Macrophage SR-BI regulates LPS-induced pro-inflammatory signaling in mice and isolated macrophages.

Scavenger receptor BI (SR-BI), an HDL receptor, plays a key role in reverse cholesterol transport. In mice, disruption of SR-BI results in hypersensitivity to lipopolysaccharide (LPS) and bacteria-induced septic shock due to adrenal insufficiency and abnormal hepatic pathogen clearance. In this study, we identify an anti-inflammatory role of macrophage SR-BI. Using bone marrow transplantation, we report an enhanced pro-inflammatory response to LPS in wild-type (WT) mice receiving SR-BI-null compared with WT bone marrow cells and a reduced response in SR-BI-null mice receiving WT compared with SR-BI-null cells. Although significant, SR-BI deficiency limited to bone marrow-derived cells promoted a relatively modest enhancement of the inflammatory response to LPS in mice compared with the effect of whole-body SR-BI deletion. Consistent with earlier findings, SR-BI-null primary macrophages exhibited a greater inflammatory cytokine response to LPS than control macro phages. In addition, we showed that overexpression of SR-BI in J774 macrophages attenuated the inflammatory response to LPS. The LPS-induced cytokine expression in both WT and SR-BI-null macrophages was dependent not only on NFκB as previously reported but also on JNK and P38 cell signaling pathways. The increased inflammatory signaling in SR-BI-null cells was not related to alterations in cellular cholesterol content. We conclude that SR-BI plays an important function in regulating the macrophage inflammatory response to LPS.

unless otherwise indicated. The MAPK inhibitors PD98059, SB203580, and SP600125 as well as LPS and LTA were obtained from Sigma-Aldrich (St. Louis, MO). The NF B inhibitor 1-pyrrolidinecarbodithioic acid (PDTC) was obtained from Calbiochem (San Diego, CA). J774 cells stably transfected with mouse SR-BI and control J774 cells were kindly provided by Dr. Theodore Mazzone (NorthShore University, Chicago, IL). The J774-SR-BI cells were transfected to stably express full-length murine SR-BI in pcDNA 3.1 zeo as previously described in detail ( 23 ). Clones were selected with zeocin, and cell viability was not altered by the expression of SR-BI.

Bone marrow transplantation
The procedure was performed as previously reported ( 24 ). Six-to 8-week-old SR-BI +/+ (WT) and SR-BI Ϫ / Ϫ (SR-BI-null) recipient mice were maintained on water containing an antibiotic (sulfratrim, 4 g/ml) for 1 week prior to irradiation. Recipient mice were irradiated with 450 rad from a cesium ( ␥ radiation) source twice with 3 h between exposures. Irradiated recipient mice were repopulated by a tail vein injection of 1 × 10 6 bone marrow cells harvested from tibia and femurs of age-matched SR-BI Ϫ / Ϫ and SR-BI +/+ mice. Following bone marrow transplantation, mice were maintained on water containing sulfratrim (4 g/ml) for 4 weeks and then switched to regular water for 1 week. Mice were then injected intraperitoneally with LPS (0.25 g/g body weight for SR-BI +/+ recipients, 0.1 g/g body weight for SR-BI Ϫ / Ϫ recipients). Mice were euthanized 2 h after LPS injection, and blood and tissues were harvested. To assess engraftment of donor hematopoietic cells, DNA was isolated from bone marrow cells of recipient mice using a DNeasy Tissue Kit (Qiagen, Valencia, CA) and analyzed by PCR for SR-BI expression.

Bone marrow macrophage isolation
Bone marrow cells were harvested from tibia and femurs of age-matched SR-BI Ϫ / Ϫ and SR-BI +/+ mice 5 days after bio-gel elicitation [injected intraperitoneally with 1 ml/mouse of polyacrylamide gel P-100 (Bio-Rad, Hercules, CA) in a 2% w/v suspension in endotoxin-free water]. Bone marrow cells were then cultured and differentiated in 12-well dishes containing RPMI-1640 supplemented with 50 IU/ml penicillin G, 50 g/ml streptomycin, 2 mM glutamine, 10% fetal bovine serum, and 15% (v/v) L929-cell conditioned medium for 7 days.

Infl ammatory response in macrophages
Primary bone marrow macrophages (BMM) were treated with LPS in the presence or absence of specifi c inhibitors of cell signaling pathways in RPMI medium supplemented with 5% fetal bovine serum. These inhibitors included MAPK inhibitors SP600125 (JNK, 20 M), SB203580 (P38, 20 M), and PD98059 (ERK, 25 M), and NF B inhibitor PDTC (50 M). BMMs were incubated overnight in serum-free RPMI, pretreated with the indicated inhibitor for 1 h, and then treated with the same conditions plus LPS for 4 h at 37°C. Media were collected for cytokine analysis by ELISA (Invitrogen, Carlsbad, CA), and cellular protein was harvested in lysis buffer containing protease inhibitor (Roche Diagnostics, Indianapolis, IN) and phosphatase inhibitor (Thermo Scientifi c, Rockford, IL) cocktails.

Western blotting of solubilized macrophage protein
Total cellular protein was harvested from BMMs using Ripa buffer (Sigma, St. Louis, MO) supplemented with protease inhibitors (Roche Diagnostics, Indianapolis, IN) and a phosphatase inhibitor cocktail (Thermo Scientifi c, Rockford, IL). Cell protein (15 g/well) was separated in 4-16% SDS polyacrylamide gels nitric oxide (NO)-induced cell toxicity ( 9,10 ). In line with these fi ndings, pro-infl ammatory cytokine production was markedly enhanced in SR-BI-null mice ( 9 ) Li et al. showed that SR-BI protected against septic death in mice and inhibited the infl ammatory response to LPS in macrophages ( 11 ).
SR-BI is expressed in a variety of cell types, most abundantly in steroidogenic cells and in the liver ( 12 ). SR-BI is also expressed in macrophages ( 13 ); however, its function in macrophages and the impact of macrophage SR-BI on the integrated infl ammatory response is not yet clear. SR-BI has been recognized as a mediator of bidirectional cholesterol fl ux between macrophages and extracellular acceptors ( 12,14 ). When fed a Western diet, mice lacking both SR-BI and apoE showed enhanced macrophage cholesterol accumulation ( 15 ). The dysregulation of cholesterol homeostasis in these cells was attributed to a disruption of macrophage cholesterol effl ux and lysosomal cholesterol traffi cking ( 15 ). Recent studies from our laboratory revealed that SR-BI can promote macrophage cholesterol effl ux to HDL ( 16 ). Deletion of SR-BI in bone marrow cells was shown to promote atherosclerosis in mice, at least during later-stage lesion development (17)(18)(19).
Macrophage infl ammatory responses have been shown to be modulated by cellular membrane cholesterol levels. In ABCA1-and ABCG1-defi cient macrophages, the increased cell membrane cholesterol content resulting from reduced cellular cholesterol effl ux contributes to the hyperinfl ammatory responses in these cells in response to stimuli ( 20,21 ). SR-BI is known to affect the content and distribution of cholesterol in plasma membranes, but the possibility that SR-BI modulates macrophage infl ammatory responses by such an alteration remains to be explored.
In this study, we investigated the role of macrophage SR-BI in the LPS-stimulated infl ammatory response. We present our fi ndings that SR-BI expression in macrophages decreases the pro-infl ammatory cytokine response to LPS in both mice and isolated macrophages. The increased cytokine expression in SR-BI-null cells involved both NF B as well as JNK and P38 cell signaling pathways. The hyperinfl ammatory response in SR-BI-null macrophages is not associated with alterations in cellular cholesterol content.

Animals and reagents
Mice were housed in the Veterans Affairs Medical Center (Lexington, KY), and all animal protocols received appropriate institutional approval (Animal Welfare Assurance Number, Veterans Affairs Medical Center, A3506-01; VMU IACUC protocols 2009-0005V, 2009-0006V). Animals were maintained in a pathogen-free facility with 12 h light/dark cycle and free access to food and water. SR-BI-null mice were obtained from Jackson Laboratory (Monty Krieger, MIT) ( 22 ) and backcrossed for 10 generations into a C57BL/6 background (MacRae F. Linton, Vanderbilt University) ( 15 ). ELISA sets for cytokine determination were purchased from BD Biosciences (San Jose, CA). Antibodies were purchased from Cell Signaling (Danvers, MA) times with PBS. The cells were then solubilized in 0.1 N NaOH for 1 h at room temperature, and the protein content and radioactivity of the lysates were determined.

Statistical analysis
Statistical signifi cance in experiments comparing only two groups was determined by 2-tailed Student t -test. The signifi cance of the difference in mean values between more than two groups was evaluated by one-way ANOVA, followed by post hoc analysis using Tukey's test. All signifi cant differences ( P < 0.05 or P < 0.01) are given in the fi gures. Statistical analyses were carried out using GraphPad Prism 4 (GraphPad Software, CA). Values are expressed as mean ± SD. A P value of less than 0.05 was considered signifi cant.

Effect of SR-BI expression in transplanted bone marrow cells on the infl ammatory response to LPS in mice
Our recent studies showed that mice lacking SR-BI are highly sensitive to LPS-induced toxic shock. The protective effect of SR-BI was largely due to its role in maintaining adrenal corticosteroid production and facilitating hepatic LPS clearance. To determine the possible role of macrophage-specifi c SR-BI in the infl ammatory response, we investigated the response to LPS in mice transplanted with either SR-BI-null or WT bone marrow cells ( Fig. 1 ). Given that SR-BI-null mice are hypersensitive to LPS ( 9 ), a lower dose of LPS was given to SR-BI-null mice (0.1 g/g body weight) than to WT mice (0.25 g/g body weight). Even at the relatively lower LPS dose, SR-BI-null mice ( Fig. 1B ) demonstrated signifi cantly elevated pro-infl ammatory cytokine secretion compared with the WT mice ( Fig. 1A ). However, the infl ammatory response was also affected by SR-BI expression in transplanted bone marrow cells. As shown in Fig. 1A , WT mice that received SR-BI-null cells had significantly elevated tumor necrosis factor (TNF) ␣ and interleukin (IL)-6 levels following LPS treatment compared with mice that received WT cells. This indicated that SR-BI expressed in macrophages reduced the pro-infl ammatory response to LPS. In line with these fi ndings, SR-BI-null mice that received WT cells showed decreased levels of proinfl ammatory cytokines and monocyte chemotactic protein-1 (MCP-1) in response to LPS compared with mice that received SR-BI-null cells ( Fig. 1B ).
Although the differences observed between mice transplanted with WT and SR-BI-null bone marrow are smaller than the differences observed between LPS-treated WT and SR-BI-null mice (about 3-to 4-fold) ( 9 ), these results confi rm that SR-BI expressed in bone marrow-derived cells makes a signifi cant contribution to the general infl ammatory response. The differences between mice are independent of adrenal function because mice receiving bone marrow cells of different genotypes showed similar plasma corticosterone levels during infl ammation ( Fig. 1C ). Successful bone marrow cell repopulation was achieved in mice receiving either WT or SR-BI-null bone marrow cells, such that expression of SR-BI in the bone marrow cells of the transplanted mice predominantly refl ected and transferred to PVDF membranes. The blots were incubated overnight at 4°C with specifi c antibodies: rabbit anti-mouse SR-BI (Novus, Littleton, CO); rabbit anti-mouse P42/44 ERK or ERK; rabbit anti-mouse P-JNK or JNK; and rabbit anti-mouse P-P38 or P38. Following 1 hr incubation with the corresponding second antibody (horseradish peroxidase-linked anti-rabbit IgG), immunoblots were briefl y treated with a chemiluminescent reagent (Thermo Scientifi c, Rockford, IL) and exposed to Kodak fi lm. Gel bands were quantifi ed using KODAK ID Image Analysis software.

Macrophage cholesterol depletion and repletion
Cholesterol depletion and repletion of macrophages was performed as previously described ( 20,25 ). Briefl y, BMMs were incubated with or without 10 mM methyl-␤ -cyclodextrin (M ␤ CD) in prewarmed Hanks buffered salt solution (HBSS) supplemented with calcium and magnesium at 37°C for 30 min. To reload BMMs with cholesterol, cholesterol-depleted cells were washed with Hanks buffer and incubated in the presence of cholesterol (80 g/ml) complexed to M ␤ CD (1.5 mM) at 37°C for 1 h. Following treatment with 100 ng/ml LPS for 4 h, BMMs were treated with hexane/isopropanol to extract lipids, and cellular cholesterol levels were determined by commercially available kits (Wako Chemicals, Richmond, VA). BMM membrane lipid distribution was analyzed using CtxB staining ( 20 ). CtxB staining was performed using commercially available kits (Invitrogen, Carlsbad, CA) ( 20 ). Fluorescent signals were visualized on an Olympus BX51 microscope, and images were captured with equal exposure time. The intensity of fl uorescent signal in BMMs was quantifi ed by ImageJ public domain software (Wayne Rasband, National Institutes of Health, Bethesda, MD) ( 26 ). Individual cells were analyzed from fi ve random fi elds.

Quantitative real-time PCR
Total RNA was isolated from mouse adipose tissue and liver using the standard TRIzol method (Invitrogen, Calsbad, CA) and was purifi ed with DNase I (Roche) and RNeasy Mini Kits (Qiagen, Valencia, CA). Two micrograms of RNA was reversetranscribed into cDNA using a reverse-transcription system (Promega, Madison, WI). After a 4-fold dilution, 5 l was used as a template for Q-PCR. Amplifi cation was done for 40 cycles using a Power SYBR Green PCR Master Mix Kit (Applied Biosystems, Carlsbad, CA) and DNA Engine Optical 2 System (MJ Research Inc., Ramsey, MN). Both an internal control (GAPDH) and a negative control (minus reverse transcriptase) were included in each analysis. Values of each RNA sample were the average of duplicate assays normalized to GAPDH RNA levels.

I radionuclide labeling of LPS and BMM LPS uptake
LPS was iodinated by the chloramine T method as described previously ( 9 ). In brief, 1 mg of LPS was dissolved in 0.05 M borate buffer (pH = 8) and incubated with 50 mM p-OH-methylbenzimidate at 37°C for 18 h. After extensive dialysis, the M-LPS was radiolabeled with Na 125 I by the chloramine T method. After iodination, unincorporated Na 125 I was removed by extensive dialysis in sterile saline at 4°C. The bioactivity of labeled LPS was determined using a Limulus assay (LAL) kit. Nearly identical bioactivity was observed between unlabeled LPS and 125 I-LPS. LPS uptake into BMMs was performed as described previously ( 9 ). In brief, BMMs were washed with phosphate-buffered saline (PBS) and incubated at 37°C for 30 min or 4°C for 2 h with 125 I-LPS in RPMI medium supplemented with 5% fetal bovine serum. Following incubation, medium was removed and the cells were washed four

Hyperinfl ammatory response in macrophages isolated from SR-BI-null mice and reduced response in macrophages overexpressing SR-BI
To further examine how the presence of SR-BI in macrophages infl uences the macrophage infl ammatory response, primary BMMs isolated from SR-BI-null and WT mice were the genotype of the donor bone marrow cells ( Fig. 1D ). As shown in Table 1 , bone marrow transplantation did not alter plasma lipid levels. Therefore, the different infl ammatory responses seen in the mice receiving different genotypes of bone marrow cells was likely not related to alterations in plasma lipids.  IL-6 and TNF ␣ in response to LPS ( Fig. 2E, F ). Together, these results provide strong evidence that SR-BI expression modulates the infl ammatory response in macrophages. To examine whether the effect of SR-BI on infl ammatory signaling is restricted to LPS-induced signaling, we examined the infl ammatory response to LTA, a ligand that induces cell signaling through its interaction with TLR2. As shown in Fig. 3 , LTA induced TNF ␣ , IL-6, IL-1 ␤ , and MCP-1 gene expression in WT BMMs and to an even greater extent in SR-BI-null BMMs, indicating that the anti-infl ammatory role of SR-BI is not restricted to LPS-induced infl ammation.
The release of pro-infl ammatory cytokines such as TNF ␣ and IL-1 ␤ in response to LPS is mediated by activation of the NF B and MAP kinase pathways, and the hyperinfl ammatory response in SR-BI-null macrophages was shown to involve increased NF B activation ( 11 ). We examined the ERK, P-38, and cJun-JNK signaling pathways. As shown in challenged with LPS. BMMs lacking SR-BI produced elevated levels of the pro-infl ammatory cytokines TNF ␣ and IL-6 compared with SR-BI-expressing BMMs ( Fig. 2A , B ). Consistent with cytokine protein expression, infl ammatory cytokine gene expression was also markedly increased in SR-BI-null BMMs in response to LPS, confi rming an earlier report ( 11 ) ( Fig. 2C ). Interestingly, the anti-infl ammatory gene TGF ␤ , typically associated with the M2 response, was slightly but signifi cantly downregulated in SR-BI-null BMMs.
To substantiate the anti-infl ammatory function of SR-BI observed in BMMs, the infl ammatory response to LPS was determined in J774 macrophages stably overexpressing SR-BI. Stable transfection with an SR-BI-expressing vector resulted in approximately 2-fold increase in SR-BI expression in J774 cells ( Fig. 2D ). In comparison to the parent J774 cells, the SR-BI-overexpressing cells (J774-SR-BI) showed a markedly reduced secretion of the pro-infl ammatory cytokines two genotypes. Baseline levels of P-38 activation were also similar in the two genotypes as observed in similar experiments (data not shown). To investigate the contribution of these pathways to enhanced cytokine production in SR-BI-null cells, specifi c inhibitors were employed. As demonstrated in Fig. 4B , inhibition of ERK activity did not    to LPS in SR-BI-null BMMs, as assessed by TNF ␣ secretion ( Fig. 6A -C ). As expected from earlier studies ( 27 ), cholesterol depletion reduced TNF ␣ expression, and cholesterol replenishment increased TNF ␣ expression. However, the difference between control and SR-BI-null cells was maintained in each condition ( Fig. 6A-C ). These results suggest that the hyperinfl ammatory response observed in SR-BI-null BMMs was not due to altered cellular or membrane cholesterol content and point to a difference in the mechanism by which ABCA1 and SR-BI modulate LPS-induced infl ammatory signaling. One possible mechanism by which SR-BI might exert its inhibitory effect on the response to LPS in macrophages is to bind and internalize LPS. Uptake and clearance of LPS could reduce the level of LPS available to activate TLR4 at the cell surface and therefore indirectly inhibit the infl ammatory response. SR-BI is known to mediate LPS uptake into cells (7)(8)(9). However, SR-BI defi ciency in BMMs did not alter LPS surface binding at 4°C or cellular uptake at 37°C ( Fig. 7A , B ). This indicates that the anti-infl ammatory effect of macrophage SR-BI is likely independent of LPS binding and clearance.

DISCUSSION
Novel fi ndings are reported in this study examining the function of macrophage SR-BI. A bone marrow transplantation experiment showed that SR-BI expression in bone marrow-derived cells reduces the plasma infl ammatory cytokine response to LPS in mice. However, the effect of SR-BI deletion in bone marrow-derived cells was substantially less than the effect of a whole-body knockout of SR-BI. We confi rmed that alter the pro-infl ammatory response in either WT or SR-BI-null BMMs as measured by IL-6 secretion. Successful inhibition of ERK activation in response to LPS by the specifi c inhibitor was confi rmed by Western blotting ( Fig. 4C ). In contrast, inhibition of JNK, P-38, or NF B activity signifi cantly reduced IL-6 production in both WT and SR-BInull cells. Inhibition of P-38 or JNK signaling also partially reduced the difference between the two cell genotypes. These results indicated that both JNK and P-38 signaling pathways contribute to the hyperinfl ammatory response observed in SR-BI null BMMs but do not fully account for the differences in infl ammatory response between the two genotypes.

Cellular cholesterol content and LPS uptake in SR-BI-null macrophages
Recent studies have shown that altered membrane cholesterol content affects the infl ammatory response to LPS in ABCA1-null macrophages ( 20,21 ). To determine whether the enhanced infl ammatory response in SR-BInull BMMs is associated with altered cellular cholesterol content, we examined cellular cholesterol levels in WT control and SR-BI-null cells. As shown in Fig. 5 , no difference was found in either cellular free cholesterol content or the level of cholesterol-rich lipid rafts in the plasma membrane between control and SR-BI-null BMMs ( Fig. 5A-C ). Plasma membrane cholesterol-rich lipid raft content was assessed by the cell association of Alexa-568labeled cholera toxin B, which binds to lipid raft-associated lipid gangliosides ( 20 ). Neither depletion nor subsequent replenishment of cellular cholesterol with methyl-␤ -cyclodextrin normalized the hyperinfl ammatory response between SR-BI-null and WT mice in their response to LPS ( 9 ), relatively smaller differences in circulating TNF ␣ and IL-6 levels were observed in the transplanted mice receiving SR-BI-null versus WT bone marrow.
The role of SR-BI in the LPS-induced infl ammatory response was examined in primary BMMs isolated from WT and SR-BI-null mice. In agreement with a recent study ( 11 ), SR-BI-null BMMs displayed markedly enhanced proinfl ammatory cytokine production. In addition, the antiinfl ammatory cytokine TGF ␤ was reduced in SR-BI-null BMMs. The regulatory role of SR-BI in the pro-infl ammatory response was also confi rmed by a gain-of-function approach in which SR-BI overexpression in J774 cells markedly reduced the pro-infl ammatory response. These results provide strong evidence that SR-BI is an important modulator of the macrophage infl ammatory response.
A previous study had shown that the hyperinfl ammatory response in SR-BI-null macrophages involved increased NF B activation ( 11 ). In the present study, we also examined the potential involvement of the MAP kinase-signaling cascade in the SR-BI-regulated macrophage infl ammatory response. We demonstrated enhanced phosphorylation of ERK, P38, and JNK signaling molecules in SR-BI-null BMMs, which implicated altered signaling through these alternative pathways in the regulation of cytokine production by SR-BI. The use of specifi c pathway inhibitors targeting ERK, P38, JNK, and NF B identifi ed JNK-and P38-mediated signaling, but not ERK, as additional key mediators responsible for SR-BI in cultured macrophages regulates the infl ammatory cytokine response to LPS by showing that the response is affected by both SR-BI ablation as well as overexpression in cells. The LPS-induced cytokine expression in both WT and SR-BI-null cells involved JNK and P38 cell signaling pathways in addition to NF B. Although the underlying mechanism of SR-BI modulation of the infl ammatory response in macrophages is not yet fully understood, modulation was not associated with altered LPS uptake or cellular cholesterol levels.
In a previous study, we showed that both adrenal insuffi ciency and reduced hepatic LPS clearance contribute to the hypersensitivity to LPS in SR-BI-null mice ( 9 ). To assess the contribution of macrophage SR-BI to the general infl ammatory response to LPS in mice, we performed bone marrow transplantation experiments in the current study. One advantage of this approach was that altered SR-BI expression in bone marrow-derived cells did not affect plasma lipoproteins or plasma corticosterone levels, alterations of which would exert signifi cant effects on the general infl ammatory response. Signifi cantly elevated infl ammatory cytokine levels were observed in WT mice receiving SR-BI-null bone marrow cells compared with WT cells in response to LPS, whereas reduced cytokine expression was found in SR-BI-null mice receiving WT bone marrow cells compared with SR-BI-null cells. These results demonstrate that SR-BI expression in bone marrow-derived cells attenuates infl ammatory cytokine production in mice. Note that compared with the dramatic difference  cellular cholesterol content. However, the unaltered cellular cholesterol and membrane cholesterol-rich raft content observed in SR-BI-null macrophages suggests that SR-BI, unlike ABCA1, may modulate infl ammatory responses independently of changes in macrophage cholesterol content.
Our fi ndings may shed light on understanding the mechanism of infl uence of SR-BI in atherosclerotic lesion formation. The protective effect of SR-BI has been mainly attributed to its role in the reverse cholesterol transport pathway ( 39 ). However, liver-specifi c SR-BI-null mice exhibit less atherosclerotic lesion formation than SR-BI whole-body-defi cient mice ( 40 ), suggesting an extra-hepatic protective effect(s) of SR-BI. Studies of SR-BI-null mice overexpressing cholesteryl ester transfer protein have also indicated that SR-BI plays an atheroprotective role that is unrelated to its role in mediating plasma HDL cholesterol clearance ( 41 ). Despite having normalized plasma HDL levels, SR-BI-null × CETP tg mice were not protected ( 41 ) or only partially protected ( 42 ) against atherosclerotic lesion formation compared with SR-BI-null mice. Given the fact that macrophage infl ammation is a key feature in the development of atherosclerosis, our fi ndings provide evidence that SR-BI may exert a protective effect against atherosclerotic lesion formation through modulating macrophage infl ammatory responses.
This study establishes that SR-BI in macrophages serves to regulate the infl ammatory response to LPS, an effect that is distinct from its known protective role in the adrenals and liver. The anti-infl ammatory function of SR-BI appears to be mediated through a dampening of LPS-induced signaling through the MAP kinase and NF B pathways, resulting in reduced infl ammatory cytokine production. Our results therefore strongly suggest that SR-BI may play a key role in limiting pathological tissue infl ammation and the macrophage response to pro-infl ammatory stimuli, such as LPS. A more defi ned understanding of macrophage SR-BI may contribute to improved strategies for the treatment of acute and chronic infl ammatory disease, such as sepsis and atherosclerosis. enhanced LPS-induced cytokine production in macrophages that also may partially contribute to the difference in infl ammatory responses between WT and SR-BI-null cells.
The mechanism(s) responsible for the enhanced infl ammatory signaling in SR-BI Ϫ / Ϫ macrophages is not known. One possibility is that an interaction between SR-BI and the TLR4 complex might modulate TLR4 activation and subsequent intracellular signaling pathways. Such an interaction between TLRs and CD36, a class B scavenger receptor that is closely related to SR-BI, has been shown to regulate TLR signaling (28)(29)(30). As a coreceptor for TLR2, CD36 facilitates the TLR2mediated NF B activation through binding and clustering of TLR2 ligands, such as LTA ( 28,29 ). CD36 was also shown to impact TLR4 signaling. CD36-dependent recognition of endogenous ligands is essential for stimulating the dimerization and activation of TLR4/TLR6, a complex required to initiate the innate immune activation of macrophages and microglia resulting in the activation of NF B signaling ( 30 ). Possible interaction of TLR4 with SR-BI, shown in the current study to attenuate the infl ammatory response to LPS, remains to be determined. In contrast to CD36, scavenger receptor A can attenuate TLR4-induced pro-infl ammatory cytokine production by mediating rapid internalization and clearance of extracellular LPS, thus reducing LPS-induced TLR4 activation and infl ammatory signaling ( 31 ). In the current study, LPS binding and cell association was not altered in SR-BI-null macrophages, suggesting that SR-BI does not modify TLR4 activation by affecting LPS clearance.
Alternatively, SR-BI may function as an independent signaling receptor capable of modulating intracellular signaling pathways upon ligand association as previously demonstrated for SR-BI expressed in nonmacrophage cell types ( 32,33 ). For example, in endothelial cells, HDL stimulates the activation of e-NOS and Src family kinase(s) via SR-BI ( 34,35 ). Src activation leads to the activation of phosphatidylinositol 3-kinase (PI3K), Akt kinase, and MAP kinase pathways ( 32,36 ). In addition, interaction of HDL and SR-BI in endothelial cells inhibits the activation of NF B and subsequent expression of the adhesion molecules vascular cell adhesion molecule 1 and intercellular adhesion molecule 1 to infl ammatory stimuli ( 37 ). However, TLR4-independent SR-BI signaling in macrophages and the implication for its role in infl ammation has yet to be examined.
In the current study, we addressed a third possible mechanism in which SR-BI impacts infl ammatory signaling through its ability to facilitate free cholesterol fl ux between cells and lipoproteins and to infl uence the levels and distribution of plasma membrane cholesterol. Macrophage lipid content, especially the membrane cholesterol distribution, has been shown to regulate infl ammatory responses in macrophages ( 20,21 ). Free cholesterol accumulation in membranes, as observed in ABCA1 or ABCG1-null macrophages as a result of defective cholesterol effl ux, stimulates the infl ammatory response to LPS in a MyD88-dependent pathway ( 20,21 ). Free cholesterol loading, especially in the endoplasmic reticulum, induces NF B and MAPK stress signaling in response to LPS, promoting IL-6 and TNF ␣ production ( 38 ). We therefore investigated whether macrophage SR-BI functions to regulate the infl ammatory response by modulating