|
Advertisement | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
J. Biol. Chem., Vol. 278, Issue 49, 49072-49078, December 5, 2003
Enlargement of High Density Lipoprotein in Mice via Liver X Receptor Activation Requires Apolipoprotein E and Is Abolished by Cholesteryl Ester Transfer Protein Expression*![]() ¶![]() ![]() ![]() ![]()
From the
Lilly Research Laboratories, Eli Lilly and Company, Indianapolis, Indiana 46285 and the
Received for publication, April 23, 2003 , and in revised form, August 22, 2003.
The factors involved in the generation of larger high density lipoprotein (HDL) particles, HDL1 and HDLc, are still not well understood. Administration of a specific synthetic liver X receptor (LXR) agonist, T0901317, in mice resulted in an increase of not only HDL cholesterol but also HDL particle size (Cao, G., Beyer, T. P., Yang, X. P., Schmidt, R. J., Zhang, Y., Bensch, W. R., Kauffman, R. F., Gao, H., Ryan, T. P., Liang, Y., Eacho, P. I., and Jiang, X. C. (2002) J. Biol. Chem. 277, 3956139565). We have investigated the roles that apoE and CETP may play in this process. We treated apoE-deficient, cholesterol ester transport protein (CETP) transgenic, and wild type mice with various doses of the LXR agonist and monitored their HDL levels. Fast protein liquid chromatography and apolipoprotein analysis revealed that in apoE knockout mouse plasma, there was neither induction of larger HDL formation nor increase of HDL cholesterol, suggesting that apoE is essential for the LXR agonist effects on HDL metabolism. In CETP transgenic mice, CETP expression completely abolished LXR agonist-mediated HDL enlargement and greatly attenuated HDL cholesterol levels. Analysis of HDL particles by electron microscope and nondenaturing gel electrophoresis revealed similar findings. In apoE-deficient mice, LXR agonist also produced a significant increase in very low density lipoprotein/low density lipoprotein cholesterol and apolipoprotein B content. Our studies provide direct evidence that apoE and CETP are intimately involved in the accumulation of the enlarged HDL (HDL1 or HDLc) particles in mice.
Epidemiological studies have firmly established that plasma HDL1 cholesterol is inversely correlated to coronary artery events (1). The biogenesis of HDL is thought to originate from the secretion of its major apolipoprotein component, apoAI, from the liver and the small intestine. ApoAI enters the circulation and interacts with and removes excessive free cholesterol from peripheral tissues, forming disc-like nascent HDL particles. ApoAI-dependent phospholipid and cholesterol efflux from peripheral tissues requires the protein ABCA1, an ATP binding cassette transporter (2). The maturation of HDL depends on both lecithin-cholesterol acyltransferase and phospholipid transfer protein (PLTP) activities. The former esterifies free cholesterol to form spherical HDL (3), and the latter transfers phospholipid from triglyceride-rich lipoproteins into the nascent HDL particles (4). Cholesteryl ester transfer protein (CETP) catalyzes transfer of cholesteryl ester from mature HDL into apoB-containing lipoproteins for catabolism through liver low density lipoprotein receptor (LDLR) (5). HDL cholesterol can also be delivered to the liver via scavenger receptor BI, the HDL receptor, through the process of selective cholesterol uptake (6). It is conceivable that the regulation of ABCA1, lecithin-cholesterol acyltransferase, PLTP, CETP, and scavenger receptor BI would have an important impact on HDL metabolism, including particle catabolic rate, lipid composition, and particle size.
Liver X receptors were initially isolated as orphan nuclear receptors. Two isoforms, LXR Administration of LXR agonist increases HDL cholesterol (13, 15) and HDL size in mice (13). The enlarged HDL particles are apoE-enriched (13). This feature is very much similar to an HDL subpopulation, HDL1 or HDLC (20). The enlargement of HDL after LXR agonist administration to mice may largely involve the induction of ABCA1 and apoE, since both gene products are intimately involved in HDL metabolism and are regulated by LXRs (10, 11, 21). ABCA1 deficiency causes hypoalphalipoproteinemia in humans (2225) and mice (26, 27). ABCA1 overexpression in mice leads to an increased plasma HDL cholesterol level (28, 29). It has also been reported that apoE plays an obligatory role in large HDL formation (20). CETP may be another factor that can influence the formation of large HDL particles. In patients with complete CETP deficiency, HDL is increased in size and enriched in apoE and cholesteryl esters (30). Conversely, in CETP transgenic rats, conversion of large HDL to small HDL particles was observed (31). In order to investigate the contribution of apoE and CETP in the LXR-mediated formation of enlarged HDL particles, we treated CETP transgenic (CETP-Tg), apoE-deficient (apoE/), and wild type mice with various doses of LXR agonist, T0901317. We show that in these mice, the enlargement of HDL particle size after administrating the LXR agonist requires apoE and is completely abolished by CETP expression.
AnimalsSeven-week-old male mice were purchased and acclimated for 1 week prior to the start of the study. C57BL/6 mice were purchased from Harlan (Indianapolis, IN). Human CETP-Tg and apoE-deficient (apoE/) mice were purchased from Taconic (Germantown, NY). Mice were provided Purina 5001 food ad libitum, and the compound was dosed once daily via oral gavage for 7 days. Animals were sacrificed by CO2 asphyxiation in the morning, 2 h after the eighth dose; blood samples were taken by cardiac puncture; and tissues were collected and frozen in liquid nitrogen. Lipoprotein Analysis by FPLCLipoprotein analysis was performed as described (13). Briefly, plasma samples from animals were prepared and pooled. 50 µl of pooled sample was applied to Superose 6 size exclusion columns and eluted with phosphate-buffered saline, pH 7.4. Cholesterol content of different fractions was measured by commercial kit. Lipoprotein Analysis by Nondenaturing Gel ElectrophoresisFive microliters of mouse plasma was loaded on a 420% nondenaturing gel (Bio-Rad), which was prerun with TBE buffer (10 mM Tris, 10 mM boric acid, and 1 mM EDTA, pH 7.4) at 4 °C, 50 V, for 1 h. The electrophoresis was carried out at 4 °C and 50 V for 12 h. The gel then was stained with 1% of Oil Red O (in isopropyl alcohol) at 58 °C for 6 h. Analysis of Isolated HDL Particles by Native Gel Electrophoresis Lipoproteins were purified by sequential ultracentrifugation using a TL 100.4 rotor with a tabletop TL 100 centrifuge (Beckman Instruments). HDL was purified between the densities of 1.063 and 1.210 g/ml in all animals. HDL particles were separated by gradient gel electrophoresis (420%; Bio-Rad) under nonreducing and nondenaturing conditions. The electrophoresis was carried out for 3 h, and the gel was stained with Coomassie Brilliant Blue R-250 and destained with a solution of methanol/acetic acid/water (30:58:12, v/v/v). Western Blot AnalysisPlasma samples were separated by FPLC size exclusion columns and different fractions were pooled (fractions 2023, 2427, 2831, 3236, and 3741) for apolipoprotein analysis. Designated FPLC fraction samples were separated on Tris/glycine gels (Novex) under denaturing conditions. Protein was transferred to nitrocellulose membrane and then blotted with antibodies to apolipoprotein AI or E (Biodesign) or apolipoprotein B48/100 (U. S. Biological). Blots were developed with ECL Western blotting detection reagents (Amersham Biosciences) and documented using X-Omat film (Eastman Kodak Co.). Isolated HDL fractions were analyzed on a gradient gel (420%; Bio-Rad) under reducing and denaturing conditions, followed by a similar procedure as described above. SDS-PAGE Apolipoprotein AnalysisPlasma HDL (density = 1.0631.21 g/ml) and LDL (density = 1.0061.063 g/ml) were separated by preparative ultracentrifugation. SDS-PAGE was performed on 320% SDS-polyacrylamide gradient gel, and the aplipoproteins were stained by Coomassie Brilliant Blue as described (4). Electron MicroscopyNegative stain electron microscopy was done using a JEM-100C electron microscope (JEOL USA Inc., Cranford, NJ). Lipoprotein samples in 125 mM ammonium acetate, 2.6 mM ammonium carbonate, and 0.26 mM EDTA (pH 7.4) were mixed with an equal volume of 2% sodium phosphotungstate (pH 7.4) on a carbon/Formvar-coated copper grid (Electron Microscopy Sciences, Fort Washington, PA). After 30 s, the excess liquid was blotted, the rest was allowed to dry, and the grid was examined within 1 h. mRNA AnalysisTotal RNAs were prepared from frozen tissue samples with TRIzol reagent (Invitrogen). Mouse ABCA1 and PLTP mRNA were measured by an RNase protection assay. A primer set was used to amplify a fragment of about 200 base pairs from mouse macrophage RAW cells for ABCA1 probe. The resulting 200-bp fragment was cloned into pGEM-T Easy (Promega) and sequenced. The resulting construct was linearized, and RNA probe was synthesized using the Promega T7/SP6 transcription kit. Specific activity was >108 dpm/µg. After column purification, the probe was used for RPA analysis by using a kit from Ambion. The probe for PLTP mRNA measurement was described previously (13). The signal was quantified with an Amersham Biosciences PhosphorImager model 51.
Previously, it was observed that treatment of C57BL/6 mice with a specific LXR agonist, T0901317, resulted in an increase of both HDL cholesterol and particle size (13). To investigate the effect of apoE on the formation of enlarged HDL particles, as well as on HDL cholesterol levels, we treated apoE/ mice with T0901317 at 10 and 50 mg/kg for 7 days. Wild type C57BL/6 mice were treated in a similar scheme for comparison. As reported previously (13), treatment of C57BL/6 mice with 50 mg/kg T0901317 resulted in dramatic increase in HDL cholesterol, as determined by FPLC (Fig. 1A). The HDL fraction included a primary peak (fractions 3241) as well as a second peak (fractions 2831), which was largely absent in vehicle-treated C57BL/6 mice. This latter peak overlapped with that of the LDL fraction, based on its size and the presence of apolipoproteins B-48 and B-100 (Fig. 2A). The apoE content of these FPLC elutes was markedly increased by treatment of 50 mg/kg T0901317 (the overall plasma apoE content was increased about 80% as judged by the quantitative Western blots), whereas the apoB level was not significantly changed (Fig. 2A), which confirmed our previous findings (13). The apoAI content of both HDL peaks was increased by T0901317.
In apoE/ mice, VLDL/LDL represent major lipoprotein components and HDL represents a minor fraction of the FPLC profile. There was no apparent increase in HDL size or HDL cholesterol after the LXR agonist treatment (Fig. 1B). To further analyze the data, we expanded the FPLC plot that covered fractions 3042 (Fig. 1B'). No HDL cholesterol increase was observed (fractions 3642), and some elevation of cholesterol upon LXR agonist treatment (fractions 3035) was closely associated with a large amount of apoB48 representing remnant lipoprotein particles. In VLDL/LDL fractions 2427 and 2831 of the apoE/ mice, there was a dose-related increase in the cholesterol content (Fig. 1B) and an increase of apoB48 (Fig. 2B). Interestingly, apoAI was present in these fractions, as was observed previously in apoE/ mice (32) and was slightly increased by treatment with T0901317 (Fig. 2B). ApoAI level was not changed in fractions 3236 and appeared to be slightly decreased in fractions 3741 in response to LXR agonist treatment. Thus, administration of the LXR agonist to apoE/ mice did not appear to cause HDL particle enlargement or an increase in HDL cholesterol levels. Instead, cholesterol in remnant particles was increased. To investigate the role of CETP in the LXR agonist-mediated changes in HDL cholesterol metabolism, we utilized a transgenic model in which the human CETP transgene was expressed under the control of the human apoAI promoter (33). LXRs are not known to regulate apoAI (15), and we found that treatment of these mice with T0901317 did not result in a change in plasma CETP activity (data not shown). Treatment of these mice with T0901317 did not lead to the formation of enlarged HDL particles (Fig. 1C), as was observed in C57BL/6 mice. A small increase in HDL cholesterol was only observed at the higher dose of the compound used. No significant increase of apoAI or apoE in the HDL fraction was observed (Fig. 2C). Thus, CETP expression greatly attenuated the LXR agonist-induced HDL cholesterol increase and abolished the increase in HDL particle size. To further confirm the HDL cholesterol and particle size changes in these different animal models after oral administration of the LXR agonist, we utilized nondenaturing gel electrophoresis to evaluate HDL particles. Mouse plasma samples were applied to the nondenaturing gels (420% gradient), which were then stained with Oil Red O after running. HDL particles were readily separated and stained on the gel, whereas VLDL/LDL particles could not get into the gel because of the size. In C57BL/6 mice, T0901317 treatment increased the intensity of Oil Red O staining of HDL in a dose-dependent manner, reflecting an increase in HDL lipid (Fig. 3A). A dose-dependent HDL particle size enlargement was also evident (Fig. 3A). In apoE/ mice, there was reduced HDL lipid content but no change in particle size after LXR agonist treatment (Fig. 3B). In CETP transgenic mice, LXR agonist administration did not change HDL particle size, and the increase of HDL lipid contents was attenuated (Fig. 3C). Thus, the analysis of HDL cholesterol and particle size with nondenaturing gels confirmed the findings of our FPLC analysis.
In order to unequivocally prove the results described above, we physically separated HDL particles from LDL particles by ultracentrifugation. The isolated HDL fractions were subjected to electron microscopic study (Fig. 4A). The HDL particle size from wild type animals was in the range of 1012 nm in diameter. T0901317 treatment resulted in an obvious enlargement of HDL particles. The size of largest HDL particle reached about 20 nm in diameter. On the contrary, T0901317 treatment in either apoE-deficient mice or CETP transgenic mice did not induce any significant change in HDL particle size (Fig. 4A). The isolated HDL fractions were then analyzed by native gel electrophoresis and Coomassie Blue staining (Fig. 4B). Consistent with electron microscopic data, T0901317 treatment of wild type mice resulted in slower migration of HDL particles in the native gel under nondenaturing conditions, suggesting the increase in HDL particle size. Treatment of apoE-deficient mice or CETP transgenic mice, however, did not result in any observable changes in the mobility of HDL particles. The presence of apoAI and apoE in these HDL particles isolated from ultracentrifugation was further proved with Western blot analysis (Fig. 4C). Increases in both apoAI and apoE content in these HDL particles after T0901317 treatment of wild type mice was evident, whereas no change was observed in either apoE-deficient mice or CETP transgenic mice. We further performed SDS-PAGE apolipoprotein analysis on isolated LDL and HDL from wild type mouse plasma with or without LXR agonist treatment. As can be seen from Fig. 5, the isolated LDL contained significant amounts of apoB48 and apoB100. The isolated HDL particles had no detectable apoB, whereas the LXR agonist treatment greatly increased apoE and apoAI levels. Since HDL particles from LXR agonist-treated animal plasma contained normal HDL and enlarged HDL, the results indicated that the later did not contain any appreciable amount of apoB.
To investigate the molecular mechanisms of the above-mentioned observations, we analyzed mRNA regulation of two important LXR target genes, ABCA1 and PLTP, in the liver. Treatment of C57BL/6 mice at 50 mg/kg resulted in a 2.7-fold induction of ABCA1 and a 5.3-fold increase of PLTP. In apoE/ mice, both ABCA1 and PLTP up-regulation was readily detectable (2.4- and 3.7-fold, respectively). In CETP transgenic mice, ABCA1 and PLTP were also up-regulated by LXR agonist (2.3- and 2.5-fold, respectively) (Fig. 6, A and B). Thus, the regulation of ABCA1 and PLTP did not appear to be drastically different in the three models, suggesting the direct roles of apoE and CETP expression in the LXR agonist-mediated HDL cholesterol metabolism.
LXR agonists are known to modulate HDL cholesterol levels in mice. In this study, we demonstrated for the first time that 1) LXR agonist-mediated increase in HDL particle size and cholesterol levels requires apoE; 2) LXR agonist-mediated HDL enlargement is completely abolished by CETP transgene expression; and 3) LXRs up-regulate ABCA1 mRNA in the liver. LXR agonist-induced enlarged HDL particles are apoE-enriched (13). This feature is similar to an HDL subpopulation, HDL1 or HDLC (20). Although HDL1 particles have been known for many years, the origin of these particles in vivo is still poorly understood. Gordon et al. (34) used canine HDL and cholesterol-loaded macrophages or cholesterol-coated Celite particles to demonstrate that small HDL particles could be converted into large HDL1 particles in the presence of excessive cholesterol in a lecithin-cholesterol acyltransferase-dependent manner. It was also reported that, in cholesterol-loaded macrophages, HDL1 formation depends on the availability of apoE, which can be newly synthesized from cells, derived from other plasma lipoproteins, or exogenously added (20). In our study, we demonstrated that the enlargement of HDL after LXR agonist treatment was absent in apoE/ mice (Figs. 1B, 2B, 3B, and 4), indicating that, similar to the in vitro situation (20), apoE is required for the formation of HDL1 in vivo. The mechanism by which HDL becomes enriched with apoE in LXR agonist-treated C57BL/6 mice is not known. Although apoE associated with enlarged HDL particles can be derived from liver in the VLDL fraction that is later transferred to HDL particle, this may represent an unlikely possibility, since liver apoE mRNA was not regulated by LXR agonist treatment (15). Alternatively it could be derived from peripheral tissues, especially macrophage cells and adipocytes, since apoE expression is up-regulated by LXRs in both adipose tissue and macrophage cells (11). Plasma CETP is another factor that can influence the HDL1 formation (35). Humans with CETP deficiency have large HDL particles that are enriched in apoE (30, 36). Cholesterol feeding of CETP-deficient animals results in a marked increase of large HDL1, which can extend to low plasma densities (d < 1.006 g/ml) and may transport as much as 50% of the plasma cholesterol (36). The enlarged HDL particles with enriched apoE in C57BL/6 mice after LXR agonist treatment are strikingly similar to the HDL1 or HDLc particles obtained from human subjects with CETP homozygous mutations. CETP catalyzes the exchange of cholesterol and triglycerides between HDL and triglyceride-rich lipoprotein particles. CETP transgenic mice displayed reduced plasma HDL cholesterol levels especially on the human apoAI transgenic background (37, 38). In our studies, CETP expression in mice resulted in total ablation of enlarged HDL production (Figs. 1B, 3B, and 4), and these enlarged HDL particles appear to be ideal and preferential substrates for CETP (Figs. 1B and 4). The effect of the LXR agonist to elevate HDL cholesterol was also attenuated in the CETP transgenic mice. These studies suggest that, in species expressing CETP, LXR agonist treatment may not result in a significant change in steady-state HDL cholesterol levels. Treatment of hamsters with T0901317 did not lead to any significant change in HDL cholesterol,2 which is consistent with this hypothesis. The potential of CETP to abrogate an LXR agonist-induced increase of HDL in humans may be compounded, considering that LXRs up-regulate CETP expression (12). Despite the failure of the LXR agonist to increase HDL significantly in the presence of CETP, we would expect it to increase the dynamic process of reversal of cholesterol transport.
The effects of HDL1 particles on atherosclerosis are unknown. HDL1 particles bind to the LDL receptor with high affinity (30, 39), and they can substitute for LDL as the major class of cholesterol-carrying plasma lipoproteins delivering lipids to peripheral tissues and the liver (36). Since these particles are still HDL, composed mainly of apoE and apoAI (13), they may have antioxidant (40) and anti-inflammatory properties (41). Indeed, compared with C57BL/6J, ICR mouse plasma contained more HDL1, which might contribute to the resistance of ICR mice and the susceptibility of C57BL/6J mice to the development of aortic fatty streak lesions when challenged with an atherogenic diet (42). A recent study in the CETP transgenic Dahl salt-sensitive hypertensive rats suggested combined hyperlipidemia, atherosclerotic lesions, coronary heart disease, and decreased survival (43). ABCA1 deficiency leads to a near total absence of HDL cholesterol both in humans and mice (2227). The molecular mechanisms of ABCA1 contributing to the HDL biogenesis, however, may need further refinement. It is presumed that peripheral tissue ABCA1 deficiency results in defective cholesterol efflux to apoAI, which is quickly cleared by the kidney as a result of poor lipidation. Although it has been firmly established in vitro that cholesterol efflux was defective with fibroblasts from Tangier patients (44), the definitive in vivo proof is still lacking that a nearly total absence of HDL cholesterol in Tangier patients and ABCA1-deficient animals is due solely to peripheral ABCA1 deficiency. In fact, hepatic ABCA1 appears to have a major role in the formation of plasma HDL. Neufeld et al. (45) demonstrated that an ABCA1-green fluorescent protein fusion protein was specifically expressed on the basolateral surface of a polarized hepatocyte cell line. This expression pattern strongly suggests that ABCA1 is involved in the lipidation of apoAI secreted by the liver. Indeed, Kiss et al. (46) have shown a significant reduction in apoAI lipidation by primary hepatocytes from ABCA1-deficient mice compared with its wild type counterpart. Brewer and colleagues have shown convincingly that specific overexpression of ABCA1 in the liver was sufficient to increase HDL cholesterol in mice (29, 47). Interestingly, overexpression of ABCA1 in the liver through adenovirus resulted in not only HDL cholesterol increase but also an increase in HDL particle size (47). We have observed that the increase of HDL cholesterol upon LXR agonist treatment was accompanied by a significant increase in liver ABCA1 mRNA, implicating that at least part of HDL cholesterol increase may be originated from liver ABCA1 regulation. To summarize our findings, we have discovered that apoE and CETP play critical roles in determining the effects of LXR activation on HDL cholesterol metabolism. Our studies provide further evidence that HDL metabolism is a dynamic process in which multiple protein components contribute in a variety of genetic, physiological, and pathological conditions.
* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
¶ Supported by National Institute of Health Grants HL-69817 and HL-64735. To whom correspondence may be addressed. Tel.: 718-270-6701; Fax: 718-270-3732; E-mail: xjiang{at}downstate.edu. || To whom correspondence may be addressed. Tel.: 317-433-3535; Fax: 317-276-1417; E-mail: Guoqing_Cao{at}lilly.com.
1 The abbreviations used are: HDL, high density lipoprotein; ABCA1, ATP binding cassette transporter A1; CETP, cholesterol ester transport protein; CETP-Tg, CETP transgenic; apoE, apolipoprotein E; apoAI, apolipoprotein AI; LXR, liver X receptor; PLTP, phospholipid transfer protein; LDL, low density lipoprotein; LDLR, low density lipoprotein receptor; FPLC, fast protein liquid chromatography.
2 T. P. Beyer, P. I. Eacho, and G. Cao, unpublished data.
We thank Dr. Lawrence Rudel for helpful discussions. We are thankful to Dr. Raymond Kauffman for discussions and for critically reading the manuscript. We are indebted to Drs. Timothy Grese, George Cullinan, and Steve Kuo-Long Yu for making the compound available for the study. We also thank Richard Tielking, Jack Cochran, Phyllis Cross, and Patrick Forler for invaluable technical assistance.
This article has been cited by other articles:
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Advertisement | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||