A Novel Efflux–Recapture Process Underlies the Mechanism of High-Density Lipoprotein Cholesteryl Ester-Selective Uptake Mediated by the Low-
http://www.100md.com
《动脉硬化血栓血管生物学》
From the Lipoprotein and Atherosclerosis Group, University of Ottawa Heart Institute, Ottawa, Ontario, Canada.
ABSTRACT
Objective— To determine the mechanism of low-density lipoprotein (LDL) receptor-related protein (LRP)-mediated selective uptake of high-density lipoprotein (HDL)-derived cholesteryl esters (CE).
Methods and Results— Apolipoprotein E (apoE) and heparin sulfate proteoglycans are required for LRP-mediated selective uptake in adipocytes. Furthermore, 2-deoxyglucose and NaN3 abolish this process, indicating that cellular energy is required. LRP-mediated selective uptake is also abolished by monensin or when clathrin-mediated internalization is inhibited (using hypotonic, K+-free medium or hyperosmolar sucrose), clearly implicating receptor endocytosis. The receptor-associated protein (RAP), an inhibitor of ligand binding to LRP, reduced the transport of CE into an intracellular compartment but not into the plasma membrane. Remarkably, the CE that is ultimately transported by LRP first enters the plasma membrane then undergoes apoE-mediated CE efflux before being recaptured and internalized by LRP.
Conclusion— According to this "efflux-recapture" model, LRP contributes to selective uptake because it recovers CE that would normally be lost by efflux mediated by apoE.
In adipocytes, the LDL receptor-related protein contributes to selective uptake when it recaptures and internalizes HDL-derived cholesteryl esters that are otherwise lost by apoE-mediated efflux. This novel "efflux-recapture" process explains some conflicting observations of selective uptake and underscores the bi-directional nature of efflux.
Key Words: LDL receptor-related protein ? HDL ? selective uptake ? cholesteryl ester ? adipocyte ? apolipoprotein E
Introduction
Adipocytes, hepatocytes, and steroidogenic tissues can selectively acquire cholesteryl esters (CE) from high-density lipoprotein (HDL) without internalization and degradation of the entire lipoprotein,1 in contrast to the classical receptor-mediated pathway of low-density lipoprotein (LDL) catabolism. This "selective uptake" process is poorly understood but was originally distinguished by 3 phases.2 First, the lipoprotein binds to the cell surface. Then, CE in the lipoprotein is passively and reversibly transferred to the plasma membrane pool. Finally, the CE in the plasma membrane is irreversibly internalized and hydrolyzed.
See page 1538
The distinction between selective and endocytic uptake2 has been blurred recently by evidence for HDL recycling during selective CE transfer in ob/ob hepatocytes.3 This implicates an endocytic step in selective uptake and prompted us to examine the role of the LDL receptor-related protein (LRP)4 in selective uptake, particularly because the LRP binds to apolipoprotein E (apoE), lipoprotein lipase, and hepatic lipase, and these molecules have been shown to facilitate selective uptake.5–7
We found8 that molecules that interfere with ligand binding to LRP, or reduction of LRP expression using antisense, inhibited HDL-CE selective uptake by human primary adipocytes and SW872 liposarcoma cells by 35% to 50%. These inhibitors included the receptor-associated protein (RAP), which inhibits all ligand binding to the LRP;4 the polysulfated drug suramin,9 and 2 ligands of LRP, namely 2-macroglobulin (2M) and lactoferrin.4 Although LRP accounted for approximately one-third of the selective uptake, these studies demonstrated that a comediator was required given that human skin fibroblasts express LRP abundantly but lack the capacity for LRP-dependent selective uptake.
The ability of LRP to mediate selective uptake rather than holoparticle uptake is surprising given that this endocytic receptor normally mediates the lysosomal catabolism of ligands. We have investigated the mechanism of LRP-dependent selective uptake and the possible contribution of proteoglycans and report that this involves a novel "efflux-recapture" process.
Methods
Materials
All common reagents were analytical grade, purchased from Fisher (Fair Lawn, NJ), Sigma (St Louis, Mo), or BDH (Poole, England). Retired breeders of homozygous apoE-null mice (strain B6.129P2-Apoetm1Unc) were obtained from Jackson Laboratory (Bar Harbor, ME).
Protein Purification and Labeling
ApoA-I was expressed, purified, and labeled with as described.8 ApoE3 was obtained from Calbiochem (San Diego, Calif). Native 2M was purified from human plasma by zinc-chelate chromatography.10 The 2M was activated with methylamine and labeled with 125I as described.11 RAP was cloned into the IMPACT-CN system (New England Biolabs, Beverly, Mass), then expressed and purified essentially as described for the purification of apoA-I.8
Lipoprotein Purification and Labeling
Total HDL or HDL2 were purified from normolipemic plasma by density gradient ultracentrifugation.12 The lipoproteins were labeled with -cholesteryl oleate (-CE) or -cholesteryl oleoyl ether (-COE), or with -apoA-I as described.8
Tissue Culture
SW872 liposarcoma cells were cultured as described.8 Primary human adipocytes were prepared and cultured as described.8 Mouse epididymal fat pads were isolated and mature adipocytes were obtained after differentiation of preadipocytes by the same method used to prepare human adipocytes. Human skin fibroblasts from foreskin were from American Type Culture Collection, repository number CRL2523, and were cultured as described.11
Selective Uptake Assay
The cellular incorporation of HDL-derived -CE or -apoA-I was measured as described.8
Internalization of -2M
-2M cellular association (internalization and cell surface binding) was measured in SW872 cells at 37°C as described.11 At 37°C, internalization of 2M constitutes the majority of cell associated 2M in SW872 cells, because cell surface binding is only 2 fmol/mg cell protein.8
Digestion and Metabolic Inhibition of Proteoglycans
To determine whether proteoglycans were involved in LRP-dependent selective uptake, we used 4 separate treatments to alter the proteoglycans, as follows. Treatment 1 involved heparinase I digestion of heparin sulfate proteoglycans (HSPG). Heparinase I (E.C.4.2.2.7) from Sigma was incubated with the cells for 2 hours at 37°C dissolved at 3 U/mL in F12 containing 5 mg/mL of bovine serum albumin and complete protease inhibitors (Roche Molecular Biochemicals). Treatment 2 involved chondroitinase ABC digestion of chondroitin and dermatan sulfate proteoglycans. Chondroitinase ABC (E.C.4.2.2.4) from Sigma was incubated with the cells at 1.5 U/mL exactly as described for Heparinase I. Treatment 3 involved growth of the SW872 cells for 2 days in 30 mmol/L chlorate, which inhibits the sulfation of all proteoglycans.11 Treatment 4 involved growth of the cells for 24 hours in 2 mmol/L 4-methyl umbelliferyl-?-D-xyloside (MX), which significantly reduces proteoglycan synthesis and appearance on the cell surface.13 The efficiency of proteoglycan digestion, chlorate, or MX treatment was attributed to a reduction in cellular -sulfate, as described.11
Inhibition of Membrane Trafficking
The cellular association of -2M and selective uptake of HDL-derived -CE was measured in SW872 cells under the following conditions: (1) the cells were depleted of metabolic energy with 20 mmol/L 2-deoxyglucose and 5 mmol/L NaN3;14 (2) clathrin-mediated endocytosis was inhibited with 10 μmol/L monensin as described;15 and (3) clathrin-mediated internalization was inhibited after a hypotonic shock followed by incubation in a K+-free medium16 or with hyperosmolar sucrose.17
Lipid Extraction and Thin-Layer Chromatography
SW872 cells were incubated at 37°C for 2 hours with 75 μg/mL of -CE or -COE-labeled HDL2 in ligand buffer. This loads the radiolabeled lipids into the reversible and irreversible compartments. After this time, the cells were washed 6 times in HBSS, 20 mmol/L Hepes pH 7.45, to remove the labeled HDL2 from the extracellular medium, and the cells were incubated for various times up to 4 hours at 37°C with 300 μL of ligand buffer in the presence (500 μg/mL) or absence of suramin. The extracellular medium was centrifuged at 6000g for 2 minutes to pellet any cell debris, and 250 μL of the supernatant (designated the "postefflux" medium) was transferred to a new microfuge tube and mixed with 83 μL methanol, 167 μL ethanol, 667 μL chloroform, and 8 μL of glacial acetic acid to extract the lipids. The mixture was mixed vigorously and stored overnight at –20°C. The samples were warmed to room temperature, then centrifuged at 6000g for 4 minutes. The chloroform phase was transferred to a fresh tube and dried under a stream of nitrogen. The lipids were resuspended in a total of 40 μL of chloroform containing 10 μg each of cholesteryl oleate and cholesterol as carrier. Each sample was spotted onto a silica gel 60 thin-layer chromatography plate (Merck, Germany) and developed in hexane:diethyl ether:glacial acetic acid (105:45:1.5). After visualization in an I2 tank, the bands corresponding to cholesterol and cholesteryl oleate were scraped into a scintillation vial and the radioactivity was measured by liquid scintillation counting using Ecolite (ICN, Costa Mesa, Calif).
Immunoprecipitation of ApoE
All the "postefflux" medium (see previous) was mixed with 5 μg each of apoE antibodies 3H1 and 6C518 for 1 hour at room temperature. Protein G separose (12 μL of 50% slurry) was added and the samples were mixed end-over-end overnight at 4°C. The protein G was pelleted and the radioactivity of the supernatant was measured. Each pellet was washed twice, each time with 1 mL of HBSS, 20 mmol/L Hepes pH 7.45, then solubilized in 0.2 mol/L NaOH, and the radioactivity of the pellet was measured.
Results
Although LRP accounted for at least one-third of the selective uptake in adipocytes, a comediator is required given that human skin fibroblasts express LRP abundantly but lack the capacity for LRP-dependent selective uptake.8 RAP inhibited the selective uptake of -CE from human HDL2 by 34% in control adipocytes but only marginally in apoE-null adipocytes, (Figure 1) confirming that apoE is the required comediator. The small but measurable RAP-inhibitable selective uptake in apoE-null cells may be attributable to apoE present on HDL2. Addition of recombinant apoE to human skin fibroblasts increased the selective uptake by 7-fold, 50% of which was inhibitable by RAP (not shown). This result shows that exogenous apoE is able to "rescue" LRP-dependent selective uptake in human skin fibroblasts. The fact that 50% of the apoE-stimulated selective uptake could not be inhibited by RAP suggests that in human skin fibroblasts, apoE stimulates selective uptake by LRP-dependent and LRP-independent mechanisms. This is in contrast to the observation in SW872 cells8 in which all the apoE-stimulated selective uptake was inhibitable by RAP.
Figure 1. Role of apoE in LRP-dependent selective uptake. Primary mouse adipocytes from C57BL/6 (control) and apoE-null mice were incubated at 37°C for 8 hours with 25 μg/mL of -CE-labeled or -apoA-1-labeled HDL2 in the absence or presence (25 μg/mL) of RAP. Selective uptake in the absence (white bars) or presence (black bars) of RAP is calculated on the cell-associated and degraded -apoA-1 subtracted from the cellular incorporation of -CE. Each value plotted the mean of 6 measurements and the standard error (SE) is shown.
Many ligands, particularly apoE,4 are internalized by LRP with the collaboration of proteoglycans. We tested whether proteoglycans were required for LRP-dependent selective uptake and found that heparinase I digestion reduced selective uptake by 26%, whereas chondroitinase ABC reduced the selective uptake by only <5% (Figure 2). Additionally, chlorate and MX treatment reduced the selective uptake to almost the levels seen in the presence of RAP. These data indicate that LRP-dependent selective uptake of HDL-CE uses HSPG and is consistent with the observation that apoE is internalized by LRP with the participation of HSPG.4 Although our result differs from Swarnakar et al,19 who demonstrated that chondroitin sulfate proteoglycans were required for apoE-dependent LDL-CE selective uptake in apoE-expressing adrenocortical cells, these divergent findings may be a consequence of differing contribution of these proteoglycan species on adrenocortical and SW872 cells.
Figure 2. Role of proteoglycans in LRP-dependent selective uptake. Cells were digested with heparinase I or chondroitinase ABC or treated with chlorate or MX, then incubated at 37°C for 8 hours with 25 μg/mL of -CE-labeled or -apoA-1-labeled HDL2, without any further additions (control) or with 25 μg/mL of RAP. Selective uptake is calculated as described for Figure 1. The efficacy of each treatment was confirmed by its effect on the cellular incorporation of -sulfate as described.11 Each value plotted is the mean of 4 measurements and the SE is shown.
When the SW872 cells were incubated in energy-depleting media, the internalization of -2M was reduced to the same level seen in the presence of RAP (not shown). Therefore, energy depletion completely blocked LRP internalization as expected. RAP reduced the cellular incorporation of HDL-derived -CE by 40%, as did energy depletion of the cells (Figure 3A). Furthermore, addition of RAP did not result in further inhibition in the energy-depleted cells; hence, energy depletion abolished the RAP-inhibitable, LRP-dependent, -CE incorporation. RAP only inhibits the selective component of -CE cellular incorporation,8 indicating that LRP-dependent selective uptake requires cellular energy. To confirm that this is because membrane trafficking is necessary, the same experiment was performed on cells that were treated with monensin. As expected, monensin reduced the internalization of -2M to the levels seen in the presence of RAP (not shown). Similarly, RAP treatment resulted in 34% inhibition of -CE incorporation (Figure 3B) and monensin reduced the -CE uptake by 41%. Addition of RAP in the presence of monensin did not result in further inhibition, indicating that LRP internalization or apoE recruitment to the cell surface (or both these events) is required for LRP-dependent selective uptake. When clathrin-mediated internalization was blocked by treating the cells with a hypotonic shock followed by incubation in a K+-free medium16 or with hyperosmolar sucrose,17 -2M degradation was reduced to the same level achieved with RAP (not shown), indicating that internalization was prevented. Blocking clathrin-mediated internalization, particularly with hyperosmolar sucrose, also reduced selective uptake to approximately the same level observed in the presence of RAP, without much further inhibition with the addition of RAP (Figure 3C). These results confirm that the LRP-dependent selective uptake requires internalization of the receptor via clathrin-coated pits.
Figure 3. Effect of cellular trafficking on LRP-dependent selective uptake. SW872 cells were incubated for 4 hours at 37°C with 25 μg/mL of HDL2 or 1 μg/mL of -2M (not shown) in the presence (25 μg/mL) or absence of RAP in ligand buffer (control) or energy depletion (ED) media (A). Under identical conditions, the cells were also incubated in ligand buffer containing 10 μmol/L monensin (B) or incubated in a K+-free buffer or hyperosmolar sucrose (C). After this incubation, the cells were washed and the cell-associated radioactivity was measured. Each value shows the mean of 6 measurements and the SE is shown. The y-axes for (B) and (C) are identical to that for (A). These experiments were performed twice with essentially the same results.
CE taken-up selectively characteristically enters a reversible and irreversible compartment,5,6 and this has been observed in SW872 cells.8 According to the classical view, the reversible compartment is proposed to include CE that has entered the plasma membrane and remains accessible to extraction by extracellular unlabeled HDL. The plasma membrane CE is subsequently transferred to an irreversible compartment as it is internalized and becomes inaccessible to extraction by extracellular unlabeled HDL. A different interpretation3 is that the reversible compartment includes CE in HDL that is in a recycling compartment. During recycling, some of the CE is transferred to nonrecycling compartments and thus accumulates inside the cell in an irreversible manner. The remainder of the CE within the HDL is recycled back to the extracellular medium, thus manifesting as a reversible phase.
When SW872 cells were incubated at 37°C for up to 4 hours with 100 μg/mL -CE labeled HDL in the presence (25 μg/mL) or absence of RAP, and the cells then washed and incubated for 2 hours at 37°C with 800 μg/mL of unlabeled HDL, we found that RAP had no effect on the -CE that was extracted by the unlabeled HDL (the reversible compartment), but reduced the -CE that remained in the cells after this incubation (the irreversible compartment) (Figure 4). This result shows that RAP blocks the internalization of -CE and confirms the requirement for LRP internalization. Importantly, this result excludes a model in which LRP mediates selective uptake by directing HDL into a recycling compartment. In this case, blocking the LRP would reduce CE accumulation into the reversible and irreversible compartments. The LRP-dependent movement of -CE directly into the irreversible compartment is consistent with a hypothesis in which cell surface apoE binds transiently to HDL and microsolubilizes some of the -CE when it dissociates, then the apoE-bound -CE is internalized by LRP, independently of the apoA-I.
Figure 4. Effect of RAP on the CE accumulation into the reversible and irreversible compartments. SW872 cells were incubated at 37°C for up to 4 hours with 100 μg/mL of -CE-labeled HDL in the presence (25 μg/mL) or absence of RAP. After this time, the cells were washed 6 times in HBSS, 20 mmol/L Hepes pH 7.45, and the cells were incubated for 2 hours at 37°C in ligand buffer containing 800 μg/mL of unlabeled HDL. After this incubation, the radioactivity in the extracellular media and in the cells were counted separately. These counts in the media were deemed to be in the reversible compartment in the absence () or presence () of RAP, and the cell-associated counts were deemed to be in the irreversible compartment in the absence () or presence (?) of RAP. The protein in the wells were not measured directly but were measured in separate identical wells after the 6 washes in HBSS, 20 mmol/L Hepes pH 7.45. Each point shows the mean of 4 measurements and the SE is shown.
Our data are also consistent with a model in which LRP-independent, -CE selective uptake into the plasma membrane is followed by apoE-mediated efflux of this -CE by membrane microsolubilization and the resultant -CE-lipidated apoE is finally recaptured and internalized by LRP. We have termed this mechanism "efflux-recapture" by analogy to the secretion-recapture model for apoE-mediated uptake of lipoproteins.20 Although apoE-mediated efflux of cholesterol has been demonstrated and proposed to occur by membrane microsolubilization,21 the efflux of CE by apoE has not been demonstrated. Nevertheless, CE is predicted to be effluxed by membrane microsolubilization, which circumvents the need for the thermodynamically unfavorable22 diffusion of CE into the aqueous phase.
To distinguish lipoprotein microsolubilization from efflux-recapture, SW872 cells were incubated with 75 μg/mL of -CE labeled HDL2 at 37°C for 2 hours to load the reversible and irreversible compartments with -CE. The cells were then washed to remove the labeled HDL (thus preventing lipoprotein microsolubilization), then the cells were incubated at 37°C for various times up to 4 hours in the presence (500 μg/mL) or absence of suramin. Suramin inhibits LRP-dependent selective uptake similarly to RAP,8 but in this experiment suramin is preferred to prevent the re-association of apoE with proteoglycans. During this incubation, some -CE is expected to traffic from the reversible (plasma membrane) compartment to the irreversible (intracellular) compartment. In addition, we observed that -CE accumulated into the extracellular medium, consistent with removal of CE by efflux (Figure 5A). Most of this -CE was accumulated in the first hour and suggested that the CE was being rapidly removed from a compartment that comprised <10% of the total cellular counts. After 15 minutes, 1.9-fold more -CE had accumulated into the ligand medium in the presence of suramin compared with ligand medium alone. This is consistent with 2 possible interpretations: either suramin accelerated the efflux of -CE or it prevented the recapture of -CE. We found that suramin did not affect the amount of -CE in the plasma membrane measured by extraction for 2 hours at 37°C with 800 μg/mL of unlabeled HDL (Figure 5B); therefore, suramin did not increase efflux from this compartment. However, the increase in -CE in the medium containing suramin was mirrored by a corresponding reduction of -CE in the irreversible compartment (Figure 5C). Although this result is consistent with the interpretation that suramin stimulates -CE efflux from an intracellular compartment (but not the plasma membrane), given our previous data that RAP blocks the movement of -CE into the irreversible compartment (Figure 4), collectively these observations support the conclusion that suramin prevents the recapture of -CE and its subsequent movement into the cell.
Figure 5. Effect of suramin on the movement of CE from the reversible to the irreversible compartment. SW872 cells were incubated at 37°C for 2 hours with 75 μg/mL of -CE-labeled HDL2 in ligand buffer. This loads -CE into both compartments. After this time, the cells were washed 6 times in HBSS, 20 mmol/L Hepes pH 7.45, to remove the labeled HDL2 from the extracellular medium, and the cells were incubated for various times up to 4 hours at 37°C with ligand buffer () or with ligand buffer containing 500 μg/mL of suramin (?). After each time-point the extracellular medium was removed for counting (A) and the cells were incubated for 2 hours at 37°C with ligand buffer containing 800 μg/mL of unlabeled HDL2. The counts extracted into this media were deemed to be in the reversible compartment (B), and the cell-associated counts were deemed to be in the irreversible compartment (C). The protein in the wells were not measured directly but were measured in separate identical wells after the 6 washes in HBSS, 20 mmol/L Hepes pH 7.45. Each point shows the mean of 4 measurements and the SE is shown.
During the preceding experiment, RAP inhibited HDL-derived CE selective uptake by 35% and by an amount corresponding to 190 ng HDL protein/mg cell protein/h (not shown). In the cells that were preloaded with -CE, the -CE appears in the medium at a rate of 50 and 98 ng HDL protein/mg cell protein during the first 15 minutes in the absence and presence of suramin, respectively (Figure 5A). Therefore, the amount of CE that is recaptured in the absence of suramin corresponds to the difference between these values, namely 48 ng HDL protein/mg cell protein/15 minutes or 192 ng HDL protein/mg cell protein/h. This corresponds to the amount of CE selective uptake inhibited by RAP and suggests that the efflux-recapture mechanism can fully account for LRP-dependent selective uptake.
The efflux-recapture model is not conditional on the efflux of CE rather than cholesterol derived from the hydrolysis of CE. However, we have shown that RAP inhibits selective uptake of cholesteryl ether, indicating that this process does not require the hydrolysis of the CE. To date, there has been little research on the efflux of CE, although the membrane microsolubilization model of efflux predicts that CE efflux should occur. When SW872 cells were incubated with 75 μg/mL of -CE or -COE-labeled HDL2 at 37°C for 2 hours to load the reversible and irreversible, the efflux (measured as described for Figure 5) was greater for the -COE-labeled cells compared with the -CE-labeled cells (Figure 6A and 6B). Although we do not know the reason for this difference, the increase in efflux in the presence of suramin was approximately the same in both cases. This result indicates that -COE can undergo efflux and indicates that CE efflux should occur. We repeated this experiment after labeling the cells with -CE, then extracted the lipids that were effluxed and measured the efflux of CE and cholesterol after separation by thin-layer chromatography (Figure 6C and 6D). Approximately 25% of the effluxed label was cholesterol, and suramin did not effect this proportion significantly. This conclusively demonstrates the efflux of CE and supports the microsolubilization model. We next addressed the question of whether the effluxed label was associated with apoE. The SW872 cells were loaded with -CE and the efflux of the labeled lipid was permitted as described for Figure 5. The extracellular medium was immunoprecipitated with antibodies directed against apoE. We first verified that our protocol could fully immunoprecipitate apoE and that the presence of suramin did not affect this result (not shown). Immunoprecipitation of apoE in the extracellular medium of the suramin-treated cells reduced the radiolabeled lipid to approximately the levels seen in extracellular medium in the absence of suramin (Figure 7). This indicates that the additional radiolabeled lipid observed in the presence of suramin is primarily associated with apoE, consistent with our model of apoE-mediated efflux and suramin-inhibitable recapture.
Figure 6. Effect of suramin on the efflux of CE, CE-derived cholesterol, and COE. SW872 cells were incubated with 75 μg/mL of -CE-labeled (A) or -COE-labeled (B) HDL2 at 37°C for 2 hours to load the reversible and irreversible compartments. The efflux of the labels into ligand buffer in the presence (?) or absence () of suramin was then measured over various times up to 4 hours at 37°C. In a separate experiment, cells were incubated with -CE-labeled HDL2 and efflux was permitted as described. The -cholesterol () derived from hydrolyzed -CE and intact -CE () that was effluxed was measured after extraction and separation of these lipids by thin-layer chromatography (Figure 6C and 6D). The y-axis for (B) and (D) are identical to those for (A) and (B), respectively. The protein in the wells was not measured directly but was measured in separate identical wells after the 6 washes in HBSS, 20 mmol/L Hepes pH 7.45. Each point shows the mean of 4 measurements and the SE is shown.
Figure 7. Role of apoE in the efflux of -CE derived lipids. SW872 cells were incubated with 75 μg/mL of -CE-labeled HDL2 at 37°C for 2 hours to load the reversible and irreversible compartments. The efflux of the labels into ligand buffer in the presence (?) or absence () of suramin was permitted for various times up to 4 hours at 37°C. The extracellular medium was immunoprecipitated with antibodies directed against apoE, and the total (apoE precipitable plus nonprecipitable) radiolabel in the medium in the presence (?) or absence () of suramin is plotted, as is the nonprecipitable label in the presence () or absence () of suramin. The protein in the wells was not measured directly but was measured in separate identical wells after the 6 washes in HBSS, 20 mmol/L Hepes pH 7.45. Each point shows the mean of 4 measurements and the SE is shown.
Discussion
We have studied the mechanism of LRP-dependent selective uptake of HDL-CE by primary adipocytes and SW872 liposarcoma cells. We found that apoE and HSPG were necessary for the process. LRP does not mediate selective uptake by recycling HDL through a compartment where selective uptake is favored. Instead, it would appear that CE is first transferred to the plasma membrane by selective uptake independently of LRP. Subsequently, apoE mediates the efflux (presumably by microsolubilization) of CE in the plasma membrane, and then the apoE-bound CE is endocytosed by LRP with the collaboration of HSPG. According to this "efflux-recapture" model (Figure 8), LRP contributes to selective uptake because it recovers CE that would normally be lost by efflux mediated by apoE. In addition, the removal of CE from the plasma membrane may maintain the concentration gradient between the HDL and plasma membrane, and this gradient drives the passive movement of CE into the membrane.
Figure 8. Efflux-recapture model of LRP-mediated selective uptake. LRP-independent selective uptake transfers HDL-derived CE to the plasma membrane. ApoE secreted by the cells mediates efflux of plasma membrane CE. The CE-lipidated apoE binds to HSPG and is then transferred to LRP and internalized by receptor-mediated endocytosis.
In a variety of cells (Hep G2, Neuro-2a, and macrophages), nascent apoE apparently arrives at the cell surface in a form that is relatively lipid-poor.23,24 This lipid-poor condition may enable apoE to acquire lipids by microsolubilization. In macrophages, for example, lipid-poor apoE at the cell surface participates in lipid efflux.24–26 Even after apoE has acquired lipids (including CE) by microsolubilization, it is still likely to be relatively lipid-poor. Nevertheless, binding and internalization of poorly lipidated apoE to the LRP (with the collaboration of HSPG) has been demonstrated.27
The efflux-recapture model may explain a number of features of selective uptake. First, CE that is internalized by selective uptake was originally thought to undergo hydrolysis in a nonacidic compartment,28 but chloroquine-sensitive hydrolysis of selectively internalized HDL-derived CE has been observed.29 It is anticipated that apoE-bound CE that is internalized by LRP would be transported to lysosomes and undergo chloroquine-sensitive hydrolysis there. Second, the involvement of apoE and HSPG in LRP-mediated selective uptake explains earlier findings that apoE and proteoglycans are able to mediate selective uptake.7,30 Other molecules in addition to LRP, apoE, and HSPG may be involved in efflux-recapture. One possible candidate is the ATP-binding cassette protein ABCA1.31 ABCA1 directly controls the secretion of apoE,32 and lipid-poor apoE has been shown to interact with ABCA1 in a manner that enhances efflux of cholesterol.33 The potential role of ABCA1 in efflux-recapture remains to be determined.
The efflux-recapture model indicates that efflux is not a unidirectional process as it is usually depicted, but rather a fraction of cholesterol that is removed by efflux, particularly if it is apoE-mediated, can be recaptured immediately and transported to an intracellular compartment. Preventing this recapture manifests as an increase in efflux of cholesterol from an intracellular compartment, as we have shown.
During selective uptake, cells accumulate disproportionately more CE from HDL compared with the apolipoproteins. This is in sharp contrast to the receptor-mediated holoparticle endocytosis of lipoproteins in which all components are internalized equally. In this article, we have shown that selective uptake is not exclusive of receptor-mediated endocytosis and have provided evidence for a novel process to account for this apparent contradiction. It is clear from this work that the mechanisms that underlie selective uptake are likely to be as diverse as the molecules that mediate the process.
Acknowledgments
We thank Nihan Gul Kavaslar, Paulina Lau, and Thet Naing for excellent technical assistance. Ruth McPherson is supported by the Canadian Institutes for Health Research (CIHR) Grant (MOP-44360) and CIHR/Wyeth-Ayerst Chair in Cardiovascular Disease.
References
Krieger M. The "best" of cholesterols, the "worst" of cholesterols: a tale of two receptors. Proc Natl Acad Sci U S A. 1998; 95: 4077–4080.
Pittman RC, Knecht TP, Rosenbaum MS, Taylor CA, Jr. A nonendocytotic mechanism for the selective uptake of high-density lipoprotein-associated cholesterol esters. J Biol Chem. 1987; 262: 2443–2450.
Silver DL, Wang N, Tall AR. Defective HDL particle uptake in ob/ob hepatocytes causes decreased recycling, degradation, and selective lipid uptake. J Clin Invest. 2000; 105: 151–159.
Willnow TE, Nykjaer A, Herz J. Lipoprotein receptors: new roles for ancient proteins. Nat Cell Biol. 1999; 1: E157–E162.
Kadowaki H, Patton GM, Robins SJ. Metabolism of high density lipoprotein lipids by the rat liver: evidence for participation of hepatic lipase in the uptake of cholesteryl ester. J Lipid Res. 1992; 33: 1689–1698.
Schorsch F, Malle E, Sattler W. Selective uptake of high-density lipoprotein-associated cholesterylesters by differentiated Ob1771 adipocytes is modulated by endogenous and exogenous lipoprotein lipase. FEBS Lett. 1997; 414: 507–513.
Swarnakar S, Reyland ME, Deng J, Azhar S, Williams DL. Selective uptake of low-density lipoprotein-cholesteryl ester is enhanced by inducible apolipoprotein E expression in cultured mouse adrenocortical cells. J Biol Chem. 1998; 273: 12140–12147.
Vassiliou G, Benoist F, Lau P, Kavaslar GN, McPherson R. The low density lipoprotein receptor-related protein contributes to selective uptake of high density lipoprotein cholesteryl esters by SW872 liposarcoma cells and primary human adipocytes. J Biol Chem. 2001; 276: 48823–48830.
Vassiliou G. Pharmacological concentrations of suramin inhibit the binding of 2-macroglobulin to its cell-surface receptor. Eur J Biochem. 1997; 250: 320–325.
Porath J, Carlsson J, Olsson I, Belfrage G. Metal chelate affinity chromatography, a new approach to protein fractionation. Nature. 1975; 258: 598–599.
Vassiliou G, Stanley KK. Exogenous receptor-associated protein binds to two distinct sites on human fibroblasts but does not bind to the glycosaminoglycan residues of heparin sulfate proteoglycans. J Biol Chem. 1994; 269: 15172–15178.
Sattler W, Mohr D, Stocker R. Rapid isolation of lipoproteins and assessment of their peroxidation by high-performance liquid chromatography postcolumn chemiluminescence. Methods Enzymol. 1994; 233: 469–489.
Lucas M, Mazzone T. Cell surface proteoglycans modulate net synthesis and secretion of macrophage apolipoprotein E. J Biol Chem. 1996; 271: 13454–13460.
Hao M, Lin SX, Karylowski OJ, Wustner D, McGraw TE, Maxfield FR. Vesicular and non-vesicular sterol transport in living cells. The endocytic recycling compartment is a major sterol storage organelle. J Biol Chem. 2002; 277: 609–617.
Basu SK, Goldstein JL, Anderson RG, Brown MS. Monensin interrupts the recycling of low-density lipoprotein receptors in human fibroblasts. Cell. 1981; 24: 493–502.
Larkin JM, Brown MS, Goldstein JL, Anderson RG. Depletion of intracellular potassium arrests coated pit formation and receptor-mediated endocytosis in fibroblasts. Cell. 1983; 33: 273–285.
Oka JA, Weigel PH. Effects of hyperosmolarity on ligand processing and receptor recycling in the hepatic galactosyl receptor system. J Cell Biochem. 1988; 36: 169–183.
Weisgraber KH, Rall SC, Jr., Mahley RW, Milne RW, Marcel YL, Sparrow JT. Human apolipoprotein E. Determination of the heparin binding sites of apolipoprotein E3. J Biol Chem. 1986; 261: 2068–2076.
Swarnakar S, Beers J, Strickland DK, Azhar S, Williams DL. The apolipoprotein E-dependent low-density lipoprotein cholesteryl ester selective uptake pathway in murine adrenocortical cells involves chondroitin sulfate proteoglycans and an 2-macroglobulin receptor. J Biol Chem. 2001; 276: 21121–21128.
Ji ZS, Fazio S, Lee YL, Mahley RW. Secretion-capture role for apolipoprotein E in remnant lipoprotein metabolism involving cell surface heparin sulfate proteoglycans. J Biol Chem. 1994; 269: 2764–2772.
Gillotte KL, Zaiou M, Lund-Katz S, Anantharamaiah GM, Holvoet P, Dhoest A, Palgunachari MN, Segrest JP, Weisgraber KH, Rothblat GH, Phillips MC. Apolipoprotein-mediated plasma membrane microsolubilization. Role of lipid affinity and membrane penetration in the efflux of cellular cholesterol and phospholipid. J Biol Chem. 1999; 274: 2021–2028.
Phillips MC, Johnson WJ, Rothblat GH. Mechanisms and consequences of cellular cholesterol exchange and transfer. Biochim Biophys Acta. 1987; 906: 223–276.
Burgess JW, Gould DR, Marcel YL. The HepG2 extracellular matrix contains separate heparinase- and lipid-releasable pools of ApoE. Implications for hepatic lipoprotein metabolism. J Biol Chem. 1998; 273: 5645–5654.
Lin CY, Huang ZH, Mazzone T. Interaction with proteoglycans enhances the sterol efflux produced by endogenous expression of macrophage apoE. J Lipid Res. 2001; 42: 1125–1133.
Zhang WY, Gaynor PM, Kruth HS. Apolipoprotein E produced by human monocyte-derived macrophages mediates cholesterol efflux that occurs in the absence of added cholesterol acceptors. J Biol Chem. 1996; 271: 28641–28646.
Smith JD, Miyata M, Ginsberg M, Grigaux C, Shmookler E, Plump AS. Cyclic AMP induces apolipoprotein E binding activity and promotes cholesterol efflux from a macrophage cell line to apolipoprotein acceptors. J Biol Chem. 1996; 271: 30647–30655.
Narita M, Holtzman DM, Fagan AM, LaDu MJ, Yu L, Han X, Gross RW, Bu G, Schwartz AL. Cellular catabolism of lipid poor apolipoprotein E via cell surface LDL receptor-related protein. J Biochem (Tokyo). 2002; 132: 743–749.
Sparrow CP, Pittman RC. Cholesterol esters selectively taken up from high-density lipoproteins are hydrolyzed extralysosomally. Biochim Biophys Acta. 1990; 1043: 203–210.
Rhainds D, Falstrault L, Tremblay C, Brissette L. Uptake and fate of class B scavenger receptor ligands in HepG2 cells. Eur J Biochem. 1999; 261: 227–235.
Seo T, Al Haideri M, Treskova E, Worgall TS, Kako Y, Goldberg IJ, Deckelbaum RJ. Lipoprotein lipase-mediated selective uptake from low-density lipoprotein requires cell surface proteoglycans and is independent of scavenger receptor class B type 1. J Biol Chem. 2000; 275: 30355–30362.
Oram JF. AT. P-binding cassette transporter A1 and cholesterol trafficking. Curr Opin Lipidol. 2002; 13: 373–381.
von Eckardstein A, Langer C, Engel T, Schaukal I, Cignarella A, Reinhardt J, Lorkowski S, Li Z, Zhou X, Cullen P, Assmann G. ATP binding cassette transporter ABCA1 modulates the secretion of apolipoprotein E from human monocyte-derived macrophages. FASEB J. 2001; 15: 1555–1561.
Remaley AT, Stonik JA, Demosky SJ, Neufeld EB, Bocharov AV, Vishnyakova TG, Eggerman TL, Patterson AP, Duverger NJ, Santamarina-Fojo S, Brewer HB, Jr. Apolipoprotein specificity for lipid efflux by the human ABCAI transporter. Biochem Biophys Res Commun. 2001; 280: 818–823.(Gerard Vassiliou; Ruth Mc)
ABSTRACT
Objective— To determine the mechanism of low-density lipoprotein (LDL) receptor-related protein (LRP)-mediated selective uptake of high-density lipoprotein (HDL)-derived cholesteryl esters (CE).
Methods and Results— Apolipoprotein E (apoE) and heparin sulfate proteoglycans are required for LRP-mediated selective uptake in adipocytes. Furthermore, 2-deoxyglucose and NaN3 abolish this process, indicating that cellular energy is required. LRP-mediated selective uptake is also abolished by monensin or when clathrin-mediated internalization is inhibited (using hypotonic, K+-free medium or hyperosmolar sucrose), clearly implicating receptor endocytosis. The receptor-associated protein (RAP), an inhibitor of ligand binding to LRP, reduced the transport of CE into an intracellular compartment but not into the plasma membrane. Remarkably, the CE that is ultimately transported by LRP first enters the plasma membrane then undergoes apoE-mediated CE efflux before being recaptured and internalized by LRP.
Conclusion— According to this "efflux-recapture" model, LRP contributes to selective uptake because it recovers CE that would normally be lost by efflux mediated by apoE.
In adipocytes, the LDL receptor-related protein contributes to selective uptake when it recaptures and internalizes HDL-derived cholesteryl esters that are otherwise lost by apoE-mediated efflux. This novel "efflux-recapture" process explains some conflicting observations of selective uptake and underscores the bi-directional nature of efflux.
Key Words: LDL receptor-related protein ? HDL ? selective uptake ? cholesteryl ester ? adipocyte ? apolipoprotein E
Introduction
Adipocytes, hepatocytes, and steroidogenic tissues can selectively acquire cholesteryl esters (CE) from high-density lipoprotein (HDL) without internalization and degradation of the entire lipoprotein,1 in contrast to the classical receptor-mediated pathway of low-density lipoprotein (LDL) catabolism. This "selective uptake" process is poorly understood but was originally distinguished by 3 phases.2 First, the lipoprotein binds to the cell surface. Then, CE in the lipoprotein is passively and reversibly transferred to the plasma membrane pool. Finally, the CE in the plasma membrane is irreversibly internalized and hydrolyzed.
See page 1538
The distinction between selective and endocytic uptake2 has been blurred recently by evidence for HDL recycling during selective CE transfer in ob/ob hepatocytes.3 This implicates an endocytic step in selective uptake and prompted us to examine the role of the LDL receptor-related protein (LRP)4 in selective uptake, particularly because the LRP binds to apolipoprotein E (apoE), lipoprotein lipase, and hepatic lipase, and these molecules have been shown to facilitate selective uptake.5–7
We found8 that molecules that interfere with ligand binding to LRP, or reduction of LRP expression using antisense, inhibited HDL-CE selective uptake by human primary adipocytes and SW872 liposarcoma cells by 35% to 50%. These inhibitors included the receptor-associated protein (RAP), which inhibits all ligand binding to the LRP;4 the polysulfated drug suramin,9 and 2 ligands of LRP, namely 2-macroglobulin (2M) and lactoferrin.4 Although LRP accounted for approximately one-third of the selective uptake, these studies demonstrated that a comediator was required given that human skin fibroblasts express LRP abundantly but lack the capacity for LRP-dependent selective uptake.
The ability of LRP to mediate selective uptake rather than holoparticle uptake is surprising given that this endocytic receptor normally mediates the lysosomal catabolism of ligands. We have investigated the mechanism of LRP-dependent selective uptake and the possible contribution of proteoglycans and report that this involves a novel "efflux-recapture" process.
Methods
Materials
All common reagents were analytical grade, purchased from Fisher (Fair Lawn, NJ), Sigma (St Louis, Mo), or BDH (Poole, England). Retired breeders of homozygous apoE-null mice (strain B6.129P2-Apoetm1Unc) were obtained from Jackson Laboratory (Bar Harbor, ME).
Protein Purification and Labeling
ApoA-I was expressed, purified, and labeled with as described.8 ApoE3 was obtained from Calbiochem (San Diego, Calif). Native 2M was purified from human plasma by zinc-chelate chromatography.10 The 2M was activated with methylamine and labeled with 125I as described.11 RAP was cloned into the IMPACT-CN system (New England Biolabs, Beverly, Mass), then expressed and purified essentially as described for the purification of apoA-I.8
Lipoprotein Purification and Labeling
Total HDL or HDL2 were purified from normolipemic plasma by density gradient ultracentrifugation.12 The lipoproteins were labeled with -cholesteryl oleate (-CE) or -cholesteryl oleoyl ether (-COE), or with -apoA-I as described.8
Tissue Culture
SW872 liposarcoma cells were cultured as described.8 Primary human adipocytes were prepared and cultured as described.8 Mouse epididymal fat pads were isolated and mature adipocytes were obtained after differentiation of preadipocytes by the same method used to prepare human adipocytes. Human skin fibroblasts from foreskin were from American Type Culture Collection, repository number CRL2523, and were cultured as described.11
Selective Uptake Assay
The cellular incorporation of HDL-derived -CE or -apoA-I was measured as described.8
Internalization of -2M
-2M cellular association (internalization and cell surface binding) was measured in SW872 cells at 37°C as described.11 At 37°C, internalization of 2M constitutes the majority of cell associated 2M in SW872 cells, because cell surface binding is only 2 fmol/mg cell protein.8
Digestion and Metabolic Inhibition of Proteoglycans
To determine whether proteoglycans were involved in LRP-dependent selective uptake, we used 4 separate treatments to alter the proteoglycans, as follows. Treatment 1 involved heparinase I digestion of heparin sulfate proteoglycans (HSPG). Heparinase I (E.C.4.2.2.7) from Sigma was incubated with the cells for 2 hours at 37°C dissolved at 3 U/mL in F12 containing 5 mg/mL of bovine serum albumin and complete protease inhibitors (Roche Molecular Biochemicals). Treatment 2 involved chondroitinase ABC digestion of chondroitin and dermatan sulfate proteoglycans. Chondroitinase ABC (E.C.4.2.2.4) from Sigma was incubated with the cells at 1.5 U/mL exactly as described for Heparinase I. Treatment 3 involved growth of the SW872 cells for 2 days in 30 mmol/L chlorate, which inhibits the sulfation of all proteoglycans.11 Treatment 4 involved growth of the cells for 24 hours in 2 mmol/L 4-methyl umbelliferyl-?-D-xyloside (MX), which significantly reduces proteoglycan synthesis and appearance on the cell surface.13 The efficiency of proteoglycan digestion, chlorate, or MX treatment was attributed to a reduction in cellular -sulfate, as described.11
Inhibition of Membrane Trafficking
The cellular association of -2M and selective uptake of HDL-derived -CE was measured in SW872 cells under the following conditions: (1) the cells were depleted of metabolic energy with 20 mmol/L 2-deoxyglucose and 5 mmol/L NaN3;14 (2) clathrin-mediated endocytosis was inhibited with 10 μmol/L monensin as described;15 and (3) clathrin-mediated internalization was inhibited after a hypotonic shock followed by incubation in a K+-free medium16 or with hyperosmolar sucrose.17
Lipid Extraction and Thin-Layer Chromatography
SW872 cells were incubated at 37°C for 2 hours with 75 μg/mL of -CE or -COE-labeled HDL2 in ligand buffer. This loads the radiolabeled lipids into the reversible and irreversible compartments. After this time, the cells were washed 6 times in HBSS, 20 mmol/L Hepes pH 7.45, to remove the labeled HDL2 from the extracellular medium, and the cells were incubated for various times up to 4 hours at 37°C with 300 μL of ligand buffer in the presence (500 μg/mL) or absence of suramin. The extracellular medium was centrifuged at 6000g for 2 minutes to pellet any cell debris, and 250 μL of the supernatant (designated the "postefflux" medium) was transferred to a new microfuge tube and mixed with 83 μL methanol, 167 μL ethanol, 667 μL chloroform, and 8 μL of glacial acetic acid to extract the lipids. The mixture was mixed vigorously and stored overnight at –20°C. The samples were warmed to room temperature, then centrifuged at 6000g for 4 minutes. The chloroform phase was transferred to a fresh tube and dried under a stream of nitrogen. The lipids were resuspended in a total of 40 μL of chloroform containing 10 μg each of cholesteryl oleate and cholesterol as carrier. Each sample was spotted onto a silica gel 60 thin-layer chromatography plate (Merck, Germany) and developed in hexane:diethyl ether:glacial acetic acid (105:45:1.5). After visualization in an I2 tank, the bands corresponding to cholesterol and cholesteryl oleate were scraped into a scintillation vial and the radioactivity was measured by liquid scintillation counting using Ecolite (ICN, Costa Mesa, Calif).
Immunoprecipitation of ApoE
All the "postefflux" medium (see previous) was mixed with 5 μg each of apoE antibodies 3H1 and 6C518 for 1 hour at room temperature. Protein G separose (12 μL of 50% slurry) was added and the samples were mixed end-over-end overnight at 4°C. The protein G was pelleted and the radioactivity of the supernatant was measured. Each pellet was washed twice, each time with 1 mL of HBSS, 20 mmol/L Hepes pH 7.45, then solubilized in 0.2 mol/L NaOH, and the radioactivity of the pellet was measured.
Results
Although LRP accounted for at least one-third of the selective uptake in adipocytes, a comediator is required given that human skin fibroblasts express LRP abundantly but lack the capacity for LRP-dependent selective uptake.8 RAP inhibited the selective uptake of -CE from human HDL2 by 34% in control adipocytes but only marginally in apoE-null adipocytes, (Figure 1) confirming that apoE is the required comediator. The small but measurable RAP-inhibitable selective uptake in apoE-null cells may be attributable to apoE present on HDL2. Addition of recombinant apoE to human skin fibroblasts increased the selective uptake by 7-fold, 50% of which was inhibitable by RAP (not shown). This result shows that exogenous apoE is able to "rescue" LRP-dependent selective uptake in human skin fibroblasts. The fact that 50% of the apoE-stimulated selective uptake could not be inhibited by RAP suggests that in human skin fibroblasts, apoE stimulates selective uptake by LRP-dependent and LRP-independent mechanisms. This is in contrast to the observation in SW872 cells8 in which all the apoE-stimulated selective uptake was inhibitable by RAP.
Figure 1. Role of apoE in LRP-dependent selective uptake. Primary mouse adipocytes from C57BL/6 (control) and apoE-null mice were incubated at 37°C for 8 hours with 25 μg/mL of -CE-labeled or -apoA-1-labeled HDL2 in the absence or presence (25 μg/mL) of RAP. Selective uptake in the absence (white bars) or presence (black bars) of RAP is calculated on the cell-associated and degraded -apoA-1 subtracted from the cellular incorporation of -CE. Each value plotted the mean of 6 measurements and the standard error (SE) is shown.
Many ligands, particularly apoE,4 are internalized by LRP with the collaboration of proteoglycans. We tested whether proteoglycans were required for LRP-dependent selective uptake and found that heparinase I digestion reduced selective uptake by 26%, whereas chondroitinase ABC reduced the selective uptake by only <5% (Figure 2). Additionally, chlorate and MX treatment reduced the selective uptake to almost the levels seen in the presence of RAP. These data indicate that LRP-dependent selective uptake of HDL-CE uses HSPG and is consistent with the observation that apoE is internalized by LRP with the participation of HSPG.4 Although our result differs from Swarnakar et al,19 who demonstrated that chondroitin sulfate proteoglycans were required for apoE-dependent LDL-CE selective uptake in apoE-expressing adrenocortical cells, these divergent findings may be a consequence of differing contribution of these proteoglycan species on adrenocortical and SW872 cells.
Figure 2. Role of proteoglycans in LRP-dependent selective uptake. Cells were digested with heparinase I or chondroitinase ABC or treated with chlorate or MX, then incubated at 37°C for 8 hours with 25 μg/mL of -CE-labeled or -apoA-1-labeled HDL2, without any further additions (control) or with 25 μg/mL of RAP. Selective uptake is calculated as described for Figure 1. The efficacy of each treatment was confirmed by its effect on the cellular incorporation of -sulfate as described.11 Each value plotted is the mean of 4 measurements and the SE is shown.
When the SW872 cells were incubated in energy-depleting media, the internalization of -2M was reduced to the same level seen in the presence of RAP (not shown). Therefore, energy depletion completely blocked LRP internalization as expected. RAP reduced the cellular incorporation of HDL-derived -CE by 40%, as did energy depletion of the cells (Figure 3A). Furthermore, addition of RAP did not result in further inhibition in the energy-depleted cells; hence, energy depletion abolished the RAP-inhibitable, LRP-dependent, -CE incorporation. RAP only inhibits the selective component of -CE cellular incorporation,8 indicating that LRP-dependent selective uptake requires cellular energy. To confirm that this is because membrane trafficking is necessary, the same experiment was performed on cells that were treated with monensin. As expected, monensin reduced the internalization of -2M to the levels seen in the presence of RAP (not shown). Similarly, RAP treatment resulted in 34% inhibition of -CE incorporation (Figure 3B) and monensin reduced the -CE uptake by 41%. Addition of RAP in the presence of monensin did not result in further inhibition, indicating that LRP internalization or apoE recruitment to the cell surface (or both these events) is required for LRP-dependent selective uptake. When clathrin-mediated internalization was blocked by treating the cells with a hypotonic shock followed by incubation in a K+-free medium16 or with hyperosmolar sucrose,17 -2M degradation was reduced to the same level achieved with RAP (not shown), indicating that internalization was prevented. Blocking clathrin-mediated internalization, particularly with hyperosmolar sucrose, also reduced selective uptake to approximately the same level observed in the presence of RAP, without much further inhibition with the addition of RAP (Figure 3C). These results confirm that the LRP-dependent selective uptake requires internalization of the receptor via clathrin-coated pits.
Figure 3. Effect of cellular trafficking on LRP-dependent selective uptake. SW872 cells were incubated for 4 hours at 37°C with 25 μg/mL of HDL2 or 1 μg/mL of -2M (not shown) in the presence (25 μg/mL) or absence of RAP in ligand buffer (control) or energy depletion (ED) media (A). Under identical conditions, the cells were also incubated in ligand buffer containing 10 μmol/L monensin (B) or incubated in a K+-free buffer or hyperosmolar sucrose (C). After this incubation, the cells were washed and the cell-associated radioactivity was measured. Each value shows the mean of 6 measurements and the SE is shown. The y-axes for (B) and (C) are identical to that for (A). These experiments were performed twice with essentially the same results.
CE taken-up selectively characteristically enters a reversible and irreversible compartment,5,6 and this has been observed in SW872 cells.8 According to the classical view, the reversible compartment is proposed to include CE that has entered the plasma membrane and remains accessible to extraction by extracellular unlabeled HDL. The plasma membrane CE is subsequently transferred to an irreversible compartment as it is internalized and becomes inaccessible to extraction by extracellular unlabeled HDL. A different interpretation3 is that the reversible compartment includes CE in HDL that is in a recycling compartment. During recycling, some of the CE is transferred to nonrecycling compartments and thus accumulates inside the cell in an irreversible manner. The remainder of the CE within the HDL is recycled back to the extracellular medium, thus manifesting as a reversible phase.
When SW872 cells were incubated at 37°C for up to 4 hours with 100 μg/mL -CE labeled HDL in the presence (25 μg/mL) or absence of RAP, and the cells then washed and incubated for 2 hours at 37°C with 800 μg/mL of unlabeled HDL, we found that RAP had no effect on the -CE that was extracted by the unlabeled HDL (the reversible compartment), but reduced the -CE that remained in the cells after this incubation (the irreversible compartment) (Figure 4). This result shows that RAP blocks the internalization of -CE and confirms the requirement for LRP internalization. Importantly, this result excludes a model in which LRP mediates selective uptake by directing HDL into a recycling compartment. In this case, blocking the LRP would reduce CE accumulation into the reversible and irreversible compartments. The LRP-dependent movement of -CE directly into the irreversible compartment is consistent with a hypothesis in which cell surface apoE binds transiently to HDL and microsolubilizes some of the -CE when it dissociates, then the apoE-bound -CE is internalized by LRP, independently of the apoA-I.
Figure 4. Effect of RAP on the CE accumulation into the reversible and irreversible compartments. SW872 cells were incubated at 37°C for up to 4 hours with 100 μg/mL of -CE-labeled HDL in the presence (25 μg/mL) or absence of RAP. After this time, the cells were washed 6 times in HBSS, 20 mmol/L Hepes pH 7.45, and the cells were incubated for 2 hours at 37°C in ligand buffer containing 800 μg/mL of unlabeled HDL. After this incubation, the radioactivity in the extracellular media and in the cells were counted separately. These counts in the media were deemed to be in the reversible compartment in the absence () or presence () of RAP, and the cell-associated counts were deemed to be in the irreversible compartment in the absence () or presence (?) of RAP. The protein in the wells were not measured directly but were measured in separate identical wells after the 6 washes in HBSS, 20 mmol/L Hepes pH 7.45. Each point shows the mean of 4 measurements and the SE is shown.
Our data are also consistent with a model in which LRP-independent, -CE selective uptake into the plasma membrane is followed by apoE-mediated efflux of this -CE by membrane microsolubilization and the resultant -CE-lipidated apoE is finally recaptured and internalized by LRP. We have termed this mechanism "efflux-recapture" by analogy to the secretion-recapture model for apoE-mediated uptake of lipoproteins.20 Although apoE-mediated efflux of cholesterol has been demonstrated and proposed to occur by membrane microsolubilization,21 the efflux of CE by apoE has not been demonstrated. Nevertheless, CE is predicted to be effluxed by membrane microsolubilization, which circumvents the need for the thermodynamically unfavorable22 diffusion of CE into the aqueous phase.
To distinguish lipoprotein microsolubilization from efflux-recapture, SW872 cells were incubated with 75 μg/mL of -CE labeled HDL2 at 37°C for 2 hours to load the reversible and irreversible compartments with -CE. The cells were then washed to remove the labeled HDL (thus preventing lipoprotein microsolubilization), then the cells were incubated at 37°C for various times up to 4 hours in the presence (500 μg/mL) or absence of suramin. Suramin inhibits LRP-dependent selective uptake similarly to RAP,8 but in this experiment suramin is preferred to prevent the re-association of apoE with proteoglycans. During this incubation, some -CE is expected to traffic from the reversible (plasma membrane) compartment to the irreversible (intracellular) compartment. In addition, we observed that -CE accumulated into the extracellular medium, consistent with removal of CE by efflux (Figure 5A). Most of this -CE was accumulated in the first hour and suggested that the CE was being rapidly removed from a compartment that comprised <10% of the total cellular counts. After 15 minutes, 1.9-fold more -CE had accumulated into the ligand medium in the presence of suramin compared with ligand medium alone. This is consistent with 2 possible interpretations: either suramin accelerated the efflux of -CE or it prevented the recapture of -CE. We found that suramin did not affect the amount of -CE in the plasma membrane measured by extraction for 2 hours at 37°C with 800 μg/mL of unlabeled HDL (Figure 5B); therefore, suramin did not increase efflux from this compartment. However, the increase in -CE in the medium containing suramin was mirrored by a corresponding reduction of -CE in the irreversible compartment (Figure 5C). Although this result is consistent with the interpretation that suramin stimulates -CE efflux from an intracellular compartment (but not the plasma membrane), given our previous data that RAP blocks the movement of -CE into the irreversible compartment (Figure 4), collectively these observations support the conclusion that suramin prevents the recapture of -CE and its subsequent movement into the cell.
Figure 5. Effect of suramin on the movement of CE from the reversible to the irreversible compartment. SW872 cells were incubated at 37°C for 2 hours with 75 μg/mL of -CE-labeled HDL2 in ligand buffer. This loads -CE into both compartments. After this time, the cells were washed 6 times in HBSS, 20 mmol/L Hepes pH 7.45, to remove the labeled HDL2 from the extracellular medium, and the cells were incubated for various times up to 4 hours at 37°C with ligand buffer () or with ligand buffer containing 500 μg/mL of suramin (?). After each time-point the extracellular medium was removed for counting (A) and the cells were incubated for 2 hours at 37°C with ligand buffer containing 800 μg/mL of unlabeled HDL2. The counts extracted into this media were deemed to be in the reversible compartment (B), and the cell-associated counts were deemed to be in the irreversible compartment (C). The protein in the wells were not measured directly but were measured in separate identical wells after the 6 washes in HBSS, 20 mmol/L Hepes pH 7.45. Each point shows the mean of 4 measurements and the SE is shown.
During the preceding experiment, RAP inhibited HDL-derived CE selective uptake by 35% and by an amount corresponding to 190 ng HDL protein/mg cell protein/h (not shown). In the cells that were preloaded with -CE, the -CE appears in the medium at a rate of 50 and 98 ng HDL protein/mg cell protein during the first 15 minutes in the absence and presence of suramin, respectively (Figure 5A). Therefore, the amount of CE that is recaptured in the absence of suramin corresponds to the difference between these values, namely 48 ng HDL protein/mg cell protein/15 minutes or 192 ng HDL protein/mg cell protein/h. This corresponds to the amount of CE selective uptake inhibited by RAP and suggests that the efflux-recapture mechanism can fully account for LRP-dependent selective uptake.
The efflux-recapture model is not conditional on the efflux of CE rather than cholesterol derived from the hydrolysis of CE. However, we have shown that RAP inhibits selective uptake of cholesteryl ether, indicating that this process does not require the hydrolysis of the CE. To date, there has been little research on the efflux of CE, although the membrane microsolubilization model of efflux predicts that CE efflux should occur. When SW872 cells were incubated with 75 μg/mL of -CE or -COE-labeled HDL2 at 37°C for 2 hours to load the reversible and irreversible, the efflux (measured as described for Figure 5) was greater for the -COE-labeled cells compared with the -CE-labeled cells (Figure 6A and 6B). Although we do not know the reason for this difference, the increase in efflux in the presence of suramin was approximately the same in both cases. This result indicates that -COE can undergo efflux and indicates that CE efflux should occur. We repeated this experiment after labeling the cells with -CE, then extracted the lipids that were effluxed and measured the efflux of CE and cholesterol after separation by thin-layer chromatography (Figure 6C and 6D). Approximately 25% of the effluxed label was cholesterol, and suramin did not effect this proportion significantly. This conclusively demonstrates the efflux of CE and supports the microsolubilization model. We next addressed the question of whether the effluxed label was associated with apoE. The SW872 cells were loaded with -CE and the efflux of the labeled lipid was permitted as described for Figure 5. The extracellular medium was immunoprecipitated with antibodies directed against apoE. We first verified that our protocol could fully immunoprecipitate apoE and that the presence of suramin did not affect this result (not shown). Immunoprecipitation of apoE in the extracellular medium of the suramin-treated cells reduced the radiolabeled lipid to approximately the levels seen in extracellular medium in the absence of suramin (Figure 7). This indicates that the additional radiolabeled lipid observed in the presence of suramin is primarily associated with apoE, consistent with our model of apoE-mediated efflux and suramin-inhibitable recapture.
Figure 6. Effect of suramin on the efflux of CE, CE-derived cholesterol, and COE. SW872 cells were incubated with 75 μg/mL of -CE-labeled (A) or -COE-labeled (B) HDL2 at 37°C for 2 hours to load the reversible and irreversible compartments. The efflux of the labels into ligand buffer in the presence (?) or absence () of suramin was then measured over various times up to 4 hours at 37°C. In a separate experiment, cells were incubated with -CE-labeled HDL2 and efflux was permitted as described. The -cholesterol () derived from hydrolyzed -CE and intact -CE () that was effluxed was measured after extraction and separation of these lipids by thin-layer chromatography (Figure 6C and 6D). The y-axis for (B) and (D) are identical to those for (A) and (B), respectively. The protein in the wells was not measured directly but was measured in separate identical wells after the 6 washes in HBSS, 20 mmol/L Hepes pH 7.45. Each point shows the mean of 4 measurements and the SE is shown.
Figure 7. Role of apoE in the efflux of -CE derived lipids. SW872 cells were incubated with 75 μg/mL of -CE-labeled HDL2 at 37°C for 2 hours to load the reversible and irreversible compartments. The efflux of the labels into ligand buffer in the presence (?) or absence () of suramin was permitted for various times up to 4 hours at 37°C. The extracellular medium was immunoprecipitated with antibodies directed against apoE, and the total (apoE precipitable plus nonprecipitable) radiolabel in the medium in the presence (?) or absence () of suramin is plotted, as is the nonprecipitable label in the presence () or absence () of suramin. The protein in the wells was not measured directly but was measured in separate identical wells after the 6 washes in HBSS, 20 mmol/L Hepes pH 7.45. Each point shows the mean of 4 measurements and the SE is shown.
Discussion
We have studied the mechanism of LRP-dependent selective uptake of HDL-CE by primary adipocytes and SW872 liposarcoma cells. We found that apoE and HSPG were necessary for the process. LRP does not mediate selective uptake by recycling HDL through a compartment where selective uptake is favored. Instead, it would appear that CE is first transferred to the plasma membrane by selective uptake independently of LRP. Subsequently, apoE mediates the efflux (presumably by microsolubilization) of CE in the plasma membrane, and then the apoE-bound CE is endocytosed by LRP with the collaboration of HSPG. According to this "efflux-recapture" model (Figure 8), LRP contributes to selective uptake because it recovers CE that would normally be lost by efflux mediated by apoE. In addition, the removal of CE from the plasma membrane may maintain the concentration gradient between the HDL and plasma membrane, and this gradient drives the passive movement of CE into the membrane.
Figure 8. Efflux-recapture model of LRP-mediated selective uptake. LRP-independent selective uptake transfers HDL-derived CE to the plasma membrane. ApoE secreted by the cells mediates efflux of plasma membrane CE. The CE-lipidated apoE binds to HSPG and is then transferred to LRP and internalized by receptor-mediated endocytosis.
In a variety of cells (Hep G2, Neuro-2a, and macrophages), nascent apoE apparently arrives at the cell surface in a form that is relatively lipid-poor.23,24 This lipid-poor condition may enable apoE to acquire lipids by microsolubilization. In macrophages, for example, lipid-poor apoE at the cell surface participates in lipid efflux.24–26 Even after apoE has acquired lipids (including CE) by microsolubilization, it is still likely to be relatively lipid-poor. Nevertheless, binding and internalization of poorly lipidated apoE to the LRP (with the collaboration of HSPG) has been demonstrated.27
The efflux-recapture model may explain a number of features of selective uptake. First, CE that is internalized by selective uptake was originally thought to undergo hydrolysis in a nonacidic compartment,28 but chloroquine-sensitive hydrolysis of selectively internalized HDL-derived CE has been observed.29 It is anticipated that apoE-bound CE that is internalized by LRP would be transported to lysosomes and undergo chloroquine-sensitive hydrolysis there. Second, the involvement of apoE and HSPG in LRP-mediated selective uptake explains earlier findings that apoE and proteoglycans are able to mediate selective uptake.7,30 Other molecules in addition to LRP, apoE, and HSPG may be involved in efflux-recapture. One possible candidate is the ATP-binding cassette protein ABCA1.31 ABCA1 directly controls the secretion of apoE,32 and lipid-poor apoE has been shown to interact with ABCA1 in a manner that enhances efflux of cholesterol.33 The potential role of ABCA1 in efflux-recapture remains to be determined.
The efflux-recapture model indicates that efflux is not a unidirectional process as it is usually depicted, but rather a fraction of cholesterol that is removed by efflux, particularly if it is apoE-mediated, can be recaptured immediately and transported to an intracellular compartment. Preventing this recapture manifests as an increase in efflux of cholesterol from an intracellular compartment, as we have shown.
During selective uptake, cells accumulate disproportionately more CE from HDL compared with the apolipoproteins. This is in sharp contrast to the receptor-mediated holoparticle endocytosis of lipoproteins in which all components are internalized equally. In this article, we have shown that selective uptake is not exclusive of receptor-mediated endocytosis and have provided evidence for a novel process to account for this apparent contradiction. It is clear from this work that the mechanisms that underlie selective uptake are likely to be as diverse as the molecules that mediate the process.
Acknowledgments
We thank Nihan Gul Kavaslar, Paulina Lau, and Thet Naing for excellent technical assistance. Ruth McPherson is supported by the Canadian Institutes for Health Research (CIHR) Grant (MOP-44360) and CIHR/Wyeth-Ayerst Chair in Cardiovascular Disease.
References
Krieger M. The "best" of cholesterols, the "worst" of cholesterols: a tale of two receptors. Proc Natl Acad Sci U S A. 1998; 95: 4077–4080.
Pittman RC, Knecht TP, Rosenbaum MS, Taylor CA, Jr. A nonendocytotic mechanism for the selective uptake of high-density lipoprotein-associated cholesterol esters. J Biol Chem. 1987; 262: 2443–2450.
Silver DL, Wang N, Tall AR. Defective HDL particle uptake in ob/ob hepatocytes causes decreased recycling, degradation, and selective lipid uptake. J Clin Invest. 2000; 105: 151–159.
Willnow TE, Nykjaer A, Herz J. Lipoprotein receptors: new roles for ancient proteins. Nat Cell Biol. 1999; 1: E157–E162.
Kadowaki H, Patton GM, Robins SJ. Metabolism of high density lipoprotein lipids by the rat liver: evidence for participation of hepatic lipase in the uptake of cholesteryl ester. J Lipid Res. 1992; 33: 1689–1698.
Schorsch F, Malle E, Sattler W. Selective uptake of high-density lipoprotein-associated cholesterylesters by differentiated Ob1771 adipocytes is modulated by endogenous and exogenous lipoprotein lipase. FEBS Lett. 1997; 414: 507–513.
Swarnakar S, Reyland ME, Deng J, Azhar S, Williams DL. Selective uptake of low-density lipoprotein-cholesteryl ester is enhanced by inducible apolipoprotein E expression in cultured mouse adrenocortical cells. J Biol Chem. 1998; 273: 12140–12147.
Vassiliou G, Benoist F, Lau P, Kavaslar GN, McPherson R. The low density lipoprotein receptor-related protein contributes to selective uptake of high density lipoprotein cholesteryl esters by SW872 liposarcoma cells and primary human adipocytes. J Biol Chem. 2001; 276: 48823–48830.
Vassiliou G. Pharmacological concentrations of suramin inhibit the binding of 2-macroglobulin to its cell-surface receptor. Eur J Biochem. 1997; 250: 320–325.
Porath J, Carlsson J, Olsson I, Belfrage G. Metal chelate affinity chromatography, a new approach to protein fractionation. Nature. 1975; 258: 598–599.
Vassiliou G, Stanley KK. Exogenous receptor-associated protein binds to two distinct sites on human fibroblasts but does not bind to the glycosaminoglycan residues of heparin sulfate proteoglycans. J Biol Chem. 1994; 269: 15172–15178.
Sattler W, Mohr D, Stocker R. Rapid isolation of lipoproteins and assessment of their peroxidation by high-performance liquid chromatography postcolumn chemiluminescence. Methods Enzymol. 1994; 233: 469–489.
Lucas M, Mazzone T. Cell surface proteoglycans modulate net synthesis and secretion of macrophage apolipoprotein E. J Biol Chem. 1996; 271: 13454–13460.
Hao M, Lin SX, Karylowski OJ, Wustner D, McGraw TE, Maxfield FR. Vesicular and non-vesicular sterol transport in living cells. The endocytic recycling compartment is a major sterol storage organelle. J Biol Chem. 2002; 277: 609–617.
Basu SK, Goldstein JL, Anderson RG, Brown MS. Monensin interrupts the recycling of low-density lipoprotein receptors in human fibroblasts. Cell. 1981; 24: 493–502.
Larkin JM, Brown MS, Goldstein JL, Anderson RG. Depletion of intracellular potassium arrests coated pit formation and receptor-mediated endocytosis in fibroblasts. Cell. 1983; 33: 273–285.
Oka JA, Weigel PH. Effects of hyperosmolarity on ligand processing and receptor recycling in the hepatic galactosyl receptor system. J Cell Biochem. 1988; 36: 169–183.
Weisgraber KH, Rall SC, Jr., Mahley RW, Milne RW, Marcel YL, Sparrow JT. Human apolipoprotein E. Determination of the heparin binding sites of apolipoprotein E3. J Biol Chem. 1986; 261: 2068–2076.
Swarnakar S, Beers J, Strickland DK, Azhar S, Williams DL. The apolipoprotein E-dependent low-density lipoprotein cholesteryl ester selective uptake pathway in murine adrenocortical cells involves chondroitin sulfate proteoglycans and an 2-macroglobulin receptor. J Biol Chem. 2001; 276: 21121–21128.
Ji ZS, Fazio S, Lee YL, Mahley RW. Secretion-capture role for apolipoprotein E in remnant lipoprotein metabolism involving cell surface heparin sulfate proteoglycans. J Biol Chem. 1994; 269: 2764–2772.
Gillotte KL, Zaiou M, Lund-Katz S, Anantharamaiah GM, Holvoet P, Dhoest A, Palgunachari MN, Segrest JP, Weisgraber KH, Rothblat GH, Phillips MC. Apolipoprotein-mediated plasma membrane microsolubilization. Role of lipid affinity and membrane penetration in the efflux of cellular cholesterol and phospholipid. J Biol Chem. 1999; 274: 2021–2028.
Phillips MC, Johnson WJ, Rothblat GH. Mechanisms and consequences of cellular cholesterol exchange and transfer. Biochim Biophys Acta. 1987; 906: 223–276.
Burgess JW, Gould DR, Marcel YL. The HepG2 extracellular matrix contains separate heparinase- and lipid-releasable pools of ApoE. Implications for hepatic lipoprotein metabolism. J Biol Chem. 1998; 273: 5645–5654.
Lin CY, Huang ZH, Mazzone T. Interaction with proteoglycans enhances the sterol efflux produced by endogenous expression of macrophage apoE. J Lipid Res. 2001; 42: 1125–1133.
Zhang WY, Gaynor PM, Kruth HS. Apolipoprotein E produced by human monocyte-derived macrophages mediates cholesterol efflux that occurs in the absence of added cholesterol acceptors. J Biol Chem. 1996; 271: 28641–28646.
Smith JD, Miyata M, Ginsberg M, Grigaux C, Shmookler E, Plump AS. Cyclic AMP induces apolipoprotein E binding activity and promotes cholesterol efflux from a macrophage cell line to apolipoprotein acceptors. J Biol Chem. 1996; 271: 30647–30655.
Narita M, Holtzman DM, Fagan AM, LaDu MJ, Yu L, Han X, Gross RW, Bu G, Schwartz AL. Cellular catabolism of lipid poor apolipoprotein E via cell surface LDL receptor-related protein. J Biochem (Tokyo). 2002; 132: 743–749.
Sparrow CP, Pittman RC. Cholesterol esters selectively taken up from high-density lipoproteins are hydrolyzed extralysosomally. Biochim Biophys Acta. 1990; 1043: 203–210.
Rhainds D, Falstrault L, Tremblay C, Brissette L. Uptake and fate of class B scavenger receptor ligands in HepG2 cells. Eur J Biochem. 1999; 261: 227–235.
Seo T, Al Haideri M, Treskova E, Worgall TS, Kako Y, Goldberg IJ, Deckelbaum RJ. Lipoprotein lipase-mediated selective uptake from low-density lipoprotein requires cell surface proteoglycans and is independent of scavenger receptor class B type 1. J Biol Chem. 2000; 275: 30355–30362.
Oram JF. AT. P-binding cassette transporter A1 and cholesterol trafficking. Curr Opin Lipidol. 2002; 13: 373–381.
von Eckardstein A, Langer C, Engel T, Schaukal I, Cignarella A, Reinhardt J, Lorkowski S, Li Z, Zhou X, Cullen P, Assmann G. ATP binding cassette transporter ABCA1 modulates the secretion of apolipoprotein E from human monocyte-derived macrophages. FASEB J. 2001; 15: 1555–1561.
Remaley AT, Stonik JA, Demosky SJ, Neufeld EB, Bocharov AV, Vishnyakova TG, Eggerman TL, Patterson AP, Duverger NJ, Santamarina-Fojo S, Brewer HB, Jr. Apolipoprotein specificity for lipid efflux by the human ABCAI transporter. Biochem Biophys Res Commun. 2001; 280: 818–823.(Gerard Vassiliou; Ruth Mc)