Portal de Periódicos da CAPES

The differential apoA-I enrichment of preβ1 and αHDL is detectable by gel filtration separation

2002; Elsevier BV; Volume: 43; Issue: 11 Linguagem: Inglês

10.1194/jlr.d200024-jlr200

1539-7262

Maud Chétiveaux, Hassan Nazih, Véronique Ferchaud‐Roucher, Gilles Lambert, Yassine Zaïr, Martial Masson, Khadija Ouguerram, Danièle Bouhours, Michel Krempf,

Diabetes Management and Research

The aim of the study was to assess the isolation of HDL by fast protein liquid chromatography (FPLC) to perform kinetics studies of apolipoprotein (apo)A-I-HDL labelled with a stable isotope. Comparison between FPLC and ultracentrifugation has been made. ApoA-I-HDL kinetics were studied by infusion of [5.5.5-2H3]leucine for 14 h in five subjects. Using FPLC, preβ1 HDL and αHDL (HDL2 and HDL3) were separated from 200 μl of plasma samples. Total HDL was isolated by sequential ultracentrifugation (HDL-UC). The tracer-to-tracee ratio was higher in preβ1 HDL than in total HDL-UC. The higher leucine enrichment found in total HDL-UC compared to αHDL suggested the existence of a mixture of apoA-I-HDL sub-classes. From this difference in enrichments, the turnover rate of total HDL-UC, usually assumed to be αHDL, was probably overestimated in previous studies. To our knowledge, this study is the first report which provides a convenient tool to distinguish enrichments of apoA-I in preβ1 HDL and αHDL from total HDL previously used for kinetic measurements.This original and new method should help to understand the kinetics of HDL in humans and the reverse cholesterol transport dynamics. The aim of the study was to assess the isolation of HDL by fast protein liquid chromatography (FPLC) to perform kinetics studies of apolipoprotein (apo)A-I-HDL labelled with a stable isotope. Comparison between FPLC and ultracentrifugation has been made. ApoA-I-HDL kinetics were studied by infusion of [5.5.5-2H3]leucine for 14 h in five subjects. Using FPLC, preβ1 HDL and αHDL (HDL2 and HDL3) were separated from 200 μl of plasma samples. Total HDL was isolated by sequential ultracentrifugation (HDL-UC). The tracer-to-tracee ratio was higher in preβ1 HDL than in total HDL-UC. The higher leucine enrichment found in total HDL-UC compared to αHDL suggested the existence of a mixture of apoA-I-HDL sub-classes. From this difference in enrichments, the turnover rate of total HDL-UC, usually assumed to be αHDL, was probably overestimated in previous studies. To our knowledge, this study is the first report which provides a convenient tool to distinguish enrichments of apoA-I in preβ1 HDL and αHDL from total HDL previously used for kinetic measurements. This original and new method should help to understand the kinetics of HDL in humans and the reverse cholesterol transport dynamics. Plasma HDL and apolipoprotein A-I (apoA-I) concentrations are inversely correlated with the risk of coronary heart disease (1Gordon T. Castelli W.P. Hjortland M.C. Kannel W.B. Dawber T.R. High Density Lipoprotein as a protective factor against coronary heart disease: the Framingham study.Am. J. Med. 1977; 62: 707-714Google Scholar, 2Forte T.M. McCall M.R. The role of apolipoprotein AI-containing lipoproteins in atherosclerosis.Curr. Opin. Lipidol. 1994; 5: 354-364Google Scholar). Kinetics studies are a good tool for a better understanding of the role that HDL play in the reverse cholesterol transport. Preβ1 HDL are considered as the initial acceptor of cellular cholesterol and are transformed into spherical HDL3 as a result of the LCAT activity. Nascent discoidal preβ1 HDL are either directly synthesized by the liver and the intestine or generated when other classes of lipoproteins including αHDL are remodeled by hepatic lipase (3Barrans A. Collet X. Barbaras R. Jaspard B. Manent J. Vieu C. Chap R. Perret B. Hepatic lipase induces the formation of pre-beta1 high density lipoprotein from triacylglycerol-rich HDL2.J. Biol. Chem. 1994; 269: 11572-11577Google Scholar). In plasma, preβ1 HDL represented 5.5 ± 3.3% and 7.2 ± 4% of total apoA-I in women and men, respectively (4O'Connor P.M. Zysow B.R. Schoenhaus S.A. Ishida S.T. Kunitake S.T. Naya-Vigne J.M. Duchateau P.N. Redberg R.F. Spencer S.J. Mark S. Mazur M. Heilbron D.C. Jaffe R.B. Malloy M.J. Kane J.P. Prebeta-1 HDL in plasma of normolipidemic individuals: influences of plasma lipoproteins, age, and gender.J. Lipid Res. 1998; 39: 670-678Google Scholar). ApoA-I, a major protein component of HDL, could be endogenously labeled using stable isotopes to study HDL metabolism in humans (5Ikewaki K. Rader D.J. Schaefer E.J. Fairwell T. Zech L.A. Brewer H.B. Evaluation of apo A-I kinetics in humans using simultaneous endogenous stable isotope and exogenous radiotracer methods.J. Lipid Res. 1993; 34: 2207-2215Google Scholar). Plasma lipoprotein and HDL are usually separated by ultracentrifugation based on differential flotation using either sequential (6Havel R.J. Eder H.A. Bragdon J.H. The distribution and chemical composition of ultracentrifugation separated lipoproteins in human serum.J. Clin. Invest. 1955; 34: 1345-1353Google Scholar) or one-step gradient methods (7Chapman M.J. Goldstein S. Lagrange D. Laplaud P.M. A density gradient ultracentrifugal procedure for the isolation of the major lipoprotein classes from human serum.J. Lipid Res. 1981; 22: 339-358Google Scholar). It appears that the combination of high salt concentrations and high centrifugal forces may cause structural damages to the lipoproteins (8Kunitake S.T. Kane J.P. Factors affecting the integrity of high density lipoproteins in the ultracentrifuge.J. Lipid Res. 1982; 23: 936-940Google Scholar) that could change results of kinetics studies (9Cheung M.C. Wolf A.C. Differential effect of ultracentrifugation on apolipoprotein A-I-containing lipoprotein subpopulations.J. Lipid Res. 1988; 29: 15-25Google Scholar). Alternatively, plasma lipoproteins may also be separated by size-exclusion chromatography on agarose. Fast Protein Liquid Chromatography (FPLC), based on the gel filtration property, is a rapid, quantitative, and non-destructive method used to separate lipoproteins. Nascent preβ1 HDL can also be separated from αHDL by FPLC (10Nanjee M.N. Brinton E.A. Very small apolipoprotein A-I-containing particules from human plasma : isolation and quantification by high-performance size-exclusion chromatography.Clin. Chem. 2000; 46: 207-223Google Scholar). The aim of the present study is to validate the separation of HDL sub-classes by FPLC in metabolic study after an endogenous labeling with stable isotopes. In this study, we have compared the leucine enrichment of apoA-I-HDL separated by FPLC to that of HDL separated by the traditional ultracentrifugation of plasma samples. The experimental protocol was approved by the ethics committee of the Nantes University Hospital and written consents were obtained from each volunteer before inclusion in the study. The kinetic protocol was already described in a previous study (11Frenais R. Ouguerram K. Maugeais C. Mahot P. Maugere P. Krempf M. Magot T. High density lipoprotein apolipoprotein AI kinetics in NIDDM: a stable isotope study.Diabetologia. 1997; 40: 578-583Google Scholar). Endogenous labelling of apoA-I was performed by administration of l-[5.5.5-2H3]leucine (99.8 atom %; Mass Trace, Woburn, MA), dissolved in a 0.9% NaCl solution, and tested for sterility and pyrogenicity before the study. All subjects fasted overnight for 12 h prior to the study, and remained fasting for the entire procedure. They received intravenously 10 μmol.kg-1 of tracer, immediately followed by a constant tracer infusion (10 μmol.kg−1.h−1) for 14 h. Venous blood was drawn into EDTA tubes (Venoject, Paris, France) at baseline every 15 min during the first hour, every 30 min during the next 2 h, and then hourly until the end of the study. Plasma was immediately separated by centrifugation for 30 min at 4°C; sodium azide, an inhibitor of bacterial growth, and Pefabloc SC (Interchim, Montluçon, France), a protease inhibitor, were added to blood samples at final concentrations of 1.5 and 0.5 mmol/l, respectively. Plasma lipoproteins were isolated by ultracentrifugation (6Havel R.J. Eder H.A. Bragdon J.H. The distribution and chemical composition of ultracentrifugation separated lipoproteins in human serum.J. Clin. Invest. 1955; 34: 1345-1353Google Scholar) using a LKB ultracentrifuge with a RP55T rotor at 40,000 rpm for 24 h at 4°C. The density of the plasma was adjusted to the density of each lipoprotein (1.006, 1.063, and 1.025 g/ml for VLDL, LDL, HDL respectively) with NaCl solutions. Then, the lipoprotein fractions were dialysed against NaCl 0.15M for 24 h at 4°C. FPLC analysis was performed at room temperature using two P-500 pumps (1 MPa), a MV-7 place multi-injection with a 200 μl loop, an ultraviolet (UV) monitor UV-1, a FRAC-100 fraction collector, and a LCC-501 Plus controller. All operations were monitored by absorbance at 280 nm. The system was controlled by FPLC DIRECTOR software (Amersham Pharmacia Biotech Inc., Orsay, France) (12Ordovas J.M. Osgood D. Preparative isolation of plasma lipoproteins using fast protein liquid chromatography.Methods Mol. Biol. 2000; 110: 105-111Google Scholar, 14Van Gent T. Van Tol A. Automated gel permeation chromatography of plasma lipoproteins by preparative fast protein liquid chromatography.J. Chromatogr. 1990; 525: 433-441Google Scholar, 15Innis-Whitehouse W. Li X. Brown W.V. Le N-A. An efficient chromatographic system for lipoprotein fractionning using whole plasma.J. Lipid Res. 1998; 39: 679-690Google Scholar, 16Le N-A. Innis-Whitehouse W. Li X. Bakker-Arkema R. Black D. Brown W.V. Lipid and apolipoprotein levels and distribution in patients with hypertriglyceridemia : effect of triglyceride reductions with atorvastatin.Metabolism. 2000; 49: 167-177Google Scholar). Elution was performed in a 1 mM EDTA, 154 mM NaCl, and 0.02% NaN3 (pH 8.2) buffer. The buffer was filtered through a 0.22 μm filter (Membrane filters type GV, Duropore®, Millipore). The column was washed with an aqueous ethanol solution (20%). Total HDL separations (total-HDL-FPLC) by FPLC were performed on two Superose 6 HR 10/30 columns in series (Amersham Pharmacia Biotech Inc., Orsay, France). Two hundred microliters of plasma was injected and elution was performed at a flow of 0.35 ml/min. Approximately 0.35 ml fraction were collected. An entire profile was completed within 105 min. The columns were washed for 50 min at the end of each run. ApoA-I-containing HDL sub-classes separations by FPLC were performed on a Superdex 200 HR 10/30 column (Amersham Pharmacia Biotech Inc., Orsay, France) (10Nanjee M.N. Brinton E.A. Very small apolipoprotein A-I-containing particules from human plasma : isolation and quantification by high-performance size-exclusion chromatography.Clin. Chem. 2000; 46: 207-223Google Scholar). Samples of 200 μl of plasma were injected and elution was performed at a flow of 0.35 ml/min. Approximately 0.2 ml was collected for each fraction. An entire profile was completed within 50 min. The columns were washed for 50 min at the end of each run. First, human VLDL, LDL, HDL2, and HDL3 isolated by ultracentrifugation were used to determine the elution volumes of the various lipoproteins isolated by FPLC. In another way, lipids and apolipoproteins levels were measured in each FPLC fractions. Total cholesterol and triglycerides contents of lipoproteins and choline containing phospholipids in HDL subpopulations were measured using enzymatic kits (Biomerieux, Marcy-L'Etoile, France). ApoA-I and apoB-100 concentrations were measured by immunoturbidimetry (Biomerieux, Marcy-L'Etoile, France). ApoA-I of HDL isolated by FPLC and ultracentrifugation was then separated from other proteins by sodium dodecylsulfate polyacrylamide gel electrophoresis (SDS-PAGE) using a 4% to 20% discontinuous gradient and 4% stacking gel. Electrophoresis was run overnight at 70 V and 4°C in Tris-Glycine buffer. ApoA-I was identified by comparing with molecular weight standards (Biorad, Ivry-Sur-Seine, France). ApoA-I bands were excised from the polyacrylamide gels and hydrolysed for 24 h in 4 N HCl at 110°C. Hydrolysates were evaporated to dryness and amino acids were purified by cation exchange chromatography using a Dowex 50WX8-200 resin (Aldrich, Saint-Quentin Fallavier, France). Amino acids were esterified with propanol/acetyl chloride and derivatized using heptafluorobutyric anhydride (HFBA, Fluka, Saint-Quentin Fallavier, France). Electron-impact ionization gas chromatography-mass spectrometry (GC-MS) was performed on a 5890 A gas chromatograph connected with a 5971 A quadrupole mass spectrometer (Hewlett-Packard, Palo Alto, CA, USA). The isotopic ratio was determined by selected ion monitoring at m/z of 282 and 285. Non-denaturing two-dimensional PAGE was realized on HDL isolated either by FPLC or by ultracentrifugation, as well as on αHDL and preβ1 HDL to validate their separations. Elution volume of albumin (67 KDa) was used as exclusion limit between these two populations. The one-dimensional electrophoresis was performed on a low endosmosity agarose gel, 0.7% (Agarose, Type I: Low EEO, Sigma, Saint-Quentin Fallavier, France) in tris-tricine buffer (25 mM, pH 8.6) containing 3 mM calcium lactate at constant voltage (250 V) at 10°C until bromophenol blue run 3.5 cm. After electrophoresis was completed in the one-dimensional, the agarose strip was cut and placed on polyacrylamide gel as well as molecular weight standards (High molecular weight electrophoresis, Amersham Pharmacia Biotech Inc., Orsay, France). These agarose strips were sealed with the same agarose. The two-dimensional electrophoresis was performed in a 2% to 36% non-denaturing, concave gradient polyacrylamide gel at 220 V for 24 h at 4°C in a buffer containing 90 mM tris, 80 mM boric acid, and 2.5 mM EDTA (pH 8.3). After electrophoresis, lipoproteins were transferred to a 0.45 μm pore size nitrocellulose membrane (Membrane Protran BA85, Schleicher and Schuell, Dassel, Germany) at 30 V for 24 h at 10°C in a Tris-Glycine buffer containing 25% methanol. After transfer, the bands with molecular weight standards were cut and stained with red Ponceau and the membrane was completely wetted and blocked in phosphate-buffer saline (PBS) containing 3% BSA for 2 h. Then, the membrane was incubated for 2 h with the rabbit anti-human apoA-I (apoA-I, Lpu 1 mg, Institut Pasteur Lille, France) at a dilution of 1:1,000 in PBS with 1% BSA. After primary antibody incubation, the membrane was washed three times (10 min each) in PBS and then incubated with a secondary goat-anti-rabbit IgG peroxidase conjugate at a dilution of 1:1,000 for 2 h in a PBS-BSA (1%) solution. The membrane was then washed in PBS and peroxidase reaction was performed with 4 Chloro-1 Naphtol method (Sigma, Saint-Quentin Fallavier, France). The purity of apoA-I contained in HDL subpopulations was determined with a denaturing two-dimensional gel electrophoresis: isoelectric focusing (IEF)/SDS-PAGE; process which allowed the separation of proteins according to their Isoelectric point (pI) and their molecular mass. All instrumentation, gels and protocols were from Amersham Biosciences. 50 ng of αHDL and 100 ng of preβ1 HDL were diluted to a final volume of 200 μl in a solution containing 8 M urea, 2% CHAPS, 18 mM DTT, 0.2% Pharmalyte 4-7, and a few granules of bromophenol blue. IEF was carried out with Immobiline DryStrip gels (IPG, pH 4–7, 11 cm; and pH 5–10, 18 cm) in the IPFphor isoelectric focusing system, according to the manufacturer's protocol. Two-dimensional SDS-PAGE was performed using ExcelGel 2-D Homogeneous 12.5% in the Multiphor II electrophoresis unit, according to the manufacturer recommendations. After completion of the two-dimensional electrophoresis, the gels were silver stained as described by Heukeshoven and Dernick (17Heukeshoven J. Dernick R. Improved silver staining procedure for fast staining in PhastSystem Development Unit. I. Staining of sodium dodecyl sulfate gels.Electrophoresis. 1988; 9: 28-32Google Scholar). When human plasma samples were analysed on two Superose 6HR gel permeation columns, four distinct peaks of lipoproteins were identified: VLDL, LDL, HDL2, and HDL3 (Fig. 1A). The elution profiles of VLDL (Fig. 1B), LDL (Fig. 1C), HDL2 (Fig. 1D), and HDL3 (Fig. 1E) isolated by ultracentrifugation and applied to FPLC in four different runs are shown in Fig. 1. In Fig. 2, the elution profiles of cholesterol, triglycerides (Fig. 2A), apoB (Fig. 2B), and ApoA-I (Fig. 2C) were analysed in order to identify and validate the fractions corresponding to VLDL, LDL, and HDL.Fig. 2FPLC elution profiles on two Superose 6 HR for a representative plasma sample as assessed by cholesterol and triglycerides (A) or apolipoprotein B (apoB) (B) or apoA-I (C).View Large Image Figure ViewerDownload (PPT) The tracer-to-tracee ratio curves in total HDL isolated by FPLC and ultracentrifugation are shown in Fig. 3. The tracer-to-tracee ratio for ApoA-I-HDL after FPLC were less than 35% compared to HDL isolated by ultracentrifugation for all subjects. To rule out a possible retention of the tracer in the FPLC columns, we measured the tracer-to-tracee ratio of HDL-UC before and after FPLC as a control experiment. The tracer-to-tracee ratio curves were similar in these two different conditions (data not shown). These data indicated that the difference in the tracer-to-tracee ratio observed between ultracentrifugation and FPLC separation methods was related to the composition of HDL. To test this hypothesis, we have performed a two-dimensional electrophoresis of our HDL samples. Figure 4shows the two-dimensional electrophoresis of total HDL isolated by ultracentrifugation and FPLC followed by apoA-I immunoblotting. After ultracentrifugation preβ HDL was not found and only HDL with α electrophoretic mobility and free apoA-I was observed. Conversely, preβ as well as αHDL were recovered after FPLC separation of human plasma. When human plasma samples were separated by FPLC on a Superdex 200 HR gel permeation, three apoA-I-HDL subclasses, as well as three distinct phospholipids containing lipoproteins, were separated (Fig. 5). The first peak represented a minor population of the largest particles >500 KDa (fraction 1 to 13). The second peak included a spherical αHDL of 100–500 KDa: HDL2 and HDL3. The third peak included a minor population of smaller particles of 40-60 KDa: preβ1 HDL. The purity of this fractions was confirmed by a two-dimensional non denaturing electrophoresis and a immunoblotting methods. With the denaturing two-dimensional gel electrophoresis, in both preβ1 HDL and αHDL fractions, we did not observe another protein having the same molecular mass as apoA-I and a different pI. We could conclude that the apoA-I band is not contaminated by another protein. The only doubt was in αHDL fraction; Ig light chains were localised above apoA-I spot and could contaminate the isotopic ratio. To extract Ig from αHDL, 2.5 μg of αHDL were incubated for 7h at 4°C with rotation with protein A-Agarose, protein that bind to immunoglobulin classes. The Ig-protein A-Agarose complex was precipitated and the supernatant was separated on SDS-PAGE. We have obtained the same isotopic ratio for αHDL incubated or not with protein A-Agarose for several time of kinetic. Thus, we could conclude that apoA-I-αHDL was not contaminated by Ig light chains (data not shown). The tracer-to-tracee ratio of αHDL, preβ1 HDL isolated by FPLC, and total HDL isolated by ultracentrifugation were established on five subjects (Fig. 6). Experimental values of tracer-to-tracee ratio of preβ1 HDL and αHDL were higher and lower, respectively, than those of total HDL. The aim of this study was to compare two methods to separate HDL for the measurements of isotopic enrichment in apoA-I-HDL. HDL was isolated by standard sequential ultracentrifugation and by FPLC from plasma samples of five subjects who received a constant infusion of deuterated leucine during 14 h. The FPLC technique allowed the isolation of preβ1 HDL from αHDL, and detectable isotopic enrichments in both fractions were found with GC-MS analysis. To our knowledge, this study is the first to distinguish enrichments data of preβ1 HDL and αHDL kinetic compared to total HDL usually isolated in previous studies for kinetic measurements. The linearity, reproducibility, and sensitivity of FPLC technique has been demonstrated previously to separate lipoproteins (16Le N-A. Innis-Whitehouse W. Li X. Bakker-Arkema R. Black D. Brown W.V. Lipid and apolipoprotein levels and distribution in patients with hypertriglyceridemia : effect of triglyceride reductions with atorvastatin.Metabolism. 2000; 49: 167-177Google Scholar, 18Kieft K.A. Bocan T.M.A. Krause B.R. Rapid on-line determination of cholesterol distribution among plasma lipoproteins after high-performance gel filtration chromatography.J. Lipid Res. 1991; 32: 859-866Google Scholar, 19Ha Y.C. Barter P.J. Rapid separation of plasma lipoproteins by gel permeation chromatography on agarose gel superose 6B.J. Chromatogr. 1985; 341: 154-159Google Scholar). With FPLC, the samples were not exposed during the separation process to the high g-forces and salt concentrations related to ultracentrifugation (12Ordovas J.M. Osgood D. Preparative isolation of plasma lipoproteins using fast protein liquid chromatography.Methods Mol. Biol. 2000; 110: 105-111Google Scholar) that may shed some of the HDL components from the particle surface (e.g., apoA-I) (9Cheung M.C. Wolf A.C. Differential effect of ultracentrifugation on apolipoprotein A-I-containing lipoprotein subpopulations.J. Lipid Res. 1988; 29: 15-25Google Scholar). There was a progressive loss of apoA-I of about 35% with repeated ultracentrifugation of HDL at a density of 1.21 g/ml (8Kunitake S.T. Kane J.P. Factors affecting the integrity of high density lipoproteins in the ultracentrifuge.J. Lipid Res. 1982; 23: 936-940Google Scholar). Moreover, this latter procedure required approximately 1 ml to 3 ml of plasma for analysis, whereas only 200 μl were necessary with FPLC to isolate preβ and αHDL and to determine lipids and apolipoproteins concentrations in the eluted fractions. With the increasing use of the transgenic and knockout mouse models of study atherosclerosis, the ability to separate lipoproteins classes from small volumes of whole plasma is interesting. However, separation using gel filtration method resulted in sample dilutions. The experimental values of tracer-to-tracee ratio for apoA-I-HDL after FPLC were less than those of HDL isolated by ultracentrifugation. This difference could be explained by the difference in the separation process with ultracentrifugation and FPLC and thus we did not analyse the same biological materials. Indeed, effects of ultracentrifugation separation on apoA-I subpopulations were dramatic. As demonstrated by Asztalos et al. (20Asztalos B.F. Sloop C.H. Woog L. Roheim P.S. Two-dimensional electrophoresis of plasma lipoproteins: recognition of new apo A-I containing subpopulations.Biochim. Biophys. Acta. 1993; 1169: 291-300Google Scholar), after ultracentrifugation only free apoA-I and HDL with α mobility were found, indicating the loss of preβ1 and preβ2 from the bulk of flotating HDL. Free apoA-I could come from denaturation of different HDL subpopulations as small particles (preβ1 HDL). After FPLC, HDL with preβ1, preβ2, and α mobilities were detected. Therefore, FPLC reflects the biological diversity of lipoproteins better than ultracentrifugation. Preβ1 and αHDL were isolated by FPLC and the purity of each subpopulation was validated by two-dimensional PAGE. The purity of apoA-I contained in these lipoproteins was determined by a denaturing two-dimensional gel electrophoresis. The tracer-to-tracee ratios in preβ1 HDL were higher than those of total HDL and αHDL. The fast and higher enrichment of preβ1 HDL could be explained by the early role of these lipoproteins in the reverse transport. Total HDL obtained by ultracentrifugation was probably composed of principally αHDL and of small quantity of preβ1 HDL. Thus, in previous studies, based on ultracentrifugation, tracer-to-tracee ratios in apoA-I-HDL could have been overestimated by the presence of apoA-I-preβ1 HDL in αHDL population. The authors thank Mrs Pascale Maugère for her excellent technical assistance and Dr Jean-François Bouhours for critical reading of the manuscript.