The role of CD4+CD25+ regulatory T cells in macrophage-derived foam-cell formation.

Cluster of differentiation (CD)4+CD25+ regulatory T cells (Tregs) exert a suppressive activity on atherosclerosis, but the underlying mechanism remains unclear. Here, we investigated whether and how Tregs affect macrophages foam-cell formation. Tregs were isolated by magnetic cell sorting-column and analyzed by flow cytometry. Macrophages were cultured with or without Tregs in the presence of oxidized LDL (oxLDL) for 48 h to transform foam cells. After co-culture with Tregs, macrophages showed a decrease in lipid accumulation, which was accompanied by a significantly downregulated expression of CD36 and SRA but no obvious difference in ABCA1 expression. Tregs can inhibit the proinflammatory properties of macrophages and steer macrophage differentiation toward an anti-inflammatory cytokine producing phenotype. Mechanistic studies reveal that both cell-to-cell contact and soluble factors are required for Treg-mediated suppression on macrophage foam-cell formation. Cytokines, interleukin-10 (IL-10), and transforming growth factor-beta (TGF-beta) are the key factors for these suppressive functions.

In recent years, a novel subtype of T cell, called regulatory T cells (Tregs), have been shown to play a critical role in the development of atherosclerosis. Natural Tregs, characterized by the expression of cluster of differentiation (CD)4, CD25, and the forkhead-winged-helix (Foxp3) transcription factor, have the capacity to contribute to the maintenance of immunological tolerance against self and non-self antigens and prevent the development of various immunoinfl ammatory diseases (4)(5)(6)(7). Several preliminary studies have described an important role for this type of regulatory T cell response in atherosclerosis (8)(9)(10)(11). In a recent study by Ait-Oufella et al., it was suggested for the fi rst time that naturally arising CD4+CD25+ regulatory T cells are powerful inhibitors of atherosclerosis in several mouse models ( 8 ). This fi nding was strongly supported by observations that Treg numbers were reduced and their functional suppressive properties were compromised in experimental atherosclerosis and in patients with atherosclerosis (9)(10)(11). Moreover, our recent research has shown that generation of HSP60-specifi c Tregs could inhibit the formation of plaques in murine atherosclerosis ( 12 ), which was consistent with similar fi ndings from Harats et al. that induction of oral tolerance with antigen-specifi c Tregs could attenuate experimental atherogenesis ( 13 ).
Tregs on the adaptive immune system and on CD4+ T cells, in particular, have been well documented ( 1,2 ). But their effects on innate immune cells are less well known. Monocytes and macrophages are essential partners in innate and acquired immunity and as such play a variety of roles in atherosclerotic plaque development and its clinical sequelae ( 14 ). The uptake of oxidized lipoproteins via scavenger receptors and the ensuing formation of foam cells are key events in atherogenesis ( 15 ). Recent studies have implicated that CD4+CD25+ regulatory T cells induce Abstract Cluster of differentiation (CD)4+CD25+ regulatory T cells (Tregs) exert a suppressive activity on atherosclerosis, but the underlying mechanism remains unclear. Here, we investigated whether and how Tregs affect macrophages foam-cell formation. Tregs were isolated by magnetic cell sorting-column and analyzed by fl ow cytometry. Macrophages were cultured with or without Tregs in the presence of oxidized LDL (oxLDL) for 48 h to transform foam cells. After co-culture with Tregs, macrophages showed a decrease in lipid accumulation, which was accompanied by a signifi cantly downregulated expression of CD36 and SRA but no obvious difference in ABCA1 expression. Tregs can inhibit the proinfl ammatory properties of macrophages and steer macrophage differentiation toward an anti-infl ammatory cytokine producing phenotype. Mechanistic studies reveal that both cell-to-cell contact and soluble factors are required for Treg-mediated suppression on macrophage foam-cell formation. Cytokines, interleukin-10 (IL-10), and transforming growth factor-␤ (TGF-␤ ) are the key factors for these suppressive functions.

INTRODUCTION
Atherosclerosis is an infl ammatory disease of the arterial wall where both innate and adaptive Th1-driven immunoinfl ammatory responses contribute to disease development. Th2-related responses have been shown to be either protective or pathogenic (1)(2)(3). The role of the immune system in atherosclerosis has received considerable interest in recent years; however, suffi cient knowledge to justify the immunomodulatory mechanisms has not yet been obtained.

Co-culture experiments
Murine peritoneal macrophages were seeded in 6-well plates (3-4 × 10 6 /well), nonadherent cells were removed, and the culture medium was changed fi rst after 8 h incubation at 37°C. For co-culture experiment, peritoneal macrophages were cultured without T cells (no T), with CD4+CD25+ T cells (CD25+), or CD4+CD25 Ϫ T cells (CD25 Ϫ ) for 40 h in the presence of anti-CD3 antibody (50 ng/ml, eBioscience, CA) ( 16 ), then the different cultures were stimulated for 48 h with oxLDL (50 g/ml) to induce foam-cell formation. After the incubation period, fl oating T cells were aspirated, lipid-loaded macrophages were harvested, and supernatants were collected for further experiments.

Macrophage lipid analysis by oil red O-staining
Immediately following 48 h incubation with oxLDL, the medium containing fl oating T cells was aspirated, and lipid-loaded cells were fi xed in the same 6-well plates used for incubation, with 4% paraformaldehyde in water, for 2-4 min. Cells were stained with 0.3% oil red O in methanol for 15 min. Cells were observed under a phase contrast microscope with 200× magnification and then photographed. The number of foam cells formed under each condition was calculated manually and presented as a percentage foam-cell formation. At least 10 microscopic fi elds were counted from three different slides for the same treatment for quantifi cation of foam cells.
Quantifi cation of lipid accumulation was achieved by extracting oil red O from stained cells with isopropyl alcohol and measuring the OD of the extracts at 510 nm.

Cellular lipid analysis
After the incubation period, medium containing fl oating T cells was abandoned, and macrophages were removed from the culture plates and washed three times with PBS. Then intracellular lipids were extracted using isopropanol/hexane. Cellular lipid concentrations were determined by enzymatic colorimetric assays using kits (Wako Chemicals, Richmond, VA) for total cholesterol (TC) and for free cholesterol (FC). Esterifi ed cholesterol (CE) mass was calculated as the difference between TC and FC. Protein concentration was measured by the method of Lowry ( 18 ).

Cholesterol effl ux
Macrophages were seeded and incubated in 1640 plus 10% FBS containing 50 g/ml oxLDL and 2 Ci/ml [3H]cholesterol as described previously ( 19 ). After a 24-h incubation period, cells were washed and co-cultured with or without Tregs in serum-free medium. Cholesterol effl ux was performed for another 24 h in the presence of ApoA-I (10 g/ml, Sigma). The cholesterol effl ux was expressed as the percentage of the radioactivity released from the cells to the medium relative to the total radioactivity in cells plus medium containing anti-CD3 antibody (50 ng/ml). Each experiment was carried out in triplicate and repeated three times.

RNA analysis
Total RNA from macrophages was isolated using Trizol reagent (Invitrogen) according to manufacturer's instruction. One microgram of total RNA was reverse-transcribed using murine Moloney leukemia virus (M-MLV) (Toyobo, Osaka, Japan), and the resulting cDNA was used as a PCR template. Semi-quantitative PCR was performed using ReverTra Dash kit (Toyobo, Osaka, Japan) with the primers for GAPDH (5 ′ -GAG GGG CCA TCC ACA GTC TTC-3 ′ and 5 ′ -CAT CAC CAT CTT CCA GGA GCG-3 ′ ), CD36 (5 ′ -GTT TTA TCC TTA CAA TGA CA-3 ′ and 5 ′ -GGA AAT GTG GAA GCG AAA TA-3 ′ ), SRA (5 ′ -AAA GGG AGA CAG AGG alternative activation of monocytes/macrophages ( 16 ). Identifi cation and characterization of Tregs regulating the foam-cell formation, particularly genes involved in lipids accumulation, could be crucial in deciphering the effects and mechanisms of Tregs in atherosclerosis. Here we assessed a previously uncharacterized function of CD4+CD25+ Tregs, namely, their ability to modulate the transition of macrophages into foam cells.

Animals
All C57BL/6 background mice were obtained from Peking University. They were housed in the specifi c pathogen-free conditions in our animal facility and administered food and water ad libitum. The investigation conforms to the Guide for the Care and Use of Laboratory Animals published by the U.S. National Institutes of Health (NIH Publication No. 85-23, revised 1996) and is approved by the Ethics Committee of Tongji Medical College, Huazhong University of Science and Technology.

Preparation of LDL and copper-oxidized LDL
Blood for lipoprotein isolation was collected in EDTA (1 mg/ ml) from normal lipidemic donors after 12 h of fasting. LDL (density = 1.03 to 1.063 g/l) was isolated from the plasma after density adjustment with KBr, by preparative ultracentrifugation at 5,000 rpm/min for 22 h, using type 50 rotor. They were dialyzed against PBS containing 0.3 mM EDTA, sterilized by fi ltration through a 0.22-mm fi lter, and stored under nitrogen gas at 4°C. Protein content was determined by the method of Lowry et al. Copper oxidation of LDL was performed by incubation of postdialyzed LDL (1 mg of protein/ml in EDTA-free PBS) with copper sulfate (10 mM) for 24 h at 37°C. Lipoprotein oxidation was confi rmed by analysis of thiobarbituric acid-reactive substances ( 17 ).

Cell isolation and purifi cation
For cells isolation, groups of 8-to 12-week-old male mice were used for all experiments. Murine peritoneal macrophages were harvested after intraperitoneal injection (5 ml per mouse) of PBS. After centrifugation at 1000 g for 5 min, the cells were resuspended into complete RPMI-1640 medium containing 10% FBS and 100 U/ml of penicillin/streptomycin, then adjusted to the concentration of 1 × 10 6 /ml and used for further experiments.
Splenocytes were prepared from murine spleens by Ficoll density gradient. CD4+ T cells were isolated from total splenocytes by negative selection using LD column (Miltenyi Biotech, Bergisch Gladbach, Germany). Purifi ed CD4+ T cells were incubated with anti-mouse CD25 magnetic beads (Miltenyi Biotech) and separated into CD4+CD25+ and CD4+CD25 Ϫ fractions by positive selection using MS column (Miltenyi Biotech). The positively selected CD25+ cell fractions were separated again over an MS column to achieve higher purities. As assessed by using fl uorescence-activated cell sorting (FACS) (Becton-Dickinson, Oxnard, NJ), purities of CD4+CD25 Ϫ T cells and CD4+CD25+ T cells were > 95% and > 90%, respectively. To further confi rm the identity of these Tregs, the primers (5 ′ -ACA CCA CCC ACC ACC GCC ACT-3 ′ and 5 ′ -TCG GAT GAT GCC ACA GAT GAA GC-3 ′ ) were used to measure the expression of Foxp3 mRNA, which were highly expressed in CD4+CD25+ T cells but not in CD4+CD25 Ϫ T cells in our work. without macrophages). After 40 h of co-culture, the top compartments (inserts) were removed, and oxLDL (50 g/ml) was added to induce macrophage differentiation for 48 h. After the incubation period, cells were harvested for lipid measurement for further experiments.
To analyze whether TGF-␤ and IL-10 are required for Tregmediated suppression on foam-cell formation, neutralizing IL-10 antibody (5 g/ml) (R and D Systems) and/or TGF-␤ antibody (5 g/ml) or nonblocking IgG anti-mouse control antibody were added at the start of the co-culture system. Cells were cultured for 48 h and tested for cellular esterifi ed cholesterol mass to measure foam-cell formation.

Statistical analyses
Results are shown as mean ± SEM of at least three independent experiments. The signifi cance of differences was estimated by ANOVA followed by Student-Newmann-Keuls multiple comparison tests. P < 0.05 was considered signifi cant. All statistical analyses were performed with SPSS software (version 11.0, SPSS Inc., Chicago, IL).

Effects of Tregs on oxLDL uptake in peritoneal macrophages
Macrophages are critical in cholesterol metabolism. Alterations in the uptake, metabolism, and effl ux of cholesterol in macrophages could affect foam-cell formation, a prerequisite for atherosclerosis development. To determine if Tregs can modulate foam-cell formation, a frequently in vitro foam-cell model system was used: peritoneal macrophages co-cultured with or without T cells (CD4+CD25+ T cells or CD4+CD25 Ϫ T cells) were treated with oxLDL for 48 h to induce foam-cell formation. Foam-cell formation, as identifi ed by oil red O staining, was readily apparent in cells treated with CD4+CD25 Ϫ T cells and without T cells ( Fig.  1A ). After treatment with Tregs, a marked decrease (13.9 ± 5.6%) in foam-cell count was observed, compared with untreated cells (52.9 ± 10.4%; P < 0.001) or CD4+CD25 Ϫ T-treated cells (53.1 ± 17.2%; P < 0.001). The similar effect of Tregs was obtained when extracted Oil Red O was measured by a spectrophotometer ( Fig. 1B ).
To explore whether there was a threshold effect for Tregs-mediated lipid loading of cells, macrophages (3-4 × 10 6 /well) were cultured without or with various concentration of Tregs (1.25 × 10 5 , 2.5 × 10 5 , 5 × 10 5 /well), then cellular esterifi ed cholesterol accumulation was measured. A dose-dependent effect on cellular esterifi ed cholesterol accumulation was noted in macrophage foam GCT CAC-3 ′ and 5 ′ -CTT GAT CCG CCT ACA CT-3 ′ ), and ABCA1 (5 ′ -ACA ACC AAA CCT CAC ACT ACT G-3 and 5 ′ -ATA GAT CCC ATT ACA GAC AGC G-3 ′ ). The amplifi ed DNA was analyzed on 2% agarose gel and then visualized with ethidium bromide staining. Quantitative real-time PCR was carried out with ABI PRISM 7900 Sequence Detector system (Applied Biosystems, Foster City, CA) according to the manufacturer's instructions. GAPDH was used as endogenous control. PCR reaction mixture contained the SYBR Green I (Takara Biotechnology), cDNA, and the primers. Relative gene expression level (the amount of target, normalized to endogenous control gene) was calculated using the comparative Ct method formula 2 Ϫ ⌬ ⌬ Ct. The sequences of primers for real-time PCR were GAPDH (5 ′ -CCA TCA CCA  TCT TCC AGG AGC

Western blot analysis
After incubation with oxLDL, the serum containing fl oating T cells was discarded. The remnant macrophages were washed with precooled PBS, centrifuged at 4°C 3000 rpm for 5 min, lysed in ice-cold buffer containing 10 mM Hepes (pH 7.4), 10 mM KCl, 1.5 mM MgCl2, 0.1 mM sodium EDTA, 0.1 mM sodium EGTA, 1.0 mM DTT, 1.0 mM PMSF, and 1 mg/ml protease inhibitor aprotinin for 30 min and centrifuged at 4°C 2000 rpm for 10 min. The protein concentration in cellular supernatants was determined by the Lowry assay. Equal amounts of protein (20 g) were separated on 9% SDS-polyacrylamide (SDS-PAGE) gels and electrophoretically transferred to nitrocellulose (NC) membrane and blocked 2 h (5% "degrease milk powder" in Tween/PBS buff er; vacillating bed; and room temperature) and, after incubation with fi rst and second antibodies, were detected by ECL (Pierce Biotech). Bands were normalized to ␤ -actin and expressed as a percent of control. Primary antibodies were used in different dilutions as follows: anti-SRA (scavenger receptor A) 1:250 (mouse monoclonal; Santa Cruz), anti-CD36 (membrane glycoprotein belonging to the class B scavenger receptor family) 1:300 (mouse monoclonal), anti-ABCA1 (ATP-binding cassette transporter A1) 1:300 (mouse monoclonal), anti-␤ -actin 1:500 (goat monoclonal).

ELISA assay for cytokines
The peritoneal macrophages were collected and co-culture experiments were performed as described in the previous section. To exclude possible contributions of T cell-derived cytokines in these assays, medium containing fl oating T cells were aspirated and fresh medium was changed before incubation with oxLDL. Supernatants were collected after 48 h incubation and kept frozen at Ϫ 80°C until the cytokine levels (TNF-␣ , MCP-1, MMP-9, IL-10, and TGF-␤ ) were determined by ELISA assays according to the manufacturer's instructions. Each sample was tested for each cytokine in triplicate.

Transwell and neutralization experiments
To assess whether cell-to-cell contact was necessary for Tregs to mediate suppression, polycarbonate 24-well Transwell inserts (Corning, Acton, MA) were used ( 20 ). Peritoneal macrophages were collected and incubated as described in the previous section. Transwell experiments were performed by culturing macrophages (3-4 × 10 6 /well) in the lower well and CD4+CD25+ T cells (5 × 10 5 /well) with anti-CD3 mAb in the inserts (with or crease in cholesterol effl ux. Therefore, the rate of cholesterol effl ux from macrophage foam cells was determined. There was no difference detected in the cholesteroleffl ux rate in untreated, CD4+CD25+ Treg-treated, and CD4+CD25 Ϫ T-treated foam cells ( P > 0.05) ( Fig. 2 ), indicating that the reduced accumulation of oxLDL in macrophage foam cells was probably due to a decrease in cholesterol uptake. cells after 48 h incubation with oxLDL ( Fig. 1C ). It seems that there was not a threshold effect on cellular esterifi ed cholesterol accumulation with suggested concentration of Tregs.

Cholesterol effl ux in macrophage-derived foam cells
A decrease in cholesterol accumulation in macrophages could be due to a decrease in cholesterol uptake or an in-

Effects of Tregs on expression of protein involved in cholesterol homeostasis in peritoneal macrophages
The protein levels of SR-A and CD36 of the macrophages were also determined by Western blot analysis. Coincident with the decreased CD36 and SR-A mRNA expression, a signifi cant reduction of CD36 and SR-A protein expression was detected in CD25+ cultures relative to no T (CD36 = P < 0.05; SR-A = P < 0.01) and CD25 Ϫ (CD36 = P < 0.05; SR-A = P < 0.001) cultures ( Fig. 4A , B ). Western blotting of these macrophage lysates did not detect any change in protein levels of ABCA1 ( P > 0.05) ( Fig. 4A, B ).

Tregs inhibit the proinfl ammatory response in oxLDL-induced macrophage foam-cell formation
Next we determined the capacity of infl ammatory response in oxLDL-induced macrophage foam-cell formation. Our data show that, after co-culture with Tregs, macrophages displayed a decrease in their capacity to produce proinfl ammatory cytokines/chemokines (TNF-␣ , MCP-1, and MMP-9) ( Fig. 5A ). In contrast, the production of the anti-infl ammatory cytokines TGF-␤ and IL-10 was enhanced ( Fig. 5B ) in Treg-treated cultures. As a control, CD4+CD25 Ϫ T-treated cultures displayed an increased production of proinfl ammatory and decreased production of anti-infl ammatory cytokines/chemokines. Results collectively indicated that rather than inhibiting a proinfl ammatory response, Tregs could induce an anti-infl ammatory response in oxLDL-induced macrophage foam-cell formation.

Treg-mediated suppression on macrophage foam-cell formation requires cell contact as well as soluble factors
To investigate whether suppression of macrophage foam-cell differentiation depended on cell contact or soluble factors, we cultured macrophages with or without Tregs in the presence of oxLDL in either a co-culture (CC) or a transwell system (TW). By disrupting physical contact between macrophages and Tregs (TW), the suppression of cellular esterifi ed cholesterol (45.30 ± 17.18 g/mg) was reversed compared with the CC system (26.68 ± 8.88 g/mg; P < 0.05) ( Fig. 6A ), suggesting that cell-to-cell contact was required in Treg-mediated suppression on cellular esterifi ed cholesterol accumulation. However, in the absence of cell-to-cell contact, cellular esterifi ed cholesterol levels (45.30 ± 17.18 g/mg) were still markedly reduced compared with macrophages alone (102.54 ± 16.67 g/mg; P < 0.001) ( Fig. 6A ), indicating that cell-tocell contact was only partly necessary and soluble factors may play a major role in Treg-mediated suppression on cellular esterifi ed cholesterol accumulation.
To explore the mechanism of Treg-mediated suppression on foam-cell formation, a neutralizing experiment was performed. Although it has been shown that Tregs are able to produce TGF-␤ and IL-10 upon stimulation ( 41,42 ), we need to confi rm that these cytokines were indeed present in our cultures. ELISA assays showed that after stimulated with anti-CD3 antibody for 40 h, CD4+CD25+ Tregs were strongly positive in the production of TGF-␤ and IL-10 (data not shown), which was a fundamental

Effects of Tregs on expression of genes involved in cholesterol homeostasis in peritoneal macrophages
Previous studies ( 20 ) demonstrated that scavenger receptor class A (e.g., SR-AI/II) and class B (e.g., CD36) are the principal receptors responsible for binding and uptake of modifi ed LDL, while the effl ux of cholesterol in macrophages mediated by ApoA-I largely depends upon the activated transporter ABCA1 ( 21,22 ).
Thus, we further investigated whether the expression of genes involved in cholesterol homeostasis was altered in these macrophage foam cells. Macrophages (3-4 × 10 6 / well) were cultured without or with various concentration of Tregs (1.25 × 10 5 , 2.5 × 10 5 , 5 × 10 5 /well). The mRNA levels of SR-A, CD36, and ABCA1 were determined by semi-quantitative PCR and quantitative real-time PCR analysis. Fig. 3 shows that Treg-treated macrophage foam cells displayed a strong downregulation of CD36 and SRA mRNA expression. A dose-dependent effect on CD36 and SRA expression was noted in macrophage foam cells after 48 h determined by RT-PCR and quantitative real-time PCR ( Fig. 3A, B ). Indeed, we found Tregs dose-dependently inhibit oxLDL-induced CD36 and SRA expression relative to untreated cells. At 1.25 × 10 5 cells, Tregs decreased expression of CD36 by 31% ( P < 0.01) and SRA by 42% ( P < 0.001) in macrophage foam cells compared with untreated cells (no T). But no signifi cant difference in ABCA1 mRNA expression was detected in cultures incubated with various concentrations of Tregs and untreated cells ( P > 0.05) (date not shown).
In a further step, the effects of neutralizing the IL-10 antibody or TGF-␤ on mRNA and protein expression of CD36 and SRA were also investigated. Our data show that study for the further neutralizing experiment. For the neutralizing experiment, neutralizing IL-10 antibody and/ or TGF-␤ antibody or nonblocking IgG anti-mouse control antibody were added to CC system. Our data show that suppression of cellular esterifi ed cholesterol accumulation was reversed to a large degree with antibodies against IL-10 or anti-TGF-␤ antibody ( P < 0.05) ( Fig. 6B ). Further, after treatment with neutralizing anti-TGF-␤ antibody, suppression was more readily restored relative to the addition of anti-IL-10 antibody (TGF-␤ antibody = 68.69 ± in the presence of oxLDL (50 g/ml) for 48 h. Total RNA was collected and CD36 and SRA mRNA were analyzed by semi-quantitative PCR. GADPH served as endogenous control. (B) Cells were cultured as described as above. Quantitative realtime PCR analysis of CD36 and SRA mRNA relative to GAPDH was performed with equal loading of samples from indicated cultures. Data are presented as mean ± SEM of triplicate wells and are representative of at least three independent experiments. (* is indicated for versus no T [* = P < 0.05; *** = P < 0.001]). Co-cultures were set up as described in Fig. 1. Total RNA was collected. CD36, SRA, and ABCA1 mRNA were analyzed by semi-quantitative PCR (C) and realtime PCR (D). Data are presented as mean ± SEM of triplicate wells and are representative of at least three independent experiments. (# is indicated for CD25+ versus no T; * is indicated for CD25+ versus CD25 Ϫ [## = P < 0.01; ### = P < 0.001; ** = P < 0.01; *** = P < 0.001]). nals" via toll-like receptors (TLRs) and other patternrecognition receptors. Understanding the mechanisms behind the homeostatic control of monocyte/macrophage function is, therefore, of fundamental importance. Previously, indirect evidence for Treg-mediated suppressive effects on monocytes/macrophages was provided after adoptive transfer of Tregs in a colitis model ( 23 ). Recent work by Taams ( 24 ) and others ( 25 ) demonstrated that in humans, CD4+CD25+ Tregs have direct inhibitory effects on the antigen-presenting function of monocytes/macrophages, as shown by their reduced capacity to stimulate antigen or allo-specifi c T-cell responses. By oil-red staining and cellular cholesterol measurement, we examined the immune effects of Tregs on macrophage differentiation into foam cells. Our data clearly showed that treatment with Tregs can signifi cantly decrease cholesterol accumulation in macrophage foam cells by reducing oxLDL uptake in these cells. In contrast, there was no distinction in lipid accumulation between macrophage foam cells alone and macrophage foam cells treated with CD4+CD25 Ϫ T cells. These results suggest that Tregs exert a suppressive effect on macrophage foam-cell formation. the downregulation of CD36 and SRA mRNA in Tregtreated macrophage foam cells was apparently reversed by adding antibodies against IL-10 or anti-TGF-␤ antibody, blocking of both IL-10 and TGF-␤ was necessary to completely abrogate suppression ( Fig. 6C ). A similar reversal of CD36 and SR-A protein expression was also detected when neutralizing anti-IL-10 and/or neutralizing anti-TGF-␤ were added (data not shown). Collectively, these data suggest that Treg-mediated suppression of macrophage foam-cell formation requires cell contact as well as soluble factors, and that soluble factors, mainly IL-10 and TGF-␤ , contribute largely to this suppression.

DISCUSSION
In the current study, we provide evidence for a previously uncharacterized role of Tregs, namely their ability to modulate the transition of macrophages into foam cells in murine macrophages. Monocytes/macrophages play a critical role in both innate and adaptive immunity through their ability to recognize pathogens and/or "danger sig- level. Previous studies showed that mice with deletion of either of the above receptors exhibited marked reduction in atherosclerosis (26)(27)(28). These observations have led to the current dogma concerning scavenger receptors and cholesterol-laden macrophage foam cells, which is that they are proatherogenic molecules. However, recent studies have clearly revealed that the effects of scavenger receptors and macrophage foam cells on atherogenesis may be more complex, and each heterogeneity of immuneassociated cells has both atherogenic and atheroprotective roles ( 29,30 ). Although the uptake of modifi ed lipoproteins by SRA is thought to be central to foam-cell formation, it is also widely believed to represent one of the major Macrophage internalization of modifi ed lipoproteins is thought to play a critical role in the initiation of atherogenesis. Two scavenger receptors, scavenger receptor A (SRA) and CD36, have been centrally implicated in this lipid uptake process. To further investigate whether CD36 and SRA were required in the suppression of macrophage foam-cell formation, we hypothesized that the mechanism of decreased foam-cell formation mediated by Tregs could be a result of Tregs regulation of one or both receptors. We then compared the scavenger receptor expression in oxLDL-induced macrophage foam cells co-cultured with or without Tregs. The effect of Tregs on scavenger receptor expression has been evaluated in a previous section. Tregs inhibited expression of both SRA and CD36. In contrast to the suppressive effects of Tregs on scavenger receptors, CD4+CD25 Ϫ T cells do not alter expression of SRA and CD36. To understand to what extent Tregs modulated expression of scavenger receptors, various concentrations of Tregs were used to treat macrophages. Our data show that Tregs decreased oxLDL-induced CD36 and SRA expression in a dose-dependent manner. But a particularly intriguing fi nding was that the responses of SRA and CD36 were not quantitatively the same. SRA was more responsive to Tregs than CD36, especially at the protein  6 /well) were cultured with or without Tregs (CD25+, 5 × 10 5 /well) as described in Fig. 1. In some experiments, Transwell inserts were used to separate macrophages from Tregs. Neutralizing IL-10 antibody and/or TGF-␤ antibody or nonblocking IgG anti-mouse control antibody were added at the start of co-culture system. (A) Cellular cholesterol ester measured from macrophages cultured alone (no T), in Transwell insert (TW) or in the co-culture system (CC). Data are presented as mean ± SEM of at least three independent experiments. (# is indicated for TW versus no T; * is indicated for TW versus CC; + is indicated for CC versus no T [* = P < 0.05; ### = P < 0.001; *** = P < 0.001]). (B) Cellular cholesterol ester was measured in the presence of nonblocking IgG anti-mouse antibody, neutralizing anti-IL-10, and/or anti-TGF-␤ antibody in the co-culture. Data are presented as mean ± SEM of at least three independent experiments. (* is indicated for versus control antibody [* = P < 0.05, ** = P < 0.01, *** = P < 0.001]). (C) Quantitative real-time PCR was performed to measure the expression of CD36 and SRA mRNA in neutralizing experiment. GADPH served as endogenous control. Data are presented as mean ± SEM of at least three independent experiments. (* is indicated for versus control antibody [* = P < 0.05, ** = P < 0.01, *** = P < 0.001]).
However, more recent research suggests that some subsets of Tregs produce cytokines IL-10 and/or TGF-␤ ( 37-39 ). We showed that Treg-treated cultures produced IL-10 and TGF-␤ and that neutralizing antibodies to these cytokines completely restored cellular cholesterol ester accumulation. The fact that cellular esterifi ed cholesterol levels in the absence of cell-to-cell contact were still markedly reduced compared with macrophages alone combined with the completely abrogated suppression of neutralizing antibodies to IL-10 and TGF-␤ 1, indicating that soluble factors play a central role in immune suppression mediated by Tregs. Interestingly, however, our data also show that by disrupting physical contact between macrophages and Tregs (TW), the suppression of cellular esterifi ed cholesterol was only partly reversed compared with the CC system, suggesting that cell-to-cell contact was required for Treg-mediated suppression of foam-cell formation. Thus, it is possible that, although cell-to-cell contact contributes to suppression, it is dependent on Treg-derived cytokines, which regulate/induce mechanisms directly responsible for blocking of macrophage cell functions. Our data suggest that mechanisms involving cytokines and also requiring cell-to-cell contact contribute to suppression mediated by Tregs.

CONCLUSION
Our study shows that CD4+CD25+ Tregs may suppress macrophage foam-cell formation in part though inhibiting the uptake of oxLDL, which is likely due to decreased scavenger receptor SRA and CD36 expression rather than an increase in cholesterol effl ux. Moreover, Tregs can inhibit the proinfl ammatory properties of macrophages and steer macrophage differentiation toward an anti-infl ammatory cytokine producing phenotype. Both cell-to-cell contact and soluble factors were required for Treg-mediated suppression of macrophage foam-cell formation; IL-10 and TGF-␤ were the key factors for these suppressive functions. This newly discovered ability of Tregs may help us to understand the mechanisms of Tregs in atherosclerosis processes and may also provide a novel tool to manipulate atherosclerosis development.
activation events stimulating the proinfl ammatory phenotype of lesional macrophages. Recently, a possible IFN ␥responsive element was identifi ed in the human SRA gene, which apparently mediates increased mRNA expression of SRA in blood monocytes and undifferentiated THP-1 cells, a human monocyte cell line, whereas it inhibits SR-A expression in mature macrophages and differentiated THP-1 cells ( 31 ). We think that macrophage heterogeneity with respect to scavenger receptor expression and function may partly explain our results. A better picture of heterogeneity in scavenger receptor expression and function will require the development of better antibodies for use in histochemistry and more detailed phenotyping of receptor knockout mice.
Normal cellular cholesterol homeostasis, however, involves not only specifi c traffi cking of lipid storage vesicles but also processing of excess intracellular cholesterol such that it is mediated by ABCA1, a transporter, and facilitated by ApoA-I, which transfers cholesterols to HDL, which then transports cholesterol back to the liver, a process termed reverse cholesterol transport ( 21,22 ).
To further confi rm the possible mechanism for decreased lipid accumulation in macrophage foam cells, ABCA1 expression and cholesterol effl ux rate in macrophages foam cells was determined. Interestingly, our data showed no signifi cant difference in cholesterol effl ux rate, which was coincident with the unaltered ABCA1 expression in these three cultures. Collectively, these fi ndings suggest that the suppression of Tregs on cholesterol accumulation in macrophage foam cells was attributed to a decrease in the expression of scavenge receptors SRA and CD36, rather than an increase in cholesterol effl ux.
Cytokines, particularly IL-10 and TGF-␤ , play major roles in the generation and functions of Tregs (37)(38)(39), although their individual contributions to differentiation of Treg subsets and suppressor functions are still debated. In the initial studies, using in vitro assay systems, it was shown that strong suppressive activity exerted by CD4+CD25+ Treg requires cell-to-cell contact, and it was thus considered likely that suppression was dependent on ligand-receptor interactions at the cell surface (40)(41)(42).