent kinetic parameters were calculated from the corresponding Line-weaver-Burk plots (Table 3⇓).
A higher apparent Km, together with a low Vmax, was calculated for all ID peptides compared with the 18A peptide. The Vmax/Km value representing the substrate efficiency was therefore lower for all ID peptides than for the 18A peptide, and it was negligible compared with that of the natural activator apoA-I. The efficiency of the 18A peptide amounted to about 40% of that of apoA-I. The major differences in Vmax are due to differences in activation of the LCAT reaction. The dissociation of LCAT from the particles is reflected by the Km value, which is comparable to that of the apoA-I complexes, suggesting that the peptide complexes have a comparable affinity for LCAT.
Cholesterol Efflux From Lipid-Laden Cells
J774 macrophages were incubated with acetylated LDL at an apoB concentration of 100 μg/mL for 16 hours, and the free and esterified cholesterol levels of the cellular lipid extracts were quantified by HPLC.
After incubation of the cells with acetylated LDL, significant amounts of cholesterol accumulate in the lipid-laden cells, as the cholesterol levels increase up to 44 μg/2×106 cells in comparison with baseline levels of 9 μg/2×106 cells,18 consisting of 70% unesterified cholesterol. Cholesterol efflux in the medium occurred only as free cholesterol, as no CE could be detected in the medium by HPLC, and spontaneous cholesterol efflux in the absence of acceptor particles was not observed. On cellular cholesterol efflux, the intracellular CE was hydrolyzed, as the amount of cellular CE and free cholesterol both decreased. The efflux properties of lipid/protein complexes, consisting of either purified human apoA-I, apoA-II, apoE, or synthetic peptides, are represented in Fig 6⇓. These data show that at the same protein concentration, the complexes generated with the 18A and the ID peptides (Fig 6A⇓) are more efficient to promote the efflux of free cholesterol into the medium than those formed with the purified human apolipoproteins (Fig 6B⇓). The cholesterol efflux induced by the apoE/phospholipid and apoA-II/phospholipid complexes is similar to and lower than that of apoA-I/phospholipid complexes. Up to 50% of the total cholesterol content of the lipid-laden cells was recovered in the cell-culture medium after incubation with the peptide/phospholipid complexes. The cholesterol efflux from the macrophages induced by the ID peptides, especially the ID3, was slightly higher than with the 18A peptide. This difference is not significant, however, as the experimental error on the efflux measurements amounts to ≈15%.
In this paper we report on the structure and properties of a new series of synthetic amphipathic helical peptides, which were designed as homologues of the helical repeats of the plasma apolipoproteins. Their physicochemical properties, LCAT activation, and ability to promote cellular cholesterol efflux were compared with those of similar peptides studied previously.17 The lipid-binding properties of these peptides were similar to those of the 18A and 18AM4 model peptides,15 17 but their physiological activity was different. The novel ID peptides do not activate the LCAT enzyme, but they are able to promote cholesterol efflux from lipid-laden macrophages to about the same extent as the 18A peptide and are even more efficient than the apoA-I, apoA-IV, and apoE proteins. Given the similarity of the structural properties of the new ID peptides with the 18A and the 18AM4 peptide and with the apoA-I helical repeats, these results suggest that LCAT activation must be more sequence specific than lipid binding or cholesterol efflux. Several steps are probably involved in the LCAT activation by its natural cofactors apoA-I or apoA-IV or by synthetic peptide analogues. The LCAT-activating domains of the peptides or apolipoproteins may (1) mediate the binding of LCAT to its substrate; (2) “activate” the lipid substrate by promoting the correct orientation of the sn-2 ester bond of phosphatidylcholine for presentation to the catalytic site of the enzyme and the cholesterol molecule such that its esterification can occur; or (3) interact directly with the LCAT through protein/protein interactions, which might induce a conformational change of the enzyme.27 As the lipid-binding properties of the peptides, due mostly to hydrophobic interactions, are similar, the differences in LCAT activation between the 18A and the ID peptides must arise from differences in charged residues. Such residues might be involved in protein/protein interactions between the peptides and a specific domain of the LCAT enzyme. Compared with the 18A and 18AM4 peptides, the ID peptides contain only one negatively charged residue, at position 10 in the central part of their polar face. The four Glu residues at positions 1, 8, 12, and 16 of the 18A peptide are directed toward the aqueous phase, whereas in the ID3, ID7, and ID9 peptides, only Glu10 points toward the outside of the hydrophilic face of the helix. The reversed 18A peptide (18AR), in which the position of the Lys and Glu residues was switched, decreased LCAT activation compared with the original 18A peptide in one study,14 while in our study, the complexes generated with the 18A (blocked peptide) reached only 40% of the activity of the apoA-I/lipid complexes.9 Most helical repeats of apoA-I have a similar topography of negatively charged residues forming a central band in the hydrophilic face of the helical repeat, as described by Brouillette and Anantharamaiah.14 Several site-directed mutagenesis studies of apoA-I28 29 and studies with synthetic30 and natural4 apoA-I fragments have suggested that the helices of the central domain of apoA-I at residues 102-183 are involved in LCAT activation, while the C-terminal helices 190-223 are essential for lipid binding.28 29 Within the central domain of apoA-I, helices 102-118 and 124-160, have the typical configuration described above; the Asp and Glu residues at position 102, 110, 111, and 113 in the first helix and at position 125 and 136 in the next one are located in the middle of the polar face and are accessible for interactions with another protein. Deletion of the pair of helices 117-160 of apoA-I28 29 decreased LCAT activation of the deletion mutant by 75%. Deletion of helix 142-158 of apoA-IV, where E145, E150, and D153 are located in the middle of the polar face of the helix and directed toward the aqueous phase, might have contributed to the decreased LCAT activity of the apoA-IV mutant.31
These results are further supported by the observations of Subbarao et al,32 who synthesized the GALA peptide, a 30-mer with an α-helical conformation and a hydrophilic face consisting of Glu residues. This peptide was able to bind DMPC and could further activate LCAT to about 80% activity compared with apoA-I. A reversed peptide with no acidic residues in the hydrophilic face had lost all LCAT activity. These authors, moreover, clearly emphasized that the lack of Lys residues at the boundary between hydrophobic and hydrophilic faces did not influence the LCAT activity of the peptide.32 Negatively charged residues in the central part of the amphipathic helices seems, therefore, critical for LCAT activation by these peptides.
The lack of selectivity of the peptide or apolipoprotein sequences for inducing cellular cholesterol efflux observed in our study agrees with recent literature reports.19 21 Yancey et al19 have shown in studies with the 18A, the 18AR, and the 18A-Pro-18A that cholesterol efflux from the lipid-laden macrophages is not affected by the length of the sequence, the number of helices, or the lipid-binding capacity of these peptides. The efficiency of the peptides to induce cholesterol efflux was also not related to their amino acid sequence, suggesting that cholesterol efflux is not mediated through a specific cell-surface binding site.21 These authors further demonstrated that lipid-free peptides are able to take up phospholipids and subsequently cholesterol from the membrane. Hara et al33 performed cholesterol efflux experiments with plasmatic apoA-I, apoA-II, apoE, apoA-IV, and apoC-III and with apolipophorin III of Manducta sexta. They concluded that at least four amphiphilic helical segments per molecule are required for efficient cholesterol efflux. According to our results, peptide/phospholipid complexes have a higher capacity as cholesterol acceptors and are able to incorporate more cholesterol than apolipoprotein/phospholipid complexes. This is probably not related to the small differences in size of the complexes but rather to differences in the apolipoprotein/phospholipid and peptide/phospholipid interactions.
In conclusion, these novel synthetic amphipathic peptides, which have fewer negatively charged residues in the hydrophilic face of the helix, can generate discoidal complexes with phospholipids of similar composition and size to those generated with other model peptides. However, the ID/phospholipid/cholesterol complexes are not optimal substrates for LCAT, in contrast to those formed with the 18A or the GALA peptide. These data, therefore, suggest that ionic interactions between negatively charged residues of the cofactor peptide or protein with basic residues of LCAT might be required for the optimal conformation of the enzyme in its activated state. According to our data, the nature of the peptide is not critical for cholesterol efflux, as all peptide/phospholipid complexes were able to induce comparable efflux of free cholesterol from the cells and were even more efficient than apolipoprotein/phospholipid complexes.
Selected Abbreviations and Acronyms
|HPLC||=||high-performance liquid chromatography|
|LCAT||=||lecithin:cholesterol acyl transferase|
Brasseur R. Differentiation of lipid associating helices by use of the 3-dimensional hydrophobicity potential. J Biol Chem. 1991;266:16120-16127.
Vanloo B, Morrison J, Fidge N, Lorent G, Brasseur R, Ruysschaert JM, Baert G, Rosseneu M. Activation of lecithin cholesterol acyl transferase (LCAT) by the CNBr fragments of human apo AI and conversion of the discoidal complexes into spherical particles. J Lipid Res. 1991;32:1253-1264.
Jonas A, Hefele Wald J, Harms Toohill KL, Krul ES, Kézdy KE. Apolipoprotein A-I structure and lipid binding properties in homogeneous reconstituted spherical and discoidal high density lipoproteins. J Biol Chem. 1990;265:22123-22129.
Jonas A, Kézdy KE, Hefele Wald J. Defined apolipoprotein conformations in reconstituted high density lipoprotein discs. J Biol Chem. 1989;264:4818-4824.
Jonas A. Lipid-properties of apolipoproteins. In: Rosseneu M, ed. Structure and Function of Plasma Apolipoproteins. NW Boca Raton, Fla: CRC Press Inc; 1992;chap 8:217-250.
Spuhler P, Anantharamaiah GM, Segrest JP, Seelig J. Binding of apolipoprotein A-I model peptides to lipid bilayers: measurement of binding isotherms and peptide-lipid head groups interactions. J Biol Chem. 1994;269:23904-23910.
Sparrow JT, Gotto AM. Apolipoprotein/lipid interactions: studies with synthetic polypeptides. Crit Rev Biochem. 1981;13:87-107.
Yokohama S, Fukushima D, Kupferberg JP, Kézdy FJ, Kaiser ET. The mechanism of activation of lecithin:cholesterol acyl transferase by apolipoprotein A-I and an amphiphilic peptide. J Biol Chem. 1980;255:7333-7339.
Rosseneu M, Devreese A, Vanloo B, Baert J. Functional properties of discoidal apoprotein-phospholipid complexes: cholesterol efflux from macrophages and conversion by lecithin cholesterol acyl transferase. In: Sirtori CR, ed. Human Apolipoprotein Mutants, III. Springer-Verlag: Berlin, Germany; 1993:21-30.
Jonas A, McHugh HT. Reaction of lecithin:cholesterol acyl transferase with micellar complexes of apo A-I and phosphatidylcholine, containing various amounts of cholesterol. J Biol Chem. 1983;258:10335-10340.
Davidson WS, Lund-Katz S, Johnson WJ, Anantharamaiah GM, Palgunachari MN, Segrest JP, Rothblat GH, Phillips MC. The influence of apolipoprotein structure on the efflux of cellular free cholesterol to high density lipoproteins. J Biol Chem. 1994;269:22975-22982.
Brasseur R, Ruysschaert JM. Conformation and mode of organisation of amphiphilic membrane compounds: a theoretical study on conformational-related activity. Biochem J. 1986;238:1-11.
Matz CE, Jonas A. Micellar complexes of human apolipoprotein A-I with phosphatidylcholines and cholesterol prepared from cholate-lipid dispersions. J Biol Chem. 1982;257:4541-4546.
Dolphin P. Lipolytic enzymes and the role of apolipoproteins in the regulation of their activity. In Rosseneu M, ed. Structure and Function of Plasma Apolipoproteins. NW Boca Raton, Fla: CRC Press Inc; 1990;chap 11:295-362.
Sorcii-Thomas M, Kearns MW, Lee JP. Apolipoprotein A-I domains involved in lecithin-cholesterol acyltransferase activation: structure-function relationship. J Biol Chem. 1993;268:21403-21409.
Holvoet P, Zhao Z, Vanloo B, Deridder E, Dhoest A, Taveirne J, Brouwers E, Demarsin E, Engelborghs Y, Rosseneu M, Collen D, Brasseur R. Phospholipid binding and lecithin: cholesterol acyltransferase (LCAT) activating properties of human apolipoprotein A-I deletion mutants. Biochemistry. 1995;34:13334-13342.
Vanloo B, Demoor L, Boutillon C, Lins L, Brasseur R, Baert J, Fruchart JC, Tartar A, Rosseneu M. Association of synthetic peptides of human apolipoprotein AI with phospholipids. J Lipid Res. 1995;36;1501-1512.
Emmanuel F, Steinmetz A, Rosseneu M, Brasseur R, Gosselet N, Attenot F, Cuin S, Seguret S, Latta M, Fruchart JC. Denèfle P. Identification of specific alpha-helical sequence of human apolipoprotein A-IV involved in lecithin cholesterol acyltransferase activation. J Biol Chem. 1994;269:29883-29890.