Regulation of Peroxisome Proliferator-activated Receptor-γ-mediated Gene Expression

Cellular cholesterol content reflects a balance of lipid influx by lipoprotein receptors and endogenous synthesis and efflux to cholesterol acceptor particles. The beneficial effect of high density lipoprotein (HDL) in protecting against the development of cardiovascular disease is thought to be mediated predominately through its induction of cellular cholesterol efflux and “reverse cholesterol transport” from peripheral tissues to the liver. We tested the hypothesis that HDL could inhibit cellular lipid accumulation by modulating expression of peroxisome proliferator-activated receptor-γ (PPARγ)-responsive genes. To this end, we evaluated expression of two PPARγ-responsive genes, CD36, a receptor for oxidized low density lipoprotein, and aP2, a fatty acid-binding protein. HDL decreased expression of macrophage CD36 and aP2 in a dose-dependent manner. HDL also decreased aP2 expression in fibroblasts, reduced accumulation of lipid, and slowed differentiation of fibroblasts into adipocytes. HDL stimulated mitogen-activated protein (MAP) kinase activity, and inhibition of CD36 expression was blocked by co-incubation with a MAP kinase inhibitor. HDL increased expression of PPARγ mRNA and protein, induced translocation of PPARγ from the cytoplasm to the nucleus, and increased PPARγ phosphorylation. Our data demonstrate that despite induction and translocation of PPARγ in response to HDL, MAP kinase-mediated phosphorylation of PPARγ inhibited expression of PPARγ-responsive genes and suggest mechanisms by which HDL may inhibit cellular lipid accumulation.

The incidence of cardiovascular disease is inversely related to circulating HDL 1 levels. The beneficial effect of HDL in protecting against the development of cardiovascular disease is thought to be mediated predominately through its induction of cellular cholesterol efflux and "reverse cholesterol transport" from peripheral tissues to the liver. HDL also has other beneficial anti-atherosclerotic effects including preventing the oxidative modification of low density lipoprotein (LDL) (1), inhibiting cytokineinduced expression of adhesion molecules by endothelial cells (2), and activating endothelial nitric-oxide synthase (3).
PPARs are lipid-activated nuclear receptors that act as transcriptional regulators of genes encoding proteins involved in glucose and lipid metabolism (4,5). PPAR␥ binds to and is activated by such diverse agents as long chain fatty acids, arachidonic and linoleic acid metabolites (6), and the thiazolidinedione class of antidiabetic drugs (7). Growth factors, such as epidermal growth factor and platelet-derived growth factor, have been shown to phosphorylate PPAR␥ via the MAP kinase signaling pathway and to decrease PPAR␥ transcriptional activity (8). The NH 2 -terminal domain of PPAR␥ contains a consensus MAP kinase site in a region conserved between PPAR␥1 and PPAR␥2 isoforms (9). PPAR␥ proteins migrate on immunoblots as closely spaced doublets, a pattern suggestive of phosphorylation (10,11). A putative MAP kinase site is phosphorylated by ERK2 and c-Jun NH 2 -terminal kinase (9). Phosphorylation significantly inhibits both ligand-independent and ligand-dependent transcriptional activation by PPAR␥ (9). This repression is mediated by MAP kinase phosphorylation of Ser-82 on PPAR␥ (8). Mutation of the phosphorylated residue (Ser-82) prevents PPAR␥ phosphorylation as well as the growth factor-mediated repression of PPAR␥-dependent transcription. Previously we showed that transforming growth factor-␤ decreased the expression of the class B scavenger receptor CD36 by phosphorylation of MAP kinase, subsequent MAP kinase phosphorylation of PPAR␥, and CD36 transcription by phosphorylated PPAR␥ (12).
aP2 belongs to a multigene family of fatty acid and retinoid transport proteins. It is expressed in adipocytes and is postulated to serve as a lipid shuttle, solubilizing hydrophobic fatty acids and delivering them to the appropriate metabolic system for utilization. Macrophages also express aP2, but the specific role that aP2 plays in regulating lipid metabolism in macrophages is unclear (31,32).
We investigated the effect of HDL on PPAR␥ transcriptional activity and expression of two PPAR␥-responsive genes, CD36 and aP2. In addition, we evaluated the effect of HDL on PPAR␥-mediated fibroblast to adipocyte differentiation and lipid accumulation. We provide evidence for the first time that HDL inhibits cellular lipid accumulation by modulating the expression of CD36 and aP2.

MATERIALS AND METHODS
Cell Culture-Macrophages (RAW264.7) and fibroblasts (3T3) were cultured in RPMI 1640 and Dulbecco's modified Eagle's medium, respectively, containing 10% fetal calf serum, 50 g/ml each of penicillin and streptomycin, and 2 mM glutamine. Macrophages were switched to serumfree medium for 3-5 h when the confluence was about 85% and then received treatments in serum-free medium. Fibroblasts were treated in complete Dulbecco's modified Eagle's medium at the first day of confluence.
Isolation of HDL-HDL (1.063-1.210 g/ml) was isolated from normal human plasma by sequential ultracentrifugation after removal of very low density lipoprotein (Ͻ1.019 g/ml) and LDL (1.019 -1.063 g/ml). HDL was dialyzed against PBS containing 0.3 mM EDTA, sterilized by filtration through a 0.22-m filter, and stored under N 2 gas at 4°C. The protein content was determined by the methods of Lowry et al. (45).
Isolation of Total RNA, Purification of Poly(A ϩ ) RNA, and Northern Blotting-Cells were lysed in RNAzol TM B (Tel-Test, Inc., Friendswood, TX) and chloroform-extracted, and total RNA was precipitated in isopropanol. After washing with 80 and 100% ethanol, the dried pellet of total RNA was dissolved in distilled water and quantified. The poly(A ϩ ) RNA was isolated from about 80 g of total RNA by using the PolyATtract ® mRNA Isolation System III (Promega, Madison, WI).
Total RNA or poly(A ϩ ) RNA was loaded on a 1% formaldehyde agarose gel. After electrophoresis, RNA was transferred to a Zeta-probe ® GT Genomic Tested Blotting Membrane (Bio-Rad) in 10ϫ SSC by capillary force overnight. The blot was UV cross-linked for 2 min and then prehybridized with Hybrisol TM I (Oncor, Inc., Gaithersburg, MD) for 30 min before addition of 32 P randomly primed labeling probe for mouse PPAR␥, CD36, aP2, or glyceraldehyde-3-phosphate dehydrogenase. After overnight hybridization, the membrane was washed for 2 ϫ 20 min with 2ϫ SSC plus 0.2% SDS and then for 2 ϫ 20 min with 0.2ϫ SSC plus 0.2% SDS at 55°C. The blot was then autoradiographed by exposure to x-ray film (X-Omat TM AR, Kodak). Autoradiograms were assessed by densitometric scanning using a UMAX (Santa Clara, CA) UC630 flatbed scanner attached to a Macintosh IIci (Apple Computer, Cupertino, CA) running NIH Image software (Bethesda, MD). Template DNA for PPAR␥ and aP2 were generated by reverse transcription-polymerase chain reaction based on the published sequences. The 5Ј-and 3Ј-sequences of oligonucleotides used for PPAR␥ were TCGGCGTTGTCATGATCCTC (nucleotides 121-141) and GGTTCATAAAAGCACGCTGG (nucleotides 551-571), respectively. The 5Ј-and 3Ј-sequences of oligonucleotides used for aP2 were GATGCCTTTGTGGGAACC (nucleotides 307-325) and AACTCTT-GTGGAAGTCACG (nucleotides 665-684), respectively.
Analysis of PPAR␥/Phospho-PPAR␥ and Phospho-p44/42 MAP Kinase by Western Blotting-After treatment, the cells were washed twice with cold PBS and then scraped and lysed in ice-cold lysis buffer (50 mM Tris, pH 7.5, 150 mM NaCl, 1% Triton X-100, 1% sodium deoxycholic acid, 1 mM phenylmethylsulfonyl fluoride, 50 mM sodium fluoride, 1 mM sodium orthovanadate, 50 g/ml aprotinin, and 50 g/ml leupeptin). Lysates were sonicated for 20 cycles and then microcentrifuged for 15 min at 4°C. Supernatants were transferred to a new test tube. Protein was separated by SDS-PAGE and then transferred to nylon-enhanced nitrocellulose membranes. Membranes were blocked with a solution of 0.1% Tween 20/PBS (PBS-T) containing 5% fat-free milk for 1 h and then incubated with rabbit polyclonal anti-PPAR␥ (1:2000) or polyclonal anti-phospho-p44/42 MAP kinase antibodies for 2 h at room temperature followed by washing for 3 ϫ 10 min with PBS-T buffer. The blot was reblocked with PBS-T containing 5% milk followed by incubation with horseradish peroxidase-conjugated goat anti-rabbit IgG for another hour at room temperature. After washing 3 ϫ 10 min with PBS-T, the membrane was incubated for 1 min in a mixture of equal volumes of Western blot chemiluminescence reagents 1 and 2 (Amersham Biosciences). The membrane was then exposed to film before development.
Determination of CD36 Protein Expression on Cell Surface by FACS-After treatment, cells were removed with trypsin and washed three times with PBS. About 2 ϫ 10 6 cells were suspended in 300 l of PBS containing 5% mouse serum and incubated for 30 min at room temperature with shaking. After incubation with 10 l of polyclonal anti-CD36 antibody conjugated to fluorescein isothiocyanate isomer 1 (Chemicon International Inc., Temecula, CA) for 2 h at room tempera-ture, the cells were washed three times with PBS and subjected to flow cytometric evaluation (Coulter FACScan).
Immunohistochemistry of PPAR␥ in 3T3 Cells-3T3 cells were cultured in multichamber wells and processed for indirect immunofluorescence as follows. Cells were fixed in chilled acetone for 20 min. Affinitypurified rabbit anti-mouse PPAR␥ antibody (in 20 mM Tris, pH 7.5, 0.5 M NaCl, 2% bovine serum albumin, and 2% nonfat dry milk) was applied at a dilution of 1:50 at 4°C overnight followed by fluorescein isothiocyanate isomer 1-labeled donkey anti-rabbit IgG diluted 1:80 for 45 min. After washing three times with buffer, cells were shielded with Vectashield (Vector Laboratories Inc., Burlingame, CA), observed, and photographed using a Nikon Labphot2 fluorescent microscope.
Oil Red O Staining of Adipocytes-Confluent fibroblasts were stimulated to differentiate by addition of insulin, 3-isobutyl-1-methylxanthine, and dexamethasone for 2 days followed by insulin alone for 2 days and complete medium for two more days. HDL or MAP kinase inhibitor PD98059 was added at the 1st day of stimulation.
For the Oil Red O staining, cells were washed with PBS twice and then fixed with prechilled methanol for 3 min at Ϫ20°C followed by washing with cold PBS briefly. Cells were stained with Oil Red O solution (6 parts Oil Red O solution and 4 parts water, Oil Red O stock solution is 0.5% Oil Red O in isopropanol) for 2 h at room temperature. After removing the stain, cells were washed with PBS twice, ethanol once, and water twice and then photographed.

HDL Decreases CD36 and aP2
Expression-We first evaluated the effect of HDL on two PPAR␥-responsive genes, CD36 and aP2, in RAW264.7 cells, a murine macrophages cell line. HDL decreased CD36 (Fig. 1a) and aP2 (Fig. 1b) mRNA expression in a dose-dependent manner. Approximately 50 -75 g of HDL caused a 50% decrease of CD36 expression. The decrease in CD36 mRNA was associated with decreased CD36 surface protein as detected by FACS (Fig. 1c).
HDL Increases PPAR␥ mRNA Expression and PPAR␥ Protein Translocation-To determine the mechanism by which HDL decreased expression of CD36 and aP2 we next evaluated the effect of HDL on PPAR␥ expression. RAW macrophages and NIH-3T3 cells were treated with various concentrations of HDL for 10 h. HDL increased PPAR␥ mRNA in both cell types in a non-dose-dependent manner (Fig. 2). PPAR␥ is primarily located within the cytosol and moves to the nucleus following FIG. 1. HDL inhibits mRNA expression of PPAR␥-responsive genes in macrophages. RAW macrophages in serum-free RPMI 1640 medium were incubated with HDL at the indicated concentrations for 12 h. Expression of CD36 was evaluated by Northern blot using a 32 P-labeled cDNA probe for CD36 (a) and aP2 (b). The blots were rehybridized with 32 P-labeled glyceraldehyde-3-phosphate dehydrogenase (GAPDH) cDNA. The decrease in CD36 mRNA was associated with decreased CD36 surface protein as detected by FACS (c). Ctrl, control.
ligand binding and activation. We also evaluated expression and cellular localization of PPAR␥ by immunocytochemistry. Incubation of NIH-3T3 cells with HDL (50 g/ml) for 16 h increased the amount of PPAR␥ and caused the translocation of PPAR␥ from a cytoplasmic to a nuclear location (Fig. 3). The amount of PPAR␥ in RAW cells was low and difficult to detect by immunocytochemistry (data not shown).
HDL Increases PPAR␥ Phosphorylation-To address the mechanism(s) and apparent paradox of decreased expression of PPAR␥-responsive genes coupled with increased expression and nuclear translocation of PPAR␥ in response to HDL, we next evaluated the phosphorylation state of PPAR␥. PPAR␥ protein levels and PPAR␥ phosphorylation were evaluated by Western blot in both HDL-treated macrophages and fibroblasts. Consistent with the Northern blot and immunohistochemistry, HDL significantly increased PPAR␥ protein in both cell types (Fig. 4). In RAW cells, HDL increased expression of both the non-phosphorylated and phosphorylated forms of PPAR␥. At the highest concentrations of HDL evaluated (75 g/ml), only phosphorylated PPAR␥ was observed (Fig. 4A, top panel). In NIH-3T3 cells, an increase in only the phosphorylated form was seen in cells incubated with HDL (Fig. 4B, top panel).
Others (and we) have previously demonstrated that PPAR␥ is phosphorylated by MAP kinase (12,33). Addition of a specific MAP kinase inhibitor, UO126, blocked MAP kinase activity and prevented phosphorylation of PPAR␥ (Fig. 4). To determine the effect of HDL on MAP kinase activity, we evaluated phosphorylation of MAP kinase by Western blot. HDL stimulated phosphorylation of both p42 and p44 isoforms of MAP kinase in RAW cells and NIH-3T3 fibroblasts (Fig. 5).

Inhibition of CD36 Expression by HDL Is Abrogated by a MAP Kinase
Inhibitor-To determine the impact of MAP kinase inhibition and PPAR␥ phosphorylation on PPAR␥-responsive expression we evaluated expression of CD36 in RAW cells incubated with HDL and UO126. In the presence of the MAP kinase inhibitor, the negative regulatory effect of HDL on CD36 mRNA (Fig.  6A) expression and cell surface protein (Fig. 6B) was abolished.
HDL Inhibits Adipocyte Differentiation and aP2 Expression-Adipogenesis is regulated by a number of factors and hormones. Activation of PPAR␥ occurs early during the differ-entiation process, and ligands of PPAR␥ induce adipocyte differentiation. To determine the effect of HDL on adipocyte differentiation, we used a well characterized system of differentiating fibroblasts into adipocytes using insulin, dexamethasone, and isobutylmethylxanthine. We evaluated the effect of HDL on lipid accumulation with a neutral lipid stain, Oil Red O. Fibroblasts incubated with insulin, dexamethasone, and isobutylmethylxanthine accumulated abundant neutral lipid (Fig. 7, panel 1). Lipid accumulation was greatly diminished in cells incubated with HDL plus insulin, dexamethasone, and isobutylmethylxanthine (Fig. 7, panel 2). However, when fibroblasts were incubated with insulin, dexamethasone, isobutylmethylxanthine, and HDL in the presence of a MAP kinase inhibitor, lipid accumulated in amounts similar to the control (Fig. 7, panel 4). Incubation with a MAP kinase inhibitor alone did not impair lipid accumulation (Fig. 7, panel 3). Finally aP2 expression in NIH-3T3 cells was evaluated by Northern blotting. aP2 was significantly reduced in cells incubated with HDL (Fig. 8). DISCUSSION Our results demonstrate that HDL inhibits the expression of PPAR␥-responsive genes CD36 and aP2 and identifies a mechanism by which HDL may inhibit cellular lipid accumulation. Our data show that HDL increases the expression, translocation, and phosphorylation of PPAR␥ and that the effects of HDL on PPAR␥-mediated signaling are inhibited by blocking MAP kinase phosphorylation of PPAR␥. These data also suggest new beneficial effects of HDL and one that may potentially play a role in the ability of HDL to prevent lipid accumulation leading to the development of cardiovascular disease.
Although the mechanism(s) by which HDL can alter cell signaling events remains undefined, several pieces of evidence suggest that it is likely that HDL may activate MAP kinase signaling and PPAR␥ phosphorylation and inhibit CD36 expression by stimulating cholesterol efflux. We have previously demonstrated that when macrophages are incubated with either bovine serum albumin or cyclodextrin (cholesterol acceptor particles), expression of CD36 is decreased (27,34). However, the mechanisms by which this occurred were not  identified. To determine whether a cholesterol acceptor particle could modulate MAP kinase activity and PPAR␥ phosphorylation, we incubated J774 macrophages with cyclodextrin and evaluated p42/p44 MAP kinase and PPAR␥ by Western blot. Both p42/p44 MAP kinase and PPAR␥ were phosphorylated in response to incubation with cyclodextrin (data not shown), supporting the concept that the removal of cholesterol, not the addition of lipid or lipoprotein, is causing these signaling events. HDL has been shown previously to activate MAP kinase activity in human skin fibroblasts and in vascular smooth muscle cells (11,35). In fibroblasts, but not smooth muscle cells, activation of MAP kinase by HDL is mediated by protein kinase C (11,35). Lipid-free apolipoproteins A-I and A-II had no effect on MAP kinase activation (35).
Inhibition of CD36 expression has been demonstrated to reduce the development of atherosclerosis in atherosclerosisprone apoE-null mice (30). However, the role that HDL might play in the regulation of PPAR␥ signaling in vivo and its subsequent effects on lipid metabolism are likely to be complicated. For example, in two murine models of atherosclerosis, the LDL receptor-null and apoE-null, treatment with PPAR␥ agonists protected against the development of atherosclerosis (36 -38). In apoE-null mice, there was reduced lesion notwithstanding an increase in CD36 mRNA expression in aortic macrophages (36). The reason for this apparent contradiction is that there is a second target of PPAR␥ activation, liver X receptor ␣ (39). This transcription factor has been shown to regulate expression of the cellular cholesterol/phospholipid efflux protein ABC-A1 (37). Mutations in ABC-A1 are the cause of Tangier disease, a condition characterized by lack of HDL, massive storage of cholesteryl esters in macrophages, and premature atherosclerosis (40,41).
The role of aP2 in the development of atherosclerosis has also been recently addressed. In apoE-null mice crossed with Ap2null mice, development of atherosclerosis was markedly reduced. Similarly when bone marrow of aP2-null/apoE-null mice is transplanted to recipient aP2 wild-type/apoE-null mice, lack of expression of aP2 by macrophages protected against the development of atherosclerosis (42). aP2 is a member of the intracellular fatty acid-binding protein family and is expressed in adipose tissue. Its expression is highly regulated during differentiation of adipocytes and is transcriptionally controlled by fatty acids (43). Macrophages express aP2, and its expression suggests overlapping functional roles of adipocytes and macrophages. Genes that are critical in adipocytes, including those encoding transcription factors, cytokines, inflammatory molecules, fatty acid transporters, and scavenger receptors, are also expressed in macrophages and have an important role in their biology (44). The precise role of aP2 in modulating lipid metabolism of macrophages is unknown; it is likely that it functions in the mobilization and trafficking of intracellular fatty acids. Cholesterol-loaded THP-1 macrophages express aP2, and its expression is stimulated by oxidized low density lipoprotein (31). This expression is thought to be regulated by saturated fatty acids present in oxidized LDL either directly or indirectly through activation of PPAR␥ (32).
In summary, HDL inhibits the expression of two PPAR␥responsive genes, CD36 and aP2. This effect of HDL is novel and suggests a role of HDL in inhibiting the uptake of oxidized lipids in macrophage and in modulating the trafficking of fatty acids. These effects are likely to enhance the protective role of HDL in preventing lipid accumulation during atherosclerosis.