The Oxysterol 24(S),25-Epoxycholesterol Attenuates Human Smooth Muscle–Derived Foam Cell Formation Via Reduced Low-Density Lipoprotein Uptake and Enhanced Cholesterol Efflux

Background Foam cell formation by intimal smooth muscle cells (SMCs) inhibits the elaboration of extracellular matrix, which is detrimental to plaque stabilization. In the present study, we examined the lipoproteins and receptors involved in human SMC foam cell formation and investigated the ability of 24(S),25-epoxycholesterol [24(S),25-EC], an oxysterol agonist of the liver X receptor, to attenuate SMC foam cell formation. Methods and Results Incubation of human internal thoracic SMCs with atherogenic lipoproteins demonstrated that low-density lipoprotein (LDL), but not oxidized or acetylated LDL, was the primary lipoprotein taken up, resulting in marked cholesteryl ester deposition (6-fold vs 1.8-fold; P<0.05; n=4). Exposure of SMCs to exogenous or endogenously synthesized 24(S),25-EC attenuated LDL uptake (−90% and −47% respectively; P<0.05; n=3) through decreased sterol regulatory element–binding protein-2 expression (−30% and −17%, respectively; P<0.001; n=3), decreased LDL receptor expression (−75% and −40%, respectively; P<0.05; n=3) and increased liver X receptor–mediated myosin regulatory light chain interacting protein expression (7- and 3-fold, respectively; P<0.05; n=4). Furthermore, exogenous 24(S),25-EC increased adenosine triphosphate–binding cassettes A1– and G1–mediated cholesterol efflux to apolipoprotein AI (1.9-fold; P<0.001; n=5) and high-density lipoprotein3 (1.3-fold; P<0.05; n=5). 24(S),25-EC, unlike a nonsteroidal liver X receptor agonist, T0901317, did not stimulate sterol regulatory element–binding protein-1c–mediated fatty acid synthesis or triglyceride accumulation. 24(S),25-EC preserved the assembly of fibronectin and type I collagen by SMCs. Conclusions The oxysterol 24(S),25-EC prevented foam cell formation in human SMCs by attenuation of LDL receptor–mediated LDL uptake and stimulation of cholesterol efflux, restoring the elaboration of extracellular matrix. In contrast to T0901317, 24(S),25-EC prevented the development of a triglyceride-rich foam cell phenotype. (J Am Heart Assoc. 2012;1:e000810 doi: 10.1161/JAHA.112.000810.)

H uman smooth muscle cells (SMCs) play an integral role in atherogenesis. 1 SMCs are stimulated to migrate and proliferate within the intima in response to cytokines and growth factors. 1,2 SMCs in culture can accumulate lipoprotein-derived lipids and adopt a foam cell-like phenotype. 3,4 Furthermore, atherosclerotic lesions contain abundant foam cells of SMC origin. 1 A vital role of vascular SMCs is to stabilize lesions through the elaboration of collagen fibrils, which protects from plaque rupture. 1 However, SMCs that assume the foam cell state are potentially detrimental to plaque stability. Recently, we reported that lipid accumulation renders cultured human SMCs incapable of efficiently assembling a fibrillar extracellular matrix. 5 Therefore, elucidation of potential mechanisms for the prevention of SMC foam cell formation has implications for plaque stability and ultimately for clinical events.
There is still uncertainty as to which atherogenic lipoproteins stimulate human SMC foam cell formation and what mechanisms are involved in their uptake. Furthermore, SMCs display phenotypic diversity, 6 and evidence has emerged that human SMC subtypes demonstrate distinct responses to atherogenic lipoproteins. 5 In contrast to macrophages, rabbit aortic SMC foam cell formation was induced primarily by uptake of native low-density lipoprotein (LDL) but not modified LDL. 3,4 We previously demonstrated that human SMCs isolated from the internal thoracic artery readily accumulated lipids when exposed to human LDL and very-low-density with human LDL or VLDL. Fluorescence micrographs of HITC6 human vascular SMCs incubated for 24 h in M199 media containing 0.4% FBS followed by a 24-h incubation with human LDL (150 μg cholesterol / mL media) or human VLDL (50 μg cholesterol / mL media). Paraformaldehyde-fixed cells were mounted in PermaFluor (Thermo Electron Corporation) containing Hoechst 33258 and were stained with boron-dipyrromethene 493/503 to identify droplets of neutral lipid. The images illustrate abundant cytoplasmic lipid droplets in LDLand VLDL-exposed SMCs. Bar=25 μm. Images are representative of 3 separate experiments. lipoprotein (VLDL) but resisted lipid accumulation when exposed to modified lipoproteins. 7 Whether this difference in lipoprotein uptake was related to differential expression of lipoprotein receptors has not been established. Furthermore, the larger, slower-growing, spindle-shaped SMC clone accumulated more lipid in response to lipoprotein exposure than did the smaller, faster-growing, epithelioid-shaped SMC clone, 7 which suggests that the former cell type represents the predominant SMC foam cell precursor.
The liver X receptor (LXR) is an important regulator of cholesterol homeostasis. 8 In macrophages, LXR activation by endogenous agonists, including oxysterols, or by exogenous synthetic nonsteroidal compounds, such as T0901317, activated the transcription of genes involved in cholesterol efflux, including adenosine triphosphate-binding cassette (ABC) A1 and ABCG1. 8 T0901317 decreased aortic lesions in mouse models of atherosclerosis, mediated in part by attenuation of macrophage foam cell formation through enhanced cholesterol efflux. 9,10 Given the abundance of SMC-derived foam cells in atherosclerotic lesions, 1 LXR agonist-stimulated cholesterol efflux in SMCs could also contribute to atheroprotection. Characterization of LXR activation in SMCs remains incomplete. Although T0901317 stimulated ABCA1-and ABCG1-mediated cholesterol efflux in human SMCs, 11,12 the impact of LXR activation on atherogenic lipoprotein uptake and lipid deposition has not been explored. Triglyceride has also been shown to contribute to foam cell formation, particularly in the setting of elevated plasma VLDL and free fatty acids (FA), a profile typical of the metabolic syndrome and diabetes. 13 An adverse effect of nonsteroidal LXR agonists, including T0901317, is increased triglyceride accumulation, induced by transactivation of genes involved in FA synthesis, including sterol regulatory element-binding protein (SREBP)-1c and fatty acid synthase (FASN). 14 In human SMCs, although T0901317 induced cholesterol efflux, 11 it stimulated SREBP-1c-induced lipogenesis and triglyceride accumulation in cells exposed to adipogeneic differentiation medium. 13 In macrophages, natural oxysterol LXR ligands such as 24(S),25-epoxycholesterol [24(S),25-EC], added either exogenously or synthesized endogenously, selectively activated ABCA1-and ABCG1-mediated cholesterol efflux. 15,16 In contrast to T0901317, 24(S),25-EC or another oxysterol-like LXR agonist (DMHCA; N,N-dimethyl-3β-hydroxy-cholenamide) had little effect on SREBP-1c-or FASN-stimulated lipogenesis or triglyceride accumulation. [15][16][17] Selectivity of 24(S),25-EC was attributed to inhibition of precursor (p)SREBP-1c processing to its active nuclear form (nSREBP-1c) by the oxysterol but not by T0901317, thus preventing lipogenic gene expression. 15,16 Oxysterols also disrupt the maturation of SREBP-2. SREBP-2 stimulated the expression of genes involved in de novo cholesterol synthesis and lipoprotein uptake, including 3-hydroxy-3methylglutaryl-coenzyme A reductase (HMGCR) and the LDL receptor (LDLR), respectively. 18 In Chinese hamster ovary cells, 24(S),25-EC (like other oxysterols) blocked the maturation of pSREBP-2, similar to the effect observed for SREBP-1c. 18,19 Whether 24(S),25-EC attenuates cholesterol accumulation in SMCs through SREBP-2-regulated gene expression has not been reported.
The objectives of the present study were 2-fold: (1) to identify the receptors responsible for lipoprotein uptake in human vascular SMCs, and (2) to examine the impact of 24(S),25-EC on SMC foam cell formation and the elaboration of extracellular matrix. We studied a clonal population of SMCs-namely, human internal thoracic (HIT) C6, which adopts a foam cell phenotype after exposure to atherogenic lipoproteins. 7 We hypothesized that exogenous 24(S),25-EC or endogenously synthesized 24(S),25-EC attenuates lipoprotein-induced cholesteryl ester (CE) accumulation in SMCs, without altering triglyceride synthesis.
We demonstrate that LDL, but not modified LDL, preferentially induced SMC foam cell formation. Exogenous or endogenously synthesized 24(S),25-EC reduced SMC uptake of native LDL by inhibiting SREBP-2-mediated LDLR expression. Concurrently, 24(S),25-EC activated ABCA1-and ABCG1-mediated cholesterol efflux but did not stimulate SREBP-1c-induced lipogenesis or triglyceride accumulation. These experiments reveal a dual role for 24(S),25-EC in attenuating CE accumulation in human SMCs: reduced LDL uptake and enhanced cholesterol efflux, resulting in preservation of SMC assembly of extracellular matrix.

Cell Culture
The HIT-SMC clone used in this study, C6, is a human clonal line derived from the media of distal internal thoracic artery,  24(S),25-EC is a potent LXR ligand and is uniquely derived from a shunt in the cholesterol biosynthetic pathway (reviewed in Huff et al 21 ). 2,3-Oxidosqualene:lanosterol cyclase (OSC), a cholesterol biosynthesis enzyme downstream of 3-hydroxy-3-methylglutaryl-coenzyme A reductase, catalyzes the conversion of 2,3monoepoxysqualene to lanosterol, the dedicated step in cholesterol biosynthesis. If 2,3-monoepoxysqualene builds up, it is converted to 2,3;22,23diepoxysqualene by squalene epoxidase. However, OSC also catalyzes cyclization of 2,3;22,23-diepoxysqualene to 24(S),25-epoxylanosterol, which is subsequently transformed into 24(S),25-EC. Synthesis of 24(S), 25-EC is favored over cholesterol under conditions of partial OSC inhibition because of the higher affinity of OSC for 2,3;22,23-diepoxysqualene than for 2,3-monoepoxysqualene. Greater inhibition of OSC eventually results in the inhibition of both 24(S),25-EC and cholesterol. 21 Lipoprotein Isolation VLDL, LDL, and high-density lipoprotein 3 (HDL 3 ) were isolated from plasma of human subjects recruited from the Lipid Clinic at the London Health Sciences Centre, University Campus (London, Ontario, Canada). This study was approved by the University of Western Ontario Institutional Review Board (protocol No. 15685). Lipoproteins were separated by differential ultracentrifugation as described previously. 22 LDL was oxidized (oxLDL) via the copper sulfate method 22 or was acetylated (acLDL) by using acetic anhydride, 23 and the extent of modification was confirmed by alterations in electrophoretic mobility. 22

Cellular Lipid Mass
Lipid mass was determined in SMCs preincubated in media plus 0.4% FBS for 24 h (6-well [35-mm] plates), followed by a further 24-h incubation in the same media with vehicle, LXR agonists, or OSCi (compounds). Cells were then incubated in fresh M199 media plus 0.4% FBS with or without compounds and with or without 150 μg/mL (micrograms lipoprotein cholesterol per milliliter media) of native LDL, acLDL, or oxLDL or 50 μg/mL (micrograms lipoprotein cholesterol per milliliter media) VLDL plus 0.1 U/mL of bovine lipoprotein lipase. 7 Lipids were extracted with hexane:isopropanol (3:2 volume:volume) and processed for lipid quantification as described previously. 22 Cell protein was assayed after solubilization in 0.1 N sodium hydroxide. Total cholesterol (TC), free cholesterol (Wako Chemicals, Richmond, VA), and triglyceride (Roche Diagnostics, Inc, Indianapolis, IN) were determined enzymatically by using commercial reagents as described previously. 22 CE mass was calculated by subtracting free cholesterol from TC. Values were normalized to total cell protein.
Cholesterol Esterification and Cholesterol, 24(S),25-EC, Fatty Acid, and Triglyceride Synthesis Confluent (80%) SMCs were preincubated in 6-well plates for 24 h in M199 media plus 0.4% FBS and then in the same media with or without compounds for a further 24 h. Subsequently, lipoproteins (LDL and acLDL, 150 μg/mL [micrograms lipoprotein cholesterol per milliliter media]; VLDL, 50 μg/mL [micrograms lipoprotein cholesterol per milliliter media]) plus radiolabeled [1-14 C]-oleic acid complexed with FA-free bovine serum albumin (FAF:BSA) were added for 5 h in fresh M199 media plus 0.4% FBS, with or without compounds. Lipids were extracted (as described above) and separated via thin-layer chromatography, and radiolabel incorporation into CE and triglyceride was determined according to previously described methods. 16,24 The synthesis of cholesterol and FA was determined in SMCs preincubated 24 h in media plus 0.4% FBS and then with vehicle, 24(S),25-EC, OSCi, or T0901317 for a further 24 h. [1-14 C]-acetate with or without compounds in fresh M199 media plus 0.4% FBS was added for an additional 5 h. 16 The synthesis of 24(S),25-EC was determined in SMCs preincubated for 24 h in M199 0.4% FBS media followed by a 24-h incubation with the OSCi (0 to 100 nmol/L) and [1-14 C]-acetate in fresh M199 media plus 0.4% FBS. Lipid extraction, saponification, thin-layer chromatography separation, and the amount of radiolabel incorporation into cholesterol, 24(S),25-EC, and FA were determined as described previously. 16 Values were normalized to total cell protein.

Lipoprotein Uptake
LDL and acLDL uptake was determined by using fluorescently labeled native and modified 1,1'-dioctadecyl-3,3,3',3'-tetramethylindocarbocyanine perchlorate (DiI)-labeled lipoproteins (Biomedical Technologies, Inc, Stoughton, MA). Confluent (80%) SMCs in 6-well (35-mm) plates were preincubated in M199 media plus 0.4% FBS for 24 h and then were incubated with or without the addition of vehicle or compounds for a further 24 h. Subsequently, cells were incubated in fresh media and compounds in the presence or absence of 10 μg/mL DiI-LDL or DiI-acLDL for 5 h. Cells were trypsinized and resuspended in phosphate-buffered saline for analysis by flow cytometry. SMCs were excited in the flow chamber of a 2-laser BD FACSCalibur (BD Biosciences, Franklin Lakes, NJ) using the 15-mW, 488-nm, air-cooled, argon-ion laser, and emission spectra were detected in the FL2 (PE/PI) channel (585/42 band-pass filter). Data were collected and analyzed with BD Cell Quest Pro Software.

Cholesterol Efflux
Cholesterol efflux was determined according to methods described previously. 15,16 SMCs (80% confluent) were initially incubated in M199 media containing 0.4% FBS, followed by incubation with fresh M199 media containing FAF:BSA (0.2%) plus the addition of acLDL (5 μg TC) and 1 μCi of [ 3 H]-cholesterol (Amersham Biosciences, Oakville, ON) per 1 mL of 0.2% FAF:BSA media to label endogenous cholesterol pools, for 24 h. Cells were then washed and incubated with fresh M199 0.2% FAF:BSA media with or without compounds for 24 h, followed by incubation with fresh M199 0.2% FAF:BSA media with or without compounds and with or without apolipoprotein (apo) A1 (10 μg/mL) or HDL 3 (100 μg/mL) for 16 h. Radioactivity was measured in aliquots of media and cell lysate by using a scintillation counter and was expressed as a percent of total radioactivity per well (media plus cell lysate). The percentage of cholesterol effluxed was normalized to total cell protein.

Real-Time Quantitative Reverse Transcription Polymerase Chain Reaction
Total RNA from confluent (80%) SMCs was isolated by using Trizol reagent (Invitrogen, Burlington, ON) after 24-h preincubation in 100-mm dishes in M199 media plus 0.4% FBS, followed by a further 24 h in the same media with or without compounds, as described above. RNA (2 μg) was reverse-transcribed as described previously, 16  For each gene, the standard curve method was used to determine mRNA abundance, which was normalized to the abundance of glyceraldehyde 3-phosphate dehydrogenase (GAPDH). Quantities of mRNA for a specific gene were interpolated from a standard curve plotted as cycle threshold versus the log of the quantity of serial dilutions of a known cDNA standard. This method also allowed for the comparison of basal expression of lipoprotein receptors relative to the LDLR. Primer and probe sets for each gene were obtained from Taqman Assays-on-Demand (Applied Biosystems), with the exception of SREBP-1c, which was designed as described previously to differentiate the 2 novel exon 1 variants generated from alternative start sites. 16 Immunoblotting SMCs (80% confluent) were plated in 6-well (35-mm) plates in triplicate, were preincubated in M199 media plus 0.4% FBS, and subsequently were incubated for a further 24 h in the same media with or without compounds. SMC lysates (10 mmol/L Tris-HCl, 10 mmol/L NaCl, 3 mmol/L MgCl 2 , 0.5% Igepal, anti-SREBP-1 (Neomarkers, Fremont, CA), followed by the appropriate secondary antibody conjugated to horseradish peroxidase (Santa Cruz Biotechnology, Inc, Santa Cruz, CA) for development with chemiluminescence substrate. Membranes were stripped and reprobed with a polyclonal anti-β-actin antibody (Cell Signaling, Danvers, MA) for postnuclear fractions and Lamin A/C (Santa Cruz Biotechnology, Inc, Santa Cruz, CA) for nuclear fractions for use as loading controls. Band intensities were quantified with an imaging densitometer as described previously. 16

Statistical Analysis
Each experimental method was performed with biological replicates of 3 to 5 (n=3-5), with 3 to 4 technical replicates within each biological replicate. Graphical results are presented as the mean ± standard error of the mean (SEM). The Shapiro-Wilk normality test was used to test for parametric distributions in each data set. P values for observed differences between treatment and control groups were calculated with the Student t test for parametric data or the Mann-Whitney U test for nonparametric data. Statistically significant observations were defined by a 2-tailed threshold of P<0.05. Statistical analyses were performed with SigmaPlot 11.0 software (Systat, Inc, San Jose, CA).

T0901317 but Not 24(S), 25-EC or OSCi Stimulates
SREBP-1c Expression, SREBP-1 Processing, and Lipogenesis in SMCs SMCs incubated with T0901317 and adipocyte differentiation media stimulate SREBP-1c-mediated FA synthesis, resulting in a foam cell phenotype. 13 In human HITC6 SMCs, incubation with LDL, acLDL, or oxLDL did not affect triglyceride mass, but incubation of cells with VLDL increased triglyceride mass >8.0-fold (P=0.01) ( Figure 10A through 10C and Figure  1). 24(S),25-EC or the OSCi did not affect triglyceride in SMCs incubated with or without lipoproteins ( Figure  10A and 10C). In contrast, T0901317 increased triglyceride mass up to 1.5-fold in the absence or presence of LDL, acLDL, or oxLDL (P<0.05 for each) ( Figure 10B) and increased triglyceride mass to an even greater extent (60%; P=0.023) than the 8-fold induction by VLDL alone ( Figure 10B).

24(S),25-EC Restores SMC Assembly of Fibonectin and Collagen Fibrils
We assessed whether SMC-dependent functions were retained by 24(S),25-EC in SMCs exposed to LDL. Elaboration of extracellular matrix was evaluated by examining microscopically the assembly of fibronectin and type I collagen fibrils in SMCs after the addition of labeled soluble precursors of either fibronectin or collagen. SMCs not exposed to LDL assembled an elaborate network of both fibronectin and collagen fibrils (Figure 12). Incubation of SMCs with LDL resulted in a marked decrease in their ability to assemble these fibrils, commensurate with increased Oil Red O-stained lipid, as we reported previously. 5 In contrast, preincubation of cells with 24(S),25-EC, followed by the addition of LDL, decreased Oil Red O-stained lipid and preserved the ability of SMCs to assemble both fibronectin and collagen.

Discussion
Human SMC-derived foam cells promote atherosclerosis; they are observed within the musculoelastic layer at early stages and within the intimal fibrotic layer of intermediate and advanced lesions. 1 Our results revealed that in human SMCs, native LDL was the predominant lipoprotein responsible for the induction of CE accumulation, whereas uptake of modified LDL was limited. This difference correlated with greater LDLR expression than scavenger receptor expression. Furthermore, we show for the first time that 24(S),25-EC attenuates CE accumulation in SMCs challenged with human LDL. Reduced LDL uptake by 24(S),25-EC was linked to inhibition of SREBP-2-mediated LDLR expression. 24(S),25-EC also increased MYLIP expression, which contributed to the decrease in LDLR proteinand LDL uptake. 24(S),25-EC stimulated cholesterol efflux via activation of ABCA1 and ABCG1, further reducing intracellular cholesterol. The impact of endogenously synthesized 24(S),25-EC was qualitatively similar. Importantly, with exposure of SMCs to 24(S),25-EC, in contrast to the nonsteroidal LXR agonist T0901317, cholesterol efflux was selectively increased without inducing a triglyceride-rich foam cell phenotype.
Recently, we demonstrated that human SMCs display impaired assembly of type I collagen and fibronectin after lipid accumulation induced by LDL or VLDL. 5 These findings directly link SMC foam cell formation with failure to elaborate extracellular matrix, potentially accounting for reduced plaque stability. These human SMCs, the same clone as those used in the present study, maintained expression of SMC-specific markers upon lipid loading. 5 We now show that attenuated LDL-induced CE accumulation and enhanced cholesterol efflux by 24(S),25-EC preserves the assembly of fibrillar collagen and fibronectin, a characteristic of a more productive SMC phenotype.
CE accumulation in SMCs challenged with acLDL or oxLDL was substantially lower than in those challenged with LDL (1.5-versus 6-fold; P<0.05). These human SMCs expressed more LDLR (>40-fold higher) than CD36 and SRAI/II, consistent with previous reports that scavenger receptor expression was undetectable in intimal or medial SMCs within human atherosclerotic lesions. 28 In cultured rabbit aortic SMCs, uptake of modified lipoproteins was low, whereas CE accumulation wasderived from native LDL. 4 Our results provide direct evidence that CE accumulation in human SMCs is primarily a consequence of LDL uptake. This differs from macrophages, in which modified lipoproteins primarily drive foam cell formation. 29 Furthermore, our results extend the concept that VLDL-induced triglyceride accumulation in SMCs results in a foam cell-like phenotype. 5,7 Expression of LDLR, PCSK9, and HMGCR is primarily regulated by the nuclear form of SREBP-2. 18, 30 We demonstrated that SMCs exposed to exogenous or endogenous 24(S),25-EC contained less SREBP-2 and LDLR mRNA, consistent with diminished LDLR expression in macrophages 15 and fibroblasts 31 incubated with 24(S),25-EC. The mechanism involved oxysterol-mediated disruption of pSREBP-2 maturation, preventing LDLR expression and the feed-forward amplification of SREBP-2 transcription. 19,31,32 The present study is the first to report that 24(S),25-EC in human SMCs decreased expression of SREBP-2-responsive genes, SREBP-2 and LDLR, and increased the expression of MYLIP, resulting in blunted LDLR expression and the attenuation of LDL uptake and CE accumulation.
The SREBP-2-mediated expression of PCSK9 32 was also decreased by 24(S),25-EC in SMCs. As PCSK9 targets the LDLR for lysosomal degradation, 26,30 decreased PCSK9 expression would be expected to impede LDLR degradation. 26,30 The ability of 24(S),25-EC to also decrease SREBP-2-mediated LDLR mRNA when combined with the increased LXR-stimulated MYLIP expression, which would promote LDLR degradation, 25 resulted in attenuation of LDLR protein and LDL uptake. 24(S),25-EC also inhibited HMGCR expression, providing further evidence that 24(S),25-EC prevented pSREBP-2 maturation. 31 As oxysterols are known to increase HMGCR enzyme degradation, 33 the reduced transcription and increased degradation of HMGCR provided a plausible mechanism for the marked reduction of cholesterol synthesis in 24(S),25-EC-treated SMCs.
In contrast to 24(S),25-EC, T0901317 increased SREBP-2, LDLR, PCSK9, and HMGCR expression, which led to enhanced cholesterol synthesis yet a significant decrease in LDL uptake (P<0.001). T0901317, a nonsteroid, did not inhibit pSREBP-2 maturation and thus stimulated expression of SREBP-2responsive genes. Enhanced LDLR and PCSK9 expression have been observed in T0901317-treated hepatocytes. 25,34 Although the LDLR promoter contains an LXR response element, 35 expression is significantly more responsive to sterol response element activation, which provides an explanation for increased LDLR expression by T0901317 but not by 24(S),25-EC. In the present study, T0901317 markedly induced SREBP-1c expression and nSREBP-1 protein, similar to a previous report in SMCs. 13 In contrast to 24(S),25-EC, the T0901317induced overabundance of nSREBP-1 likely increased nonselective binding to SREs and activation of both SREBP-1c-and SREBP-2-target genes. This has been demonstrated in T0901317treated Huh7 cells 35 and in tissues in which nSREBPs are overexpressed. 36 PCSK9 expression is primarily regulated by SREBP-2; however, a recent report revealed marked SREBP-1c stimulation of PCSK9 mRNA in T0901317-treated hepatoctyes. 32,34 Although an LXR response element has not been identified in the PCSK9 promoter, the increase we observed in PCSK9 mRNA in T0901317treated SMCs but not in 24(S),25-EC-treated SMCs indicated that PCSK9 expression was highly responsive to nSREBP-1c. Despite increased LDLR mRNA by T0901317, the ability of T0901317 to also increase the expression of both MYLIP 25 and PCSK9 resulted in diminished LDLR protein and decreased LDL uptake.
Activation of ABCA1 and ABCG1 and cholesterol efflux by synthetic LXR agonists has been documented in rat aortic SMCs 27 and human coronary 12 and airway 11 SMCs. Furthermore, T0901317-induced cholesterol efflux in human airway SMCs was mediated exclusively by ABCA1. 11 In the present study, we demonstrate in HITC6 SMCs that exogenous 24(S),25-EC increased LXRα, ABCA1, and ABCG1 expression, which led to enhanced cholesterol efflux. These observations are consistent with our previous studies in macrophages. 15,16 Furthermore, we show that 24(S),25-EC-mediated cholesterol efflux contributed to the attenuation of SMC foam cells. Endogenous synthesis of 24(S),25-EC, induced by partial OSC inhibition, increased ABCA1 and ABCG1 expression to levels ≈20% of those in cells exposed to exogenous 24(S),25-EC yet did not increase cholesterol efflux. Therefore, although SMCs synthesize endogenous 24(S),25-EC, LXR-induced gene expression was likely below the threshold for facilitating cholesterol removal.
LXR has been shown to regulate de novo FA biosynthesis through expression of SREBP-1c and FASN. 14 T0901317 stimulated FA synthesis and triglyceride accumulation in cells and mice. 9,15,16 In human SMCs, T0901317 increased de novo FA synthesis when incubated in adipocyte differentiation media. 13 Increased lipogenesis appears to be exclusive to nonsteroidal LXR agonists such as T0901317, because oxysterols such as 24(S),25-EC and 22(R)-OH did not affect lipogenesis in macrophages. 15,16 Our results clearly show that in human SMCs, 24(S),25-EC did not stimulate SREBP-1c or FASN mRNA and did not increase nSREBP-1 protein or synthesis of FA and triglyceride. Furthermore, triglyceride accumulation in SMCs exposed to lipoproteins was unaffected, whereas T0901317 stimulated FA synthesis and amplified triglyceride accumulation, especially in SMCs incubated with VLDL. This is consistent with a mechanism in which 24(S),25-EC, but not T0901317, retains pSREBP-1c within the endoplasmic reticulum, preventing nSREBP-1 formation and nSREBP-1c-stimulated gene expression, including the feed-forward activation of SREBP-1c itself. 37 Thus, 24(S),25-EC acts as a safety valve to diminish cellular cholesterol and prevent excessive FA synthesis and triglyceride storage. In recent studies, synthetic oxysterol-like LXR agonists reduced atherosclerosis in mice but, like steroidal agonists, did not induce FA synthesis. 38 Therefore, the present study in SMCs and our previous studies in macrophages 15,16 demonstrate that 24(S),25-EC has the potential to selectively activate LXR-regulated cholesterol efflux from cells involved in atherogenesis, without inducing lipogenesis.
SMCs participate in plaque development and have the propensity to develop into foam cells through acquisition of lipoprotein-derived lipids. Controlling cholesterol homeostasis in SMCs represents an attractive mechanism to maintain SMCs in a reparative phenotype. Exposure of human SMCs to the oxysterol LXR agonist 24(S),25-EC reduced native LDL uptake, the primary mechanism for SMC foam cell formation. Furthermore, 24(S),25-EC promoted cholesterol efflux without inducing lipogenesis. Diversion of SMCs from a foam cell state restored their ability to elaborate extracellular matrix, which has the potential to enhance lesion stabilization.