SEARCH

SEARCH BY CITATION

Keywords:

  • cholesterol;
  • cholesterol absorption;
  • HDL;
  • intestine;
  • liver X receptor

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References
  10. Supporting Information

Abstract.  Hu X, Steffensen KR, Jiang Z-Y, Parini P, Gustafsson J-Å, Gåfvels M, Eggertsen G (Karolinska Institutet, Huddinge, Stockholm, Sweden; Karolinska Institutet, Huddinge, Stockholm, Sweden; Shanghai Institute of Digestive Surgery, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China; and University of Houston, Houston, TX, USA). LXRβ activation increases intestinal cholesterol absorption, leading to an atherogenic lipoprotein profile. J Intern Med 2012; 272: 452–464.

Objectives.  Liver X receptors (LXRs) are essential for the regulation of intestinal cholesterol absorption. Because two isoforms exist, LXRα and LXRβ, with overlapping but not identical functions, we investigated whether LXRα and LXRβ exert different effects on intestinal cholesterol absorption.

Design.  Wild-type (WT), LXRα−/− and LXRβ−/− mice were fed control diet, 0.2% cholesterol-enriched diet or 0.2% cholesterol-enriched diet plus the LXR agonist GW3965.

Results.  When fed a control diet, all three genotypes showed similar levels of cholesterol absorption. Of interest, a significant increase in cholesterol absorption was found in the LXRα−/− mice, but not in the WT or LXRβ−/− animals, when fed a diet enriched with 0.2% cholesterol or 0.2% cholesterol + GW3965. Reduced faecal neutral sterol excretion and a hydrophobic bile acid profile were also observed in LXRα−/− mice. Greater increases in the apolipoprotein (apo)B-containing lipoproteins in serum were seen in the LXRα−/− mice. A 0.2% cholesterol + GW3965 diet suppressed intestinal Npc1l1 protein expression to the same extent for all genotypes, while Abca1 and Abcg5 were elevated to the same degree.

Conclusions.  In the intestine, LXRα and LXRβ seem to exert similar effects on expression of cholesterol-transporting proteins such as Npc1l1. Selective activation of LXRβ may generate effects such as increased cholesterol absorption and elevated serum levels of apoB-containing lipoproteins, which seem to be counteracted by LXRα. Therefore, an intestinal LXRβ-specific pathway might exist in terms of cholesterol transportation in addition to the main pathway.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References
  10. Supporting Information

Intestinal absorption of cholesterol represents an important process in cholesterol homeostasis. Studies in several animal models demonstrated that reduction in atherogenic lipoproteins by inhibiting intestinal cholesterol absorption was associated with a clear atheroprotective effect [1, 2]. In humans, inhibition of cholesterol absorption by administration of ezetimibe, sometimes in combination with statins, resulted in a significant reduction in LDL cholesterol [3–5]. A clear reduction in major atherosclerotic events consistent with the achieved decrease in LDL cholesterol level has been seen in patients with chronic kidney disease [plasma or serum creatinine of >150 μmol L−1 (1.7 mg dL−1) in men and >130 μmol L−1 (1.5 mg dL−1) in women, regardless of dialysis] [6]. However, in other studies, such associations have been a matter of debate [7].

Intestinal cholesterol absorption is affected by several factors, and the regulatory mechanisms are to a large extent still unclear. Free cholesterol is present in the small intestine as part of micelles, which also contain phospholipids, fatty acids and bile acids. The biliary composition of bile acids is crucial for micelle formation. Hydrophobic bile acids such as cholic acid (CA) strongly favour micelle formation, whereas hydrophilic bile acids such as muricholic acid (MCA) form micelles less efficiently [8–11]. Cholesterol is taken up at the brush border membranes of enterocyte by Niemann–Pick C1-Like1 (Npc1l1) protein and transported into the cell [12]. Several pathways contribute to the further processing of the absorbed cholesterol. It can be transferred to chylomicrons following esterification by the enzyme Acat2, or effluxed by Abca1 and assembled into HDL particles, or resecreted back to the intestinal lumen by the apical transporters Abcg5 and Abcg8. The regulation of these processes is not well understood, but it is known that liver X receptors (LXRs) exert a powerful influence, at least in preclinical models [13].

The nuclear receptors LXRα (NR1H3) and LXRβ (NR1H2) are oxysterol-activated transcription factors with different tissue expression patterns. LXRs act as cholesterol sensors by stimulating a number of genes involved in cholesterol metabolism. Studies in genetically modified mice have demonstrated that LXRα is required for the maintenance of natural resistance towards dietary cholesterol; by contrast, LXRβ seems to be important for glucose metabolism, energy utilization and maintenance of body weight [14]. In several preclinical studies, activation of LXRs was reported to have atheroprotective effects [15]. It has been demonstrated that LXRβ may substitute for LXRα in reversing atherosclerotic lesions by raising plasma HDL cholesterol levels and promoting reverse cholesterol transport (RCT) [16]. More importantly, several animal studies have showed that activation of the LXRs helps to reduce intestinal cholesterol absorption, possibly by decreasing the expression of Npc1l1 and inducing expression of Abcg5/g8 [17–19]. Recently, Lo Sasso et al. demonstrated that intestinal-specific overexpression of LXRα in mice reduced intestinal cholesterol absorption by 50%. These mice also showed enhanced RCT, reduced cholesterol esters (CEs) in plasma very low-density lipoprotein (VLDL) and LDL particles and increased resistance to atherosclerosis [20]. However, none of the above-mentioned studies compared the different effects mediated by the two subtypes of LXR in the intestine; therefore, the individual effects on cholesterol absorption of LXRα and LXRβ are still not well defined. In the present study, we have investigated the significance of LXRα and LXRβ on the absorption of cholesterol in mice and evaluated the subsequent effects on cholesterol and bile acid metabolism.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References
  10. Supporting Information

Reagents

The synthetic LXR agonist GW3965 was kindly supplied by Tim Willson (GlaxoSmithKline, NC, USA). [5,6-3H] β-sitostanol was purchased from American Radiolabel Chemicals Inc. (St. Louis, MO, USA), and [4-14C] cholesterol from Amersham Biosciences (Uppsala, Sweden). Antibodies against Abca1, Idol, LDL receptor, Sr-bI, GADPH and β-actin were purchased from Abcam (Cambridge, UK), and antibodies against Cyp8b1 and Npc1l1 were from Santa Cruz, (CA, USA), and Novus Biologicals (Littleton, CO, USA), respectively. Dr Liqing Yun (Wake Forest University, NC, USA) kindly supplied the antibody against Abcg5. All primers for real-time PCR were purchased from CyberGene AB (Stockholm, Sweden).

Animals and diets

Six- to nine-week-old male wild-type (WT), LXRα knockout (LXRα−/−) and LXRβ knockout (LXRβ−/−) mice, bred as described [21], were housed in a temperature-controlled and pathogen-free room under a 12-h light/dark cycle with free access to food and water until the start of the experiments. During the study, mice were fed for 2 weeks with standard chow (0.025% cholesterol, wt/wt) (diet 1), a cholesterol-enriched diet (0.2% cholesterol, wt/wt; TD.07436; Harlan Teklad, Madison, WI, USA) (diet 2) or a cholesterol-enriched diet supplemented with 40 mg kg−1 day−1 of the LXR agonist GW3965 administered by gavage (diet 3). The mice were given either vehicle or GW3965 once a day for 4 days. At the end of the study period, all mice were fasted for 4 h with free access to water and then sacrificed. All animal experiments were approved by the ethical committee of Karolinska Institutet.

Tissue samples

Samples of liver tissue were snap-frozen in liquid nitrogen and stored at −70 °C. The small intestine was divided into four different regions, denoted S1–S4 from the most proximal to the distal part, and the mucosal cells were scraped off and immediately frozen in liquid nitrogen with TRIzol for RNA preparation.

Fractional intestinal cholesterol absorption

Intestinal cholesterol absorption was measured using the faecal dual-isotope ratio method essentially as described [22]. Briefly, mice were gavaged intragastrically with 100 μL peanut oil containing 1 μCi [4-14C]cholesterol and 2 μCi [5,6-3H]β-sitostanol. After gavage, each mouse was housed individually in a metabolic cage, with food and water available ad libitum. Faeces were collected for 24 h, and cholesterol was purified using the Folch method (chloroform/methanol 2 : 1, v/v).

Bile acid analysis

Analysis of the bile acid composition was conducted as previously described [8]. Briefly, the filled gallbladders were quickly removed and minced in saline with 1 mol L−1 KOH for overnight hydrolysis at 120 °C to cleave conjugated bile acids. After hydrolysis, the mixture was extracted twice with diethyl ether to remove the neutral steroids. The water phase was then acidified with 6 mol L−1 HCl, and the bile acids were extracted with diethyl ether. After evaporation of ether, bile acids were methylated by diazomethane and silylated with pyridine/hexamethylsilyl/disilazane (3 : 2 : 1, v/v/v) for 30 min at 60 °C. The solvent was evaporated under N2, and the material was dissolved in hexane. Analysis of the bile acid composition was performed using gas chromatography–mass spectrometry.

Faecal neutral sterol and bile acid quantification

Faeces from individual mice were collected for 24 h and allowed to dry before analysis. Neutral sterols and bile acids were extracted from 0.5 g faeces for each mouse. Before extraction, the faeces were homogenized and hydrolysed with 1 mol L−1 KOH at 120 °C. After hydrolysis, the mixture was extracted twice with hexane and diethyl ether for neutral sterols and bile acids, respectively. Gas chromatography was used to analyse the neutral sterols and bile acids [8].

Liver, bile and serum cholesterol analysis

Liver free and total cholesterol were quantified using the Folch method essentially as described previously [23]. In brief, 100 mg liver tissue was extracted overnight. Then, the organic phase was removed, dried and resolubilized in methanol/water (4 : 1, v/v) before being applied to an MFC-18 column (ISOLUTE; Sorbent, AB, Sweden). The solvent was then dried in N2, and cholesterol was silylated and redissolved in hexane. The remaining portion of the liver was dissolved in 1 mol L−1 NaOH for protein measurement as a reference for free and total cholesterol. Data were expressed in milligrams cholesterol per gram protein. For quantification of free and total cholesterol in bile, 2 μL bile from each mouse was used for the Folch extraction procedure as described above. Blood was collected by cardiac puncture and centrifuged twice at 15 000 g for serum collection. The lipoprotein profiles in the serum were analysed using size-exclusion chromatography as previously described [24].

Protein extraction and western blotting

Liver samples from individual animals were lysed in RIPA buffer in the presence of protease inhibitors [(1 protease inhibitor tablet/10 mL buffer (Roche Diagnostics, Indianapolis, IN, USA)] for whole-cell protein extraction. The lysates were centrifuged at 12 000 g for 2 min at 4 °C. After centrifugation, the supernatants were transferred to clean tubes and used for western blotting at a final concentration of 4 μg μL−1 in a mixture of loading buffer and dithiothreitol (DTT). Protein mixtures were separated by 4–12% SDS/polyacrylamide gel with MES or MOPS running buffer (Invitrogen, Life Technology, Stockholm, Sweden). Pooled protein mixtures for each treatment group were loaded at a volume of 10 μL per well to the gel (40 μg protein per well) and separated with electrophoresis for 1.5 h before blotting. For detection of target proteins, the blots were then incubated with specific primary and secondary antibodies in 5% fat-free milk blocking solution, and the results were recorded using a Molecular Imager® camera (Bio-Rad, Hercules, CA, USA). Densitometry analyses were performed for quantification of the results.

RNA isolation and real-time PCR analysis

Frozen tissue slices were homogenized in TRIzol (Life Technologies, Sweden) and purified using the MagMAX-96 for Microarrays Total RNA Isolation Kit (Applied Biosystems, Carlsbad, CA, USA). Oligo-dT-primed cDNA synthesis was carried out with 1 μg total RNA using the High Capacity Reverse Transcriptase Kit (Applied Biosystems). Quantitative real-time assays for mRNA measurements were performed using SYBRGreen in a 7500 Fast Real-Time PCR System (Applied Biosystems), and calculation of the threshold cycle (Ct) values was based on the average results of duplicate experiments. All final expressions of target genes were calculated relative to the housekeeping gene HPRT. Primer sequences for real-time PCR can be obtained upon request.

Statistics

All data are presented as mean ± SEM. Differences between groups were assessed by two-way analysis of variance (ANOVA) followed by post hoc comparison according to Fisher’s LSD test (STATISTICA software; StatSoft, Tulsa, OK, USA). Values of < 0.05 were considered statistically significant.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References
  10. Supporting Information

Activation of LXRβ increases the intestinal cholesterol absorption rate in LXRα−/− mice

To identify the individual effects of LXRs on the cholesterol absorption rate, WT, LXRα−/− and LXRβ−/− mice were fed a diet of standard chow (diet 1), a 0.2% cholesterol-enriched diet for 2 weeks (diet 2) or a cholesterol-enriched diet for 2 weeks supplemented with the LXR agonist GW3965 for the last 4 days (diet 3). The rate of cholesterol absorption was measured on the last day before the end of the experiments.

As expected, all three groups of mice showed comparable levels of cholesterol absorption with diet 1 (Fig. 1). It is interesting that LXRα−/− mice, but not WT or LXRβ−/− mice, had significantly enhanced cholesterol absorption when fed diet 2 and diet 3 (68.6% and 76.5%, respectively, compared to 53.4% with diet 1). By contrast, diet 2 (but not diet 3) resulted in a moderate reduction in cholesterol absorption in the LXRβ−/− mice compared to the WT mice (38.6% vs. 57.7%).

image

Figure 1.  Intestinal cholesterol absorption in wild-type (WT), LXRα−/− and LXRβ−/− mice. Fractional intestinal cholesterol absorption was measured using the faecal dual-isotope method in WT, LXRα−/− and LXRβ−/− mice. Animals were fed a standard chow diet, cholesterol diet for 14 days or cholesterol diet for 14 days with GW3965 in addition for the last 4 days. Data represent the mean ± SEM (n = 5–6).*Statistically significant (< 0.05) compared to mice of the same genotype fed standard chow diet;#statistically significant (P < 0.05) compared to WT mice fed the same diet. LXR, liver X receptors.

Download figure to PowerPoint

Activation of LXRβ by GW3965 reduces faecal neutral sterol excretion without affecting biliary cholesterol content

The results of faecal neutral sterol excretion analysis supported the notion that LXRβ activation increased intestinal cholesterol absorption (Fig. 2a). Faecal neutral sterol excretion increased significantly in WT and LXRβ−/− mice fed diet 2 or diet 3. By contrast, there were no changes observed in LXRα−/− mice fed diet 2. Instead, a significant reduction in faecal neutral sterol was observed with diet 3, which was in line with the increased rate of cholesterol absorption in these mice.

image

Figure 2.  Faecal neutral sterol excretion, liver cholesterol content and gallbladder bile cholesterol content in wild-type (WT), LXRα−/− and LXRβ−/− mice. (a) Faecal neutral sterol excretion, (b) cholesterol content in the liver and (c) total cholesterol in gallbladder bile were analysed individually by gas chromatography–mass spectrometry in WT, LXRα−/− and LXRβ−/− mice. Data represent mean ± SEM (= 4–6).*Statistically significant (P < 0.05) compared to mice of the same genotype fed standard chow diet; Δstatistically significant (< 0.05) compared to mice of the same genotype fed 0.2% cholesterol diet;#statistically significant (P < 0.05) compared to WT mice fed the same diet. LXR, liver X receptors.

Download figure to PowerPoint

Quantification of liver CE content showed that diet 2 led to accumulation of CE in the liver of WT and LXRβ−/− mice, which could be reversed by diet 3 (Fig. 2b). However, LXRα−/− mice showed a prominent elevation of liver CE content in response to diet 2, but no reduction was seen with diet 3.

We then analysed the cholesterol content in the gallbladder bile and the liver, as biliary cholesterol output also contributes to faecal excretion of neutral sterols. As shown in Fig. 2c, total cholesterol content in the gallbladder bile was significantly increased in the WT and LXRβ−/− mice, whereas no differences were observed in the LXRα−/− mice treated with diet 2 or diet 3. These results are probably dependent on changes in Abcg5 in the liver with diet 2 and diet 3 (see Fig. S1).

LXRα and LXRβ compensate for each other in the transcriptional regulation of intestinal Npc1l1 and Abcg5

Previous studies have demonstrated that cholesterol absorption depends on several factors such as the expression of the intestinal transporter Npc1l1, the formation of intraluminal bile acid micelles and the activity of the heterodimer Abcg5/g8 [12, 25]. In all three genotypes, similar mRNA levels of Npc1l1 were obtained in the proximal small intestine after diet 1 or diet 2 (Fig. 3a). Diet 3 resulted in a significant reduction in Npc1l1 transcripts in WT mice, indicating that pan-activation of LXRs might downregulate the transcription of Npc1l1. It is surprising that this reduction was also achieved in LXRα−/− and LXRβ−/− mice to a similar extent when only the remaining isoform of LXR was activated (LXRα or LXRβ). The above findings were confirmed by western blot analysis of Npc1l1 (Fig. 3b,c). These findings indicate that LXRα and LXRβ are able to compensate for each other in the downregulation of Npc1l1.

image

Figure 3.  mRNA and protein expression of Npc1l1 and Abcg5 in the proximal intestine in wild-type (WT), LXRα−/− and LXRβ−/− mice. (a) Npc1l1 and (d) Abcg5 mRNA expressions in the proximal small intestine in WT, LXRα−/− and LXRβ−/− mice were measured by real-time PCR. HPRT was used as a housekeeping gene. Data represent mean ± SEM (= 5–6). (b,c) show the protein expression of Npc1l1 in the proximal intestine analysed by western blotting. (e,f) show the western blot analysis of Abcg5 in the proximal intestine. All western blot analyses were performed with pooled protein samples from each treatment group. β-Actin was used as an internal loading control. *Statistically significant (< 0.05) compared to mice of the same genotype fed standard chow diet; Δstatistically significant (< 0.05) compared to mice of the same genotype fed 0.2% cholesterol diet. LXR, liver X receptors.

Download figure to PowerPoint

We also found that LXRα and LXRβ seemed to exert redundant functions for the transcriptional upregulation of Abcg5 and Abcg8. As shown in Fig. 3d, diet 2 and diet 3 resulted in elevated levels of Abcg5 mRNA in all three groups of mice; likewise, no differences were seen between the three groups with diet 1. Because Abcg5 and Abcg8 are usually transcribed in the same manner, a similar level would be expected for Abcg8 mRNA as for Abcg5. This was confirmed by quantitative analysis of pooled Abcg8 mRNA (data not shown). Western blotting showed an induction in all three genotypes with diet 2 (Fig. 3e,f); these protein results were consistent with the mRNA expression. It is interesting that LXRα−/− and LXRβ−/− mice fed diet 3 failed to induce Abcg5 protein expression, which is not in line with the mRNA data.

Separate activation of LXRβ increases cholesterol content of serum apolipoprotein (apo)B-containing particles

Previous studies have shown that pan-activation of LXRs was able to increase the plasma levels and size of HDL cholesterol [26]. In the present study, diet 2 generated a modest but significant elevation of the total cholesterol in WT and LXRα−/− mice, whereas no increase was detected in LXRβ−/− mice. In comparison with diet 2, diet 3 led to a dramatic increase in total cholesterol in all three groups (Fig. 4a).

image

Figure 4.  Serum lipoprotein profiles in WT, LXRα−/− and LXRβ−/− mice. Total serum cholesterol (a), and serum lipoprotein profiles of WT (b) LXRα−/− (c) and LXRβ−/− (d) animals receiving diets 1, 2 and 3. Data represent mean ± SEM (= 5–6).*Statistically significant (< 0.05) compared to mice of the same genotype fed standard chow diet;Δstatistically significant (< 0.05) compared to mice of the same genotype fed 0.2% cholesterol diet. LXR, liver X receptors.

Download figure to PowerPoint

Separation of lipoproteins by size-exclusion chromatography revealed different response patterns among the three genotypes when treated with diet 2 or diet 3. In WT and LXRβ−/− mice, diet 2 mainly induced larger particles such as VLDL and, to a less extent, LDL particles. As expected, diet 3 led to a dramatic increase in the level of HDL particles (Fig. 4b,d). By contrast, LXRα−/− mice displayed a different pattern: both diet 2 and diet 3 produced significant increases in apoB-containing lipoprotein particles (VLDL and LDL), an effect that resulted in more-atherogenic lipoprotein profiles (Fig. 4c).

According to previous studies in mice, hepatic Abca1 mediates the production of 70–80% of the total HDL pool in plasma, while the additional 20–30% is probably generated by intestinal Abca1 [27]. Analysis of Abca1 mRNA levels in the proximal intestine did not reveal any genotype-dependent differences when the animals were fed either diet 1 or diet 2. Diet 3 resulted in a significant induction of Abca1 mRNA to a similar extent in all three strains (Fig. 5a). Of interest, hepatic Abca1 mRNA levels (Fig. 5b) did not differ between the three genotypes, but clear increases in liver Abca1 protein were seen in the LXRα−/− mice fed diet 2 or diet 3 (Fig. 5c).

image

Figure 5.  Abca1 expression of liver and proximal intestine in wild-type (WT), LXRα−/− and LXRβ−/− mice. (a) Proximal intestinal Abca1 mRNA expression was measured by real-time PCR. HPRTwas used as a housekeeping gene. Data represent mean ± SEM (= 5–6). (b) Liver Abca1 mRNA expression was measured by real-time PCR.HPRTwas used as a housekeeping gene. Data represent mean ± SEM (= 5–6). (c) Liver Abca1 protein expression was analysed by western blotting using pooled protein samples from each treatment group. β-Actin was used as internal loading control. Liver (d) SR-BI and (e) LDLr mRNA expressions were analysed by real-time PCR.HPRTwas used as a housekeeping gene. Data represent mean ± SEM (= 5–6).*Statistically significant (< 0.05) compared to mice of the same genotype fed standard chow diet;Δstatistically significant (< 0.05) compared to mice of the same genotype fed 0.2% cholesterol diet.#statistically significant (< 0.05) compared to WT mice fed the same diet. LXR, liver X receptors.

Download figure to PowerPoint

Cholesterol transfer from plasma HDL particles into the hepatocytes is mediated in part by the hepatic SR-B1 transporter. Here, we found an induction of Sr-bI mRNA in WT and LXRβ−/− mice with diet 2, whereas no difference was seen in LXRα−/− mice. Diet 3 reduced SR-B1 mRNA in WT and LXRβ−/− mice, but no change was seen in LXRα−/− animals (Fig. 5d). Analysis of protein expression of Sr-bI confirmed the mRNA data (Fig. S2). With regard to expression of LDLr mRNA, all three groups showed reduced levels with diet 3 (Fig. 5e). Western blot analysis demonstrated that LDLr protein expression was reduced in LXRα−/− mice (Fig. S3), which might contribute to the increased levels of serum LDL cholesterol in these animals.

LXRα and LXRβ modify bile acid composition and excretion in response to stimulation by LXR ligands

The relative proportion of CA and β-MCA, the two major bile acids in mice, is considered to influence the formation of luminal micelles and thereby cholesterol absorption [8]. We found that diet 2 and diet 3 modified the bile acid composition in all three genotypes in different ways. The bile acid composition was similar between the different groups of mice with diet 1 (Fig. 6a). Diet 2 significantly reduced the ratio of CA/β-MCA in WT and LXRβ−/− mice, whilst only a moderate reduction was seen in LXRα−/− mice. Diet 3 resulted in a further decrease in the CA/β-MCA ratio in WT mice, but not in the other two groups. Thus, LXRα−/− mice showed the highest CA/β-MCA ratio with diet 2 and diet 3 compared to WT and LXRβ−/− animals.

image

Figure 6.  Bile acid metabolism in wild-type (WT), LXRα−/− and LXRβ−/− mice. (a) Fractional bile acid composition (CA/β-MCA) in the gallbladder bile of WT, LXRα−/− and LXRβ−/− mice was measured by gas chromatography–mass spectrometry. Data represent mean ± SEM (= 5–6). (b) Faecal bile acids were analysed and quantified by gas chromatography in WT, LXRα−/− and LXRβ−/− mice. Data represent mean ± SEM (= 4–5). DCA, deoxycholic acid; UDCA, ursodeoxycholic acid; LCA, lithocholic acid; CDCA, chenodeoxycholic acid; CA, cholic acid; MCA, muricholic acid. (c) Cyp8b1 mRNA expression in the liver was measured by real-time PCR.HPRTwas used as a housekeeping gene. Data represent mean ± SEM (= 5–6). (d) Cyp8b1 protein expression in the liver was quantified by western blotting using pooled samples from each treatment group.β-Actin was used as internal loading control. (e) Cyp7a1 mRNA expression in the liver was measured by real-time PCR.HPRTwas used as a housekeeping gene. Data represent mean ± SEM (= 5–6). (f) Ostα, Ostβ and Ibat mRNA expressions in the distal small intestine were analysed by real-time PCR. Data represent pooled samples from each treatment group.HPRTwas used as a housekeeping gene. Statistical analysis in this figure is not applicable.*Statistically significant (< 0.05) compared to mice of the same genotype fed standard chow diet;Δstatistically significant (< 0.05) compared to mice of the same genotype fed 0.2% cholesterol diet;#statistically significant (< 0.05) compared to WT mice fed the same diet. LXR, liver X receptors.

Download figure to PowerPoint

The total bile acid excretion in faeces was enhanced in WT and LXRβ−/− mice fed diet 2 and increased further with diet 3 (Fig. 6b), indicating an increased bile acid synthesis. It is of particular interest that LXRα−/− mice seemed to create a more hydrophobic bile acid profile than the other two genotypes, as the excretion of MCA was much lower in the LXRα−/− mice fed diet 2 or diet 3. This would favour a higher cholesterol absorption rate in these mice.

Selective inactivation of LXRα or LXRβ affects the transcription of genes coding for bile acid synthetic enzymes and transporters

The rate-limiting enzyme for the synthesis of CA is Cyp8b1, the expression of which is regulated by nuclear receptors (i.e. PPARα and HNF4α), thyroid hormones and insulin. We found a significant reduction in Cyp8b1 mRNA in both LXRα−/− and LXRβ−/− mice compared to WT animals fed diet 1 (Fig. 6c). A significant reduction was observed in the WT mice fed diet 2 or diet 3, and these changes were in line with those in bile acid composition. The LXRα−/− mice showed lower levels of Cyp8b1 mRNA when fed diet 2, but no further reduction was observed with diet 3. The mRNA levels matched the amounts of Cyp8b1 protein as shown in Fig. 6d. It is interesting that, although a clear reduction in Cyp8b1 expression was observed in LXRα−/− and LXRβ−/− mice compared to WT mice fed diet 1, no differences in bile acid composition were seen amongst these animals.

The rate-limiting enzyme for bile acid synthesis is considered to be the P-450 cytochrome Cyp7a1, which is demonstrated as a direct target gene for LXRα in mice. In this study, we found significantly reduced mRNA levels of Cyp7a1 in LXRα−/− and LXRβ−/− mice fed diet 1, compared to WT mice (Fig. 6e). Cyp7a1 levels were reduced in WT and LXRα−/− mice fed diet 2, with no further changes recorded with diet 3. Diet 2 did not significantly change the Cyp7a1 mRNA levels in LXRβ−/− mice, but a stronger induction was seen with diet 3.

The retrieval of bile acids from the distal intestine is another important component of bile acid metabolism, as more than 95% of the bile acids are reabsorbed through the enterohepatic circulation [28]. Ibat, Ostα and Ostβ are the main transporters in this process. Pooled mRNA data showed that the LXRα−/− and LXRβ−/− mice had lower expression levels of Ibat than the WT mice (Fig. 6f), which might indicate diminished reabsorption of bile acids in the distal ileum. In addition, higher levels of Ostα/β in LXRα−/− and LXRβ−/− mice were observed with diet 1, with decreased mRNA levels following diet 2 and diet 3. Thus, there could be differences in the bile acid reabsorption process between LXRα−/− and LXRβ−/− mice.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References
  10. Supporting Information

In the present study, we aimed to define the individual roles of LXRα and LXRβ in the regulation of intestinal cholesterol absorption. We utilized mice genetically depleted of either LXRα or LXRβ and challenged them with dietary cholesterol (0.2%) with or without LXR pan-agonist GW3965. This allowed separate activation of each LXR subtype and their individual roles to be examined. We found the following: (i) stimulation of LXRβ alone increased intestinal cholesterol absorption, reduced faecal neutral sterol excretion and increased hepatic CE content in LXRα−/− mice; (ii) activation of LXRβ by GW3965 reduced faecal neutral sterol excretion without affecting biliary cholesterol excretion in LXRα−/− mice; (iii) LXRα and LXRβ compensated for each other in the transcriptional regulation of intestinal Npc1l1; (iv) increased cholesterol absorption in LXRα−/− mice was accompanied by higher levels of apoB-containing lipoprotein particles; and (v) compared to the other two genotypes, LXRα−/− mice had a more hydrophobic bile acid profile in faeces and gallbladder bile.

The absorption of cholesterol is regarded as an important determinant of plasma cholesterol levels, and increased cholesterol absorption has been recognized as a proatherogenic factor. Reduced cholesterol absorption, even to a moderate extent, has been associated with large beneficial effects with respect to plasma lipid levels and development of atherosclerosis in animal models [2, 29].

In the present study, however, we did not find significant changes in cholesterol absorption in WT mice treated with 0.2% cholesterol ± GW3965. By contrast, a clear induction of cholesterol absorption was achieved when LXRβ was exclusively activated as demonstrated in the LXRα−/− mice fed 0.2% cholesterol ± GW3965 (diet 2 and diet 3). We also found a reduction in the absorption rate when LXRα was activated by 0.2% cholesterol alone (diet 2) in the LXRβ−/− mice. This observation is in agreement with the results of a previous study in which a 40% reduction in cholesterol absorption was reported in an activated LXRα intestinal-specific transgenic mouse model [20]. Taken together, the above results suggest that LXRα and LXRβ might counteract each other in the process of cholesterol absorption, whereby selective activation of LXRβ enhances cholesterol uptake in the small intestine. Therefore, the absence of significant changes in cholesterol absorption in the WT mice might be due to the counterbalancing effects of both stimulated LXR subtypes. On the other hand, the observed effects of GW3965 on cholesterol absorption might depend on the treatment period and doses used. In other studies using synthetic LXR agonists (GW3965 and T0901317), no significant changes in cholesterol absorption rates were found [30, 31].

Our results regarding cholesterol absorption were supported by the finding of significant reduction in faecal neutral sterols exclusively in the LXRα−/− mice fed diet 3. There was no indication of a change in biliary cholesterol output. Based on the above finding, we assume that the reduction in neutral sterols in the faeces is attributable to increased uptake of cholesterol by the small intestine. Analysis of the liver cholesterol content confirmed previous results [32], showing that LXRα−/− mice were unable to withstand dietary cholesterol loading. The fact that esterified cholesterol accounted for more than 90% of the total accumulated cholesterol might partly be explained by increased cholesterol absorption in the LXRα−/− mice, causing increased cholesterol transport to the liver in the chylomicron remnants.

Intestinal cholesterol absorption involves several cholesterol transporters in the small intestine, including Npc1l1 and Abcg5/g8 [33, 34]. Administration of dietary cholesterol and GW3965 led to reduced expression of Npc1l1, regardless of whether both or either isoform of LXRs is activated. It has been reported that LXRs downregulate Npc1l1, although the exact mechanisms are unclear [34].

Several animal models have convincingly shown that deficiency of the protein Npc1l1 leads to a significant reduction in intestinal cholesterol absorption [12, 35, 36]. However, we found comparable mRNA levels of Npc1l1 in both the experimental genotypes despite significant differences in the cholesterol absorption rate. One explanation for this discrepancy could be that cholesterol uptake does not correlate with transcriptional expression of Npc1l1. Several recent studies have demonstrated that critical steps in cholesterol uptake also include the shuttling of Npc1l1 protein from the endocytic recycling compartment to the plasma membrane, and internalization and transportation of Npc1l1-cholesterol within the cell [37, 38]. Therefore, the rate of cholesterol absorption would be affected not only by the expression of Npc1l1 but also by modifications of the other steps in this process. LXRα and LXRβ might influence Npc1l1 trafficking, but, to date, no experimental evidence has been provided to support this possibility. In addition, as targeted disruption of Npc1l1 in mice does not completely eliminate intestinal uptake of cholesterol, it is also possible that a hitherto unknown pathway for cholesterol absorption might exist, which may respond to selective stimulation of LXRβ (Fig. 7).

image

Figure 7.  Proposed mechanisms of the effects of LXRα and LXRβ on the regulation of cholesterol transport processes in the small intestine. LXRα and LXRβ similarly regulate the expression of cholesterol transporters, reducing the mRNA level of Npc1l1 whilst elevating mRNA levels of Abcg5/g8 and Abca1. This leads to reduced cholesterol uptake and enhanced cholesterol excretion and efflux, which ultimately prevent cholesterol accumulation in the enterocytes. However, selective activation of LXRβ increases cholesterol uptake (as shown by the dotted line with a question mark), and this process seems to be counteracted by LXRα when both isoforms are functionally active. LXR, liver X receptors.

Download figure to PowerPoint

We also found that stimulation of LXRβ alone by dietary cholesterol and GW3965 was associated with enhanced levels of apoB-containing lipoproteins (VLDL and LDL) and hepatic accumulation of cholesterol, probably due to a combined effect of increased intestinal cholesterol absorption, failure to upregulate Cyp7a1 and the subsequent hepatic CE accumulation [32]. In contrast to LXRβ stimulation, the activation of LXRα induced a distinct elevation of HDL particles. Hepatic Abca1 is regarded as a major factor in the supply of the plasma HDL pool [39]. As activation of either LXRα or LXRβ increased the mRNA levels of intestinal Abca1, activation of either subtype should be expected to stimulate the production of HDL from the intestine. However, without a functional LXRα, larger quantities of cholesterol are absorbed and transferred into chylomicrons. The increased CE will lead to increased cholesterol secretion into VLDL, and finally, the reduced expression of hepatic LDL receptors might further contribute to the increased LDL cholesterol levels seen in LXRα−/− mice.

The bile acid composition also plays an important role in cholesterol absorption as hydrophobic bile acids are key elements for intraluminal micelle formation [28, 40]. When fed diet 3, LXRα−/− mice showed a more hydrophobic gallbladder bile acid profile compared to the other two genotypes, although the quantities of CA and β-MCA were reduced. Accordingly, a more prominent hydrophobic faecal bile acid profile was also detected, as significantly less MCA was identified compared to the other two genotypes. This might explain the increased cholesterol absorption rate in the LXRα−/− animals.

Our results indicate that the effects of LXRα and LXRβ on the transcription of certain genes vary owing to their responses to specific ligands and which organs are involved. LXRs have gained considerable attention as putative candidates to protect against atherosclerosis owing to their ability to increase RCT [26, 41]. However, preclinical studies aimed at activating LXRs have been largely unsuccessful as hepatic activation of LXRα was found to increase triglyceride levels – a risk factor for the development of cardiovascular disease. LXRβ-specific agonists have provided promising results in mice by increasing RCT without increasing the production of triglycerides in liver [42]. However, whether a therapeutic intervention targeting LXRs should be directed against LXRβ is debatable, as our observations indicate that selective stimulation of LXRβ might have detrimental effects. Consequently, it is of great importance that the individual effects of LXRα and LXRβ are carefully defined and evaluated in order to avoid unwanted side effects in future therapeutic approaches.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References
  10. Supporting Information

We are grateful to Mrs Lilian Larsson and Mrs Anita Lövgren Sandblom for excellent technical assistance. K.R.S. holds a research assistant professorship from the Swedish Research Council (contract no. 522-2008-3745). This study was financed by grants from the Swedish Research Council – Medical Branch, Swedish Heart and Lung Foundation, Karolinska Institutet Foundations, Novo Nordic Research Foundation, Swedish Diabetes Foundation and KaroBio Research Foundation, Robert A. Welch Foundation and The Emerging Technology Fund of Texas.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References
  10. Supporting Information

Figures S1–S4. Protein expressions in the liver of Abcg5 (1), Sr-bI (2), Ldlr (3) and Idol (4) were analysed by western blotting in WT, LXRα−/− and LXRβ−/− mice.

Figure S5. Food consumption was recorded daily for each treatment group of 4–5 mice for 2 weeks.

FilenameFormatSizeDescription
joim2529_sm_supplementary-legands.docx404KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.