Minimal Lipidation of Pre-ß HDL by ABCA1 Results in Reduced Ability to Interact with ABCA

【摘要】  Objectives- The aim of this study was to determine the role of ATP binding cassette transporter A1 (ABCA1) on generation of different-sized nascent HDLs.

Methods and Results- HEK293 cells stably-transfected with ABCA1 (HEK293-ABCA1) or non-transfected (control) cells were incubated with lipid free 125 I-apoA-I for 24 hours. Incubation of apoA-I with HEK293-ABCA1 cells, but not control cells, led to the formation of heterogeneous-sized, pre-ß migrating nascent HDL subpopulations (pre-ß1 to -4) that varied in size (7.1 to 15.7 nm), lipid, and apoA-I content. Kinetic studies suggested that all subpopulations were formed simultaneously, with no evidence for a precursor-product relationship between smaller and larger-sized particles. When isolated nascent pre-ß HDLs (pre-ß1 to -4) were added back to HEK293-ABCA1 cells, their ability to bind to ABCA1 and efflux lipid was severely compromised. Heat-denaturation of pre-ß1 HDL resulted in partial recovery of ABCA1 binding, suggesting that initial interaction of apoA-I with ABCA1 results in a constrained conformation of apoA-I that decreases subsequent binding.

Conclusions- Interaction of apoA-I with ABCA1 results in the simultaneous generation of pre-ß HDLs of discrete size and chemical composition. These nascent particles are poor substrates for subsequent lipidation by ABCA1 and presumably require additional non-ABCA1-mediated lipidation for further maturation.

Incubation of apoA-I with ABCA1-expressing cells generated heterogeneous-sized nascent pre-ß HDLs, which appear to form simultaneously. These particles are poor substrates for ABCA1, suggesting that the initial interaction of apoA-I with ABCA1 results in a constrained conformation of apoA-I and that further maturation of these nascent HDL occurs via non-ABCA1-mediated pathways.

【关键词】  ATP binding cassette transporter A apolipoprotein AI high density lipoprotein

Introduction

The inverse relationship between plasma high-density lipoprotein (HDL) cholesterol concentration and the risk of premature atherosclerotic vascular diseases has generated interest in understanding the steps involved in HDL assembly and catabolism. HDLs are proposed to be antiatherogenic because of their ability to accept excess cellular cholesterol in peripheral tissues and transport it to the liver in a process denoted as reverse cholesterol transport (RCT). 1

HDLs are classified into two subpopulations based on electrophoretic mobility on agarose gels: -HDL (90% to 95% total HDL in plasma) and pre-ß HDL (5% to 10%). 2 Heterogeneity is found among pre-ß and -HDLs. Using 2-dimensional gel electrophoresis (2D-gel), several pre-ß and -HDL subpopulations have been identified as a function of increasing size. 3 Although heterogeneity among pre-ß and -HDLs is well-documented, little is known about the mechanism of formation and metabolism of these pre-ß and -HDL subpopulations.

ABCA1 is critical for maintaining normal plasma HDL levels and nascent HDL biogenesis. A critical role for ABCA1 in HDL metabolism was demonstrated in patients with Tangier disease, a genetic disorder in which the ABCA1 gene is mutated. 4-6 These patients have plasma HDL-cholesterol and apoA-I concentrations <5% of normal, accumulation of CE in macrophages, and increased risk of atherosclerosis. 7,8 Fibroblasts from these patients also have a significant reduction in lipid efflux to apoA-I compared with controls, suggesting ABCA1 was essential for mediating transport of intracellular lipid to apoA-I. 9 Additionally, ABCA1 knockout mice recapitulate the Tangier disease phenotype, verifying the role of ABCA1 in cellular lipid homeostasis and plasma HDL maintenance. 10

Plasma pre-ß HDLs exist as several subpopulations. 3 Several investigators have shown that nascent HDLs that resemble plasma pre-ß HDLs in size can be generated by incubation of apoA-I with cells in which ABCA1 have been upregulated by cAMP or liver X receptor (LXR) agonists. 11-14 These particles exhibit or pre- mobility on agarose gels depending on acidic-PL content. Despite the well-documented size heterogeneity of in vitro-generated nascent HDL, detailed biochemical characteristics of ABCA1-assembled nascent HDLs are just beginning to appear. 11 Furthermore, the role of ABCA1 in the size-speciation of nascent HDLs is inadequately defined.

The goal of our study was to define the role of ABCA1 in the size-speciation of nascent HDL formation. Our initial hypothesis was that small nascent HDLs are precursors of larger nascent HDLs that underwent multiple interactions with ABCA1 to acquire additional lipid and apoA-I. Alternatively, nascent HDLs of varying size and composition may be formed simultaneously. Although there is evidence for the latter hypothesis, the data were generated for only 2 nascent HDL populations at 3 time-points. 11 In addition, the ability of isolated ABCA1-generated nascent HDL subpopulations to interact with ABCA1 to gain additional lipid and apoA-I has not been investigated. Thus, we sought to address these questions using cells that expressed ABCA1, but not other HDL remodeling factors that might catalyze the inter-conversion of nascent HDLs and confound the interpretation of a precursor-product relationship among these particles. We found that ABCA1 expression alone was necessary and sufficient to generate 4 to 5 nascent pre-ß HDL subpopulations of discrete size, and the appearance of these particles was not dependent on a precursor-product relationship. These results likely explain the observations that isolated ABCA1-generated nascent HDLs bound to ABCA1 poorly and did not efflux lipid efficiently compared with apoA-I.

Materials and Methods

Overview of Experimental Procedures

HEK293 ABCA1-expressing and control cells were plated in 5 x 150-mm dishes until cells reached 95% confluence. Cells were then washed 3 times with PBS and incubated with 10 µg/mL of lipid-free 125 I-apoA-I (10 5 cpm/µg) isolated by ultracentrifugation from human plasma in serum-free medium for 24 hours for most experiments. Immediately before incubation, 125 I-apoA-I was heated to 60°C for 30 minutes and then cooled to RT to standardize the conformational state of apoA-I. The conditioned media from ABCA1 and control cells was harvested, and analyzed by 4% to 30% nondenaturing gradient gel (NDGGE), agarose gel, and 2-dimensional gel (2D-gel) electrophoresis. The remainder of conditioned media was concentrated using an Amicon Ultra-10 concentrator and fractionated by fast protein liquid (FPLC). The particles were eluted and individual fractions were analyzed for 125 I-radioactivity and the 125 I-profile was plotted. An aliquot of each fraction was resolved on NDGGE gels and visualized with a phosphorimager to determine the elution position of homogeneous-sized pre-ß HDLs, which were then pooled and used for subsequent studies including compositional analyses, ABCA1 binding, and lipid efflux experiments.

A full description of experimental procedures can be found in the supplemental materials, available online at http://atvb.ahajournals.org.

Results

ABCA1 Expression in Control and ABCA1-Expressing Cells

Using RT-PCR, we demonstrated significant expression of ABCA1 mRNA in HEK293-ABCA1 cells, whereas no expression was detected in control cells (supplemental Table II). Control and ABCA1-expressing cells did not express significant amounts of mRNA for ABCG1, SR-B1, LCAT, apoM, and PLTP relative to human liver, suggesting that these HDL modifying factors did not contribute to the formation or modification of nascent HDLs during our experiments. Western blot analysis verified the presence of ABCA1 protein in HEK293-ABCA1 cells, with no detectable expression in control cells (supplemental Figure IA). Efflux experiments confirmed the function of ABCA1 protein in the presence of apoA-I with ABCA1-expressing cells demonstrating an increased efflux of PL and FC relative to control cells (supplemental Figure IB).

Heterogeneity of Nascent HDLs Formed by ABCA1

Nascent HDLs were formed by incubating 125 I-apoA-I with ABCA1-expressing cells for 24 hours. The electrophoretic mobility and the size of nascent HDLs formed during incubation were determined by agarose-, 1D- and 2D-gel electrophoresis of media ( Figure 1 ). Agarose gel electrophoresis showed that 125 I-apoA-I incubated in the absence of cells or with control cells had pre-ß migration, suggesting no change in electrophoretic mobility of 125 I-apoA-I. However, the electrophoretic mobility of nascent HDLs formed from 125 I-apoA-I in the presence of ABCA1 ran slightly faster than apoA-I, but did not migrate into the -position. On the basis of agarose gel electrophoretic behavior, we designated these as nascent pre-ß HDLs. One and 2D-gel electrophoresis demonstrated that incubation of apoA-I in the presence of ABCA1 protein led to the formation of heterogeneous-sized pre-ß HDLs, which were not seen in control media (compare lanes 3 and 4 for control versus ABCA1, Figure 1 B). We designated these pre-ß HDLs, from smallest to largest, as pre-ß1, pre-ß2, pre-ß3, and pre-ß4 HDL. Pre-ß1 HDLs, the smallest particles observed in ABCA1 cell-conditioned media, were similar in size to lipid-free 125 I-apoA-I. We also noted that pre-ß2 HDLs ran slightly faster, toward the anode, in the first dimension of the 2D-gel, suggesting that the faster migration of HDL particles observed in the agarose gel ( Figure 1 A) was primarily attributable to pre-ß2 HDL.

Figure 1. Analysis of conditioned media from control and ABCA1-transfected cells by agarose and 2D-gel electrophoresis. 125 I-apoA-I was incubated with control or HEK293-ABCA1 cells for 24 hours. Aliquots of media were subjected to agarose gel, NDGG, and 2D-gel electrophoresis. A, Agarose gel. a, 125 I-apoA-I as pre-ß marker; b, 125 I-apoA-I+media, not incubated with cells; c, Control cell-conditioned media; d, ABCA1-expressing cell conditioned media; e, 125 I-HDL as marker. B, 1D- and 2D-gel of conditioned media. 1, 1D-separation of 125 I-HMW marker; 2, 1D-separation of 125 I-apoA-I media; 3, 1D-separation of 125 I-conditioned media from control (left) or ABCA1-expressing cells (right); and 4, 2D-separation of control (left) or ABCA1-conditioned (right) media.

To further study these nascent HDLs, each pre-ß HDL was isolated to apparent homogeneity by FPLC. Concentrated conditioned media was injected onto Superdex-200HR FPLC columns and fractions were collected to monitor elution of 125 I-radioactivity. Figure 2 A shows the 125 I-radioactivity profile of 125 I-apoA-I, and media from control and ABCA1-expressing cells. The elution profiles for 125 I-apoA-I and control cell media were very similar, supporting the results shown in Figure 1 B. These suggest that there was minimal HDL particle formation in the absence of ABCA1. However, the elution profile of ABCA1-expressing cell media showed a decrease in radioactivity distributed in the pre-ß1 region and the appearance of discrete peaks of radioactivity from fractions 100 to 150, suggesting the formation of larger-sized HDL subfractions. Fractions were pooled and designated as pre-ß1, 2, 3, and 4 HDL on the basis of the elution positions ( Figure 2 A). The apparent homogeneity of each pre-ß HDL fraction was verified by NDGGE and phosphorimager analysis ( Figure 2 B). All pre-ß subfractions appeared as homogeneous-sized subpopulations, except for pre-ß1 HDL.

Figure 2. Fractionation of pre-ß HDL formed by ABCA1-expressing cells by Superdex-200HR FPLC. 125 I-apoA-I or media from control or ABCA1-expressing cells radiolabeled with 3 H-cholesterol and incubated with unlabeled or 125 I-radiolabeled apoA-I for 24 hours were separated by FPLC. A, Profile of 125 I-(solid line, solid symbols) and 3 H-cho- lesterol (dashed line, open symbols) radioactivity from fractions 70 to 200 is shown. Pre-ß1, 2, 3, and 4 HDL were pooled as designated by the dashed vertical lines. B, NDGGE of pre-ß HDL fractions isolated by FPLC as shown in panel A. Aliquots of pre-ß1 to 4 HDL were resolved on NDGG and visualized by phosphorimager analysis.

A previous study demonstrated that interaction of apoA-I with J774 macrophages stimulated with cAMP or human fibroblasts stimulated with LXR agonist to upregulate ABCA1 expression resulted in microparticle accumulation in the medium. 11,13 These microparticles 20 nm in diameter, devoid of apoA-I, and contained about half of the cholesterol carried in the medium. To determine whether microparticles were present in our system, we incubated unlabeled apoA-I with 3 H-cholesterol-labeled control and ABCA1-expressing cells for 24 hours. The conditioned media was fractionated by FPLC and fractions were quantified for 3 H-radioactivity. Fractionation of control media showed that 3 H-cholesterol was only associated with pre-ß1, with no significant 3 H-radioactivity eluting in the void volume region ( Figure 2 A; fractions 80 to 90). In ABCA1-conditioned media, there was minimal 3 H-radioactivity in the pre-ß1 region of the column but a significant amount of radiolabel was observed in pre-ß2 to 4 regions. Thus, we found scant 3 H-radioactivity in the void volume region, suggesting that microparticle formation was minimal.

For each isolated pre-ß HDL, we determined the size, density, number of apoA-I molecules per particle, and lipid composition ( Table ). The pre-ß HDLs ranged in diameter from <7.2 to 13 nm. The density of the particles was inversely proportional to size and ranged from 1.1687 (pre-ß1) to 1.0833 (pre-ß4) g/mL. As pre-ß HDL size increased, the number of apoA-I molecules per particle increased from 1 to 4. Lipids were extracted from pre-ß HDLs to quantify PL and cholesterol by mass spectrometry and GLC, respectively. The molar ratio of choline-containing PLs (phosphatidylcholine and sphingomyelin) to apoA-I was similar for pre-ß2 to 4 HDL (132 to 145 molecules of PC per molecule of apoA-I), but was much less for pre-ß1 (9 molecules). Analysis of the pre-ß1 region from FPLC separation of control cell medium resulted in a PC/AI molar ratio of 5. Because the apoA-I used for the study contained <1 PC molecule per apoA-I, the results suggest a small amount of PL associated with apoA-I in an ABCA1-independent manner during the incubation or FPLC isolation for control cells. We did not detect a significant amount of lysophosphatidylcholine, phosphatidylethanolamine, phosphatidylinositol, or phosphatidylserine mass associated with pre-ß HDLs by mass spectrometric analysis (data not shown). This likely explains why our particles migrated pre-ß, whereas other investigators have observed ABCA1-generated particles that contain small amounts of acidic PLs and migrate in the -position on agarose gels. 11,12 The molar ratio of FC to apoA-I also increased as pre-ß HDL size increased, from background levels in pre-ß1 to 10 mole % for the largest particles. However, there was no detectable CE in the pre-ß HDLs (data not shown).

Characterization of Pre-ß HDLs Formed by ABCA1-Expressing HEK293 Cells

Kinetics of Pre-ß HDL Formation by ABCA1

There are 2 hypothetical pathways by which pre-ß HDLs of different size may be formed by ABCA1. In the first, which we refer to as the sequential pathway, apoA-I interacts with ABCA1, resulting in the formation of pre-ß1 HDLs. Through repeated interactions with ABCA1, these small pre-ß1 HDLs gain additional lipid and apoA-I and are converted in a sequential, step-wise manner into larger pre-ß HDL particles (ie, pre-ß2, 3, and 4). The second possibility, the simultaneous pathway, is hypothesized to result in the simultaneous formation of pre-ß1, 2, 3, 4, and larger HDL by a single interaction with ABCA1. To determine which mechanism might prevail in our experimental system, we performed a series of kinetic studies. 125 I-apoA-I was incubated with ABCA1-expressing and control cells for 30 hours and, periodically, aliquots of medium were removed to monitor pre-ß HDLs formation using NDGGE and phosphorimager analysis ( Figure 3 ). There was no evidence for the formation of particles larger than pre-ß1 in control cell media over the 30 hour time course ( Figure 3 A, top). However, the presence of pre-ß2, 3, and 4 HDL in ABCA1-conditioned media was noticed between 1 to 2 hours of incubation, and radioactivity in these fractions increased gradually during the remainder of the experiment. Figure 3 B represents the quantification of phosphorimager pixel density in each pre-ß subfraction shown in Figure 3 A as the percentage of total radioactivity at each time point. Pre-ß1 radioactivity remained constant at about 100% throughout the study in control medium. Radioactivity in pre-ß1 of ABCA1-conditioned media decreased during the 30-hour incubation, whereas radioactivity in pre-ß2, 3, and 4 HDL increased. Radioactivity in pre-ß2 HDL increased gradually and reached saturation after 18 hours. A similar trend was observed for pre-ß3 and pre-ß4 HDL, which reached saturation after 24-hour incubation. Analysis of the kinetics of nascent pre-ß HDL formation by the SAAM II program suggested that pre-ß2, 3, 4, and 5 were generated simultaneously within the limits of our detection system, with no evidence for a precursor-product relationship among pre-ß HDLs (data not shown). The experimental data suggest that there is a short delay in the conversion of apoA-I to larger pre-ß HDLs, and there is differential conversion with more of the apoA-I converted to pre-ß2, less to pre-ß3, and least to pre-ß4.

Figure 3. Kinetics of pre-ß HDL formation. A, NDGGE of media during kinetic experiments. 125 I-apoA-I was incubated with control or ABCA1-expressing cells and aliquots of media were removed at indicated times for fractionation by NDGGE. Phosphorimager analysis of representative gels is shown. B, Phosphorimager quantification of pre-ß HDLs. Phosphorimager intensity of radiolabel in regions of the NDGG in which pre-ß HDLs migrated in Figure 3A was measured by ImageQuant Software. Data (mean±SD, n=3) are presented as % total intensity for each pre-ß HDL for each time point.

Minimal Lipidation of ApoA-I by ABCA1 Reduces Pre-ß HDL Interaction With ABCA1

Previous studies have shown that plasma and recombinant HDL interact poorly with ABCA1 compared with apoA-I. 15,16 To determine whether isolated pre-ß HDLs from our system could bind to ABCA1 and serve as acceptors for cellular lipid, we performed binding and efflux studies using isolated pre-ß HDLs. Compared with apoA-I, pre-ß2 to 4 HDL formed by ABCA1 showed dramatically reduced ability to interact with ABCA1-expressing cells ( Figure 4 A, left). Chemical cross-linking experiments of apoA-I with ABCA1 using DSP demonstrated that radiolabeled apoA-I was cross-linked with a protein of the apparent MW of ABCA1 ( Figure 4 A, inset-lane 2), whereas no cross-link product was observed for control cells ( Figure 4 A, inset-lane 1). Cleavage of the cross-linked product with ß-mercaptoethanol resulted in a radiolabeled protein that migrated in the position of apoA-I ( Figure 4 A, inset-lane 4), suggesting that binding of 125 I-apoA-I to ABCA1-expressing cells was specific for ABCA1. We also demonstrated that recombinant HDL particles containing 5 and 9 molecules of PC per molecule of apoA-I (5- and 9rpre-ß HDL, respectively) bound poorly to ABCA1, confirming that a small amount of PC associated with apoA-I decreases binding to ABCA1 ( Figure 4 A, right). Additionally, the ability of these pre-ß HDLs to promote lipid efflux from ABCA1-expressing cells was severely compromised relative to apoA-I ( Figure 4 B). To rule out the possibility that this outcome was attributable to the in vitro generation of pre-ß HDLs, we isolated small and pre-ß HDL from human plasma using anti-human apoA-I immunoaffinity and size-exclusion chromatography. 17 We also observed a diminished ability of small and pre-ß HDLs from human plasma to promote lipid efflux via ABCA1 compared with apoA-I ( Figure 4 C). These data suggest that pre-ß HDL formed by ABCA1 are poor substrates for subsequent interaction with ABCA1 and must receive additional lipid and apoA-I through non-ABCA1 pathways to complete their maturation.

Figure 4. ABCA1-mediated binding and lipid efflux to pre-ß HDL. A, Binding study. Binding studies were performed with 1 µg/mL of 125 I-apoA-I, 125 I-pre-ß2, 3 and 4 HDL, or 125 I-5- or 9rpre-ß HDL for 2 hours on ice as explained in the Methods section. Values represent the mean of duplicate observations. Inset, Chemical cross-linking study. One µg/mL of 125 I-apoA-I was incubated for 30 minutes at 37°C with control (lanes 1 & 3) and ABCA1-expressing (lanes 2 & 4) cells in the presence of DSP and media were subjected to SDS-PAGE and phosphorimager analysis. The cross-linked product was reduced by ß-mercaptoethanol in control (lane 3) and ABCA1 cells (lane 4). Migration positions for ABCA1 and apoA-I are indicated. B, Phospholipid and cholesterol efflux to pre-ß HDL generated by ABCA1-expressing cells. ABCA1-expressing and control cells were radiolabeled with 2µCi/mL of 3 H-choline chloride and 3 H-cholesterol. Lipid efflux was performed for 24 hours in the presence of 10 µg/mL of 125 I-apoA-I or 125 I-pre-ß2, 3, or 4 HDL, isolated from the medium of nonradiolabeled ABCA1-expressing cells. At the end of the efflux period, cells and media were lipid extracted and PL and cholesterol radiolabel were quantified after TLC separation. Percentage PL (left) and cholesterol (right) efflux (mean±SD, n=3) was calculated as radiolabel in medium divided by radiolabel in medium+cells X 100%. C. Phospholipid and cholesterol efflux to human plasma HDL. Pre-ß and small HDL were isolated from human plasma by anti-human A-I immunoaffinity and size-exclusion chromatography. Efflux studies were performed as described in panel B. Values represent mean±range, n=2.

Pre-ß1 HDLs isolated from ABCA1-expressing cell medium was indistinguishable from apoA-I on the basis of size ( Figures 1 to 3 ), but contained a small amount of PL ( Table ). We have previously described a pre-ß particle isolated from the plasma of humans as well as human apoA-I transgenic mice that were similar in size and lipid content to the pre-ß1 HDLs formed by ABCA1-expressing cells. 17 Pre-ß HDL isolated from plasma had a very rapid removal rate from plasma, suggesting it was a terminal particle that could no longer interact with ABCA1. To determine the ability of pre-ß1 HDL to interact with ABCA1, we performed binding and efflux studies using 125 I-apoA-I and 125 I-pre-ß1 HDLs, harvested from ABCA1-expressing or control cell medium ( Figure 5 ). The specific binding of pre-ß1 isolated from ABCA1-conditioned medium to ABCA1 was very low (0.08±0.08 ng/mg pro) compared with that observed with apoA-I (2.28±0.73 ng/mg pro) and pre-ß1 from control medium (1.55±0.97 ng/mg pro; Figure 5 A not heated). We reasoned that the poor binding of pre-ß1 HDLs from ABCA1-expressing cells may be attributable to a conformational constraint in apoA-I induced by binding to ABCA1 or to the small amount of PL associated with the particle. To test this hypothesis, we heated pre-ß and apoA-I to 60°C, just above the transition temperature of apoA-I, resulting in denaturation of the protein, and tested their ability to bind to ABCA1-expressing cells. Binding of apoA-I and pre-ß1 HDL was similar after heating, demonstrating that the poor binding of ABCA1-generated pre-ß1 HDL to ABCA1 could be partially recovered by heat-denaturation of apoA-I ( Figure 5 A). We also tested the ability of pre-ß1 HDL to efflux PL ( Figure 5 B) and cholesterol ( Figure 5 C) from ABCA1-expressing cells. Analogous to the binding results, pre-ß1 from ABCA1-expressing cells had background levels of PL and cholesterol efflux, whereas pre-ß1 from control cells showed considerable efflux for both lipids, though not to the level of that observed from apoA-I.

Figure 5. ABCA1 binding and efflux to pre-ß1 HDL. A, Binding study. Binding studies were performed in the presence of 1 µg/mL 125 I-apoA-I or 125 I-pre-ß1, from ABCA1-expressing or control cells, as described in Figure 4 A. The radiolabeled ligands were kept on ice (not heated) or heated to 60°C for 30 minutes and cooled on ice just before the binding experiments. Data represent mean±SEM, n=3. B, Phospholipid, and C, Cholesterol efflux. Efflux studies were performed as explained in Figure 4 B legend in the absence (no acceptor) or presence of 10 µg/mL of 125 I-apoA-I or 125 I-pre-ß1 from control and ABCA1-expressing cells. Values represent mean±SD, n=3.

Discussion

The interaction of apoA-I with ABCA1 is critical for nascent HDL formation as demonstrated by the lack of plasma HDLs in Tangier disease subjects or in ABCA1-KO mice. 10,18 The goal of this study was to investigate the initial steps in HDL particle assembly by ABCA1 without other HDL modifying proteins. To accomplish this, we investigated nascent HDL formation using HEK293 cells expressing ABCA1. The results of our study support 3 main conclusions. First, ABCA1 is necessary and sufficient to form multiple discrete-sized nascent HDLs. Second, these particles appeared to form simultaneously. Finally, once they are formed, these particles have a compromised ability to bind to and acquire lipid via ABCA1. These results suggest that the initial interaction of apoA-I with ABCA1 may result in a stabilized conformation of apoA-I which prevents further interaction with ABCA1 and the subsequent lipidation and maturation of these nascent HDL particles must involve other proteins at the cell surface or in plasma.

Historically, HDL assembly was thought to occur intracellularly through lipidation of apoA-I in the secretory pathway of the liver and intestine as it is synthesized, processed, and secreted. However, the inability to detect discoidal, nascent HDL-sized particles in the secretory pathway, 19 the demonstration that addition of apoA-I to the media of cells results in PL and cholesterol efflux, 9,20-22 and the discovery of ABCA1 4-6,9,23 led to the concept that HDLs are assembled in the extracellular space by the interaction of apoA-I with ABCA1. Several studies with hepatocytes have shown that the majority of HDL particle assembly occurs in the extracellular space, 14,24-27 and nearly all of the plasma pool of HDLs is derived from the liver and intestine. 28-30 In addition, many studies have demonstrated that hepatocytes or ABCA1-transfected or stimulated nonhepatic cells in the present of exogenously added or endogenously synthesized apoA-I, accumulate multiple HDL-sized particles in the medium. 11-13,21,24 In contrast, studies on the chemical nature of nascent HDL assembled by ABCA1 are limited. We were particularly interested in isolating and characterizing nascent HDLs assembled by ABCA1 in the absence of other HDL modifying proteins that are present in or secreted from hepatocytes and macrophages.

In our experimental system, we show that at least 4 distinct-sized nascent pre-ß HDLs could be isolated from ABCA1-expressing cell medium ( Figure 2 ). To our knowledge, only one other study has reported the isolation and physical-chemical characteristics of nascent HDLs from cells in which ABCA1 activity was stimulated. 11 Our nascent pre-ß3 and 4 HDL were similar in size to the 9- and 12-nm nascent HDLs characterized by Duong et al. 11 We also isolated 2 smaller nascent HDLs (pre-ß1 and 2, <7.2 and 7.2 to 8.2 nm in diameter, respectively) that were not detected in the Duong study. 11 Comparing the migration on agarose electrophoresis, all of our nascent HDLs migrated pre-ß, whereas nascent HDLs reported by Duong migrated alpha. This difference can be explained by the presence of acidic PL on the particles of the Duong study, whereas our particles contained no detectable acidic PL. The lipid composition of our particles (pre-ß3 and 4) is in the range of that reported by Duong and the PC/AI molar ratio and hydrodynamic size of our nascent HDLs are within the range of values reported in the literature (supplemental Figure II), except for pre-ß2, which had a greater PC/apoA-I molar ratio than would be predicted for its size. The reason for this anomalous outcome is not known, but perhaps is related to a conformation state of apoA-I protein that results in a smaller apparent hydrodynamic size. It is interesting that this subpopulation of nascent HDL also exhibited a faster migration on agarose gels, which might also reflect a conformational difference of apoA-I on pre-ß2 ( Figure 1 ). Duong et al 11 also observed microparticle formation in the medium, whereas we did not. This is likely attributable to different cell types and experimental conditions used in these studies. The studies of Liu, 13 Duong, 11 and Denis 15 suggest that there was no precursor-product relationship between the smaller and larger nascent HDL. We also found no evidence for a precursor-product relationship between pre-ß1, 2, 3, and 4 over a 30-hour time-course kinetic study. One explanation for this outcome is that apoA-I binds to ABCA1 at different sites on the cell membrane that vary in lipid availability, resulting in discrete-sized nascent HDLs that are assembled independently of 1 another. Another possibility is that ABCA1 recycles through an endosomal compartment and may assemble more or less lipid on pre-ß HDL particles compared with ABCA1 localized on the extracellular membrane surface. Finally, pre-ß HDL particles may fuse with one another after assembly by ABCA1 to generate larger sized pre-ß HDL, but our kinetic data do not support this possibility. However, if smaller pre-ß HDL particles fuse to form larger pre-ß HDL immediately on release from ABCA1, we would be unable to differentiate that from simultaneous formation of different-sized pre-ß HDL in our experimental system. In summary, using a cell system in which ABCA1 is overexpressed and expression of other known HDL modifying factors is minimal (supplemental Table II), we show that multiple discrete-sized nascent HDLs are formed. Thus, ABCA1 expression is necessary and sufficient to produce the size-speciation of nascent HDLs.

Whereas apoA-I is a substrate for ABCA1-mediated lipidation, plasma HDL are poor substrates for ABCA1. 9,16 In addition, our data as well as those of Denis el al 15 suggest that even relatively small lipid-poor recombinant HDLs interact poorly with ABCA1. Since most evidence suggests that apoA-I binds directly to ABCA1, 16,31,32 mature plasma HDL may not have the appropriate conformation of apoA-I for binding to ABCA1. However, no studies have been performed to determine whether nascent HDLs assembled by ABCA1 are substrates for repeated rounds of lipidation by ABCA1 or whether additional lipidation must occur through non-ABCA1-mediated pathways. We demonstrated that pre-ß HDL generated by ABCA1-expressing cells have a compromised ability to subsequently bind to ABCA1 and mediate lipid efflux compared with apoA-I. On the basis of these results, we suggest that initial interaction of apoA-I with ABCA1 results in a conformational change of apoA-I that reduces its subsequent binding to ABCA1 and impedes additional lipidation by this pathway. This idea is supported by our results demonstrating that pre-ß1 isolated from the medium of ABCA1-expressing cells contain a few molecules of PL, but still bind poorly to ABCA1-expressing cells. The reduction in binding of pre-ß1 from ABCA1-expressing cells does not appear to be an artifact of the isolation procedures, because pre-ß1 from control cells binds to ABCA1-expressing cells and mediates lipid efflux, albeit not to the level of that observed for apoA-I. Because pre-ß1 from control cells also contained a few molecules of PL relative to apoA-I, which likely resulted from non-specific lipid addition to apoA-I during incubation or isolation, it suggests that the interaction of apoA-I with ABCA1, per se, but not lipid binding, may change the conformation of apoA-I. However, ABCA1-mediated lipid transfer to apoA-I may help stabilize the conformation induced by interaction with ABCA1. Fielding et al 33 have also suggested that on interaction with ABCA1, apoA-I undergoes a conformational rearrangement independently of lipid transfer and this rearrangement is necessary for subsequent lipid transfer to apoA-I. Further support for our hypothesis comes from the heat-denaturation of the pre-ß1 HDL, which eliminates the "conformational restraint". This results in partial restoration of binding to ABCA1, suggesting that the conformation of apoA-I is important for interaction with ABCA1. The decreased interaction of nascent HDLs with ABCA1-expressing cells also agrees with lack of an observable precursor-product relationship among nascent HDLs in the kinetic studies. If pre-ß1 HDL increased their lipid content by repeated interactions with ABCA1, we would predict a precursor-product relationship between pre-ß1 and larger nascent HDL. In addition, the inability to detect pre-ß HDL larger than pre-ß1 during the first hour of the kinetic experiment as well as the reported t1/2 (1.4±0.66 hour) for release of apoA-I from ABCA1, 15 suggests that pre-ß particle formation occurs through one round of binding of apoA-I to ABCA1 and release of a nascent HDL (ie, pre-ß1 to 4). Taken together, these suggest that apoA-I binds to ABCA1, becomes lipidated to varying extents, and is released as a nascent HDL that can no longer interact with ABCA1. Therefore, these nascent HDLs must undergo further maturation by ABCG1, 34,35 PLTP, 36,37 SR-BI, 38 and perhaps other factors before final maturation to spherical HDL by LCAT in plasma. 39-41

Several studies have suggested that pre-ß1 is the initial acceptor of cellular cholesterol in RCT. 42,43 However, our previous studies have demonstrated that pre-ß HDL isolated from plasma has a very rapid in vivo catabolic rate from plasma and is degraded preferentially by the kidney rather than the liver, 17 which would decrease its utility for RCT. If this is the case, how can pre-ß1 survive to be an effective mediator of RCT? We suggest that plasma pre-ß1 HDL is a mixture of lipid-free apoA-I and poorly-lipidated HDL similar to the pre-ß1 isolated from ABCA1-expressing cells. We hypothesize that the pre-ß1-like particle in plasma is a terminal particle that failed to become sufficiently lipidated by ABCA1. If our hypothesis is correct, the main function of ABCA1 is to sufficiently lipidate newly secreted apoA-I before it enters plasma to prevent its hypercatabolism from plasma and degradation by the kidney. We suggest this concept is supported by the observation that the only organs that quantitatively secrete apoA-I, liver and intestine, are also the organs that account for the production of the plasma HDL pool, 28,29 at least in mice. In summary, although some apoA-I becomes lipidated in the secretory pathway of the liver in an ABCA1-dependent or -independent manner, 14,25,27 most HDL assembly occurs in the extracellular space of liver, and to a lesser extent, the intestine by ABCA1. Thus, we hypothesize that newly secreted apoA-I has at least 3 potential fates. First, it can interact with ABCA1 at the hepatocyte or enterocyte basolateral membrane to generate nascent HDLs with sufficient lipid to prevent their 7 nm in diameter; pre-ß2 to 4). These nascent HDLs would then be substrates for additional maturation by non-ABCA1 pathways and eventually become substrates for RCT. Second, apoA-I that interacts with ABCA1 but is poorly lipidated because of limited lipid availability or limited ABCA1 activity is subject to rapid removal from the plasma and degradation by the kidney. Third, apoA-I that fails to interact with ABCA1 and gains no lipid before entering the circulation is rapidly incorporated into preexisting plasma HDLs.

Acknowledgments

We thank Dr Michael Hayden from University of British Columbia, Vancouver, Canada for providing us with ABCA1-expressing cells.

Sources of Funding

This work was supported by National Institutes of Health Grants HL 049373 (to J.S.P.) and HL 054176 (to J.S.P.) and an AHA Mid-Atlantic Affiliate Predoctoral Fellowship 0515420U (to A.M.).

Disclosures

None.

【参考文献】
  Fielding CJ, Fielding PE. Molecular physiology of reverse cholesterol transport. J Lipid Res. 1995; 36: 211-228.

O?Connor PM, Zysow BR, Schoenhaus SA, Ishida BY, Kunitake ST, Naya-Vigne JM, Duchateau PN, Redberg RF, Spencer SJ, Mark S, Mazur M, Heilbron DC, Jaffe RB, Malloy MJ, Kane JP. Prebeta-1 HDL in plasma of normolipidemic individuals: influences of plasma lipoproteins, age, and gender. J Lipid Res. 1998; 39: 670-678.

Asztalos BF, Sloop CH, Wong L, Roheim PS. Two-dimensional electrophoresis of plasma lipoproteins: recognition of new apo A-I-containing subpopulations. Biochim Biophys Acta. 1993; 1169: 291-300.

Bodzioch M, Orso E, Klucken J, Langmann T, Bottcher A, Diederich W, Drobnik W, Barlage S, Buchler C, Porsch-Ozcurumez M, Kaminski WE, Hahmann HW, Oette K, Rothe G, Aslanidis C, Lackner KJ, Schmitz G. The gene encoding ATP-binding cassette transporter 1 is mutated in Tangier disease. Nat Genet. 1999; 22: 347-351.

Brooks-Wilson A, Marcil M, Clee SM, Zhang LH, Roomp K, van Dam M, Yu L, Brewer C, Collins JA, Molhuizen HO, Loubser O, Ouelette BF, Fichter K, Ashbourne-Excoffon KJ, Sensen CW, Scherer S, Mott S, Denis M, Martindale D, Frohlich J, Morgan K, Koop B, Pimstone S, Kastelein JJ, Hayden MR. Mutations in ABC1 in Tangier disease and familial high-density lipoprotein deficiency. Nat Genet. 1999; 22: 336-345.

Rust S, Rosier M, Funke H, Real J, Amoura Z, Piette JC, Deleuze JF, Brewer HB, Duverger N, Denefle P, Assmann G. Tangier disease is caused by mutations in the gene encoding ATP-binding cassette transporter 1. Nat Genet. 1999; 22: 352-355.

Asztalos BF, Brousseau ME, McNamara JR, Horvath KV, Roheim PS, Schaefer EJ. Subpopulations of high density lipoproteins in homozygous and heterozygous Tangier disease. Atherosclerosis. 2001; 156: 217-225.

Serfaty-Lacrosniere C, Civeira F, Lanzberg A, Isaia P, Berg J, Janus ED, Smith MP Jr, Pritchard PH, Frohlich J, Lees RS. Homozygous Tangier disease and cardiovascular disease. Atherosclerosis. 1994; 107: 85-98.

Francis GA, Knopp RH, Oram JF. Defective removal of cellular cholesterol and phospholipids by apolipoprotein A-I in Tangier Disease. J Clin Invest. 1995; 96: 78-87.

McNeish J, Aiello RJ, Guyot D, Turi T, Gabel C, Aldinger C, Hoppe KL, Roach ML, Royer LJ, de Wet J, Broccardo C, Chimini G, Francone OL. High density lipoprotein deficiency and foam cell accumulation in mice with targeted disruption of ATP-binding cassette transporter-1. Proc Natl Acad Sci U S A. 2000; 97: 4245-4250.

Duong PT, Collins HL, Nickel M, Lund-Katz S, Rothblat GH, Phillips MC. Characterization of nascent HDL particles and microparticles formed by ABCA1-mediated efflux of cellular lipids to ApoA-I. J Lipid Res. 2006; 47: 832-843.

Krimbou L, Hajj HH, Blain S, Rashid S, Denis M, Marcil M, Genest J. Biogenesis and speciation of nascent apoA-I-containing particles in various cell lines. J Lipid Res. 2005; 46: 1668-1677.

Liu L, Bortnick AE, Nickel M, Dhanasekaran P, Subbaiah PV, Lund-Katz S, Rothblat GH, Phillips MC. Effects of apolipoprotein A-I on ATP-binding cassette transporter A1-mediated efflux of macrophage phospholipid and cholesterol: formation of nascent high density lipoprotein particles. J Biol Chem. 2003; 278: 42976-42984.

Maric J, Kiss RS, Franklin V, Marcel YL. Intracellular lipidation of newly synthesized apolipoprotein A-I in primary murine hepatocytes. J Biol Chem. 2005; 280: 39942-39949.

Denis M, Haidar B, Marcil M, Bouvier M, Krimbou L, Genest J Jr. Molecular and cellular physiology of apolipoprotein A-I lipidation by the ATP-binding cassette transporter A1 (ABCA1). J Biol Chem. 2004; 279: 7384-7394.

Wang N, Silver DL, Costet P, Tall AR. Specific binding of ApoA-I, enhanced cholesterol efflux, and altered plasma membrane morphology in cells expressing ABC1. J Biol Chem. 2000; 275: 33053-33058.

Lee JY, Lanningham-Foster L, Boudyguina EY, Smith TL, Young ER, Colvin PL, Thomas MJ, Parks JS. Prebeta high density lipoprotein has two metabolic fates in human apolipoprotein A-I transgenic mice. J Lipid Res. 2004; 45: 716-728.

Fredrickson DS. The inheritance of high density lipoprotein deficiency (Tangier Disease). J Clin Invest. 1964; 43: 228-236.

Hamilton RL, Moorehouse A, Havel RJ. Isolation and properties of nascent lipoproteins from highly purified rat hepatocytic Golgi fractions. J Lipid Res. 1991; 32: 529-543.

Forte TM, Bielicki JK, Goth-Goldstein R, Selmek J, McCall MR. Recruitment of cell phospholipids and cholesterol by apolipoproteins A-II and A-I: formation of nascent apolipoprotein-specific HDL that differ in size, phospholipid composition, and reactivity with LCAT. J Lipid Res. 1995; 36: 148-157.

Forte TM, Goth-Goldstein R, Nordhausen RW, McCall MR. Apolipoprotein A-I-cell membrane interaction: extracellular assembly of heterogeneous nascent HDL particles. J Lipid Res. 1993; 34: 317-324.

Oram JF, Yokoyama S. Apolipoprotein-mediated removal of cellular cholesterol and phospholipids. J Lipid Res. 1996; 37: 2473-2491.

Lawn RM, Wade DP, Garvin MR, Wang X, Schwartz K, Porter JG, Seilhamer JJ, Vaughan AM, Oram JF. The Tangier disease gene product ABC1 controls the cellular apolipoprotein-mediated lipid removal pathway. J Clin Invest. 1999; 104: R25-R31.

McCall MR, Forte TM, Shore VG. Heterogeneity of nascent high density lipoproteins secreted by the hepatoma-derived cell line, Hep G2. J Lipid Res. 1988; 29: 1127-1137.

Chisholm JW, Burleson ER, Shelness GS, Parks JS. ApoA-I secretion from HepG2 cells: evidence for the secretion of both lipid-poor apoA-I and intracellularly assembled nascent HDL. J Lipid Res. 2002; 43: 36-44.

Tsujita M, Wu CA, be-Dohmae S, Usui S, Okazaki M, Yokoyama S. On the hepatic mechanism of HDL assembly by the ABCA1/apoA-I pathway. J Lipid Res. 2005; 46: 154-162.

Kiss RS, McManus DC, Franklin V, Tan WL, McKenzie A, Chimini G, Marcel YL. The lipidation by hepatocytes of human apolipoprotein A-I occurs by both ABCA1-dependent and -independent pathways. J Biol Chem. 2003; 278: 10119-10127.

Timmins JM, Lee JY, Boudyguina E, Kluckman KD, Brunham LR, Mulya A, Gebre AK, Coutinho JM, Colvin PL, Smith TL, Hayden MR, Maeda N, Parks JS. Targeted inactivation of hepatic Abca1 causes profound hypoalphalipoproteinemia and kidney hypercatabolism of apoA-I. J Clin Invest. 2005; 115: 1333-1342.

Brunham LR, Kruit JK, Iqbal J, Fievet C, Timmins JM, Pape TD, Coburn BA, Bissada N, Staels B, Groen AK, Hussain MM, Parks JS, Kuipers F, Hayden MR. Intestinal ABCA1 directly contributes to HDL biogenesis in vivo. J Clin Invest. 2006; 116: 1052-1062.

Basso F, Freeman L, Knapper CL, Remaley A, Stonik J, Neufeld EB, Tansey T, Amar MJ, Fruchart-Najib J, Duverger N, Santamarina-Fojo S, Brewer HB Jr. Role of the hepatic ABCA1 transporter in modulating intrahepatic cholesterol and plasma HDL cholesterol concentrations. J Lipid Res. 2003; 44: 296-302.

Fitzgerald ML, Morris AL, Rhee JS, Andersson LP, Mendez AJ, Freeman MW. Naturally occurring mutations in the largest extracellular loops of ABCA1 can disrupt its direct interaction with apolipoprotein A-I. J Biol Chem. 2002; 277: 33178-33187.

Chroni A, Liu T, Fitzgerald ML, Freeman MW, Zannis VI. Cross-linking and lipid efflux properties of apoA-I mutants suggest direct association between apoA-I helices and ABCA1. Biochemistry. 2004; 43: 2126-2139.

Chau P, Nakamura Y, Fielding CJ, Fielding PE. Mechanism of prebeta-HDL formation and activation. Biochemistry. 2006; 45: 3981-3987.

Wang N, Lan D, Chen W, Matsuura F, Tall AR. ATP-binding cassette transporters G1 and G4 mediate cellular cholesterol efflux to high-density lipoproteins. Proc Natl Acad Sci U S A. 2004; 101: 9774-9779.

Vaughan AM, Oram JF. ABCA1 and ABCG1 or ABCG4 act sequentially to remove cellular cholesterol and generate cholesterol-rich HDL. J Lipid Res. 2006; 47: 2433-2443.

Tu AY, Nishida HI, Nishida T. High density lipoprotein conversion mediated by human plasma phospholipid transfer protein. J Biol Chem. 1993; 268: 23098-23105.

Jauhiainen M, Metso J, Pahlman R, Blomqvist S, van TA, Ehnholm C. Human plasma phospholipid transfer protein causes high density lipoprotein conversion. J Biol Chem. 1993; 268: 4032-4036.

Ji Y, Jian B, Wang N, Sun Y, Moya ML, Phillips MC, Rothblat GH, Swaney JB, Tall AR. Scavenger receptor BI promotes high density lipoprotein-mediated cellular cholesterol efflux. J Biol Chem. 1997; 272: 20982-20985.

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 catalyzes the esterification and transfer of cell-derived cholesterol. J Biol Chem. 1989; 264: 7066-7072.

Nakamura Y, Kotite L, Gan Y, Spencer TA, Fielding CJ, Fielding PE. Molecular mechanism of reverse cholesterol transport: reaction of pre-beta-migrating high-density lipoprotein with plasma lecithin/cholesterol acyltransferase. Biochemistry. 2004; 43: 14811-14820.

Glomset JA, Mitchell CD, King WC, Applegate KA, Forte T, Norum KR, Gjone E. In vitro effects of lecithin:cholesterol acyltransferase on apolipoprotein distribution in familial lecithin:cholesterol acyltransferase deficiency. Ann N Y Acad Sci. 1980; 348: 224-243.

Castro GR, Fielding CJ. Early incorporation of cell-derived cholesterol into pre-beta-migrating high-density lipoprotein. Biochemistry. 1988; 27: 25-29.

Fielding PE, Nagao K, Hakamata H, Chimini G, Fielding CJ. A two-step mechanism for free cholesterol and phospholipid efflux from human vascular cells to apolipoprotein A-1. Biochemistry. 2000; 39: 14113-14120.

作者单位：Departments of Pathology/Section on Lipid Sciences (A.M., J.-Y.L., A.K.G., J.S.P.) and Biochemistry (M.J.T.), Wake Forest University Health Sciences, Winston-Salem, NC; the Division of Gerontology (P.L.C.), University of Maryland School of Medicine, and the Department of Veterans Affairs and Veteran