Ubiquitination is associated with lysosomal degradation of cell surface-resident ATP-binding cassette transporter A1 (ABCA1) through the endosomal sorting complex required for transport (ESCRT) pathway

Authors

  • Tadahaya Mizuno,

    1. Laboratory of Molecular Pharmacokinetics, Graduate School of Pharmaceutical Sciences, University of Tokyo, Tokyo, Japan
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Hisamitsu Hayashi,

    Corresponding author
    1. Laboratory of Molecular Pharmacokinetics, Graduate School of Pharmaceutical Sciences, University of Tokyo, Tokyo, Japan
    • Laboratory of Molecular Pharmacokinetics, Department of Medical Pharmaceutics, Graduate School of Pharmaceutical Sciences, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan
    Search for more papers by this author
    • These authors contributed equally to this work.

    • fax: +81-3-5841-4766

  • Sotaro Naoi,

    1. Laboratory of Molecular Pharmacokinetics, Graduate School of Pharmaceutical Sciences, University of Tokyo, Tokyo, Japan
    Search for more papers by this author
  • Yuichi Sugiyama

    Corresponding author
    1. Laboratory of Molecular Pharmacokinetics, Graduate School of Pharmaceutical Sciences, University of Tokyo, Tokyo, Japan
    • Laboratory of Molecular Pharmacokinetics, Department of Medical Pharmaceutics, Graduate School of Pharmaceutical Sciences, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan
    Search for more papers by this author
    • fax: +81-3-5841-4766


  • Potential conflict of interest: Nothing to report.

  • Supported by the Program for Promotion of Fundamental Studies in Health Sciences of the National Institute of Biomedical Innovation (NIBIO) and Grant-in-Aid for Young Scientists (B) (21790146).

Abstract

ATP-binding cassette transporter A1 (ABCA1) plays an essential role in the biogenesis of high-density lipoprotein in liver and in the prevention of foam cell formation in macrophages by mediating the efflux of cellular cholesterol and phospholipids to apolipoprotein A-I (apoA-I). Our current study investigated the mechanism of degradation of cell surface-resident ABCA1, focusing on ubiquitination. A coimmunoprecipitation study indicated the presence of ubiquitinated ABCA1 in the plasma membrane of the human hepatoma cell line, HuH-7, of cells from mouse liver, and of macrophages differentiated from the human acute monocytic leukemia cell line, THP-1 (THP-1 macrophages). In HuH-7 cells, degradation of cell surface-resident ABCA1 was inhibited by the overexpression of a dominant-negative form of ubiquitin. Moreover, the disruption of the endosomal sorting complex required for transport (ESCRT) pathway, a dominant mechanism for ubiquitination-mediated lysosomal degradation, by the knockdown of hepatocyte growth factor-regulated tyrosine kinase substrate (HRS), significantly delayed the degradation of cell surface-resident ABCA1. This was accompanied by an increase in ABCA1 expression as well as in apoA-I-mediated [3H]-cholesterol efflux function. The effect of HRS knockdown was also observed after calpain inhibitor treatment, which is reported to retard ABCA1 degradation. The induction of ABCA1 by HRS knockdown was confirmed in THP-1 macrophages. Conclusion: Together with the fact that lysosomal inhibitor treatments increased ABCA1 expression in HuH-7 and THP-1 macrophages, these results suggest that ubiquitination mediates the lysosomal degradation of cell surface-resident ABCA1 through the ESCRT pathway, and thereby controls the expression and cholesterol efflux function of ABCA1. This mechanism seems to mediate ABCA1 degradation independently of the calpain-involving pathway. The modulation of ABCA1 ubiquitination could thus be a potential new therapeutic target for antiatherogenic drugs. (HEPATOLOGY 2011;)

ATP-binding cassette transporter A1 (ABCA1) mediates the efflux of cellular cholesterol and phospholipids to apolipoprotein A-I (apoA-I).1, 2 Its physiological importance has been demonstrated by Tangier disease, which is caused by a mutation in the ABCA1 gene and shows near absence of high-density lipoprotein (HDL) in the plasma and increased susceptibility to atherosclerotic cardiovascular disease.1 Although ABCA1 is expressed in various tissues,3 a growing body of evidences derived from tissue-specific loss-of-function and gain-of-function studies of ABCA1 in mice indicates its essential contribution to HDL biogenesis in the liver.4, 5 Plasma HDL has various antiatherosclerotic effects including the promotion of reverse cholesterol transport, which mediates the transport of excess cholesterol from peripheral tissues to the liver,6 and NO-dependent vasorelaxation in endothelial cells.7 Indeed, an inverse relationship of plasma HDL concentration with atherosclerotic cardiovascular disease risk has been reported.8 In addition, studies of transplantation of ABCA1-deficient bone marrow showed that, in macrophages, ABCA1 plays a role in cholesterol extrusion, preventing foam cell formation and subsequent progression of atherogenesis.9 This background has generated great interest in modulating atherosclerotic cardiovascular disease by increasing ABCA1 expression and function.

Because the expression level of ABCA1 protein is discordant with that of the corresponding messenger RNA (mRNA)3 and the turnover of ABCA1 protein is rapid,10 ABCA1 expression can be tightly controlled at the protein level as well as mRNA level.11-13 Despite several articles reporting contradictory results,14 ABCA1 on the plasma membrane, but not in intracellular compartments, appears to play a dominant role in apoA-I-mediated cholesterol efflux.15 Thus, the elucidation of the mechanism of degradation of cell surface-resident ABCA1 is considered especially important for the understanding of regulation of ABCA1 function. Calpain protease has been reported to regulate the degradation of ABCA1 expressed on the plasma membrane,10, 16 but the contribution of other mechanisms to the degradation of cell surface-resident ABCA1 is unclear.

Ubiquitination is a regulated posttranslational modification that conjugates ubiquitin (Ub) to lysine residues of targeted proteins and determines their intracellular fate.17 Although the canonical role of ubiquitination is to mediate degradation by the proteasome of the proteins that carry a single or polymeric chain of Ub, it has recently been revealed that single Ub modifications at one, two, or several different sites in substrate proteins also act as a signal for degradation of cell surface-resident membrane proteins by the direction of ubiquitinated membrane protein to lysosomal degradation in the early endosome.17, 18 A set of proteins called endosomal sorting complex required for transport (ESCRT), which consists of ESCRT-0, -I, -II, and -III, plays an essential role in this ubiquitination-mediated lysosomal degradation through the recognition of the ubiquitinated cargo in the early endosome and its recruitment to and incorporation into multivesicular bodies (MVBs), which lead to lysosomal degradation.19

We and other groups have previously reported that this type of ubiquitination is associated with the degradation of cell surface-resident bile salt export pump (BSEP) and cystic fibrosis transmembrane conductance regulator (CFTR), both of which belong to the ABC transporter family, as does ABCA1.20, 21 Moreover, it has been demonstrated that ABCA1 is localized not only on the plasma membrane, but also in the lysosome.22 On the basis of these data, in the present study we examine the mechanism of degradation of cell surface-resident ABCA1, mainly using the human hepatoma cell line, HuH-7. We focus on ubiquitination and show that ubiquitination is associated with lysosomal degradation of cell surface-resident ABCA1 through the ESCRT pathway.

Abbreviations

ABCA1, ATP-binding cassette transporter A1; ABCA1-Ub, ubiquitinated ABCA1; apoA-I, apolipoprotein A-I; ESCRT, endosomal sorting complex required for transport; HDL, high-density lipoprotein; HRS, hepatocyte growth factor-regulated tyrosine kinase substrate; mLPM, mouse plasma membrane fractions from liver; stRNA, Stealth RNAi; TSG101, tumor susceptibility gene 101; Ub, ubiquitin; UbΔGG, in which the two last Ub glycines were deleted; UbΔGG/I44A, incorporating Ub I44A in addition to deletion of the last two glycines.

Materials and Methods

Materials.

Human recombinant apoA-I was purchased from BioDesign (Saco, ME). [3H]-Cholesterol was obtained from PerkinElmer (Boston, MA). The antibodies and Stealth RNA interference (RNAi) sequences (Invitrogen, Carlsbad, CA) used in this study are described in the Supporting Information. All other chemicals were of analytical grade.

HeLa and HuH-7 cells were cultured in Dulbecco's modified Eagle's medium (DMEM, Invitrogen) containing 10% fetal bovine serum (FBS, Invitrogen) in the presence (HeLa) or absence (HuH-7) of 100 μM nonessential amino acids (Invitrogen). THP-1 cells were cultured in Roswell Park Memorial Institute 1640 (RPMI 1640, Invitrogen) medium supplemented with 10% FBS at 37°C with 5% CO2 and 95% humidity. Cells were differentiated into macrophages (THP-1 macrophages) by 72-hour treatment with phorbol 12-myristate 13-acetate (PMA, 160 nM).

Construction of Plasmid Vector.

The complementary DNAs (cDNAs) of ABCA1 and Ub were amplified by way of polymerase chain reaction (PCR) with Ex-Taq (Takara Bio, Shiga, Japan) from a commercially available cDNA library (ClonTech, Palo Alto, CA) and subcloned into pcDNA3.1(+) (Neomycin) (Invitrogen). The sequence of the ABCA1 cDNA was identical to that published (Access. No., NM_005502.2). The method for construction of the other plasmid vector is described in the Supporting Information.

Animals.

ddY mice (6-7 weeks old) were purchased from Nippon SLC (Shizuoka, Japan). All animals were maintained under standard conditions with a reverse dark-light cycle and were treated humanely. Food and water were available ad libitum. The studies reported in this article were carried out in accordance with the guidelines provided by the Institutional Animal Care Committee (Graduate School of Pharmaceutical Sciences, University of Tokyo, Tokyo, Japan).

Immunoprecipitation from Cultured Cells and Mouse Liver.

Prepared crude membrane fractions from HeLa cells expressing ABCA1, containing the FLAG epitope linearly conjugated three times at the N-terminus (3×FLAG-ABCA1), from HuH-7 cells,23 and from THP-1 macrophages and plasma membrane fractions from the liver of male ddY mice (mLPM),20, 23 were used for immunoprecipitation to verify the presence of ubiquitinated ABCA1 (ABCA1-Ub). In the experiment to investigate the effect of apoA-I, HuH-7 cells were treated with 10 μg/ml apoA-I for 3 hours before sample preparation. Immunoprecipitation was performed as described20 with minor modifications. The detailed protocol is provided in the Supporting Information.

To examine the interaction of ABCA1 with hepatocyte growth factor-regulated tyrosine kinase substrate (HRS), 3×FLAG-ABCA1-expressing HeLa cells were solubilized for 1 hour at 4°C in 1 mL lysis buffer (20 mM Tris-HCl [pH 7.5], 150 mM NaCl, 1 mM EDTA, 1% Nonidet P-40, 10 mM N-ethylmaleimide, and a protease inhibitor cocktail tablet), then centrifuged at 12,000g for 10 minutes at 4°C. The precleared supernatants were first incubated with anti-FLAG antibody for 2 hours at 4°C, and then with protein G-Sepharose beads (Roche) overnight at 4°C. Immune complexes were precipitated, washed, and eluted as described.20 The specimens were separated by 6% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and subjected to western blot analysis.

Cell Surface Biotinylation and Determination of the Rate of Degradation of Cell Surface Expressed Protein.

HuH-7 cells were seeded in 6-well plates at a density of 5.0 × 105 cells per well. THP-1 cells were seeded in 6-well plates at a density of 1.5 × 106 cells per well and subsequently differentiated into macrophages with PMA treatment for 72 hours. These cells were transfected with Stealth RNAi for hepatocyte growth factor-regulated tyrosine kinase substrate (HRS) (HRS stRNA), tumor susceptibility gene 101 (TSG101) (TSG101 stRNA), or Negative Control Medium GC Duplex (Negative stRNA) using Lipofectamine RNAiMax (Invitrogen) according to the manufacturer's instructions. Seventy-two hours after stRNA transfection, cell surface biotinylation was performed to investigate the level of expression on the plasma membrane and the rate of degradation of cell surface-resident protein, as described.23 In the experiments where it was necessary to inhibit calpain activity, the cells were incubated with DMSO or calpeptin (30 μM, Calbiochem, Darmstadt, Germany) for the indicated time before cell surface biotinylation. To examine the effect of the expression of UbΔGG (in which the two last Ub glycines were deleted) and UbΔGG/I44A (incorporating Ub I44A in addition to deletion of the last two glycines) on ABCA1 degradation rate, HuH-7 cells were seeded in 6-well plates at a density of 1.0 × 106 cells per well, and subsequently transfected with empty pcDNA3.1(+), pcDNA3.1(+) containing N-terminally HA-tagged-UbΔGG (HA-UbΔGG) cDNA, or HA-UbΔGG/I44A cDNA using Lipofectamine 2000. Forty-eight hours after the transfection, cell surface biotinylation was performed. Isolated biotinylated protein was separated by 6% SDS-PAGE and subjected to western blot analysis.

Western Blot Analysis.

Specimens were loaded into wells of a 6% SDS-PAGE plate with a 3.75% stacking gel and subjected to western blot analysis as described.23 Immunoreactivity was detected with an ECL Advance Western Blotting Detection Kit (Amersham Biosciences, Piscataway, NJ). The intensity of the band indicating ABCA1 was quantified by Multi Gauge software v. 2.0 (Fujifilm, Tokyo, Japan).

[3H]-Cholesterol Efflux Assay.

HuH-7 cells were seeded in 24-well plates at a density of 1.25 × 105 cells per well and subsequently transfected with HRS stRNA, TSG101 stRNA, or Negative stRNA using Lipofectamine RNAiMax. Seventy-two hours after stRNA transfection, cells were labeled with 10% FBS/DMEM containing 2 μCi/mL [3H]-cholesterol for 24 hours, then washed with phosphate-buffered saline (PBS) three times and incubated in fresh medium for 2 hours followed by replacement of incubation medium with 0.02% BSA/DMEM with or without 10 μg/mL apoA-I. After a 4-hour culture, the medium was collected. Cells were lysed with lysis buffer (0.1% SDS and 0.1 M NaOH) after the efflux assay. The radioactivity in medium and cell lysate was determined by scintillation counting (LS 600SC, Beckman, Fullerton, CA). Cholesterol efflux was expressed as the percentage of the radioactivity released from the cells into the medium relative to the total radioactivity in cells plus medium. The results are expressed as apoA-I-mediated cholesterol efflux, which was calculated by subtraction of the cholesterol efflux value during incubation with apoA-I free medium from that during incubation with apoA-I-containing medium.

Statistical Analysis.

Experiments were repeated at least three times, and graphs include means ± standard error (SE). P-values between two variables and multiple variables were calculated at the 95% confidence level with Student's t test and analysis of variance (ANOVA), using Prism software (GraphPad Software, La Jolla, CA).

Results

Ubiquitination of ABCA1 In Vitro and In Vivo.

HuH-7 cells originated from a human hepatoma and endogenously express ABCA1. To determine whether ABCA1 on the cell surface is ubiquitinated, ABCA1 was immunoprecipitated with anti-ABCA1 antibody from crude membrane fractions of HuH-7 cells. Subsequent western blot analysis for Ub detected an ABCA1-Ub band at a slightly higher molecular weight than ABCA1 itself, which was predominantly detected as a 260-kDa band, its reported molecular weight (Fig. 1, left panel).10 The presence of ABCA1-Ub was also confirmed in vivo. In immunoprecipitates from mLPM with anti-ABCA1 antibody, ABCA1-Ub and ABCA1 were predominantly detected as 300-kDa and 260-kDa bands (Fig. 1, right panel), respectively. Although the bands corresponding to ABCA1-Ub (300 kDa) were not detected by anti-ABCA1 antibody, this may be accounted for by the low amount of ABCA1-Ub compared with nonubiquitinated ABCA1 in the steady state. Given that the molecular weight of a Ub is 8 kDa, these results show that ABCA1 on the plasma membrane can be modified with a few Ub molecules during the steady state. This implies the presence of cell surface-resident ABCA1 carrying not a polyubiquitin chain, but single Ub modifications at several different sites, which can function as the lysosomal degradation signal for membrane proteins.17, 18

Figure 1.

Ubiquitination of ABCA1. ABCA1 was immunoprecipitated from solubilized crude membrane fractions of HuH-7 cells (left panel) and solubilized mLPM (right panel) with polyclonal rabbit anti-ABCA1 antibody or control IgG. Immunoprecipitates were separated by 6% SDS-PAGE and subjected to western blot analysis.

Some ABCA1s in mouse liver were detected as higher molecular weight bands than previously reported (260 kDa); these were not the same size as ABCA1-Ub (Fig. 1). These bands disappeared after reduction (data not shown), indicating that, as reported,24 they were the result of disulfide bonds, not ubiquitination.

To investigate the ubiquitination of ABCA1 in extrahepatic tissues, THP-1 cells, a human acute monocytic leukemia cell line, was used after it was differentiated into macrophages by PMA treatment. This cell line is commonly employed as a macrophage model in studies of ABCA1.10 Immunoprecipitation of ABCA1 from crude membrane fractions of THP-1 macrophages with anti-ABCA1 antibody and subsequent western blot analysis showed the presence of ABCA1-Ub and ABCA1 at the same molecular weight as in HuH-7 cells (Fig. 1 and Supporting Fig. 1A), indicating that ABCA1 on the plasma membrane is ubiquitinated in macrophages as well as in hepatocytes.

Increase in ABCA1 Expression by Lysosomal Inhibitor Treatment.

One of the representative roles of ubiquitination is as a target signal for degradation of cell surface-resident membrane proteins by way of sorting from endosomes to lysosomes.17, 18 To explore the involvement of this pathway in degradation of cell surface-resident ABCA1, the effect of the typical lysosomal inhibitors, chloroquine and NH4Cl, on the level of ABCA1 expression in HuH-7 cells and THP-1 macrophages was examined by western blot analysis using whole cell lysates. Incubation with these inhibitors for 12 hours (chloroquine) or 24 hours (NH4Cl) at the commonly used concentrations25 caused an approximately 2-fold increase in ABCA1 expression (Fig. 2 and Supporting Fig. 1B), consistent with our hypothesis and with previous reports that showed the lysosomal localization of ABCA122 and ABCA1 induction with NH4Cl treatment.25

Figure 2.

Effect of lysosomal inhibitor on ABCA1 expression. (Upper panel) HuH-7 cells were treated with 30 μM chloroquine for 12 hours or 2.5 mM NH4Cl for 24 hours. After the treatment, whole cell lysates were prepared. Prepared specimens were separated by 6% SDS-PAGE and subjected to western blot analysis. (Lower panel) The density of the band representing ABCA1 in the upper panel was quantified by Image Gauge software and expressed as a percentage of the ABCA1 present in vehicle-treated HuH-7 cells. Each bar represents the mean ± SE from three independent experiments. Asterisks represent significant differences between vehicle- and lysosomal inhibitor-treated HuH-7 cells (**P < 0.01).

Prolongation of the Half-Life of Cell Surface-Resident ABCA1 by Expression of Ub Mutants.

The degradation rate of cell surface-resident ABCA1 was determined in HA-UbΔGG- or HA-UbΔGG/I44A-overexpressing HuH-7 cells. UbΔGG is a Ub-mutant that lacks the glycine-glycine sequence at the C-terminus (Fig. 3A). Because the C-terminal glycines are essential amino acid residues for Ub conjugation to a targeted protein, this deletion mutant can inhibit the ubiquitination reaction.26 Moreover, the overexpression of UbΔGG disrupts ubiquitination-mediated processes through an inhibitory effect on the interaction of Ub with the Ub-binding adaptor proteins that recruit the ubiquitinated cargoes from the cell surface for lysosomal degradation.21, 26 Therefore, UbΔGG can act as a dominant-negative form of Ub in two independent ways. In contrast, because the isoleucine at the 44th amino acid residue of Ub has been reported to be essential for the interaction with the Ub-binding domain (Fig. 3A), contained in the Ub-conjugating enzyme and in the Ub-binding adaptor proteins,20, 21, 26 UbΔGG/I44A has no effect on the ubiquitination reaction and ubiquitination-mediated pathway. Therefore, its overexpression was used as a negative control.20

Figure 3.

Effect of Ub mutants on the rate of degradation of cell surface-resident ABCA1. (A) Diagram illustrating the constructs of Ub and the mutants used in this study. (B) Expression of the mutated form of Ub in HuH-7 cells. Whole cell lysates were prepared from HuH-7 cells expressing HA-UbΔGG, HA-UbΔGG/I44A, or empty vector (E.V.). The prepared specimens were separated by 6% SDS-PAGE and subjected to western blot analysis. (C) Rate of degradation of cell surface-resident ABCA1 in HuH-7 cells expressing HA-UbΔGG, HA-UbΔGG/I44A, or E.V. HuH-7 cells transfected with the indicated plasmid were biotinylated and incubated for the indicated time at 37°C, as described in Materials and Methods. The remaining biotin-labeled proteins isolated with streptavidin beads were separated by 6% SDS-PAGE and subjected to western blot analysis. (D) Quantification of the density of the band representing ABCA1 in (C). The intensity of the band was quantified by Image Gauge software and expressed as a percentage of the ABCA1 present at 0 hours. Open circles, closed circles, and open squares represent remaining biotin-labeled ABCA1 in E.V.-, HA-UbΔGG-, and HA-UbΔGG/I44A-expressing HuH-7 cells, respectively. Each bar represents the mean ± SE of three independent experiments. Asterisks represent significant differences between E.V.- and HA-UbΔGG-expressing HuH-7 cells (*P < 0.05).

After the construction of HA-UbΔGG and HA-UbΔGG/I44A-overexpressing HuH-7 cells (Fig. 3B), the rate of degradation of cell surface-resident ABCA1 was determined by cell surface biotinylation. Overexpression of HA-UbΔGG significantly prolonged the half-life of cell surface-resident ABCA1 more than 2-fold, whereas HA-UbΔGG/I44A overexpression had no effect on it (Fig. 3C,D), suggesting that the ubiquitination-mediated sorting machinery is implicated in degradation of cell surface-resident ABCA1.

Involvement of HRS and TSG101 in Degradation of Cell Surface-Resident ABCA1.

The ESCRT pathway acts in a cascade from ESCRT-0 to ESCRT-III, and HRS and TSG101 play central roles in the function of ESCRT-0 and ESCRT-I, as demonstrated in the previous report showing that EGFR degradation was delayed by the depletion of HRS or TSG101.27

The importance of the ESCRT machinery in ubiquitination-mediated degradation of plasma membrane-associated ABCA1 was investigated using HRS- or TSG-depleted HuH-7 cells (Fig. 4A). Cell surface biotinylation demonstrated 2- to 4-fold prolongation of the half-life of cell surface-resident ABCA1 by HRS and TSG knockdown (Fig. 4B,C). Next, to examine whether the ESCRT pathway is directly involved in lysosomal degradation of ABCA1, a coimmunoprecipitation study was performed using 3×FLAG-ABCA1-overexpressing HeLa cells. Endogenous HRS coimmunoprecipitated with FLAG antibody was detected at the reported molecular weight by western blot analysis,27 indicative of the interaction between ABCA1 and HRS (Fig. 4D). It seems that ABCA1-Ub is directly recognized by HRS, leading to its incorporation into MVBs and to lysosomal degradation. We used 3×FLAG-ABCA1-overexpressing HeLa cells in the immunoprecipitation experiment, because it was too difficult to detect the interaction between ABCA1 and HRS in HuH-7 cells due to low endogenous expression level of ABCA1 and low transfection efficiency. HeLa cells are widely used for membrane sorting studies27 and the ubiquitination of exogenously expressed 3×FLAG-ABCA1 was also confirmed in this cell line (Fig. 4D). Therefore, we selected it as the host cell line.

Figure 4.

Involvement of HRS and TSG101 in the pathway for degradation of cell surface-resident ABCA1. (A) Knockdown of HRS and TSG101 in HuH-7 cells. Whole cell lysates were prepared from HuH-7 cells transfected with HRS stRNA, TSG101 stRNA, or Negative stRNA. The prepared specimens were separated by 6% SDS-PAGE and subjected to western blot analysis. (B,C) Rate of degradation of cell surface-resident ABCA1 in HRS- or TSG101-depleted HuH-7 cells. (Left panel) HuH-7 cells were transfected with Negative stRNA, HRS stRNA (B), or TSG101stRNA (C). StRNA-transfected HuH-7 cells were biotinylated and incubated for the indicated time at 37°C, as described in Materials and Methods. The remaining biotin-labeled proteins isolated with streptavidin beads were separated by 6% SDS-PAGE and subjected to western blot analysis. (Right panel) Quantification of the density of the band indicating ABCA1 in the left panel. The intensity of the band was quantified by Image Gauge software and expressed as a percentage of the ABCA1 present at 0 hours. Open circles and closed circles represent the remaining biotin-labeled ABCA1 in Negative stRNA-, and HRS stRNA- (B) or TSG101 stRNA- (C) transfected HuH-7 cells, respectively. Each bar represents the mean ± SE of three independent experiments. Asterisks represent significant differences between Negative stRNA- and HRS stRNA- (B) or TSG101 stRNA- (C) transfected HuH-7 cells (*P < 0.05). (D) Interaction of ABCA1 with HRS. 3×FLAG-ABCA1 was immunoprecipitated from crude membrane fractions (left panel) or whole cell lysates (right panel) of 3×FLAG-ABCA1-expressing HeLa cells with mouse-monoclonal anti-FLAG antibody or control IgG. Immunoprecipitated specimens were separated by 6% SDS-PAGE and subjected to western blot analysis. Asterisk indicates nonspecific band. E.V., empty vector.

Effect of HRS or TSG101 Knockdown on ABCA1 Expression and Function.

To examine the contribution of the ESCRT machinery to ABCA1 expression and function, these were evaluated under HRS or TSG101 knockdown condition. HRS knockdown in HuH-7 cells induced ABCA1 expression in both whole cell lysates and cell surface fractions (Fig. 5A), with a greater induction being detected in the cell surface fractions than in the whole cell lysates (Fig. 5B). ABCA1 expression increased 5.7-fold in the cell surface fractions, and 2.7-fold in the whole cell lysates. In contrast, TSG101 knockdown induced ABCA1 only in the whole cell lysates, but not in the cell surface fractions (Fig. 5C,D). A 1.5-fold increase in ABCA1 was detected in the whole cell lysates. An effect of HRS knockdown on ABCA1 expression was also observed in THP-1 macrophages. HRS stRNA transfection significantly induced ABCA1 expression 1.6-fold in whole cell lysates and 1.9-fold in cell surface fractions (Supporting Fig. 1C,D).

Figure 5.

Effect of HRS or TSG101 on ABCA1 expression and function. (A-D) Effect of HRS (A,B) or TSG101 (C,D) knockdown on ABCA1 expression in whole cell lysates and cell surface fractions. HuH-7 cells were transfected with Negative stRNA, HRS stRNA (A,B) or TSG101 stRNA (C,D). Seventy-two hours after stRNA transfection, whole cell lysates were prepared and surface biotinylation reactions were performed. (A,C) These specimens were separated by 6% SDS-PAGE and subjected to western blot analysis. (B,D) Quantification of the density of the bands indicating ABCA1 in (A,C). The intensity of the bands was quantified by Image Gauge software and expressed as a percentage of the ABCA1 present in Negative stRNA-transfected HuH-7 cells. Each bar represents the mean ± SE of three independent experiments. Asterisks represent significant differences between Negative stRNA-transfected HuH-7 cells and HRS stRNA- (B) or TSG101 stRNA- (D) transfected HuH-7 cells, and daggers represent significant differences between the variation in the whole cell lysates and on the cell surface (*P < 0.05; **††P < 0.01). (E) Effect of HRS or TSG101 knockdown on apoA-I-mediated cholesterol efflux. HuH-7 cells transfected with Negative stRNA, HRS stRNA, or TSG101 stRNA were labeled with [3H]-cholesterol and incubated with or without 10 μg/mL apoA-I, as described in Materials and Methods. After a 4-hour incubation, radioactivity in the medium and in the cells was measured. Results were expressed as apoA-I-mediated cholesterol efflux, as described in Materials and Methods. Each bar represents the mean ± SE of three independent experiments. Asterisks represent significant differences between Negative stRNA- and HRS stRNA-transfected HuH-7 cells (***P < 0.001).

ApoA-I-mediated [3H]-cholesterol efflux was measured as described for other cell lines16 to evaluate ABCA1 function in HuH-7 cells, because it was confirmed from the disappearance of apoA-I-mediated [3H]-cholesterol efflux under ABCA1 knockdown conditions (data not shown) that this parameter reflects ABCA1 function in HuH-7 cells. Knockdown of HRS, but not TSG101, induced a significant (1.5-fold) increase in apoA-I mediated [3H]-cholesterol efflux (Fig. 5E).

Relationship of the Degradation of Cell Surface-Resident ABCA1 by the Ubiquitination-Mediated Lysosomal Pathway with That by the Calpain-Mediated Pathway.

It has been reported that calpain-mediated cleavage is implicated in the degradation of ABCA1 and that this mechanism is strongly inhibited by apoA-I.16 We investigated the relationship between the degradation of cell surface-resident ABCA1 by ubiquitination-mediated lysosomal pathway and that by the calpain-mediated pathway. Because human apoA-I concentration in the culture medium of HuH-7 cells was lower than the concentration required to inhibit the calpain-mediated cleavage of ABCA1 (Supporting Fig. 2),16 the association of the two mechanisms can be examined using this cell line. Calpeptin, a typical inhibitor of calpain,10 was employed for this study. Consistent with a previous report28 and the observation of low apoA-I concentration in culture media (Supporting Fig. 2), ABCA1 expression in HuH-7 cells was induced by calpeptin in both whole cell lysates and cell surface fractions, the optimal incubation being for 3 hours (Fig. 6A). After culturing Negative stRNA- or HRS stRNA-transfected HuH-7 cells with calpeptin for 3 hours, the degradation rate of ABCA1 on plasma membrane in the presence of calpeptin was measured using cell surface biotinylation. The half-life of cell surface-resident ABCA1 was longer in HRS stRNA-transfected cells than in Negative stRNA-transfected cells and the amount of biotin-labeled ABCA1 remaining after a 4-hour incubation was significantly higher in HRS stRNA-transfected cells than in Negative stRNA-transfected cells (Fig. 6B). This retarded degradation was reflected in the ABCA1 expression. Even after calpeptin treatment under optimal conditions, ABCA1 expression was higher in HRS stRNA-transfected cells than in Negative stRNA-transfected cells, with a greater induction being detected in the cell surface fractions than in the whole cell lysates (Fig. 6C D). A 2.6-fold increase in the cell surface fractions and a 1.8-fold increase in the whole cell lysates were observed in HRS stRNA-transfected cells compared with Negative stRNA-transfected cells. Taken together, these results suggest that the calpain-mediated pathway and the ubiquitination-mediated lysosomal pathway are independently associated with the degradation of ABCA1.

Figure 6.

Investigation of the implications of calpain in ubiquitination-mediated degradation of cell surface-resident ABCA1. (A) Time-dependent effect of calpeptin treatment on ABCA1 expression. HuH-7 cells were treated with 30 μM calpeptin for the times indicated before preparation of the whole cell lysates. The prepared specimens were separated by 6% SDS-PAGE and subjected to western blot analysis. (B) Effect of calpeptin treatment on degradation rate of cell surface-resident ABCA1 under HRS-depleted conditions. (Left panel) Negative stRNA- or HRS stRNA-transfected HuH-7 cells were treated with 30 μM calpeptin for 3 hours, then biotinylated, and incubated for the indicated time with 30 μM calpeptin-containing media at 37°C. The remaining biotin-labeled proteins were analyzed, as described in Materials and Methods. (Right panel) Quantification of the density of the band indicating ABCA1 in the left panel. The intensity of the band was quantified and expressed in the same way as in Fig. 4B. Each bar represents the mean ± SE of three independent experiments. Asterisks represent significant differences between Negative stRNA- and HRS stRNA-transfected HuH-7 cells (*P < 0.05). (C,D) Effect of calpeptin treatment on ABCA1 expression in HRS-depleted conditions. (C) Whole cell lysates and cell surface fractions were prepared from Negative stRNA- or HRS stRNA-transfected HuH-7 cells after 30 μM calpeptin treatment for 3 hours as described in Materials and Methods. These specimens were analyzed by western blot analysis. (D) Quantification of the density of the bands indicating ABCA1 in (C), in the same way as in Fig. 5C. Each bar represents the mean ± SE of three independent experiments. Asterisks represent significant differences between Negative stRNA- and HRS stRNA-transfected HuH-7 cells, and daggers represent significant differences between the variation in the whole cell lysates and on the cell surface (*P < 0.05; ††P < 0.01; ***P < 0.001).

Discussion

ABCA1 plays an essential role in HDL biogenesis in liver and in the prevention of foam cell formation in macrophages through its function of mediating the efflux of cellular cholesterol and phospholipids to apoA-I, thus protecting against atherosclerotic cardiovascular disease. It has been reported that the expression level of ABCA1 protein is discordant with that of the corresponding mRNA3 and the turnover of ABCA1 protein is rapid,10 implying that the posttranslational regulation plays an essential role in the determination of ABCA1 expression. However, the underlying mechanism of this process is still elusive. Our current study was designed to investigate the mechanism of degradation of cell surface-resident ABCA1. The experiments mainly used the human hepatoma cell line, HuH-7 cells, as model cells and showed that ubiquitination is associated with this degradation through the ESCRT pathway, as demonstrated by the finding that the overexpression of dominant-negative Ub (Fig. 3C,D) and the knockdown of HRS or TSG101 (Fig. 4B,C) significantly reduced the rate of degradation of cell surface-resident ABCA1. Given that the inhibitory effect of HRS knockdown on cell surface-resident ABCA1 degradation was accompanied by an increase in ABCA1 expression both in the whole cell lysates and on the plasma membrane, thereby enhancing apoA-I-mediated cholesterol efflux (Fig. 5A,B,E), this degradation machinery contributes to the determination of ABCA1 expression and function in the liver.

Although both HRS and TSG101 are components of the ESCRT pathway, TSG101 knockdown showed effects on ABCA1 expression and function that were distinct from those of HRS knockdown, in that ABCA1 expression was increased only in the whole cell lysates, not on the plasma membrane, and its function was not enhanced (Fig. 5C-E). However, this is not inconsistent with our current conclusion. HRS consists of ESCRT-0, the initial component of the ESCRT pathway, which recognizes and captures the ubiquitinated cargoes in the early endosome, whereas TSG 101 consists of ESCRT-I, which recruits the subsequent complex, ESCRT-II, and consequently contributes to gating the ubiquitinated cargoes within invaginations.19 Taking this into consideration, it is conceivable that, under HRS knockdown conditions, ABCA1-Ub can recycle to the plasma membrane by escaping the recognition mechanism of ESCRT, and thereby ABCA1 expression was elevated to a greater extent on plasma membrane than that in the whole cell lysates (Fig. 5B). In contrast, under TSG101 knockdown conditions, ABCA1 accumulates in the early endosome (Supporting Fig. 3B) through HRS-mediated arrest and subsequent inhibition of its degradation. Moreover, the difference of the effects of HRS and TSG101 on ABCA1 function can be accounted for by a previous report showing that the recycling and accumulation of ABCA1 on the cell surface play a pivotal role in apoA-I-mediated cholesterol efflux,15 although several studies have reported contradictory results.14

In THP-1 macrophages and in HuH-7 cells, ABCA1-Ub was detected in the membrane-enriched fraction, and either lysosome inhibitor or HRS knockdown increased ABCA1 expression (Supporting Fig. 1), suggesting that ubiquitination-mediated lysosomal degradation through the ESCRT pathway regulates ABCA1 expression and function in both macrophages and liver. The only difference in the results of the studies in THP-1 macrophages from those in HuH-7 cells is that in THP-1 macrophages the increasing rate of ABCA1 by knockdown of HRS was equivalent between whole cell lysates and cell surface fractions (Supporting Fig. 1D). A possible reason for this is a difference between HuH-7 cells and THP-1 macrophages in the cellular localization of ABCA1, although we could not confirm this using immunocytochemistry because of the low ABCA1 expression in THP-1 macrophages. If the ratio of ABCA1 expression at cell surface to that in the whole cell is higher in THP-1 macrophages than in HuH-7 cells, the increased expression in cell surface fractions by HRS depletion should be reflected in that of whole cell lysates. Alternatively, it is possible that in THP-1 macrophages, ABCA1 has sorting pathways other than recycling to plasma membrane after escaping lysosomal targeting at the early endosome, meaning that the increased rate of ABCA1 expression at the cell surface after HRS knockdown is lower in THP-1 macrophages than in HuH-7 cells.

The ubiquitination of substrate proteins is mediated by the covalent attachment of Ub by way of the sequential action of three enzymes: a Ub-activating enzyme E1, a Ub-conjugating enzyme E2, and a Ub ligase E3.17 Among these three enzymes, E3s are considered to ensure the correct timing, localization, and specificity of the ubiquitination reaction, because genomic information suggests that there are several hundreds of E3s in eukaryotic cells.29 The molecular mechanism regulating ABCA1 ubiquitination on the cell surface remains to be established. However, because it was demonstrated in a study of epidermal growth factor receptor (EGFR) that autophosphorylation of EGFR by ligand binding promotes its recognition by the ring-type E3, casitas B-lineage lymphoma (Cbl),30 it is possible that phosphorylation is an initial key event in ubiquitination of cell surface-resident ABCA1, because ABCA1 has serine phosphorylation sites that negatively regulate ABCA1 expression.31 The attractive hypothesis proposed is that phosphorylation causes a conformational change in ABCA1: dissociation of ABCA1 from syntrophin, which tethers ABCA1 to the actin cytoskeleton through a protein-protein interaction and stabilizes it near the plasma membrane,32 whereby E3 is recruited to and associates with ABCA1, leading to ABCA1 ubiquitination. Alternatively, it is also conceivable that ABCA1-Ub on the cell surface is derived from ABCA1 ubiquitin-labeled in the endoplasmic reticulum that escaped proteasomal degradation. In that case, the lysosomal degradation through the ESCRT pathway, described in this present study, may function as the secondary quality control system of ABCA1. Recently, the COP9 signalosome complex has been reported to regulate the ubiquitination status of ABCA1,33 although it remains unknown whether this complex has the ability to ubiquitinate ABCA1. Further studies to explore in detail the function of the COP9 signalosome complex and to identify and characterize the E3, which enables it to ubiquitinate ABCA1, will expand the understanding of the molecular mechanism of ubiquitination-mediated ABCA1 degradation in the lysosome, and to assist in the development of new antiatherogenic drugs.

Calpain protease has been reported to regulate the degradation of ABCA1 expressed on the plasma membrane.10, 16 The ubiquitination-mediated degradation pathway through the ESCRT mechanism is likely to modulate ABCA1 expression independently from the calpain-mediated pathway, because even after calpain inhibitor treatment under optimal conditions, the depletion of HRS retarded the degradation of cell surface-resident ABCA1 in HuH-7 cells (Fig. 6B). This hypothesis is also supported by the observation that HRS directly interacts with ABCA1 and promotes the lysosomal degradation of cell surface-resident ABCA1 (Figs. 4, 5), whereas calpain protease has the ability to directly cleave ABCA1 in a manner dependent on a domain with a sequence rich in proline, glutamic acid, serine, and threonine (PEST domain).16 The contribution of each mechanism to the degradation of cell surface-resident ABCA1 remains to be elucidated. However, ubiquitination-mediated degradation through the ESCRT machinery may be the major mechanism, at least in liver, because hepatocytes produce and secrete apoA-I, which strongly interferes with calpain-mediated cleavage of ABCA1.16 The amount of ABCA1-Ub was still present in somewhat decreased amounts in crude membrane fractions of HuH-7 cells even when the cells were cultured in media containing apoA-I in conditions where calpain-mediated cleavage of ABCA1 was abolished (Supporting Fig. 4).16 Therefore, in liver, the cell surface-resident ABCA1 could be protected from calpain by an autocrine reaction with apoA-I and predominantly degraded by the ubiquitination-mediated pathway through the ESCRT machinery. Further studies will shed more light on this point and on the contribution of each mechanism to the degradation of cell surface-resident ABCA1 in pathophysiological conditions. In particular, from a clinical viewpoint of type 2 diabetes mellitus, in which the fatty acid concentration in the plasma is elevated, it is interesting to explore the role of the ubiquitination-mediated lysosomal pathway through the ESCRT mechanism in the fatty acid-mediated reduction in ABCA1 expression,31 which can lower HDL biogenesis, and subsequently cause concomitant illness with progressive atherosclerosis. Because calpeptin treatment has no effect on this fatty acid-mediated reduction in ABCA1, it is likely mediated by the ubiquitination-mediated lysosomal degradation of ABCA1 through the ESCRT pathway.

In conclusion, our present study suggests that ubiquitination mediates the lysosomal degradation of cell surface-resident ABCA1 through the ESCRT pathway, and thereby controls the expression level and cellular cholesterol efflux function of ABCA1. This information should be useful in the development of novel therapeutic agents designed to inhibit the progression of atherosclerotic cardiovascular disease.

Acknowledgements

We thank Dr. Larissa Kogleck for advice on the article.

Ancillary