Tissue-Specific Liver X Receptor Activation Promotes Macrophage Reverse Cholesterol Transport In Vivo
Objective— We previously reported that a systemic liver X receptor (LXR) agonist promoted macrophage reverse-cholesterol transport (mRCT) in vivo. Because LXR are expressed in multiple tissues involved in RCT (macrophages, liver, intestine), we analyzed the effect of tissue-specific LXR agonism on mRCT.
Methods and Results— In initial studies, the systemic LXR agonist GW3965 failed to promote mRCT in a setting in which LXR was expressed in macrophages but not in liver or intestine. To evaluate the effect of LXR activation specifically in small intestine on mRCT, wild-type mice were treated with either intestinal-specific LXR agonist (GW6340) or systemic LXR agonist (GW3965). Both GW3965 and GW6340 significantly promoted excretion of [3H]-sterol in feces by 162% and 52%, respectively. To evaluate the requirement for macrophage LXR activation, we assessed the ability of GW3965 to promote mRCT in wild-type mice using primary macrophages deficient in LXRα/β vs wild-type macrophages. Whereas GW3965 treatment promoted fecal excretion compared with vehicle, its overall ability to promote mRCT was significantly attenuated using LXRα/β knockout macrophages.
Conclusion— We demonstrate that intestinal-specific LXR agonism promotes macrophage RCT in vivo and that macrophage LXR itself plays an important, but not predominant, role in promoting RCT in response to an LXR agonist.
Reverse-cholesterol transport (RCT) is a physiological process that helps to prevent accumulation of excess cholesterol in peripheral tissues.1 Efflux of cholesterol from vessel wall macrophages is thought to be a critical first step by which the RCT pathway protects against atherosclerosis.2 Liver X receptors (LXR) are nuclear receptors activated by endogenous oxysterols that play an important role in the control of cellular and whole-body cholesterol homeostasis.3 The main target tissues of LXR agonists relevant to RCT are macrophages, liver, and small intestine. LXR agonists increase ATP-binding cassette transporters A1 (ABCA1) and G1 (ABCG1) expression in macrophages, resulting in increased cholesterol efflux,4 ABCG5/83 and cholesterol 7 α-hydroxylase (CYP7A1)5 expression in liver,5,6 resulting in increased sterol excretion into bile, and ABCG5/8 expression in small intestine,5 resulting in increased cholesterol excretion in feces. We developed an assay that specifically traces RCT from the macrophage to the feces in vivo1 and previously reported that a systemic LXR agonist promoted macrophage RCT in vivo.7 However, the relative roles of LXR agonism in macrophages, liver, and intestine in RCT are not known. Therefore, we tested the effects of tissue-specific LXR agonism on macrophage RCT.
Materials and Methods
Phosphate-buffered saline and RPMI/HEPES were purchased from Invitrogen. [1,2-3H]Cholesterol and cholesteryl oleate (cholesteryl-[1,2,6,7-3H]) were purchased from Perkin-Elmer Life Science. Hydroxypropyl methyl cellulose (HMPC K100) was purchased from Shin-etsu. Other reagents without citation in this article were purchased from Fisher Scientific. A systemic LXRα/β agonist (GW3965) and an intestine-specific LXRα/β agonist (GW6340) were obtained from GlaxoSmithKline Medicinal Chemistry Department. In vitro GW3965 is highly specific for LXR receptors,8 and GW6340 is the ester form of GW3965. GW6340 is described in a patent (PCT/US01/27622; GW6340=example 4; http://www.wipo.int/pctdb/en/). The structure of the intestinal-specific LXR agonist GW6340, a close analog of GW3965, is shown in Supplemental Figure I (available online at http://atvb.ahajournals.org). This compound has a pEC50=7 in vitro potency in LXR transactivation assay measuring ABCA1 promoter activity, a value comparable to the potency of GW3965. The dose for GW6340 was chosen to achieve a similar intestinal gene induction compared to GW3965 (Figure 2B).
Wild-type C57BL/6 mice were obtained from the Jackson Laboratory (Bar Harbor, Me), and LXRα/β double-knockout (LXR DKO) mice were obtained from Taconic (Germantown, NY). Mice were fed a standard chow diet (laboratory diet 5010; 4.8% fat and 267 ppm cholesterol) ad libitum before and during the study. For plasma lipid analyses, animals were fasted for 4 hours and then bled from the retro-orbital plexus. All animals were housed according to guidelines of the Institutional Animal Care and Usage Committee of the University of Pennsylvania. All protocols were considered and approved by the Institutional Animal Care and Usage Committee.
Total cholesterol, high-density lipoprotein cholesterol (HDL-C), phospholipid, and triglyceride levels were measured with the use of diagnostic reagents from Thermo Scientific. Pooled plasma from each group (150 μL) was separated by fast protein liquid chromatography gel filtration. Total cholesterol in each fraction was measured with the total cholesterol E kit from Wako Pure Chemicals.
Real-Time Polymerase Chain Reaction
Total RNA was isolated from mouse tissue using TRIzol Reagent from Invitrogen. The cDNA was produced from total RNA with reverse-transcription using high-capacity cDNA reverse-transcription kit from Applied Biosystems. Quantification of mRNA expression using TaqMan assay systems was performed by ABI 7300 real-time polymerase chain reaction system from Applied Biosystems. All reagents necessary for performing a TaqMan reverse-transcription polymerase chain reaction assay including primers and probes were purchased from Applied Biosystems and used according to the manufacturer’s instructions.9
Macrophage RCT Study
Experiments were performed in wild-type mice and LXR DKO mice. For each experiment, each group was dosed with 30 mg/kg body weight GW3965, GW6340, or vehicle (HPMC K100 0.5% and Tween 80 1% in phosphate buffer 60 mmol/L; pH, 7) once per day for 10 days by oral gavage. On day 10, RCT studies were performed as previously described.10,11 GW3965, GW6340, or vehicle were continued to be gavaged during the 48-hour experiment. J774 macrophages were obtained from American Type Culture Collection and were grown in suspension in RPMI 1640 medium supplemented with 10% fetal bovine serum, followed by being labeled with [3H]-cholesterol and loaded with acetylated low-density lipoprotein. Bone marrow macrophages (BMM) were isolated from femurs and tibias and cultured in Dulbecco modification of Eagle media supplemented with 10% fetal bovine serum and 30% L929 conditioned medium for 7 days.12 BMM were grown in 100-mm Petri dishes, labeled with 5 μCi/mL [3H]-cholesterol, and loaded with 25 μg/mL acetylated low-density lipoprotein. Twenty-four hours later, cells were washed and equilibrated overnight in Dulbecco modification of Eagle media 0.2% bovine serum albumin with 10 μmol/L GW3965. Before the injection, cells were collected in phosphate-buffered saline 10 mmol/L EDTA, washed, and suspended in RPMI/HEPES.
In Vivo HDL Turnover Studies
Wild-type C57BL/6 were fed a control or LXR agonist GW3965-supplemented (10 mg/kg/d) chow diet ad libitum for 2 weeks. GW3965 was grinded and added to the chow diet at 100 mg/kg diet (Research Diets). HDL was prepared from pooled human plasma by sequential ultracentrifugation (density 1.063<density<1.21 g/mL). After extensive dialysis against dialysis buffer (0.15 mol/L NaCl, EDTA 1 mmol/L; pH 7.4), HDL was exchange-labeled with cholesteryl oleate as previously described.13 After labeling, the HDL was re-isolated by ultracentrifugation (density 1.063<density<1.21 g/mL), extensively dialyzed, filter-sterilized, and stored at 4°C until injection. [3H]-cholesteryl oleate-labeled HDL (1–2 million cpm/animal) was injected intravenously via tail veins into mice (n=5–6/group). Blood samples were drawn by retro-orbital bleeding at 2 minutes, 1 hour, 3 hours, 6 hours, 9 hours, 24 hours, and 48 hours (≈25 μL at each time point). Feces were collected continuously from 0 to 48 hours and stored at 4°C before extraction of cholesterol and bile acids. At study termination (48 hours after injection), mice were exsanguinated, perfused with ice-cold phosphate-buffered saline, and liver samples were collected for analysis. Plasma decay curves for both tracers were normalized to radioactivity at the initial 2-minute time point after tracer injection. To measure [3H]-tracer recovery in liver, lipid extraction was performed according to the procedure of Bligh and Dyer.7 Briefly, a 50-mg piece of tissue was homogenized in water, and then lipids were extracted with a mixture of chloroform/methanol 2:1 (vol/vol). The lipid layer was collected, evaporated, resuspended in toluene, and counted in a liquid scintillation counter. Results were expressed as a percentage of the injected dose. The injected dose was calculated by multiplying the initial plasma counts (2-minute time point) by the estimated plasma volume (3.5% of the body weight).
All values are presented as mean±SD. Comparisons between control and agonist-treated mice were made with the use of the Student t test (2-tailed) and ANOVA with the use of GraphPad Prism Software.
LXR Agonist in Macrophages Alone Is Insufficient to Substantially Promote RCT in the Absence of Hepatic and Intestinal LXR Expression
To test whether activation of LXR in macrophages is, by itself, sufficient to promote macrophage RCT in the absence of LXR in all other tissues, wild-type and LXR DKO mice were gavaged for 10 days with GW3965 or vehicle before injection of cholesterol-labeled J774 macrophages, and they continued to be gavaged during the 48-hour experiment. At day 5, plasma HDL-C levels were increased in GW3965-treated wild-type mice (111±12 vs 73±10 mg/dL) but did not change in GW3965-treated LXR DKO mice (Supplemental Table I, available online at http://atvb.ahajournals.org). [3H]-Cholesterol counts in plasma of GW3965-treated mice showed a trend toward an increase in wild-type mice (Figure 1A), which was not observed in LXR DKO mice (Figure 1B). There was no difference in [3H]-cholesterol in the liver at the end of the 48-hour time period among 4 groups (Figure 1C). GW3965 treatment promoted significantly greater fecal excretion of total tracer and cholesterol tracer but not bile acid tracer (Figure 1D) in wild-type mice but not in LXR DKO mice. These data suggest that LXR activation in macrophages alone is insufficient to promote RCT and that LXR activation in the liver or the intestine is also required for promoting macrophage RCT.
An Intestinal-Specific LXR Agonist Increased Macrophage RCT in Wild-Type Mice
To determine the role of intestinal LXR activation in promoting RCT, we used an intestinal-specific LXR agonist, GW6340. Wild-type mice were gavaged with GW6340, the systemic LXR agonist GW3965, or vehicle for 10 days before injection of cholesterol-labeled J774 macrophages, and they continued to be gavaged during the 48-hour experiment. To confirm the intestinal-specific effect of GW6340, gene expression in liver and small intestine of vehicle, GW3965-treated, and GW6340-treated wild-type mice were assessed by quantitative polymerase chain reaction (Figure 2A, 2B). GW3965-treated mice had significantly increased LXR target gene expression in liver and small intestine, whereas GW6340-treated mice had increased LXR target gene expression exclusively in small intestine. In addition, treatment with GW3965 significantly increased hepatic triglyceride content, whereas treatment with GW6340 did not (Supplementary Table II). These observations confirm that, in vivo, GW6340 is an intestinal-specific LXR agonist and, unlike the systemic LXR agonist GW3965, does not modulate hepatic gene expression or triglyceride content.
At day 5, plasma HDL-C levels were significantly higher in GW3965-treated mice compared with vehicle-treated mice (100±16 vs 57±7 mg/dL; Supplemental Table III). Whereas there was a trend toward an increased HDL-C level in GW6340-treated mice, it was not significant (Supplemental Table III). In the fast protein liquid chromatography analysis, GW3965 substantially increased cholesterol in the usual HDL fraction as well as in a large HDL fraction, whereas GW6340 modestly increased cholesterol in the usual HDL fraction (Figure 3A). The [3H]-cholesterol counts in plasma after J774 macrophage injection, expressed as a percentage of total [3H]-counts injected, were significantly increased in GW3965-treated mice compared with control mice, whereas they trended higher but were not significantly different in the GW6340-treated mice compared with the vehicle-treated mice (Figure 3C). The [3H]-cholesterol in plasma tracked closely with the cholesterol mass in plasma lipoprotein distribution and were consistent with HDL [3H] counts highest in GW3965-treated mice, next highest in GW6340-treated mice, and lowest in vehicle-treated mice (Figure 3B). There were no differences in [3H]-cholesterol in the liver at the end of 48-hour time period among the 3 groups (Figure 3D). As expected, GW3965-treated mice excreted a significantly 162% higher amount of macrophage-derived tracer in the feces compared with the control mice (Figure 3E). Interestingly, GW6340-treated mice also excreted significantly more (52%) macrophage-derived tracer in the feces compared with the control mice (Figure 3E). Thus intestine-specific LXR activation promotes macrophage RCT, although it may be less effective than systemic LXR activation.
Macrophage LXR Contributes to the Ability of a Systemic LXR Agonist to Promote Macrophage RCT
We tested whether macrophage LXR is critical for the ability of a systemic LXR agonist to promote RCT. As expected, in BMM from wild-type mice, both GW3965 and acetylated low-density lipoprotein treatment increased cholesterol efflux, whereas in BMM from LXR DKO mice GW3965 and acetylated low-density lipoprotein treatment had no effect on efflux ex vivo (Supplemental Figure II). Next, we evaluated RCT in vivo using primary BMM from LXR DKO and wild-type mice. Wild-type mice were gavaged for 10 days with GW3965 or vehicle before injection of BMM from either wild-type or LXR DKO mice and continued to be gavaged during the 48-hour experiment. At day 5, plasma HDL-C levels were increased in GW3965-treated mice (Supplemental Table IV). The [3H]-cholesterol counts in plasma were significantly increased by GW3965 treatment in mice injected with wild-type BMM (Figure 4A). In contrast, there was no effect of GW3965 treatment on plasma [3H]-cholesterol counts in mice injected with LXR DKO BMM (Figure 4B). There was no difference in [3H]-cholesterol in the liver at the end of the 48-hour time period among 4 groups (Figure 4C). GW3965 treatment significantly increased fecal excretion of total tracer in mice injected with wild-type BMM but not in mice injected with LXR DKO BMM (Figure 4D). Excretion of cholesterol tracer was markedly increased by GW3965 in wild-type BMM and was increased, modestly but significantly, by GW3965 in LXR DKO BMM (Figure 4D). In contrast, GW3965 had no effect on fecal excretion of bile acid tracer in either group (Figure 4D). These data confirm that macrophage LXR is an important determinant of the ability of a systemic LXR agonist to promote macrophage RCT, but that LXR agonist promotes the fecal excretion of macrophage-derived cholesterol even in the absence of macrophage LXR.
To confirm that activation of LXR in the liver and intestine promotes the fecal excretion of HDL-derived cholesterol, we performed a [3H]-HDL-cholesteryl-ester kinetics study with GW3965-treated mice. Compared to the control group treated with vehicle, GW3965 treatment did not alter the rate of plasma turnover of [3H]-HDL-cholesteryl-ester (Figure 5A) or the amount of [3H]-cholesterol in the liver (Figure 5B), but it significantly increased the fecal cholesterol excretion of the HDL-derived cholesterol (Figure 5C). These data confirm the macrophage RCT study results and suggest that LXR agonism promotes RCT, at least in part, by accelerating the rate by which HDL-C is targeted for biliary and fecal excretion.
In these studies, we provide evidence that LXR in all 3 major LXR-expressing tissues, macrophages, liver, and intestine, play a role in promoting macrophage RCT in response to LXR activation. We show that that ability of an LXR agonist to promote macrophage RCT requires LXR expression in tissues other than just the macrophage. We provide the first evidence to our knowledge that an intestine-specific LXR agonist promotes macrophage RCT in vivo. Finally, we show that macrophage LXR is an important, but not the sole, component of the ability of a systemic LXR agonist to promote RCT and that an LXR agonist promotes the fecal excretion of HDL-derived cholesterol. These observations are potentially relevant to therapeutic targeting of LXR for reduction of atherosclerotic vascular disease.
The small intestine plays a critical role in cholesterol homeostasis. It regulates cholesterol absorption and excretion and also produces apolipoprotein A-I and nascent HDL. Intestine-specific ABCA1 deficiency resulted in a 30% reduction of plasma HDL levels,14 whereas a systemic LXR agonist significantly increased plasma HDL levels even in liver-specific ABCA1-deficient mice.15 These data indicate that intestinal ABCA1 is a major determinant of plasma HDL-C levels. In this study, we used an intestine-specific LXR agonist GW6340 to address the question of how intestine-specific LXR agonist influences macrophage RCT and showed that it significantly promoted macrophage RCT. GW6340 significantly upregulated ABCA1, ABCG5, and ABCG8 expression in the small intestine but not in the liver. Fast protein liquid chromatography analysis suggested that it modestly increased cholesterol mass and macrophage-derived cholesterol tracer in the HDL fraction, providing 1 possible mechanism by which it promoted macrophage RCT. In addition, ABCG5/8 promotes cholesterol excretion in the small intestine.16 Recently, Veen et al17 reported that direct intestinal secretion of plasma cholesterol significantly contributes to fecal neutral sterol loss in mice. This transintestinal cholesterol excretion was upregulated by a LXR agonist and impaired in ABCG5-deficient mice. Thus, the intestine-specific LXR agonist might stimulate transintestinal excretion of plasma cholesterol partially via activating expression of the ABCG5/ABCG8 heterodimer. Therefore, the intestine-specific LXR agonist may promote RCT through at least 2 independent pathways: increasing intestinal HDL production and promoting intestinal excretion of HDL-derived cholesterol.
We previously reported that ABCA1-deficient and ABCG1-deficient macrophages showed a significant reduction of macrophage-derived fecal cholesterol excretion.11 LXR agonist treatment is known to upregulate ABCA1 and ABCG1 and cholesterol efflux from macrophages.18 These data suggest that macrophage LXR is crucial for the ability of an LXR agonist to promote RCT. We show here that macrophage-specific LXR deficiency resulted in a significant blunting, but not abolishment, of the ability of the systemic LXR agonist GW3965 to promote RCT. However, even in mice injected with labeled macrophages lacking LXR, GW3965-treated mice still increased fecal cholesterol excretion compared with vehicle group. Once cholesterol efflux occurred from macrophages to plasma through an LXR-independent pathway, activated LXR in liver and small intestine likely promoted cholesterol excretion from plasma to feces. Using HDL labeled with cholesteryl ester we confirmed that GW3965 promoted the fecal excretion of HDL-derived cholesterol in a macrophage-independent manner.
Some systemic LXR agonists activate SREBP-1c in the liver, leading to elevated plasma triglycerides and fatty liver.19 This has slowed the clinical development of LXR agonists as a potential treatment for atherosclerosis. Based on our studies, we suggest that intestinal-specific LXR agonism, which avoids undesired effects of activating hepatic LXR, might be expected to promote macrophage RCT and thus potentially might be antiatherogenic.
The authors thank Dawn Marchadier, Debra Cromley, Amrith Rodrigues, Michelle Joshi, Aisha Wilson, Edwige Edouard, and Mao-Sen Sun for their excellent technical assistance.
Source of Funding
This work was supported by an Alternative Drug Discovery Initiative award funded by GlaxoSmithKline and grant P01-HL22633 from the NHLBI. Tomoyuki Yasuda was supported by the grant from the Japan Heart Foundation.
Didier Grillot and Stephane Huet are employees of GlaxoSmithKline. Philippe Delerive was previously employed by GlaxoSmithKline.
Received on: September 15, 2009; final version accepted on: January 20, 2010.
Rader DJ, Alexander ET, Weibel GL, Billheimer J, Rothblat GH. Role of reverse cholesterol transport in animals and humans and relationship to atherosclerosis. J Lipid Res. 2009; 50 (Suppl): S189–S194.
Venkateswaran A, Laffitte BA, Joseph SB, Mak PA, Wilpitz DC, Edwards PA, Tontonoz P. Control of cellular cholesterol efflux by the nuclear oxysterol receptor LXR alpha. Proc Natl Acad Sci U S A. 2000; 97: 12097–12102.
Repa JJ, Berge KE, Pomajzl C, Richardson JA, Hobbs H, Mangelsdorf DJ. Regulation of ATP-binding cassette sterol transporters ABCG5 and ABCG8 by the liver X receptors alpha and beta. J Biol Chem. 2002; 277: 18793–18800.
Naik SU, Wang X, Da Silva JS, Jaye M, Macphee CH, Reilly MP, Billheimer JT, Rothblat GH, Rader DJ. Pharmacological activation of liver X receptors promotes reverse cholesterol transport in vivo. Circulation. 2006; 113: 90–97.
Collins JL, Fivush AM, Watson MA, Galardi CM, Lewis MC, Moore LB, Parks DJ, Wilson JG, Tippin TK, Binz JG, Plunket KD, Morgan DG, Beaudet EJ, Whitney KD, Kliewer SA, Willson TM. Identification of a nonsteroidal liver X receptor agonist through parallel array synthesis of tertiary amines. J Med Chem. 2002; 45: 1963–1966.
Lagor WR, Brown RJ, Toh SA, Millar JS, Fuki IV, de la Llera-Moya M, Yuen T, Rothblat G, Billheimer JT, Rader DJ. Overexpression of apolipoprotein F reduces HDL cholesterol levels in vivo. Arterioscler Thromb Vasc Biol. 2009; 29: 40–46.
Tanigawa H, Billheimer JT, Tohyama J, Zhang Y, Rothblat G, Rader DJ. Expression of cholesteryl ester transfer protein in mice promotes macrophage reverse cholesterol transport. Circulation. 2007; 116: 1267–1273.
Schiller NK, Black AS, Bradshaw GP, Bonnet DJ, Curtiss LK. Participation of macrophages in atherosclerotic lesion morphology in LDLR−/− mice. J Lipid Res. 2004; 45: 1398–1409.
Tietge UJ, Maugeais C, Cain W, Grass D, Glick JM, de Beer FC, Rader DJ. Overexpression of secretory phospholipase A(2) causes rapid catabolism and altered tissue uptake of high density lipoprotein cholesteryl ester and apolipoprotein A-I. J Biol Chem. 2000; 275: 10077–10084.
Brunham LR, Kruit JK, Pape TD, Parks JS, Kuipers F, Hayden MR. Tissue-specific induction of intestinal ABCA1 expression with a liver X receptor agonist raises plasma HDL cholesterol levels. Circ Res. 2006; 99: 672–674.
Duan LP, Wang HH, Wang DQ. Cholesterol absorption is mainly regulated by the jejunal and ileal ATP-binding cassette sterol efflux transporters Abcg5 and Abcg8 in mice. J Lipid Res. 2004; 45: 1312–1323.
van der Veen JN, van Dijk TH, Vrins CL, van Meer H, Havinga R, Bijsterveld K, Tietge UJ, Groen AK, Kuipers F. Activation of the liver X receptor stimulates trans-intestinal excretion of plasma cholesterol. J Biol Chem. 2009; 284: 19211–19219.
Sparrow CP, Baffic J, Lam MH, Lund EG, Adams AD, Fu X, Hayes N, Jones AB, Macnaul KL, Ondeyka J, Singh S, Wang J, Zhou G, Moller DE, Wright SD, Menke JG. A potent synthetic LXR agonist is more effective than cholesterol loading at inducing ABCA1 mRNA and stimulating cholesterol efflux. J Biol Chem. 2002; 277: 10021–10027.
Repa JJ, Liang G, Ou J, Bashmakov Y, Lobaccaro JM, Shimomura I, Shan B, Brown MS, Goldstein JL, Mangelsdorf DJ. Regulation of mouse sterol regulatory element-binding protein-1c gene (SREBP-1c) by oxysterol receptors, LXRalpha and LXRbeta. Genes Dev. 2000; 14: 2819–2830.