MCP-1 impacts RCT by repressing ABCA1, ABCG1, and SR-BI through PI3K/Akt posttranslational regulation in HepG2 cells.

Monocyte chemoattractant protein-1 (MCP-1) plays crucial roles at multiple stages of atherosclerosis. We hypothesized that MCP-1 might impair the reverse cholesterol transport (RCT) capacity of HepG2 cells by decreasing the cell-surface protein expression of ATP binding cassette A1 (ABCA1), ATP binding cassette G1 (ABCG1), and scavenger receptor class B type I (SR-BI). MCP-1 reduced the total protein and mRNA levels of ABCA1 and SR-BI, but not of ABCG1. MCP-1 decreased the cell-surface protein expression of ABCA1, ABCG1, and SR-BI in dose-dependent and time-dependent manners, as measured using cell-surface biotinylation. We further studied the phosphoinositide 3-kinase (PI3K)/serine/threonine protein kinase Akt pathway in regulating receptor trafficking. Both the translation and transcription of ABCA1, ABCG1, and SR-BI were not found to be regulated by the PI3K/Akt pathway. However, the cell-surface protein expression of ABCA1, ABCG1, and SR-BI could be regulated by PI3K activity, and PI3K activation corrected the MCP-1-induced decreases in the cell-surface protein expression of ABCA1, ABCG1, and SR-BI. Moreover, we found that MCP-1 decreased the lipid uptake by HepG2 cells and the ABCA1-mediated cholesterol efflux to apoA-I, which could be reversed by PI3K activation. Our data suggest that MCP-1 impairs RCT activity in HepG2 cells by a PI3K/Akt-mediated posttranslational regulation of ABCA1, ABCG1, and SR-BI cell-surface expression.

diluted 1:1,000), anti-Ser 473 -phosphorylated Akt (p-Akt) (Cell Signaling Technology; diluted 1:750), anti-Akt (Cell Signaling Technology; diluted 1:750), or mouse monoclonal anti-␤ -actin antibody (Sigma; diluted 1:2,000) overnight at 4°C. After three washes with TBS containing 0.1% Tween-20, the membranes were incubated with horseradish peroxidase-conjugated secondary antibody for 1 h at room temperature. To visualize the immunoreactive bands, enhanced chemiluminescence (ECL) Western blotting detection reagents and medical X-ray fi lms were used according to the manufacturers' suggestions. The band intensity was analyzed with Quantity One. The data were normalized to the ␤ -actin expression.

Real-time PCR
Total RNA was purifi ed from the cultured cells using TRIzol (Invitrogen) according to the manufacturer's protocol and treated with DNase I to remove any residual DNA contamination. First-strand cDNA synthesis was performed using a PrimeScript II 1st Strand cDNA Synthesis Kit (Takara Bio, Japan) according to the manufacturer's protocol. Real-time PCR was performed using a SYBR Green PCR Master Mix Kit (Applied Biosystems, Foster, CA). The gene expression of ABCG1, ABCA1, and SR-BI was normalized to that of ␤ -actin. The primers used for quantitative RT-PCR are presented in supplementary Table I.

Cell-surface protein assays using biotinylation
For the cell-surface ABCA1, ABCG1, and SR-BI analyses, cells were fi rst biotinylated with 0.5 mg/ml EZ-Link Sulfo-NHS-SS-Biotin (Pierce Chemical Co., Rockford, IL) at 4°C for 30 min. The cells were then lysed with RIPA buffer at 4°C. After centrifugation, the supernatants were incubated with anti-FLAG agarose beads overnight at 4°C. Following centrifugation and washing, the supernatants and pellets (the collected agarose beads) represented intracellular and surface proteins, respectively. The proteins were dissociated from the pellets by boiling with SDS loading buffer and were analyzed with SDS-PAGE and immunoblotting using mouse monoclonal anti-ABCA1 antibody (Novus Biologicals; diluted 1:1,000), rabbit polyclonal anti-ABCG1 antibody (Novus Biologicals; diluted 1:1,000), goat polyclonal anti-SR-BI antibody (Novus Biologicals; diluted 1:1,000), or anti-Na + /K + ATPase antibody (Pierce Biotechnology; diluted 1:1,000).

1,1 Ј -dioctadecyl-3,3,3 Ј ,3 Ј -tetramethylindocarbocyanine perchlorate (Dil) -HDL lipid uptake assay
Cells were grown on glass coverslips until 60-70% confl uent and were then incubated in DMEM containing 0.5% BSA with or without MCP-1 (40 ng/ml) for 48 h following pretreatment with a PI3K activator (insulin) or a PI3K inhibitor (wortmannin) for 45 min. Dil-HDL (Biomedical Technologies Inc., Stoughton, MA) was added to a serum-free medium containing 0.5% BSA to obtain a fi nal concentration of 5 g/ml. After a 4 h incubation with Dil-HDL, the cells states. The acute-phase response, which can be induced by infection or infl ammation, impairs the capacity of the human HUH-7 hepatoma cell line to deliver cholesteryl ester and diminishes the cholesterol effl ux capacity of macro phages ( 21,22 ). Irina et al. ( 23 ) demonstrated that lipopolysaccharide treatment resulted in the downregulation of SR-BI and ABCA1 expression, a significant inhibition of HDL-mediated cholesterol effl ux, compared with untreated RAW 264.7 cells . Many pro-infl ammatory factors, including IFN-␥ , lipopolysaccharide, tumor necrosis factor, interleukin-1, and interleukin-6, have been found to modulate the expression of HDL receptors and, in turn, alter RCT activity (24)(25)(26).
However, the impact of MCP-1 on HDL-mediated RCT activity and whether this effect contributes to the expression or redistribution of the relevant transporters have not been investigated. In this report, we used the HepG2 cell line to provide evidence that the pro-atherogenic effects of MCP-1 may refl ect, at least in part, altered cholesterol metabolism through RCT. We then focused on the involvement of the phosphoinositide 3-kinase (PI3K)/serine/ threonine protein kinase Akt pathway, which posttranslationally regulates the recruitment of receptors to the plasma membrane.

Cell culture and treatment
The HepG2 cells were a generous gift from the medical school of Sun Yat-sen University. The cells were grown in Dulbecco's modifi ed Eagle medium (DMEM) and supplemented with 10% fetal bovine serum (FBS), 100 U/ml penicillin, and 100 g/ml streptomycin, in a humidifi ed atmosphere containing 5% CO 2 and 95% air at 37°C. The cells were grown until 60-70% confl uent and were then incubated in a serum-free medium containing 0.5% BSA for 6 h. To investigate the dose effect of MCP-1, cells were incubated in a serum-free medium containing 0.5% BSA with MCP-1 (R and D Systems, Inc., Abingdon, UK) at different concentrations (0-80 ng/ml) for 48 h. Cells treated with MCP-1 at 40 ng/ml for increasing times (0-72 h) were used to study the time effect of MCP-1. To further examine the involvement of PI3K, we fi rst pretreated the cells with either the PI3K activator insulin (100 nM; Sigma, St. Louis, MO) or the PI3K inhibitor wortmannin (100 nM; Cell Signaling Technology, Beverly, MA) for 45 min. The cells were then incubated with or without MCP-1 (40 ng/ml) for 48 h.

Western blotting
Following the incubation, the cells were harvested, washed with phosphate-buffered saline (PBS) (pH 7.4), and lysed in RIPA buffer (Roche Molecular Biochemicals) for 30 min at 4°C. The proteins were fractionated on 4-10% gradient SDS/polyacrylamide gels and electrophoretically transferred to polyvinylidene difl uoride membranes (Invitrogen, Carlsbad, CA). The membranes were incubated with a TBS blocking solution (200 mM Tris-HCl, 150 mM NaCl, 5% nonfat dry milk) for 1 h at room temperature . The membranes were immunoblotted with the appropriate antibody: mouse monoclonal anti-ABCA1 antibody (Novus Biologicals, Oakville, CA; diluted 1:1,000), rabbit polyclonal anti-ABCG1 antibody (Novus Biologicals; diluted 1:1,000), goat polyclo nal anti-SR-BI antibody (Novus Biologicals; approximately 80% in the dose-effect group. Fig. 2B shows the time-effect response to MCP-1 of the gene expression of the three acceptors. Treatment with MCP-1 (40ng/ml) for 48 h resulted in signifi cant decreases in the mRNA levels of ABCA1 and SR-BI, to 17 and 48%, respectively, of the levels of the untreated cells . The addition of MCP-1 did not alter the ABCG1 mRNA levels. Most importantly, the treatment of the HepG2 cells with MCP-1 repressed the ABCA1 and SR-BI mRNA levels with no changes in the ABCG1 gene expression, which was similar to the effects of MCP-1 on the total protein levels of the three genes. The impaction of MCP-1 on mouse primary hepatocytes ABCA1, ABCG1, and SR-B1 mRNA (supplementary Fig. II) were similar to that in HepG2 cells.
The subcellular localization and cell-surface protein expression of ABCA1, SR-BI, and ABCG1 in HepG2 were regulated by MCP-1 Many studies have indicated that the subcellular localization of ABCA1, ABCG1, and SR-BI can be posttranslationally modulated by certain substances to allow them to move to the plasma membrane and, in turn, affect RCT (18)(19)(20)27 ). We have shown that treatment with MCP-1 resulted in reductions in the ABCA1 and SR-BI total proteins while leaving the ABCG1 level unchanged. However, it was unknown whether MCP-1 could cause corresponding reductions in the cell-surface levels of the ABCA1 and SR-BI receptors, or could even affect the ABCG1 surface expression. The numbers of cell-surface receptors were directly measured using cell-surface biotinylation. As shown in Fig. 3A , the cell-surface expression levels of ABCA1 and SR-BI were reduced in a dose-dependent and time-dependent manner by up to 87 and 75%, respectively, after treatment with 40 ng/ml MCP-1 for 72 h. This result was not completely parallel to the changes in total protein expression. The cell-surface ABCG1 level of the HepG2 cells incubated with MCP-1 at 80 ng/ml for 48 h also decreased to 14% of the level of the untreated cells.
To confi rm the distribution of the receptors in the HepG2 cells treated with MCP-1, the cells were examined under confocal microscopy. After reaching 60-70% confl uence, the cells were equilibrated for 6 h and incubated in a serum-free medium in the presence or absence of MCP-1 (40 ng/ml) for 48 h. As shown in Fig. 3B , ABCA1, ABCG1, and SR-BI were found to be distributed throughout the cytoplasm and cell surface in the untreated cells. We found that the cell-surface proteins of ABCA1 and SR-BI were inhibited by MCP-1, which was in line with the changes in the numbers of cell-surface receptors. Strikingly, a 48 h incubation with 40 ng/ml MCP-1 resulted in both a marked redistribution of ABCG1 to the cell nucleus and a decreased distribution at the cell surface. This fi nding is in accordance with the change in the numbers of the cell-surface ABCG1 receptor.

-phosphorylated Akt was regulated after treatment with MCP-1 in HepG2 cells
Because PI3K plays a key role in cell transporter traffi cking, we explored whether MCP-1 could alter PI3K/Akt were washed with PBS, fi xed, and subjected to laser confocal microscopy. For each experiment, all the pictures were identically exposed and processed.

Cholesterol effl ux from HepG2 cells
HepG2 cells were seeded on collagen-coated 24-well plates at a density of 1 × 10 5 cells per well in DMEM supplemented with 10% FBS. After a 6 h serum starvation, the cells were washed with PBS and labeled by incubation in DMEM supplemented with 0.5% BSA containing [ 3 H]cholesterol (0.5 uCi/ml; Perkin Elmer, CA) for 48 h. The cells were pretreated with a PI3K activator (insulin, 100 nM) or a PI3K inhibitor (wortmannin, 100 nM) for 45 min. Cellular cholesterol effl ux was initiated by the addition of DMEM containing 0.2% BSA with 20 g/ml human apoA-I in the presence or absence of MCP-1 (40 ng/ml). After a 48 h incubation, the radioactivity of the medium and cells was measured with a liquid scintillation counter. The cholesterol effl ux was expressed as the percentage of counts in the medium relative to the total counts in the medium and cells together.

Statistical analysis
All the data were expressed as means ± SEM. The statistical signifi cance of differences was determined using Student's t -test or a one-way ANOVA followed by Bonferroni's post hoc test, as appropriate. Statistical signifi cance was defi ned as a two-tailed probability of less than 0.05.

MCP-1 decreased the total protein expression of ABCA1 and SR-BI but did not affect the total protein expression of ABCG1 in HepG2 cells
To investigate whether MCP-1 could alter the total protein expression levels of ABCA1, ABCG1, or SR-BI, we treated HepG2 cells for 48 h with increasing doses of MCP-1 (0-80 ng/ml) in the dose-effect study. We also treated HepG2 cells with MCP-1 at a fi xed concentration of 40 ng/ ml for increasing durations (0-72 h) in the time-course study. As shown in Fig. 1A , 40 ng/ml MCP-1 decreased the total protein expression of ABCA1 and SR-BI by up to 44 and 32%, respectively, compared with the untreated group. Fig. 1B shows the time-course responses of the ABCA1, SR-BI, and ABCG1 protein expression to 40 ng/ml MCP-1. The MCP-1 treatment resulted in a signifi cant inhibition (79%) of the ABCA1 expression at 72 h compared with the untreated cells. In addition, the SR-BI expression was reduced to 44% of the control level at 48 h. However, the MCP-1 intervention did not alter the total protein expression of ABCG1. Furthermore, mouse primary hepatocytes were also incubated with MCP-1 (40 ng/ml) for 48 h to test the effects of MCP-1 on mouse primary hepatocytes, and similar results are shown in supplementary Fig. II.

MCP-1 decreased the mRNA expression of ABCA1 and SR-BI but did not alter the mRNA expression of ABCG1
We next studied whether MCP-1 could induce the gene expression of the three receptors. HepG2 cells were treated in the same manner as for the total protein detection. As shown in Fig. 2A , MCP-1 repressed the gene expression of both ABCA1 (at 40 ng/ml) and SR-BI (at 80 ng/ml) by

MCP-1 reduced the total protein and gene expression of ABCA1 and SR-BI without PI3K involvement
To further examine the regulation of the ABCA1, ABCG1, and SR-BI receptors by PI3K, we assessed the effects of PI3K activity on the total protein expression and mRNA levels of ABCA1, ABCG1, and SR-BI. Cells were treated with a PI3K activator (insulin) or inhibiter (wortmannin) for 45 min. The cells were then incubated with or without MCP-1 (40 ng/ml) for 48 h. As shown in Fig. 5 , in the HepG2 cells without MCP-1 treatment, neither the total protein expression ( Fig. 5A ) nor the gene expression ( Fig. 5B, C, D ) of ABCA1, ABCG1, and SR-BI was regulated by the PI3K inhibition or activation. In addition, PI3K activity could not alter the total protein expression or the mRNA levels of ABCA1, ABCG1, or SR-BI after incubation with MCP-1. These results suggest that PI3K activity activity to result in acceptor redistributions. p-Akt is generated from the phosphorylation of Akt by PI3K; therefore, we tested p-Akt expression with an antibody specifi c for Ser 473 -phosphorylated Akt. HepG2 cells were treated with a PI3K activator (insulin at 100 nM) and/or a PI3K inhibitor (wortmannin at 100 nM) in the presence or absence of MCP-1 for 45 min. As shown in Fig. 4 , the expression of p-Akt was induced by the PI3K activation mediated by insulin. In contrast, p-Akt was not detected when the cells were treated with the PI3K inhibitor, either in the absence or presence of insulin. In the cells that were coincubated in a medium containing insulin (100 nM) and MCP-1 (40 ng/ml) for 45 min, p-Akt was repressed; specifically, MCP-1 repressed the insulin-induced p-Akt. This phenomenon suggests that MCP-1 downregulates PI3K/ Akt activity. Total proteins were extracted from the cultured cells, and the protein levels were analyzed using Western blot analysis, as described in the Materials and Methods. The relative expression of ABCA1, ABCG1, and SR-B1 is expressed as the ratios of ABCA1, ABCG1, and SR-B1 to the corresponding ␤ -actin expression. The error bars indicate the standard deviations. Each experiment was performed three times. *** P < 0.001, ** P < 0.01, * P < 0.05 compared with the untreated group. The upper panel shows a Western blot and is representative of one experiment. To assess whether PI3K activity affects the subcellular localization of the three acceptors, the surface receptor levels were measured using cell-surface biotinylation. The biotinylation was performed after the treatment with MCP-1 and the following PI3K activation (by insulin) or inhibition (by wortmannin). The results in Fig. 6 show that PI3K activation signifi cantly increased the cell-surface expression of ABCA1, ABCG1, and SR-BI. In contrast, PI3K inhibition, in both the absence and the presence of insulin, markedly decreased the numbers of receptors at the cell surface. PI3K activation with insulin could restore the numbers of ABCA1, ABCG1, and SR-BI receptors at the cell surface following treatment with MCP-1.

MCP-1 decreased Dil-HDL lipid uptake, which could be reversed by PI3K activation
Hepatocytes play a pivotal role in RCT, especially in SR-BI-mediated lipid uptake. Because inhibitions in the expression of total and cell-surface SR-BI proteins were observed in HepG2 cells after MCP-1 treatment, we assessed whether incubation with MCP-1 could alter Dil-HDL lipid uptake by HepG2 cells. We examined lipid uptake by incubating cells with Dil-HDL for 4 h following treatment with 40 ng/ml MCP-1 for 48 h. The cells were then subjected to confocal microscopy after fi xation. Following incubation with MCP-1, cells had a greatly impaired lipid uptake capacity (red in Fig. 7B ) compared with the cells with no MCP-1 treatment, as shown in Fig. 7A . To assess the ability of PI3K activity to infl uence Dil-HDL lipid uptake, cells pretreated with insulin or wortmannin for 45 min were incubated in the presence or absence of MCP-1 (40 ng/ml) for 48 h. As indicated in Fig. 7E , PI3K activation increased the Dil-HDL lipid uptake compared with the untreated controls. In contrast, PI3K inhibition, in both the absence ( Fig. 7C ) and the presence ( Fig. 7G ) of insulin, markedly decreased the lipid uptake. Pretreatment with does not alter the inhibitory action of MCP-1 in terms of both the gene expression and the total protein expression of the ABCA1 and SR-BI receptors in HepG2 cells. Furthermore, PI3K activity did not affect either the gene expression or the total protein expression of ABCG1 in the absence or presence of MCP-1.

PI3K activation corrected the MCP-1-induced reduction of the numbers of ABCA1, ABCG1, and SR-BI cellsurface receptors
Our previous results showed that MCP-1 repressed insulin-induced p-Akt expression. This fi nding led us to speculate that the MCP-1-induced changes in the cell-surface numbers or redistribution of the ABCA1, ABCG1, and SR-BI transporters are likely to be regulated by PI3K activity because the MCP-1-induced changes in the ABCA1 and SR-BI total protein and gene expression are not regulated by PI3K. Fig. 3. The effects of MCP-1 on the cell-surface expression and subcellular localization of ABCA1, ABCG1, and SR-BI in HepG2 cells. A: HepG2 cells pretreated with MCP-1 at different concentrations for the indicated times were cell-surface biotinylated as described in the Materials and Methods, and the cell-surface receptor was analyzed with SDS-PAGE. The ABCA1, ABCG1, and SR-BI cellsurface receptors were reduced in a dose-dependent and time-dependent manner. *** P < 0.001, ** P < 0.01, * P < 0.05 compared with the untreated cells. B: HepG2 cells grown on glass coverslips were serum-starved for 6 h after reaching 60-70% confl uence, followed by incubation for 48 h in a serum-free medium in the absence (a) or presence (b) of MCP-1 (40 ng/ml). The subcellular localization of ABCA1, ABCG1, and SR-BI was analyzed using confocal microscopy, as described in the Materials and Methods. The images were captured with confocal microscopy after ABCA1, ABCG1, and SR-BI were labeled with Alexa 546 (red), Alexa 488 (green), and Alexa 633 (pink), respectively. The images indicate decreased distributions of ABCA1, ABCG1, and SR-BI at the cell surface after treatment with MCP-1. In particular, ABCG1 was markedly traffi cked to the cell nucleus. Fig. 4. MCP-1 repressed the activity of PI3K/Akt. To investigate the effect of MCP-1 on PI3K activity, we treated HepG2 cells with (+) or without ( Ϫ ) a PI3K activator (insulin, 100 nM) or inhibitor (wortmannin, 100 nM) and coincubated them with (+) or without ( Ϫ ) MCP-1 (40 ng/ml) for 45 min. Total proteins were extracted from the cultured cells, and the protein levels were analyzed using Western blot analysis, as described in the Materials and Methods. The relative expression of p-Akt was expressed as the ratio of p-Akt expression to the corresponding Akt expression. Following the PI3K inhibition by wortmannin, p-Akt was not detected, whereas the PI3K activation by insulin induced p-Akt, which could be repressed by MCP-1 coincubation. an important step for HDL formation. Based on the fi nding that MCP-1 reduced the expression and cell localization of the HDL transporters, we then investigated the effects of MCP-1 on the cholesterol effl ux to apoA-I. After treating HepG2 cells with MCP-1 (40 ng/ml), we detected the cholesterol effl ux to apoA-I, as indicated in the Materials and Methods. As shown in Fig. 8A , the cholesterol effl ux to apoA-I in cells treated with MCP-1 was reduced by 54% compared with the untreated cells. However, whether insulin ( Fig. 7F ) improved the lipid uptake compared with the MCP-1 group ( Fig. 7B ).

MCP-1 decreased cholesterol effl ux to apoA-I from HepG2 cells, which could be reversed by PI3K activation, and also reduced cholesterol effl ux to apoA-I from mouse primary hepatocytes in an ABCA1-dependent manner
Cholesterol effl ux from hepatocytes via HDL transporters, namely, ABCA1, ABCG1, and SR-BI, or diffusional effl ux is Fig. 5. MCP-1 regulates ABCA1 and SR-BI total protein and gene expression without PI3K involvement. Cells were serum-starved for 6 h after reaching 60-70% confl uence, followed by the pretreatment of the cells with (+) or without ( Ϫ ) PI3K activation (by insulin) or inhibition (by wortmannin) for 45 min. The cells were then incubated with (+) or without ( Ϫ ) MCP-1 (40 ng/ml) for 48 h. Western blot analysis and real-time PCR were used to analyze the total protein (A) and mRNA levels (B, C, D), respectively, of ABCA1, ABCG1, and SR-BI, as described in the Materials and Methods. After incubation with or without MCP-1, both the total protein expression and the gene expression of ABCA1 and SR-BI were found to not be regulated by pretreatment with a PI3K inhibitor (wortmannin) or activator (insulin). The PI3K activity did not affect either the gene expression or the total protein expression of ABCG1 in the absence or presence of MCP-1. Fig. 6. PI3K activation corrected the MCP-1-induced decreases in the numbers of the ABCA1, ABCG1, and SR-BI cell-surface receptors. HepG2 cells were equilibrated for 6 h in DMEM containing 0.5% BSA after reaching 60-70% confl uence. The cells were then treated with (+) or without ( Ϫ ) MCP-1 after incubating with (+) or without ( Ϫ ) PI3K activation (by insulin) or inhibition (by wortmannin). The ABCA1, ABCG1, and SR-BI cell-surface receptor levels and Na + /K + ATPase were directly extracted using cell-surface biotinylation and then measured by Western blotting, as described in the Materials and Methods. * P < 0.05 compared with the untreated group; ‡ P < 0.05 compared with MCP-1 group.
of ABCA1 by si-ABCA1 resulted in a 52% reduction in the cholesterol effl ux to apoA-I, whereas the knockdown of SR-BI resulted in a slight but signifi cant increase in the cholesterol effl ux . In addition, the knockdown of ABCG1 had little effect on the cholesterol effl ux. The treatment with MCP-1 in hepatocytes transfected with si-ABCG1 or si-SR-BI still caused signifi cant reductions in the cholesterol effl ux compared with the cells that were not treated with MCP-1, whereas MCP-1 had no impact on the cholesterol effl ux in mouse primary hepatocytes that were transfected with si-ABCA1 compared with the nonMCP-1 treatment. These results suggest that MCP-1 reduces the cholesterol effl ux by decreasing the expression of the transporter ABCA1.
Because hepatocyte-secreted extracellular apoA-I is another determinant of cholesterol effl ux, we detected the impact of MCP-1 on the apoA-I mRNA levels and secretion. As shown in supplementary Fig. I, MCP-1 did not affect either the level of apoA-I mRNA or the extracellular secretion of apoA-I, suggesting that the alteration in the cholesterol effl ux to apoA-I is not affected by apoA-I secretion but is primarily affected by the amount of ABCA1 localized to the cell surface.

DISCUSSION
MCP-1 causes cholesterol accumulation in hepatocytes ( 28 ), and HDL receptors in hepatocytes contribute to cholesterol metabolism through the RCT process, including the generation of HDL and cholesterol ester uptake from HDL-cholesterol. Thus, we evaluated the effects of MCP-1 on the activities of HDL receptors in hepatocytes and the this effect was the result of a reduction in the expression of the total protein or in the cell-surface expression of the HDL transporters was unknown. To further investigate this issue, we pretreated HepG2 cells with either the PI3K inhibitor wortmannin (100 nM) or the PI3K activator insulin (100 nM) for 45 min. As shown in Fig. 8A , the cells treated with wortmannin displayed a reduced cholesterol effl ux to apoA-I; in contrast, the cells treated with insulin exhibited an enhanced cholesterol effl ux, and pretreatment with insulin improved the reduction in the cholesterol effl ux caused by MCP-1, which was in accordance with the observed changes in the cell-surface expression of the HDL transporters. Therefore, we suggest that the reduced cholesterol effl ux to apoA-I by MCP-1 resulted from a parallel decrease in the cell localization of the HDL transporters.
Because hepatoma cells may differentially respond to chemokines compared with normal cells, we tested the effects of MCP-1 on mouse primary hepatocytes. Incubating mouse primary hepatocytes with MCP-1 (40 ng/ml) for 48 h also resulted in reductions in the total expression and cellsurface expression of ABCA1, ABCG1, and SR-BI (supplementary Fig. II) and in a 46% decrease in the cholesterol effl ux to apoA-I compared with untreated mouse primary hepatocytes ( Fig. 8B ), which is similar to what was observed in HepG2 cells.
To further confi rm which HDL transporters mediate the reduced cholesterol effl ux to apoA-I by MCP-1, we knocked down ABCA1, ABCG1, and SR-BI by transfecting mouse primary hepatocytes with small interference (si ) RNA (supplementary Fig. II). As shown in Fig. 8B , compared with the si-negative control (NC) group, the knockdown Fig. 7. The effects of MCP-1 and PI3K activity on Dil-HDL lipid uptake. HepG2 cells were equilibrated in DMEM containing 0.5% BSA for 6 h after reaching 60-70% confl uence and were then treated with MCP-1 following PI3K activation (by insulin) or inhibition (by wortmannin). The cells were incubated for 4 h in a serum-free medium containing Dil-labeled HDL. After being washed and fi xed, the cells were subjected to confocal microscopy. The lipid components of the HDL taken up by the cells are shown in red. After incubation with MCP-1, the cells (B) had a greatly impaired capacity for lipid uptake from the medium compared with the untreated cells (A). The PI3K activation (E) increased the cholesterol uptake compared with that of the untreated controls. In contrast, the PI3K inhibitor, both in the absence (C, D) and presence (G, H) of insulin, markedly decreased the lipid uptake. Pretreatment with insulin (F) could improve the lipid uptake compared with the MCP-1 group . effl ux, and the knockdown of SR-BI resulted in a slight increase in the effl ux . These data are in accordance with a recent study that used liver-specifi c ABCA1-knockout mice to prove the role of ABCA1 in cholesterol effl ux in hepatocytes ( 31,32 ). In addition, extracellular apoA-I is another determinant for ABCA1-mediated cholesterol effux ( 33 ). Because the apoA-I mRNA or secretion into the medium is not altered by MCP-1, we speculated that the reduced cholesterol effl ux to apoA-I by MCP-1 resulted from suppression of ABCA1 expression in hepatocytes. The ABCA1 transporter resides on the cell surface and in intracellular compartments, and ABCA1 functions in lipid effl ux and HDL biogenesis at the cell surface rather than in the intracellular compartments ( 16,19 ). Our results showed a parallel suppression of both cholesterol effl uxes to apoA-I and ABCA1 cell-surface localization by MCP-1 . In addition, both of these effects can be corrected by PI3K activation, whereas the ABCA1 total protein was not regulated by PI3K activity, indicating that the cholesterol effl ux was directly associated with ABCA1 cell-surface localization. Therefore, we suggest that the decreased cholesterol effl ux by MCP-1 may be due to defi cient ABCA1 traffi cking to the cell surface at the posttranslation level, which results in cholesterol deposits in hepatocytes and the impaired lipidation of apoA-I to form HDL .
Unexpectedly, the regulation of ABC transporters ABCA1 and ABCG1 by MCP-1 differed in HepG2 cells. The mechanism responsible for the downregulation of ABCA1 by MCP-1 involves both transcription with a corresponding decrease in ABCA1 mRNA levels and posttranslation via the PI3K/Akt pathway. In contrast, ABCG1 is only posttranslationally regulated with a reduction in its cell-surface localization. The transcriptional regulation of ABCA1 and ABCG1 appears to be different in the cells. Other evidence also supports this phenomenon. Lipopolysaccharide treatment reduces ABCG1 mRNA expression but not ABCA1 mRNA expression in mouse hepatocytes. In contrast, tumor necrosis factor ␣ markedly decreases ABCA1 gene expression by attenuating the ABCA1 promoter activity transcriptionally via the nuclear factor kappa B pathway, but not the liver X receptor (LXR) pathway, and posttranslationally enhances the rate of ABCA1 degradation without attenuating the expression of LXR target genes, such as ABCG1 ( 24,34 ). In Npc1-null hepatocytes, the upregulation of ABCG1 expression is mainly transcriptional without changes on LXR ␣ mRNA, whereas ABCA1 expression is largely dependent on posttranscriptional mechanisms, including an increased translation rate and decreased degradation of ABCA1 by cathepsin D ( 35 ). Thus, the transcriptional factors regulating the expression of ABC transporters are not identical, except for the known LXR pathway. The differential regulation of ABC transporters at the transcription level by MCP-1 suggest that some transcriptional pathways other than LXR may be involved in the downregulation of ABCA1 by MCP-1.
The PI3K/Akt pathway has been extensively studied. PI3K phosphorylates phosphatidylinositol on the 3-OH position of the inositol ring, thereby generating PI3P, namely PtdIns(3,4)P2 and PtdIns (3,4,5)P3, which in turn RCT capacity, and we selected the HepG2 cell line for these experiments in the present study. These data suggest that MCP-1 suppresses hepatic ABCA1 and SR-BI expression both transcriptionally and posttranslationally but only decreases ABCG1 cell-surface expression posttranslationally. We also further demonstrated the similar effects of MCP-1 on mouse primary hepatocytes (supplementary Fig. II). The PI3K/Akt pathway participates in the MCP-1mediated posttranslational suppression of the cell-surface localization of ABCA1, ABCG1, and SR-BI, and the PI3K activator restores the impaired RCT activity caused by MCP-1.
The effl ux of cholesterol to apoA-I is the fi rst stage of biogenesis of HDL in hepatocytes ( 29 ). Several different potential cellular cholesterol effl ux pathways have been described: diffusional effl ux, ABCA1-, ABCG1-, and SR-BImediated cholesterol effl ux pathways ( 30 ). The experiments in which mouse primary hepatocytes were transfected with siRNA indicated that the knockdown of ABCA1 induced a signifi cant reduction in cholesterol effl ux to apoA-I, whereas the knockdown of ABCG1 did not affect the cholesterol phosphorylate the serine/threonine protein kinase Akt ( 36,37 ). A key role of PI3K is its involvement in vesicular traffi cking ( 38 ), including the recruitment of regulatory proteins such as the insulin-responsive glucose transporter-4 ( 39 ) to the plasma membrane. Our study illustrated that PI3K activity regulated the cell-surface protein expression, but not the total protein expression level, of ABCA1, ABCG1, and SR-BI in HepG2 cells. Meanwhile, p-Akt was repressed by MCP-1 in HepG2 cells that were pretreated with insulin, and PI3K activation could correct the MCP-1induced decrease in the amounts of ABCA1, ABCG1, and SR-BI at the cell surface, which indicated that MCP-1 behaved like a PI3K inhibitor (such as wortmannin). However, another study ( 40 ) indicated that MCP-1 stimulates two separate PI3K isoforms, p85/p110 PI3-kinase and PI3K-C2a, in THP-1 cells. The variations in PI3K activity in response to MCP-1 may result from the existence of multiple isoforms of PI3K and the different cell types investigated ( 41 ). Moreover, there are different beliefs about whether the cellular effects of MCP-1 are mediated independently of the C-C chemokine receptor type 2 ( 42 ).
MCP-1 may potentially reduce reverse cholesterol transport in two ways and consequently increase the risk for atherosclerosis. On one hand, the cell-specific deletion of ABCA1 demonstrates that hepatocytes generate 70-80% of the plasma nascent HDL pool ( 43 ), which contributes signifi cantly to the antiatherogenic process of RCT by regulating extrahepatic cellular cholesterol effl ux. Because the cholesterol effl ux to apoA-I is essential for generating nascent HDL ( 44,45 ), which is the fi rst step of RCT, the reduction of RCT by MCP-1 may be due to the impaired cholesterol effl ux to apoA-I from hepatocytes, which is dependent on cell-surface expression of ABCA1. On the other hand, the reduction in RCT by MCP-1 may result from the impaired terminal step of RCT, namely, the reduced HDL-cholesterol lipid uptake by HepG2 cells for biliary secretion. Previous studies have demonstrated that the bulk of SR-BI-mediated lipid uptake occurs at the plasma membrane. The present study indicates that PI3K posttranslationally regulates the MCP-1-induced reduction of cell-surface expression of SR-BI and Dil-HDL lipid uptake but does not modify the total protein expression of SR-BI, which indicates the alternation in lipid uptake is mostly caused by the changes in the cell-surface expression of SR-BI . We suggest that MCP-1 impairs RCT activity in hepatocytes through the posttranslational regulation of ABCA1 and SR-BI cell-surface expression by PI3K/Akt, which may be an important mechanism that underlies the pro-atherogenic effects that are associated with infl ammation.
The HepG2 cells were a generous gift from the medical school of Sun Yat-set University.