Delineation of the Role of Pre-β1-HDL in Cholesterol Efflux Using Isolated Pre-β1-HDL
Objective— The role of pre-β1-high density lipoprotein (pre-β1-HDL) in cholesterol efflux was investigated by separating human plasma into purified pre-β1-HDL and pre-β1-HDL–deficient plasma by using a monoclonal antibody specifically reacting with pre-β1-HDL.
Methods and Results— When compared with whole plasma, pre-β1-HDL–deficient plasma was equally efficient in promoting cholesterol efflux from human skin fibroblasts and THP-1 human macrophage cells. When added at the same apolipoprotein A-I concentration, pre-β1-HDL was less effective than whole plasma in promoting cholesterol efflux from fibroblasts but equally effective in promoting cholesterol efflux from THP-1 cells. However, pre-β1-HDL–deficient plasma reconstituted with 16% pre-β1-HDL was more active than whole plasma, demonstrating that pre-β1-HDL does promote cholesterol efflux actively. The amount of cellular cholesterol present in reisolated pre-β1-HDL was 1.5- to 2-fold greater after incubation of the cells with whole plasma than after incubation of the cells with pre-β1-HDL–deficient plasma or plasma treated with the anti–pre-β1-HDL antibody. However, the anti–pre-β1-HDL antibody did not inhibit cholesterol efflux.
Conclusions— We conclude that whereas pre-β1-HDL is capable of taking up cellular cholesterol, its presence in plasma is not essential for cholesterol efflux, at least in vitro. Instead, pre-β1-HDL may be the first product of apolipoprotein A-I lipidation during the formation of HDL but may not play a major role in transferring cellular cholesterol to HDL.
Cholesterol efflux is the first step in reverse cholesterol transport, a pathway removing excess cholesterol from extrahepatic tissues, putatively protecting against the development of atherosclerosis (see review1). The preferred acceptor of cholesterol released from cells is HDL. One HDL subfraction, the lipid-poor discoid pre-β1-HDL, is believed to be the initial acceptor. On taking up cellular cholesterol, these particles enlarge into pre-β2-HDL particles, within which esterification of cholesterol progresses, leading to the formation of larger α-HDL.1–4⇓⇓⇓ Catabolism of α1-HDL occurs through the activities of hepatic lipase, cholesterol ester transfer protein, and phospholipid transfer protein to regenerate smaller α3-HDL particles and lipid-free apoA-I.5 The cycle is then renewed with relipidation of apoA-I by cellular phospholipid and cholesterol to form pre-β1-HDL particles. Although it is generally accepted that pre-β1-HDL participate actively in removing cellular cholesterol, these particles constitute only ≈7% to 10% of total apoA-I.6,7⇓ Furthermore, in cholesterol efflux experiments, substantial amounts of cellular cholesterol are transferred early to other lipoprotein subfractions.7–9⇓⇓ This could be interpreted as reflecting extremely fast cholesterol efflux, leading to rapid transformation of pre-β1-HDL into spherical HDL particles. Alternatively, spherical HDL particles also appear to be primary acceptors of cholesterol, and the role of pre-β1-HDL might be that of an intermediate in the pathway of the formation of spherical HDL particles. Therefore, pre-β1-HDLs may not be essential constituents of cholesterol transport from cells. For further understanding of this issue, we separated human plasma into pre-β1-HDL and pre-β1-HDL–deficient plasma by using a monoclonal antibody (MoAb) specifically recognizing pre-β1-HDL.10 Using 2 cell types, human skin fibroblasts and THP-1 human macrophages, we confirmed that isolated pre-β1-HDL is an active acceptor of cellular cholesterol but is not essential for cholesterol efflux, at least in vitro.
Human skin fibroblasts were grown in a CO2 incubator (5% CO2/95% air) in 75-cm2 flasks or 24-well cell culture clusters. Cells were confluent at the beginning of the experiments.
THP-1 monocytes were grown in a CO2 incubator in 75-cm2 flasks or 24-well cell culture clusters and maintained in RPMI 1640 medium containing 10% FCS, 2 mmol/L l-glutamine, 100 U/mL penicillin, 100 μg/mL streptomycin, and 0.02 μmol/L mercaptoethanol. To differentiate monocytes into macrophages, phorbol myristate acetate (Sigma) was added to a final concentration of 150 ng/mL, and cells were incubated for 72 hours at 37°C. Cells were confluent at the beginning of the experiments.
Isolation of Pre-β1-HDL and Pre-β1-HDL–Deficient Plasma
Blood from healthy volunteers was collected in saline containing streptokinase (final concentration 150 U/mL, Sigma), and plasma was isolated by repeated centrifugation at 4°C. To inhibit endogenous lecithin-cholesterol acyltransferase (LCAT), 5,5′-dithio-bis-(2-nitrobenzoic acid) (Sigma) was added to the final concentration of 2 mmol/L. Plasma (40 mL) was then passed through the Sepharose–mouse IgG column (10 mL) to remove fractions nonspecifically binding to mouse IgG. The eluant was then passed through the column containing Sepharose bound to MoAb 55201 (10 mL).10 The unbound fraction was designated pre-β1-HDL–deficient plasma. The material bound to the column was eluted with a solution containing 1 mol/L acetic acid and 0.15 mol/L NaCl; peak fractions were collected and designated pre-β1-HDL. Pre-β1-HDL was then transferred to PBS by passing through a Sephadex G25 column. Finally, to remove possible contamination with MoAb, pre-β1-HDL fractions were passed through a protein G–Sepharose column (10 mL). All procedures were carried out at 4°C. ApoA-I concentrations in whole plasma, pre-β1-HDL, and pre-β1-HDL–deficient plasma were determined by single radial immunodiffusion assay (Daiichi). Approximately 0.6% of serum apoA-I was recovered as pre-β1-HDL. Pre-β1-HDL concentrations were determined by using a pre-β1-HDL sandwich ELISA.10
HDL (density 1.063 to 1.21) was isolated by sequential ultracentrifugation, and human plasma lipid-free apoA-I was isolated as described previously.11
To label cellular cholesterol, cells were incubated in serum-containing medium with [1α,2α(n)-3H]cholesterol (specific radioactivity 1.81 TBq/mmol, final radioactivity 0.5 MBq/mL; Amersham-Pharmacia-Biotech) for 48 hours in a CO2 incubator. After they were labeled, cells were washed 6 times with Hanks solution containing 1% BSA. Cells were incubated for indicated periods of time at 37°C with serum-free medium containing plasma, apoA-I, or HDL subfractions at a final apoA-I concentration of 30 μg/mL. The medium was then collected and centrifuged for 15 minutes at 4°C at 30 000g, and aliquots of supernatant were counted in a β-counter. Cells were harvested and counted.
Nondenaturing 2D Electrophoresis
Nondenaturing 2D electrophoresis, transblotting, and Western blotting were performed as described previously6,9⇓ with few modifications. In brief, 20 μL of plasma sample was run on a 0.75% agarose gel in 50 mmol/L Merbital buffer (Merck), pH 8.6, at 4°C for 12 hours at 50 V. Strips of agarose gel were cut out, laid on the top of vertical slabs of 2% to 15% polyacrylamide gels, and fixed in position with agarose, and the second dimension was performed in 25 mmol/L Tris and 200 mmol/L glycine buffer, pH 8.3, for 2.5 hours at 200 V. Lipoproteins were then transferred to nitrocellulose filters, and Western blotting was performed by using rabbit polyclonal anti–apoA-I antibody and goat anti-rabbit IgG horseradish peroxidase–labeled secondary antibody (Chemicon).
Mapping the Epitope of MoAb
Human plasma apoA-I, recombinant mature and pro-apoA-I, and 5 mutated forms of recombinant pro-apoA-I, pro-apoA-I(Δ222-243), pro-apoA-I(Δ210-243), pro-apoA-I(Δ150-243), pro-apoA-I(Δ135 to 243), and pro-apoA-I(Δ140 to 150), were separated on a 12% SDS-PAGE gel, transferred to nitrocellulose membranes, and probed with MoAb 55201.
All experiments were performed in quadruplicate (ie, determination from 4 dishes). Background values for the cholesterol efflux, ie, the amount of radioactivity released to the medium in the absence of an acceptor, were subtracted. Mean±SEM values are presented. Statistical significance of differences was determined by the Student 2-tailed t test.
Affinity chromatography with use of the MoAb reacting specifically with pre-β1-HDL was used to separate plasma into pre-β1-HDL and pre-β1-HDL–deficient plasma. Native plasma, pre-β1-HDL, and pre-β1-HDL–deficient plasma were analyzed by using nondenaturing 2D electrophoresis. In native plasma, pre-β1-HDL constituted ≈8% of total apoA-I in HDL (Figure 1A). There was no pre-β1-HDL detected in pre-β1-HDL–deficient plasma (Figure 1B), whereas in the pre-β1-HDL preparation, apoA-I was found almost exclusively in pre-β1-HDL (Figure 1C). In addition, pre-β1-HDL–deficient plasma was analyzed by pre-β1-HDL ELISA, and no pre-β1-HDL was detected in pre-β1-HDL–deficient plasma by this method. Thus, within the sensitivity limits of 2 methods, pre-β1-HDL–deficient plasma did not contain pre-β1-HDL.
Cholesterol Efflux With Pre-β1-HDL–Containing and Pre-β1-HDL–Deficient Acceptors
To assess the activity of different HDL subfractions in promoting cholesterol efflux from fibroblasts, whole plasma, isolated pre-β1-HDL, pre-β1-HDL–deficient plasma, isolated HDL (composed mainly of α-HDL subfractions), and lipid-free apoA-I were added to human skin fibroblasts at the same apoA-I concentration (30 μg/mL) and incubated for 15 minutes. The short incubation time has been chosen to minimize the possible effect of HDL remodeling during incubation. Isolated pre-β1-HDL and lipid-free apoA-I (fractions that contained no spherical HDL particles) were able to promote cholesterol efflux, which was approximately half of that observed with whole plasma (P<0.05, Figure 2A). Pre-β1-HDL–deficient plasma and isolated HDL, fractions composed of predominantly spherical HDL particles, promoted cholesterol efflux that was not significantly different from that detected with whole plasma (Figure 2A).
On the other hand, in experiments with THP-1 cells, efflux of cholesterol to pre-β1-HDL and to pre-β1-HDL–deficient plasma was not significantly different from efflux to whole plasma (Figure 2B). However, in no instance was isolated pre-β1-HDL more active or pre-β1-HDL–deficient plasma less active than whole plasma when added at the same apoA-I concentration.
Time Course and Dose Dependence of Cholesterol Efflux
Time-course experiments of cholesterol efflux from human skin fibroblasts to whole plasma and pre-β1-HDL are shown in Figure 3A. The time course of cholesterol efflux consisted of 2 phases, a fast phase (between 0 and 20 minutes) and a slow phase (>20 minutes). During the fast phase, pre-β1-HDL was equal to or only slightly less effective in promoting cholesterol efflux than was whole plasma. In the slow phase, pre-β1-HDL was considerably less effective in promoting cholesterol efflux. The lesser activity of pre-β1-HDL in the slow phase may result from either inability to promote efflux from certain cholesterol pools or from the lower capacity of pre-β1-HDL to accept cellular cholesterol. It could also result from the loss of pre-β1-HDL activity during incubation.
To distinguish between these possibilities, the dose dependence of cholesterol efflux was studied (Figure 3B). At all concentrations, the efflux to pre-β1-HDL was less effective than that to whole plasma, and saturation was achieved at lower concentrations when pre-β1-HDL was used as an acceptor. This suggests that more cholesterol is available for efflux to plasma than to pre-β1-HDL.
Cholesterol Efflux to Reconstituted Plasma
The possibility cannot be excluded that the lower activity of pre-β1-HDL in cholesterol efflux is due to damage or remodeling of pre-β1-HDL during isolation. To investigate this possibility, fibroblasts or THP-1 cells were incubated with whole plasma, with pre-β1-HDL–deficient plasma, and with reconstituted plasma. Reconstituted plasma consisted of pre-β1-HDL–deficient plasma containing 25 μg/mL apoA-I, to which was added pre-β1-HDL, lipid-free apoA-I, or whole HDL to a final concentration of 30 μg/mL apoA-I. Cholesterol efflux from fibroblasts to pre-β1-HDL–deficient plasma reconstituted with pre-β1-HDL or lipid-free apoA-I was slightly but not significantly higher than cholesterol efflux to whole plasma or to pre-β1-HDL–deficient plasma added at the same apoA-I concentration (30 μg/mL, Figure 4A). However, reconstitution of pre-β1-HDL–deficient plasma with whole HDL increased its capacity to promote cholesterol efflux significantly (P<0.05) compared with whole plasma (Figure 4A). A similar experiment with THP-1 cells showed more clearly that reconstitution of pre-β1-HDL–deficient plasma with pre-β1-HDL or lipid-free apoA-I increased cholesterol efflux significantly (P<0.05), although the effect of reconstitution with HDL was less pronounced (Figure 4B). The results of these experiments demonstrate that isolated pre-β1-HDL is active when reconstituted with the rest of the plasma. Interestingly, the efficiency of pre-β1-HDL as a cholesterol acceptor appeared greater with THP-1 cells than with fibroblasts.
Reisolation of Pre-β1-HDL After Cholesterol Efflux
THP-1 cells were incubated for 1 hour with equal amounts of apoA-I in the form of whole plasma, pre-β1-HDL, pre-β1-HDL–deficient plasma, reconstituted plasma, or plasma preincubated with anti–pre-β1-HDL antibody. Pre-β1-HDL in the incubation medium was then reisolated by passing through the affinity column containing Sepharose–MoAb 55201. This column should retain pre-β1-HDL with [3H]cholesterol taken up during incubation with the cells, whereas other lipoprotein fractions should pass through the column. Approximately twice as much [3H]cholesterol was found to be associated with reisolated pre-β1-HDL in whole plasma than with pre-β1-HDL–deficient plasma or with plasma treated with the antibody (Figure 5A), confirming that plasma pre-β1-HDL was active in taking up cellular cholesterol. The amount of [3H]cholesterol in reisolated pure pre-β1-HDL was less than the amount in whole plasma. The capacity of pre-β1-HDL to remove cholesterol was also shown in the experiment with reconstituted plasma: addition of pre-β1-HDL to pre-β1-HDL–deficient plasma at a level of 16% resulted in a 50% increase in the amount of [3H]cholesterol residing in reisolated pre-β1-HDL, although this did not reach the level observed with whole plasma (Figure 5A).
The fact that a sizable amount of [3H]cholesterol was found in pre-β1-HDL after reisolation even from pre-β1-HDL–deficient plasma and from plasma treated with antibody may have resulted from nonspecific sorption of lipids on the column or may indicate that pre-β1-HDL had been formed during incubation and participated in cholesterol efflux. The proportion of pre-β1-HDL reisolated from plasma after incubation with cells was similar to that before incubation, indicating that there was no substantial net loss of pre-β1-HDL particles during incubation.
In another experiment, after incubation with cells, whole plasma or plasma preincubated with anti–pre-β1-HDL antibody was passed through a protein G–Sepharose column, which should retain complexes of pre-β1-HDL with antibody. Twice as much [3H]cholesterol was retained in the column when plasma was preincubated with the antibody (Figure 5B), again confirming that plasma pre-β1-HDL is an active acceptor of cellular cholesterol.
Effect of Anti–Pre-β1-HDL Antibody on Cholesterol Efflux
MoAb 55201 was also tested for its ability to affect cholesterol efflux. When added at a concentration of 500 μg/mL (3.3-fold molar excess), the antibody failed to inhibit the cholesterol efflux to whole plasma from fibroblasts or THP-1 cells (Figure 6) and the cholesterol efflux to pre-β1-HDL or pre-β1-HDL–deficient plasma (not shown). Fab fragments of the antibody also failed to inhibit cholesterol efflux from fibroblasts to whole plasma (not shown). To investigate why specific recognition of pre-β1-HDL was not accompanied by a reduction in cholesterol efflux (although as shown in previous experiments, the antibody could bind and remove pre-β1-HDL),10 we mapped the epitope for MoAb 55201. In the Western blot, the antibody reacted with human plasma apoA-I as well as with recombinant mature and pro forms of apoA-I (please see online Figure I, available at http://atvb.ahajournals.org). The antibody also reacted with apoA-I truncated to the residues 222 and 210 but failed to react when the truncation was extended to residues 150 and 135 (online Figure I), indicating that the carboxyl end of the recognized epitope is located between residues 150 and 210. The antibody also did not react with the mutant apoA-I(Δ140-150), indicating that the amino end of the epitope is most likely located between amino acids 140 and 150. Thus, the epitope for MoAb 55201 is located between residues 140 and 210, which is consistent with the position of the epitope of another anti–pre-β1-HDL MoAb.12 This region is situated away from 2 known lipid-binding sites of apoA-I, 220 to 24313–15⇓⇓ and 63 to 73,16 sites that are also responsible for the initial stages of cholesterol efflux,17 which may explain the lack of inhibition of cholesterol efflux.
A key hypothesis exploited in the present study was whether pre-β1-HDL was essential to cholesterol efflux, at least when tested in the conventional in vitro model. The corollary to this hypothesis was that pre-β1-HDL was an active acceptor of cellular cholesterol but was no more effective than other HDL subfractions. Taken together, our experiments also show that whereas pre-β1-HDL actively stimulated cholesterol efflux, its removal from plasma did not significantly reduce the cholesterol efflux capacity of the plasma. It follows that pre-β1-HDL does not appear to be essential for cholesterol efflux. Either it does not represent a quantitatively significant pathway for cholesterol efflux, or other nascent HDL fractions, such as γ-LpE18 or α-HDL, can take up the role of pre-β1-HDL in its absence. This finding is in contrast to the results of Kawano et al,3 who showed that a substantial inhibition of the ability of plasma to facilitate cholesterol efflux after pre-β1-HDL was removed by the incubation of plasma at room temperature. Lee at al19 also observed that depletion from pre-β1-HDL in plasma after incubation with chymase was accompanied by reduced cholesterol efflux. However, long incubations at room temperature, especially in the presence of a protease, might have also affected other HDL subfractions as well as other components of reverse cholesterol transport, such as cholesterol ester transfer protein, LCAT, or phospholipid transfer protein.
A further confirmation of the nonessential role of pre-β1-HDL was our finding that isolated pre-β1-HDL was less efficient in facilitating cholesterol efflux than was whole plasma when added at the same apoA-I concentration. The lower activity of pre-β1-HDL in cholesterol efflux might be due to inactivation or remodeling of pre-β1-HDL during isolation. However, several findings point against this possibility and, at the same time, demonstrate the capacity of pre-β1-HDL to promote cholesterol efflux. First, when plasma was reconstituted by combining pre-β1-HDL–deficient plasma and isolated pre-β1-HDL to a proportion of HDL resembling that in whole plasma, the reconstituted plasma was more effective than whole plasma in promoting cholesterol efflux. Second, reconstitution of pre-β1-HDL–deficient plasma with pre-β1-HDL resulted in increased flux of cellular cholesterol to pre-β1-HDL. Third, reisolated pre-β1-HDL attracted significant amounts of cellular cholesterol. A more likely explanation is that poorly lipidated pre-β1-HDL particles are biologically less active in promoting cholesterol efflux than fully lipidated α-HDL particles.
Additional insight into a possible mechanism underlying this difference was found from analyzing time and dose dependencies of efflux. The time course of cholesterol efflux involved fast and slow phases, as reported by Gaus et al20; only the slow phase showed a significant difference between whole plasma and pre-β1-HDL in cholesterol uptake capacity. In the dose-dependence experiments, the amount of cellular cholesterol fluxing to pre-β1-HDL was less than the amount fluxing to whole plasma. Together, these findings suggest that certain pools of cholesterol, specifically those that are slowly released and that probably represent intracellular cholesterol,20 might not be as readily available for efflux to pre-β1-HDL as to whole plasma.
An initially unexpected finding was that the antibody against pre-β1-HDL failed to inhibit cholesterol efflux to whole plasma. However, the epitope of the antibody was found to be located between residues 140 and 210 of apoA-I, away from the known lipid-binding sites of apoA-I. Thus, although reacting specifically with pre-β1-HDL, the antibody may not block lipid binding nor interfere with cholesterol efflux. By contrast, another MoAb against pre-β1-HDL with an epitope between amino acids 137 and 144 did reduce cholesterol efflux.12
The function of pre-β1-HDL has been assumed to be an initial acceptor of cholesterol from cells, shuttling cholesterol from cells to larger α-HDLs in the main cholesterol efflux pathway.2,4⇓ This assumption was based on in vitro experiments demonstrating that during early stages of efflux, cellular cholesterol fluxed to pre-β1-HDL sooner than to other lipoprotein fractions2,9⇓ and that in pulse-chase experiments, cholesterol taken up by pre-β1-HDL later appeared in pre-β2-HDL and α-HDL fractions.2,4⇓ In agreement with these findings, we have found that pre-β1-HDL was an active acceptor of cellular cholesterol, although its contribution to overall cholesterol efflux might be less significant and possibly not essential. An alternative role for pre-β1-HDL in cholesterol efflux might be that of an essential intermediate in cholesterol efflux during the formation of mature HDL from lipid-free apoA-I. Pre-β1-HDL may be the first product of apoA-I lipidation, presumably by ATP-binding cassette transporter A1–dependent and/or caveolin-dependent pathways.21 During this initial lipidation of apoA-I, pre-β1-HDLs are formed and may take up some cellular cholesterol, which eventually appears in the α-HDL fraction. However, it is not clear whether this pathway plays a quantitatively major role in transferring cellular cholesterol to HDL. We have previously observed that the capacity of human plasma to facilitate cholesterol efflux in vitro does not correlate with the pre-β1-HDL concentration in plasma.7 Instead, pre-β1-HDL may be a useful marker of the efficiency of HDL formation. The concentration of pre-β1-HDL in plasma is frequently correlated inversely with the concentration of HDL (see review22), so that a low concentration of an intermediate may indicate faster formation of a product. On the other hand, Francone et al23 demonstrated that pre-β-HDL contains the majority of LCAT and that cellular cholesterol in pre-β-HDL is a preferred substrate for esterification. Consequently, cholesterol taken up by pre-β-HDL during HDL formation may be more effectively trapped than cholesterol taken up by α-HDL via exchange. Thus, the pre-β1-HDL–dependent pathway of HDL formation may also play a significant role in cholesterol efflux.
A significant limitation of the present study is that these are in vitro experiments and do not necessarily reflect the situation in vivo. Isolation of pre-β1-HDL may remove essential factors required for its metabolism, thus affecting its ability to take up cellular cholesterol. Further transformations or transfer reactions may require participation of factors absent in isolated pre-β1-HDL. Plasma from patients with Tangier disease was unable to convert lipid-free apoA-I into α-HDL,24 indicating that there might be as-yet-unknown plasma factors linked to the mutation in ATP-binding cassette transporter A1. Furthermore, after being incubated with cells, pre-β1-HDL may be remodeled into different particles or even taken up by cells through retroendocytosis. Identification of the factors required for efficient cholesterol efflux may link the pre-β1-HDL concentration to overall efficiency of reverse cholesterol transport and be of diagnostic value or a target for pharmacological intervention.
We conclude that whereas pre-β1-HDLs are clearly capable of taking up cellular cholesterol and promoting cholesterol efflux, their presence in plasma is not essential to this process, at least in vitro.
This work was supported by a grant from the National Health and Medical Research Council of Australia and by a grant from Daiichi Pure Chemicals Ltd.
Received February 27, 2002; revision accepted June 24, 2002.
- ↵Fielding CJ, Fielding PE. Molecular physiology of the reverse cholesterol transport. J Lipid Res. 1995; 36: 211–228.
- ↵Huang Y, von Eckardstein A, Assmann G. Cell-derived unesterified cholesterol cycles between different HDLs and LDL for its effective esterification in plasma. Arterioscler Thromb. 1993; 13: 445–458.
- ↵Liang H-Q, Rye K-A, Barter PJ. Dissociation of lipid-free apolipoprotein A-I from high density lipoproteins. J Lipid Res. 1994; 35: 1187–1199.
- ↵Sasahara T, Yamashita T, Sviridov D, Fidge N, Nestel P. Altered properties of high density lipoprotein subfractions in obese subjects. J Lipid Res. 1997; 38: 600–611.
- ↵Sasahara T, Nestel P, Fidge N, Sviridov D. Cholesterol transport between cells and high density lipoprotein subfractions from obese and lean subjects. J Lipid Res. 1998; 39: 544–554.
- ↵Huang Y, von Eckardstein A, Wu S, Langer C, Assmann G. Generation of pre-beta 1-HDL and conversion into alpha-HDL: evidence for disturbed HDL conversion in Tangier disease. Arterioscler Thromb Vasc Biol. 1995; 15: 1746–1754.
- ↵Miyazaki O, Kobayashi J, Fukamachi I, Miida T, Bujo H, Saito Y. A new sandwich enzyme immunoassay for measurement of plasma pre-beta1-HDL levels. J Lipid Res. 2000; 41: 2083–2088.
- ↵Ji Y, Jonas A. Properties of an N-terminal proteolytic fragment of apolipoprotein AI in solution and in reconstituted high density lipoproteins. J Biol Chem. 1995; 270: 11290–11297.
- ↵Davidson WS, Hazlett T, Mantulin WW, Jonas A. The role of apolipoprotein AI domains in lipid binding. Proc Natl Acad Sci U S A. 1996; 93: 13605–13610.
- ↵Sviridov D, Hoang A, Sawyer W, Fidge N. Identification of a sequence of apolipoprotein A-I associated with activation of lecithin:cholesterol acyltransferase. J Biol Chem. 2000; 275: 19707–19712.
- ↵Sviridov D, Pyle L, Fidge N. Efflux of cellular cholesterol and phospholipid to apolipoprotein A-I mutants. J Biol Chem. 1996; 271: 33277–33283.
- ↵Huang Y, von Eckardstein A, Wu S, Maeda N, Assmann G. A plasma lipoprotein containing only apolipoprotein E and with gamma mobility on electrophoresis releases cholesterol from cells. Proc Natl Acad Sci U S A. 1994; 91: 1834–1838.
- ↵Lee M, von Eckardstein A, Lindstedt L, Assmann G, Kovanen PT. Depletion of pre beta1 LpA1 and LpA4 particles by mast cell chymase reduces cholesterol efflux from macrophage foam cells induced by plasma. Arterioscler Thromb Vasc Biol. 1999; 19: 1066–1074.
- ↵Francone OL, Gurakar A, Fielding C. Distribution and functions of lecithin:cholesterol acyltransferase and cholesteryl ester transfer protein in plasma lipoproteins: evidence for a functional unit containing these activities together with apolipoproteins A-I and D that catalyze the esterification and transfer of cell-derived cholesterol. J Biol Chem. 1989; 264: 7066–7072.
- ↵von Eckardstein A, Huang Y, Kastelein JJ, Geisel J, Real JT, Kuivenhoven JA, Miccoli R, Noseda G, Assmann G. Lipid-free apolipoprotein (apo) A-I is converted into alpha-migrating high density lipoproteins by lipoprotein-depleted plasma of normolipidemic donors and apo A-I-deficient patients but not of Tangier disease patients. Atherosclerosis. 1998; 138: 25–34.