Liver X receptor-α (LXRα) is considered a master regulator of hepatic lipid metabolism; however, little is known about the link between LXR activation, hepatic insulin signaling, and very low-density lipoprotein (VLDL)-apolipoprotein B (apoB) assembly and secretion. Here, we examined the effect of LXRα activation on hepatic insulin signaling and apoB-lipoprotein production. In vivo activation of LXRα for 7 days using a synthetic LXR agonist, TO901317, in hamsters led to increased plasma triglyceride (TG; 3.6-fold compared with vehicle-treated controls, P = 0.006), apoB (54%, P < 0.0001), and VLDL-TG (eightfold increase compared with vehicle). As expected, LXR stimulation activated maturation of sterol response element binding protein-1c (SREBP-1c) as well as the SREBP-1c target genes steroyl CoA desaturase (SCD) and fatty acid synthase (FAS). Metabolic pulse-chase labeling experiments in primary hamster hepatocytes showed increased stability and secretion of newly synthesized apoB following LXR activation. Microsomal triglyceride transfer protein (MTP) mRNA and protein were unchanged, however, likely because of the relatively short period of treatment and long half-life of MTP mRNA. Examination of hepatic insulin-signaling molecules revealed LXR-mediated reductions in insulin receptor (IR)β subunit mass (39%, P = 0.014) and insulin receptor substrate (IRS)-1 tyrosine phosphorylation (24%, P = 0.023), as well as increases in protein tyrosine phosphatase (PTP)1B (29%, P < 0.001) protein mass. In contrast to IRS-1, a twofold increase in IRS-2 mass (228%, P = 0.0037) and a threefold increase in IRS-2 tyrosine phosphorylation (321%, P = 0.012) were observed. In conclusion, LXR activation dysregulates hepatic insulin signaling and leads to a considerable increase in the number of circulating TG-rich VLDL-apoB particles, likely due to enhanced hepatic assembly and secretion of apoB-containing lipoproteins.
- liver X receptor
- very low-density lipoprotein
nuclear liver x receptors (LXRs) act as intracellular sensors for sterols (22) and, in response to ligands, induce transcriptional responses to maintain cholesterol, lipid, and glucose homeostasis (23, 25, 51). LXR is a central player in energy homeostasis, as indicated by its putative involvement in lipogenesis, gluconeogenesis, lipoprotein metabolism, and glucose uptake (46). The initial rationale for studies of LXR and LXR agonists was based on observations of their beneficial effects on reverse cholesterol transport, a process whereby cholesterol is transported from extrahepatic tissues to the liver for eventual excretion through the bile (9). Synthetic LXR agonists have therefore been designed with the intention of treating disorders such as atherosclerosis and diabetes (8, 17).
The two isoforms of LXR, α and β, are highly homologous and form obligate heterodimers with retinoid X receptor (RXR) (11, 55). LXR is part of a large family of ligand-activated nuclear transcription factors that includes the farnesoid X receptor (FXR). Hepatic lipogenic genes known to be under direct LXR transcriptional control include fatty acid synthase (Fas), stearoyl CoA desaturase (Scd), and the key lipogenesis regulator steroid response element binding protein (Srebp)-1c (39).
Recent studies have shown that LXR activation leads to enhanced de novo lipogenesis via induction of Srebp-1c, a key regulator of lipid biosynthesis, in both mice and hamsters (3, 23, 39). Importantly, it has been shown that both LXRα and insulin stimulate Srebp-1c transcription; however, full induction of the transcriptionally active mature form of Srebp-1c protein requires insulin (7, 19, 41). At present it is unclear whether activation of LXR perturbs hepatic insulin sensitivity, leading to downstream changes such as those observed following Srebp-1c activation. The mature nuclear forms of SREBP control a coordinated program of fatty acid biosynthesis in the liver, as evidenced by the observation that SREBP overexpression induces transcription of Fas and Scd-1 mRNA (41). Unfortunately, the potent lipogenic effects of LXRα have precluded the therapeutic usage of LXR agonists in the treatment of hypercholesterolemia.
LXR's role as a key regulator of lipid metabolism suggests a potential role for LXR in regulating hepatic lipoprotein metabolism. There is little evidence as to whether LXR activation may have a direct effect on regulation of apoB or apoB-containing lipoproteins. Since insulin is known to directly regulate both the hepatic production of apoB-containing lipoproteins (32) as well as directly increase levels of Lxrα mRNA in a dose-dependent manner (51), a link between the activation state of LXRα and hepatic apoB assembly and secretion is likely. There is one report suggesting that LXR activation may increase the size of secreted VLDL (16). It is, however, unclear whether LXR can modulate hepatic apoB biosynthesis or the number of apoB-containing VLDL particles.
Considering the important link between LXRα, SREBP-1c, and increased hepatic lipogenesis, we postulated that LXRα activation may influence the hepatic assembly and secretion of apoB-containing lipoproteins via changes in hepatic lipogenesis as well as hepatic insulin signaling cascades. Following initial activation of the insulin receptor (IR), hepatic insulin signaling is dependent on IR-mediated tyrosine phosphorylation of the insulin receptor substrates IRS-1 and IRS-2. Once activated, these molecules link to downstream targets to mediate the metabolic effects of insulin. Although these proteins have been demonstrated to serve complementary roles in insulin signaling and glucose metabolism, IRS-2 has been suggested to be of particular importance to hepatic regulation of these processes (35) and thus may be an important target for LXR-mediated regulation of hepatic insulin signaling. Furthermore, the ability of LXR to modulate the insulin-regulatory protein tyrosine phosphatase (PTP)1B has not been demonstrated. PTP1B is a key negative regulator that attenuates insulin signaling through dephosphorylation of the active forms of IR, IRS-1, and IRS-2. (26). In the present study, LXRα was stimulated in vivo in a Syrian golden hamster model by use of the LXR agonist, TO901317. Evidence is provided for LXRα-mediated stimulation of hepatic production of apoB-containing lipoproteins as well as complex perturbations of hepatic insulin signaling cascades.
MATERIALS AND METHODS
Tissue culture medium (DMEM) was purchased from Wisent (Montreal, QC, Canada). FBS, liver perfusion media, liver digest media, hepatocyte wash media, and Williams E media were obtained from Invitrogen Life Technologies (Burlington, ON, Canada). Trasylol (aprotinin) was from Bayer (Etobicoke, ON, Canada). All chemicals used for SDS-PAGE were from Bio-Rad (Mississauga, ON, Canada). All fine chemicals were obtained from Sigma-Aldrich (St. Louis, MO).
All animal experiments were approved by the animal ethics committee of the Hospital for Sick Children and were conducted according to national guidelines. Male Syrian golden hamsters (Charles River, Wilmington, MA) weighing between 80 and 100 g were housed individually on alternating 12-h light and dark cycles with free access to food (standard chow) and water. After a week of acclimatization and subsequent collection of baseline blood samples following a 5-h fast, animals were treated by oral gavage with either 1% carboxymethylcellulose (CMC) vehicle or 50–75 mg/kg LXR agonist (TO901317; dissolved in 1% CMC) once daily for 4–7 days. Hamster weights were monitored throughout the treatment period. All surgical procedures, including hepatocyte isolation, were conducted under 4% isoflurane-induced general anesthesia immediately following the final agonist dose as described previously (50). Briefly, livers were isolated from the circulatory system; the thoracic aorta, the caudal vena cava, the abdominal aorta, and the abdominal vena cava were tied off with sutures. The portal vein was then severed, and livers were perfused via the inferior vena cava with 50 ml of liver perfusion medium followed by 25 ml of liver digest medium at 42°C. Following perfusion, the liver was excised and minced in hepatocyte wash medium. Digested liver tissue was filtered through a cell strainer (100 μm), and the released hepatocytes were pelleted by centrifugation (60 g, 3 min), washed three times in hepatocyte wash medium, and resuspended in attachment media (Williams E containing 5% FBS, 1 μg/ml insulin, 0.1% penicillin-streptomycin). Cells were seeded on Primaria cell culture plates (BD Biosciences, Bedford, MA) at a density of 1–1.5 million cells/35-mm dish and incubated for ∼3 h (37°C, 5% CO2) to facilitate attachment. Viability of hepatocytes ranged from 75 to 90% as measured by Trypan blue dye exclusion; hepatocytes were not used for experiments if the viability was less than 70%. For experiments performed using liver tissue, tissues were snap frozen in liquid N2 immediately following collection and stored at −80°C for subsequent analyses.
Following a 5-h fast, blood was collected from anesthetized animals in heparin-coated tubes and centrifuged at 3,000 g for 10 min at 4°C to obtain plasma. Plasma glucose, cholesterol, alanine aminotransferase (ALT), aspartate aminotransferase (AST), and TG levels were determined on a clinical chemistry analyzer (VITROS 950, Ortho-Chemical Diagnostics, Rochester, NY).
Triglyceride and cholesterol mass measurements.
Approximately 300 mg of liver tissue was added to 20 volumes of a 2:1 chloroform-methanol mixture and incubated for 24 h at room temperature. Following the incubation period, 0.2 volumes of 0.9% NaCl were added to the solvent mixture. The samples were thoroughly vortexed then centrifuged at 2,000 rpm for 3 min. The upper aqueous phase was removed and the combined solvent layer was allowed to evaporate. The dried lipids were resuspended in 1 ml of 100% ethanol and TG and cholesterol concentrations were determined by use of commercially available kits from Randox (Mississauga, ON, Canada) as per the manufacturer's instructions. Lipid data are expressed in milligrams lipid per gram of liver tissue.
Analysis of lipoproteins by fast-pressure liquid chromatography.
Lipoprotein subclasses were separated by automated size exclusion chromatography. Plasma samples (100–200 μl) were injected onto a Superose 6 10/200 GL column (Amersham Pharmacia Biotechnology, Piscataway, NJ), at a flow rate of 0.5 ml/min. Fractions were eluted from the column with use of an eluant containing 10 mM Tris, 150 mM NaCl, 2 mM CaCl2, 100 μM DTPA, 0.02% NaN3, pH 7.4. Fractions were analyzed for cholesterol, TG, and glycerol blank (Bayer Diagnostics). Standards were derived from fresh human plasma with known cholesterol and TG concentrations. Total lipid concentrations for each lipoprotein interval were determined as the sum of the lipids in each fraction of that interval.
For SREBP experiments, whole liver was homogenized in buffer B (3 mM imidazole, 250 mM sucrose at pH 7.4 with 1 mM PMSF, 100 kallikrein-inactivating units/ml aprotinin). Liver homogenates were then centrifuged (2,200 g, 4°C) to isolate the nuclei. Pelleted nuclei were lysed in buffer C [phosphate-buffered saline (PBS) containing 1% NP40, 1% deoxycholate, 5 mM EDTA, 1 mM EGTA, 1 mM PMSF, 100 kallikrein-inactivating units/ml aprotinin]. The supernatant was used to isolate microsomes. For all other immunoblotting experiments, liver tissue was lysed in buffer D (150 mM NaCl, 10 mM Tris, pH 7.4, 1 mM EDTA, 1 mM EGTA, 2 mM PMSF, 10 μg/ml aprotinin). Lysates containing equivalent amounts of total protein were resolved by SDS-PAGE and transferred onto PVDF membranes (PerkinElmer Life Sciences, Woodbridge, ON, Canada). IRβ, IRS-1, IRS-2, SREBP-1, and p-Tyr (py99) antibodies were from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-FAS was from Novus Biologicals (Littleton, CO). Anti-PPARα was from Research Diagnostics (Concord, MA), and anti-PTP1B was from Upstate Cell Signaling (Lake Placid, NY). Anti-MTP antibody was a gift from Dr. Andre Theriault, University of Hawaii; rabbit anti-hamster apoB antiserum was prepared for our laboratory by Lampire Biological Laboratories (Pipersville, PA) by use of purified hamster LDL. Bound horseradish peroxidase-linked secondary antibodies were incubated with ECL (PerkinElmer Life Sciences, Boston, MA) and proteins were visualized following exposure of the membranes to Kodak X-Omat Blue XB-1 film (Eastman Kodak, Rochester, NY). Films were developed and quantitative analysis was performed by densitometry (Gel Doc Documentation System, Bio-Rad Laboratories).
Determination of tyrosine phosphorylation of the insulin receptor, IRS-1, and IRS-2.
Hamsters were treated in vivo with 5 U/kg of insulin for 2 min via the inferior vena cava. Livers were excised and homogenized in solubilizing buffer D (as outlined above) supplemented with phosphatase inhibitors (100 mM sodium fluoride, 10 mM sodium pyrophosphate, and 2 mM sodium vanadate). Hepatic cell lysates were immunoprecipitated with specific antibodies for the IR β-subunit, IRS-1, or IRS-2 (1 μg antibody/500 μg total protein from cell lysate). Immunoprecipitated proteins were analyzed by immunoblotting with an anti-pTyr (pY99) antibody as described previously (2). Membranes were reprobed and the level of phosphorylated tyrosine was normalized to the mass of the particular immunoprecipitated protein.
Metabolic labeling and immunoprecipitation.
Isolated primary hepatocytes were preincubated in cysteine-free and methionine-free medium (Sigma-Aldrich, Oakville, ON, Canada) at 37°C for 1 h followed by incubation with 100 μCi/ml of [35S]methionine (PerkinElmer Life Sciences) for 45 min. The medium was discarded and cells were washed twice with PBS. Chase medium (20% FBS DMEM) was then added and cells were incubated for an additional 2 h. The medium was collected and cells were lysed in buffer C. ApoB was immunoprecipitated from both cells and media as described previously by use of an anti-hamster apoB antibody. Samples were resuspended in 100 μl SDS-PAGE Laemmli sample buffer (28), and apoB100 was resolved on 5% SDS-PAGE gels. The gels were fixed, incubated with Amplify (Amersham, Baie d'Urfe, QC, Canada), and then dried and exposed to Kodak X-Omat Blue XB-1 film at −80°C. The radioactivity in the apoB100 band was quantified in a beta liquid scintillation counter following digestion of the excised gel bands by addition of 200 μl 60% perchloric acid and 400 μl hydrogen peroxide and incubation at 60°C for 18 h.
Ex vivo labeling of newly synthesized lipids.
Primary hepatocytes were incubated with 5 μCi/ml [3H]acetate for 18 h to assess synthesis and secretion of cholesterol ester. TG synthesis and secretion were analyzed by labeling primary hamster hepatocytes for 18 h with 5 μCi/ml [9,10-3H]oleic acid conjugated to BSA (47). Following labeling, cells and media were extracted with hexane-isopropanol (3:2), and the total lipid extract was dried, dissolved in hexane, and applied to a thin layer chromatography (TLC) plate (EMD Chemicals, Gibbstown, NJ). The TLC plates were developed with petroleum ether-ethyl ether-acetic acid (90:10:1). The lipids were stained with iodine vapor and identified on the basis of comparison to lipid standards (Sigma-Aldrich). The spots identified on the TLC plates were excised and quantified with a liquid scintillation counter.
Tissue collection, RNA extraction, and cDNA synthesis.
Total RNA was extracted from homogenized liver samples by using a commercially available RNA isolation kit (RNeasy Mini kit; Qiagen). All samples were treated with DNase during the extraction protocol. RNA quality was assessed visually by resolving extracted RNA on MOPS-buffered 1% agarose gels. RNA was deemed to be of acceptable quality if both the 28S rRNA and 18S rRNA were visible as clear, sharp bands in an approximate ratio of 2:1. Total RNA (0.2 μg) was converted to single-stranded cDNA using TaqMan Reverse Transcription (RT) Reagents (Applied Biosystems, Foster City, CA). Random hexamers were used to prime the RT reaction, with 18S rRNA as an internal reference. Samples were preincubated at 25°C for 10 min, RT performed at 37°C for 60 min and the enzyme inactivated by heating the samples at 95°C for 5 min.
Real-time quantitative PCR analysis.
Messenger RNA levels of a number of genes (Table 1) were assessed by real-time quantitative RT-PCR. All PCR reactions were performed in a total volume of 50 μl and included the following components: cDNA derived from 10 ng of total RNA, forward and reverse primers at a final concentration of 400 nM, and 25 μl of SYBR Green PCR Master Mix (Applied Biosystems). PCR cycling parameters were as follows: denaturation for 10 min at 95°C, then 15 s at 95°C followed by annealing and elongation at 60°C for 1 min repeated for 40 cycles. Relative quantities of mRNA were calculated from threshold cycle for amplification (CT) values by the comparative CT method (31) using 18S rRNA as an internal reference. Primer pairs for real-time PCR were designed using Primer3 software (NIH) and sequence information obtained from GenBank. Since hamster-specific sequences were unavailable for many of the genes measured in this study, primers were designed to highly conserved regions of individual genes identified through multiple sequence alignments between other species (e.g., rat, mouse, or human). To ensure specificity of amplification during real-time PCR, amplified products were subjected to agarose gel electrophoresis to visually confirm the presence of a single amplicon of the expected size.
Protein content was measured using the DC Protein Assay Kit from Bio-Rad according to the manufacturer's protocol.
Statistical differences were analyzed by the Student's t-test with the level of significance set at P < 0.05.
Metabolic changes induced following in vivo administration of TO901317.
To assess the pharmacological response upon treatment with the LXR agonist, TO901317, whole body metabolic alterations and hepatic function were assessed. Blood was collected from vehicle- and TO901317-treated hamsters prior to treatment and at the 7-day endpoint following a 5-h fasting period. Plasma levels of glucose, cholesterol, TG, ALT, and AST were measured. No significant differences were observed between the two groups prior to treatment (data not shown). A 7-day TO901317 treatment led to moderate, but nonsignificant, increases in body weight (Fig. 1A) while inducing a 3.6-fold elevation in plasma TG (P = 0.006) and increasing plasma cholesterol by 20% (P = 0.0005) compared with vehicle-treated control animals (Fig. 1B). No significant changes in plasma glucose were observed. This is in contrast to previous studies that have observed LXR agonists to have a lowering effect on plasma glucose (29). Plasma ALT and AST in the T0901317-treated group were three- and twofold higher, respectively, than in the control group, suggesting that some hepatic stress was induced by TO901317 in these animals.
The plasma lipoprotein profile was also examined by FPLC (Fig. 1, C and D). Compared with vehicle-treated control hamsters, there was a greater than eightfold elevation of the VLDL-TG fraction, a 10-fold increase in HDL-TG and a 2.4-fold increase in VLDL-cholesterol observed in the TO901317-treated animals.
Hepatic lipid changes following in vivo TO901317 treatment.
To assess changes in hepatic lipid mass, TG and cholesterol mass were measured in hepatic tissue of vehicle- and TO901317-treated hamsters after the 7-day treatment period. There were no significant changes in hepatic cholesterol mass (Fig. 1E), but there was a significant increase in the hepatic mass of TG (approximately twofold; Fig. 1E, P < 0.01) compared with vehicle treated controls, suggesting that LXR-activation promotes hepatic TG storage.
Ex vivo analysis of newly synthesized lipids (treatment of hamsters with TO901317 as above, followed by isolation of hepatocytes and ex vivo lipid labeling with [3H] acetate or oleate), showed significant increases in hepatic cholesteryl ester accumulation and secretion (Fig. 1F) in TO901317-treated hamsters (P = 0.05). There was also an increase in the amount of secreted TG following agonist treatment (Fig. 1G), although this was not statistically significant (P = 0.17).
In vivo TO901317 treatment increases plasma apoB100.
To assess the effect of TO901317 on apoB production in vivo, hamsters were treated with a lower concentration of the drug (50 mg/kg) for 4 days to limit hepatic stress induced by LXR-activation at higher doses. Plasma from vehicle- and TO901317-treated hamsters was immunoblotted for apoB100 mass (Fig. 2A) before and after treatment. Baseline plasma apoB100 mass was similar between all hamsters, and the level of apoB100 was unchanged after 4 days of vehicle treatment. Following 4 days of TO901317 administration, however, circulating plasma apoB100 levels were increased by 54% compared with the vehicle-treated group (P < 0.0001). When examined with respect to their baseline levels (before treatment), plasma apoB100 increased by 28% in TO901317-treated hamsters compared with a 6% decrease in the vehicle-treated group (P = 0.01). An increase in plasma apoB concentration was also observed following administration of lower doses of TO901317 (5 mg·kg−1·day−1 and 10 mg·kg−1·day−1; data not shown). Such profound changes in plasma apoB levels (following a relatively short period of LXR activation) suggest potent regulation of apoB metabolism by LXR. The ability of LXR activation to strongly promote increases in apoB has not been previously reported in mice and may indicate differences in LXR-mediated regulation of apoB between the two species.
Increases in hepatic apoB100 accumulation and secretion upon in vivo TO901317 treatment in hamsters.
Since posttranslational mechanisms are primarily responsible for regulation of apoB, metabolic labeling experiments were performed to assess changes in degradation or stability of the apoB protein. Primary hepatocytes from hamsters treated with TO901317 or vehicle were pulsed with [35S]methionine for 45 min, and the radiolabel was chased for up to 2 h with cold methionine. Cells and media were collected at 0-, 1-, and 2-h chase time points to determine levels of total newly synthesized apoB100 (cells + media) and secreted apoB100 (media). At 1 and 2 h, hepatocytes from TO901317-treated hamsters exhibited a greater amount of both cellular (Fig. 2B) and secreted (Fig. 2C) apoB100 compared with hepatocytes from vehicle-treated animals. Significant increases in cellular apoB100 of 3.4-fold and 3.9-fold over vehicle-treated levels were observed following TO901317 treatment at 1-h and 2-h chase times, respectively (P < 0.05). Increases in secreted apoB with TO310917 treatment were also observed (1.6-fold at 1-h chase and 2.1-fold at 2-h chase; P < 0.05). These data, together with results showing increased plasma apoB100 levels, suggest that activation of LXR results in increased apoB stability, which in turn leads to increases in secretion of apoB-containing lipoproteins.
Alterations in hepatic mRNA levels upon LXR activation.
In TO901317-treated hamsters, increases in mRNA levels of the LXR-responsive Fas and Scd genes were observed, with increases in Fas being almost 7-fold (P < 0.05) and Scd almost 35-fold above vehicle levels (Fig. 3; P < 0.005). Increases in Srebp-1c and Srebp-2 were also observed, but these were not significant. Hepatic mRNA expression of Dgat2, an important enzyme regulating TG synthesis, was increased by 45% in TO901317-treated hamsters relative to the vehicle-treated group (P < 0.02). Fxr, a nuclear receptor that regulates bile acid metabolism and is also known to be stimulated by LXR agonists (20), was elevated over twofold by TO901317 treatment (P < 0.04). Interestingly, mRNA levels for Mtp, a gene involved in the lipid loading and packaging of VLDL, were unchanged, and ApoB mRNA levels were reduced after TO901317 treatment (30%, P < 0.002). Treatment with a lower dose of TO901317 (10 mg·kg−1·day−1) also induced marked increases in the mRNA levels of Fas, Scd, and Fxr (data not shown).
Alterations in hepatic protein levels upon LXR activation.
In addition to mRNA levels, the protein mass of a few select hepatic molecules were also measured (Fig. 4). We assessed hepatic levels of lipogenic proteins (SREBP, FAS, SCD), as well as microsomal triglyceride transfer protein (MTP), an enzyme involved in apoB lipidation; PPARα, which is involved in fatty acid oxidation; and several proteins involved in the negative regulation of insulin signaling (PTP1B, PTEN). After LXR agonist treatment, hepatic mass of SCD was significantly increased by 58% (P = 0.0087), and FAS was increased by 41% (P = 0.0033). The level of immature SREBP-1, a master regulator of cellular lipid metabolism, as measured following an immunoblot of cytosol (including microsomes but excluding nuclei), appears to be unchanged. When the level of mature, nuclear SREBP-1 was measured by immunoblotting of isolated nuclei, an increase of 25% (P = 0.0073) was observed compared with controls. PPARα protein levels following TO901317 treatment were unchanged, as were MTP and ChREBP protein mass. Interestingly, TO901317 treatment increased hepatic levels of PTP1B by 29% (P = 0.0005) but had no significant effect on the protein mass of PTEN.
Hepatic insulin receptor and IRS-1 phosphorylation is reduced whereas IRS-2 phosphorylation is increased with in vivo TO901317 treatment.
Following LXR activation and a 2-min in vivo infusion of insulin (5 U/kg), hepatic insulin receptor-β (IR-β) mass decreased by 39% (P = 0.0014), accompanied by an equal decrease in tyrosine phosphorylation (36%), as measured by immunoblotting (Fig. 5B). Insulin receptor substrate 1 (IRS-1) showed a net decrease in phosphorylation: total mass of this protein remained unchanged, whereas total tyrosine phosphorylation decreased by 24% (P = 0.023) (Fig. 5C). IRS-2, on the other hand, showed a more than twofold increase in mass (228%, P = 0.0037) (Fig. 5D) and a threefold increase in phosphorylation (321%, P = 0.012), resulting in a net increase in phosphorylation of 40% (P = 0.044) (Fig. 5D).
LXR activation is known to increase plasma TG levels by activating expression of hepatic lipogenic genes (12, 16, 34, 41) including stimulation of lipogenesis via Srebp and induction of Scd and Fas gene transcription (23, 39, 41, 52). In the present study, 4–7 days of TO901317 administration in the hamster resulted in elevated plasma TG, which was concentrated in the VLDL density fraction. This was accompanied by a concomitant increase in plasma apoB mass, which may have resulted from either increased hepatic apoB production or reduced VLDL-apoB clearance and catabolism. Further investigation revealed increased protein and mRNA levels of SCD and FAS, as well as increased levels of the processed form of SREBP-1c in the TO901317-treated hamster livers. Alterations in activity of SREBP-1c and expression of SCD and FAS may thus underlie the increased hepatic output of TG and VLDL observed. Fas, a central enzyme in de novo lipogenesis (23), was likely upregulated by LXR both directly, via interaction with the Fas promoter (23), and indirectly via elevated SREBP-1c activity. Since there was no significant increase in Srebp-1 mRNA following LXR activation, at least part of the LXR-mediated stimulation of lipogenic genes may be SREBP-1c independent. This is supported by previous studies that show that, whereas LXR−/− mice are obesity resistant when challenged with high fat and cholesterol, SREBP-1−/− mice are not (24). Increased Dgat2 mRNA levels may also partly explain the increased plasma TG, since Dgat2 is a key enzyme in the conversion of free fatty acids to TG (6, 30, 54, 56).
Hepatic lipid measurements indicated that hepatic TG mass increased and cholesterol mass remained constant after LXR agonist treatment. However, lipid labeling experiments performed on isolated hepatocytes demonstrated large increases in the amount of cholesteryl ester (both cellular and secreted). This apparent discrepancy (increased cholesterol production by hepatocytes without accumulation in liver tissue) may be partially explained by increased LXR-induced cholesterol efflux, although this was not directly measured. Since Fxr mRNA was found to be increased, hepatic cholesterol levels may have normalized as a result of increased bile efflux. FXR can modify the transcriptional activity of factors involved in controlling lipogenesis and simultaneously control bile acid and lipid metabolism (10, 45). LXR also activates transcription of the gene responsible for the synthesis of bile acids from cholesterol, Cyp7a1 (25), which would contribute to the observed decreased hepatic cholesterol.
Only one prior study has directly examined the link between LXR activation and hepatic apoB-lipoprotein production. Grefhorst et al. (16) administered the synthetic LXR agonist T0901317 to C57BL/6J mice and showed considerable induction of SREBP-1c and FAS, resulting in hepatic steatosis and an increase in VLDL-TG secretion. Enhanced VLDL secretion was found to be due to the formation of large TG-rich particles with no change in particle number (16). Our data showing increased plasma VLDL-TG are in agreement with these findings in the C57BL/6J mice; however, our findings of increased plasma apoB100 also suggest an increase in VLDL particle number (although we have no direct evidence to support this). Grefhorst et al. found no appreciable change in apoB100 or apoB48 in C57BL/6J mice treated with T0901317. This may be due to species-specific differences in hepatic lipid mobilization. Since LXR activation in the mouse model led to hepatic TG accumulation, hepatic assembly and secretion of apoB-containing lipoproteins may not have been optimally stimulated. In contrast, our hamster model showed marked increases in hepatic TG but also showed increased plasma TG and apoB levels, suggesting that hamsters may be more efficient at assembling and secreting VLDL. It is tempting to suggest that this difference in hepatic TG mobilization may relate to predominant expression of apoB48 in the mouse liver, as opposed to hamster liver, which (like that of humans) expresses and secretes lipoproteins containing apoB100. ApoB48 may be less efficient in mobilizing the large lipid mass induced via enhanced de novo lipogenesis. Further studies are clearly needed to examine this hypothesis.
Interestingly, Mtp mRNA remained unchanged, and ApoB mRNA decreased slightly after LXR activation, despite the increased amounts of newly synthesized cellular and secreted apoB observed. It has been well established that MTP activity is required for VLDL synthesis and secretion (14, 21, 42). VLDL synthesis, however, occurs in two steps. As ApoB is translated on ER-bound ribosomes, it is cotranslationally lipidated in a process that is MTP dependent (38). Following the initial lipidation, the nascent VLDL particle merges with a preformed lipid droplet. This second step is not dependent on MTP presence or activity (15). Because of this two-step process of VLDL assembly, it is possible that an increase in hepatic lipid mass (as observed with LXR activation) could stimulate VLDL secretion at the second step, without a need for a parallel increase in MTP mass or activity. Additionally, the level of surviving apoB is determined predominantly by its lipidation state, rather than by its level of transcription or translation. It has been previously shown that 40% of translated apoB is degraded before being assembled into a mature VLDL particle (50). Therefore, a decrease in apoB mRNA is not necessarily followed by a reduction in VLDL production. The observed increase in plasma apoB and VLDL is likely due to decreased intracellular degradation and/or increased protein stability and enhanced VLDL assembly.
The insulin signaling alterations, including reduced IRβ subunit mass and reduced IRS-1 tyrosine phosphorylation, may have contributed to the hepatic overproduction of apoB particles. It has now been well established that VLDL production and secretion increase in the insulin-resistant state, partly because VLDL production becomes insensitive to the suppressive effects of insulin (27, 53). Previous studies from our laboratory using a model of diet-induced insulin resistance in hamsters have shown a strong relationship between insulin resistance and increased apoB production. We have shown that feeding hamsters a high-fructose diet not only causes insulin resistance, characterized by decreased phosphorylation of the insulin receptor as well as IRS-1 and IRS-2, but also causes increased production and secretion of VLDL and total apoB (48, 49). Furthermore, in this same model, treatment with rosiglitazone, a drug known to improve insulin sensitivity, resulted in lower VLDL secretion and reduced MTP mass, indicating that improved insulin signaling decreases VLDL assembly and apoB secretion (5). In addition, PTP1B activity has been linked to increased apoB release. Studies in PTP1B knockout mice have shown that knockout of PTP1B reduces plasma apoB48 and 100, while blocking PTP1B production with short interfering RNA reduces hepatic apoB100 secretion (36). In addition, insulin-resistant hamsters have increased hepatic PTP1B protein levels, contributing to reduced phosphorylation of insulin receptor, IRS-1, and IRS-2 (5). The increased hepatic mass of PTP1B observed after TO901317 treatment may be a contributing factor to observed perturbations in insulin signaling. Enhanced Ptp-1B expression has been previously shown to cause increases in cellular and secreted apoB, inducing hepatic VLDL overproduction and a dyslipidemic state (36). Tyrosine phosphatases, including PTP1B, can regulate the biological effects of insulin (4, 44). Therefore, the LXR agonist-induced elevation in PTP1B may have contributed to insulin insensitivity through IRβ and IRS-1 dephosphorylation (36, 37).
Interestingly, despite these reductions in IRβ mass and IRS-1 mass and tyrosine phosphorylation, LXR stimulation caused an increase in both IRS-2 mass and phosphorylation. One possible explanation for this is that LXR-mediated activation of Pparγ causes IRS-2 upregulation. Administration of TO901317 has been shown to activate hepatic Pparγ transcription via a conserved LXR binding site within the promoter (40). Previous studies have shown that the PPARγ activation specifically increases Irs-2 transcription and translation via a direct effect of PPARγ on the Irs-2 promoter. These effects have not been shown for Irs-1 (43), potentially explaining the differential effects of LXR activation on insulin signaling through IRS-1 vs. IRS-2. Since the IRS-1 and IRS-2 have been found to have divergent functions in breast cancer metastasis (13), this may be also true for their liver function. This differential effect could be reflective of the relatively short duration of treatment (4 days); chronic LXR agonist treatment may lead to more profound signaling changes and induction of an insulin-resistant state in the liver (supported by preliminary data in hamsters treated for 2–4 wk with the LXR agonist at a concentration of 10 mg·kg−1·day−1; data not shown).
It is interesting that there was no significant difference in plasma glucose following LXR activation in this study, since LXR agonists have previously been shown to decrease plasma glucose (29). This may be due to species differences, since all studies showing a glucose-lowering effect have so far been conducted in mice. LXR activation has been shown to regulate its targets differentially in different species (33). LXR induces Cyp7a1 in rats and mice, but not in humans or rabbits (33). Other LXR targets are equally responsive in humans and mice, indicating that only certain genes are differentially regulated between species. In addition, a slight lowering of plasma glucose was observed in some hamsters following LXR activation, even though the mean change was not significant compared with vehicle.
Finally, our findings raise the possibility that LXR activation may be a causative factor in the induction of insulin resistance through reductions in IR and IRS-1 signaling, which is contrary to previous findings that suggest an antidiabetic role for LXR agonists (17, 46). IRS-1 has an important role as an early intermediary between the insulin receptor and downstream molecules involved in insulin signal transduction, since IRS-1 phosphorylation is required for insulin-stimulated responses (18); the inhibition of IRS-1 by LXR activation could therefore lead to insulin resistance (1). Further long-term studies are needed to assess whether chronic LXR activation can lead to the development of hepatic and/or whole body insulin resistance and how this may exacerbate perturbations in lipoprotein metabolism.
In conclusion, data from acute activation of LXR in the hamster suggest a critical link between LXR activation, perturbations in insulin signaling cascades, and stimulation of the assembly and secretion of apoB-containing lipoproteins. Understanding the molecular mechanisms linking this master regulator to factors involved in the hepatic VLDL assembly process will require further investigation. Especially useful would be to induce chronic low-level activation of LXR and monitor the sequence of events leading to perturbations in insulin signaling cascades, alterations in hepatic lipid mobilization, and the factors promoting the hepatic assembly of VLDL-apoB100 particles.
This work was supported by an operating grant from the Heart and Stroke Foundation of Ontario (T-6041) to K. Adeli. A. E. Miller was funded by a Restracomp scholarship from the Hospital for Sick Children. H. Basciano was funded by an Ontario Graduate studentship.
↵* H. Basciano and A. Miller contributed equally to work presented in this manuscript.
- Copyright © 2009 the American Physiological Society