Astrocyte Elevated Gene-1 (AEG-1) Regulates Lipid Homeostasis^{<a href="#FN1" rid="FN1" class=" fn">*</a>}

Introduction

The role of astrocyte elevated gene-1 (AEG-1)⁴ (1, 2), also known as MTDH (metadherin) (3) and LYRIC (4), as an oncogene is well established (5,–8). AEG-1 plays a crucial role in regulating tumor cell proliferation, invasion, and metastasis (3, 5, 8, 9) and positively contributes to tumor-associated angiogenesis, chemoresistance, and protection from apoptosis (6, 10,–12). However, the physiological function of AEG-1 is still not known. The AEG-1 gene is present only in vertebrates, in which the protein is highly conserved, suggesting that AEG-1 evolved to perform specific functions in higher organisms (13). Analysis of AEG-1 expression during mouse development demonstrated AEG-1 expression in mid-to-hindbrain, fronto-nasal processes, limbs, and pharyngeal arches in the early development period from embryonic day 8.5 to 9.5 (14). At stages of embryonic day 12.5–18.5, AEG-1 expression was detected in the brain, the olfactory and skeletal systems, the skin, and the liver. Co-localization of AEG-1 expression was observed with the proliferation marker Ki-67, and it was hypothesized that AEG-1 might play a role in normal mouse development in the context of cell proliferation and differentiation (14).

Cholesterol and fatty acid absorption from the intestine is a complicated process involving specific transporters and efflux pumps, the expression of which is regulated by nuclear receptors, such as liver X receptor (LXR) and peroxisome proliferator-activated receptor (PPAR)-α (15, 16). In addition, metabolism of cholesterol and fatty acids in enterocytes determines their absorption into the lymph and blood and availability to other organs. Unesterified free cholesterol is taken up into the enterocytes by the cholesterol transporter NPC1L1, which can have three fates (15). Free cholesterol can be transported to the endoplasmic reticulum, where it is esterified by acyl-CoA:cholesterol acyltransferase 2 enzyme and then conjugated with ApoB48 to form chylomicron, which enters into the lymphatic system. Unesterified free cholesterol can be transported back to the intestinal lumen by apically located cholesterol pump ABCG5/ABCG8 heterodimer. Free cholesterol may be transported into the circulation as a constituent of HDL via ABCA1 located on the basolateral membrane of the enterocytes. When cellular oxysterols accumulate as a result of increasing concentrations of cholesterol, LXR activation, by cholesterol metabolites such as oxysterols (17), causes a net loss of cholesterol through down-regulation of NPC1L1 and reduction of acyl-CoA:cholesterol acyltransferase 2 enzyme activity as well as induction of ABCG5/ABCG8 and ABCA1 (18,–20).

Important intestinal fatty acid transporters are: 1) fatty acid-binding protein, plasma membrane (FABPpm/Got2), located peripherally on the plasma membrane; 2) fatty acid translocase (FAT/CD36), an integral membrane glycoprotein; and 3) fatty acid transport protein subtype 4 (FATP4/Slc27a4), an integral membrane protein (21). In the intestinal lumen, triglyceride (TG) is broken down into free fatty acids by lipases. Once fatty acids enter into the enterocytes, cytoplasmic fatty acid-binding protein (FABPc) may transport them to the mitochondria for β-oxidation (22). The majority of fatty acids are re-esterified into TG and excreted into chylomicrons. In the enterocytes, PPARα promotes β-oxidation of fatty acids, thereby decreasing fatty acid absorption in the circulation (23, 24).

The present study unravels a novel physiological function of AEG-1 in lipid homeostasis by analyzing an AEG-1 knock-out (AEG-1KO) mouse. We previously documented that AEG-1 interacts with retinoid X receptor (RXR) (25), a master regulator of cell growth metabolism and development mediating the action of hormones, vitamins, and lipids (26, 27). RXR heterodimerizes with one-third of the 48 human nuclear receptor superfamily members, including LXR and PPARα (27). In the absence of their corresponding ligands, RXR heterodimers interact with co-repressors that maintain histones in a deacetylated state and inhibit transcription. Upon ligand binding, there is a conformational change so that the co-repressors are replaced by co-activators inducing histone acetylation and transcriptional activation. Co-activators, such as SRC-1, interact with nuclear receptors via an LXXLL motif (28). We documented that AEG-1 contains an LXXLL motif, through which it interacts with RXR that prevents co-activator recruitment and inhibits RXR function even in the presence of ligand (25). We now document augmented LXR and PPARα activity in enterocytes of AEG-1KO mice that hinders cholesterol and fatty acid absorption that may be one mechanism that leads to a lean phenotype. By regulating LXR and PPARα function, AEG-1 may significantly modulate the effects of a high fat diet and thus play a role in obesity-associated diseases.

Materials and Methods

Generation of AEG-1KO Mouse

The AEG-1KO mouse was generated in a C57BL/6:129/Sv background (25, 29). We have back-crossed the line to C57BL/6 for 10 generations and obtained similar results, which are described in this paper, for both the WT and AEG-1KO mice on the C57BL/6 background as on the C57BL/6:129/Sv background. The data presented in the figures were obtained from mice on the C57BL/6 background. All animal studies were approved by the institutional animal care and use committee at Virginia Commonwealth University and were conducted in accordance with the Animal Welfare Act, the Public Health Service Policy on Humane Care and Use of Laboratory Animals, and the United States Government Principles for the Utilization and Care of Vertebrate Animals Used in Testing, Research, and Training.

Measurement of Food Consumption and Body Weight

A measured amount of food was given to 4-week-old age-matched WT and AEG-1KO mice (n = 8/group) housed in individual cages. The amount of food consumed per week was determined by measuring the leftover food. Body weight was also measured weekly. The experiment was performed for 12 weeks. Food consumption was calculated by the amount of food (in g) consumed/g of body weight. The mice were fed regular chow. For the high fat diet (HFD) experiment, 8-week-old mice were fed a high fat and cholesterol-containing diet (Harlan; TD.88137) for 5 weeks. This diet contains 0.2% total cholesterol and 21% total fat by weight, which provides 42% kcal (4.5 kcal/g).

MRI of Total Body Fat Mapping

Male WT and AEG-1KO mice at 12 weeks of age were used (n = 6/group). Using a Bruker-Biospin Biospec console and a 7-tesla, 30-cm free bore magnet (Bruker Biospin), T2 weighted multislice spin echo images with and without fat suppression were acquired, covering the entire body from the base of skull to the tail, with a field of view of 4 × 4 cm and 256 × 256 matrix size. Animals were anesthetized with isoflurane during preparation (4%) and imaging (1%). The imaging parameters used were as follows: repetition time of 6.5 s, time of echo of 56 ms. The resultant two-dimensional images were exported and analyzed using ImageJ software (National Institutes of Health, Bethesda, MD). By subtracting the fat-suppressed images from the non-fat-suppressed images, fat-only images were produced, which were then used for region of interest analysis of the fat depots. Multiplying the area of the region of interest by slice thickness gave the volume.

Indirect Calorimetry

Male WT and AEG-1KO mice at 12 weeks of age were used (n = 6/group). VCO₂, VO₂, and respiratory quotient were measured using a four-chamber Columbus Instruments Comprehensive Lab Animal Monitoring System CLAMS/Oxymax (Columbus, OH) indirect calorimetry machine, which was calibrated based on the O₂ and CO₂ levels of the room dedicated for indirect calorimetry only. The mice were acclimated to the indirect calorimetry room for 1 week prior to the experiment. The oxygen consumption is calculated by taking the difference between the input oxygen flow and the output oxygen flow. Similarly, the carbon dioxide production is calculated by taking the difference between the output and input carbon dioxide flows. The calculations were analyzed as VO₂ = V_iO2i − V_oO2o, where V_i and V_o are the input and output ventilation rates (liters/min), and O2i and O2o are oxygen fractions at the input and output, and as VCO₂ = V_oCO2o − V_iCO2i, where V_i and V_o are the input and output ventilation rates (liters/min), and CO2i and CO2o are carbon dioxide fractions at the input and output. The respiratory exchange ratio (RER) is simply the ratio between the carbon dioxide production and the oxygen consumption; thus, the equation is the following, RER = VCO₂/VO₂. Energy expenditure or heat is calculated before the application of any normalization or correction and, thus, reflects the exact heat of the subject. The calorific value (CV) is based on the observed respiratory exchange ratio. CV is then used with the observed oxygen consumption (VO₂) to calculate heat; the equations are as follows, heat = CV × VO₂ and CV = 3.815 + 1.232 × RER.

Measurement of Triglyceride Concentrations and Synthesis by Stable Isotopes

The Case Mouse Phenotyping Center measured the triglyceride content and newly synthesized triglyceride levels. In order to enrich the body water with ∼2% ^{H, intraperitoneal injection of labeled water (20 μl × g of body weight of 9 g/liter NaCl in 99% atomic percentage excess 2H}₂O) was administered to adult male mice, and they were returned to their cages and maintained on 5% ^{H-labeled drinking water for 5 days. The mice were sacrificed, and blood and tissue samples were collected and flash-frozen in liquid nitrogen. The samples were stored at −80 °C until analysis. Triglyceride concentrations and} de novo lipogenesis were determined as described previously (30). Briefly, triglyceride from tissues was isolated, and labeled glycerol and palmitate were analyzed after derivatization by mass spectrometry. The ^{H-labeled triglyceride covalently linked to glycerol measures the amount of newly synthesized triglyceride, whereas the H-labeled triglyceride covalently attached to palmitate indicates the amount of new palmitate. In mice given 2H}₂O for 5 days, the contribution of de novo lipogenesis to the pool of triglyceride and palmitate was calculated using the following equation, % newly made palmitate = ((total ^{H-labeled palmitate × (H-labeled body water ×} n)) − 1) × 100, where n is the number of exchangeable hydrogens, which is assumed to be 22 (31, 32).

The percentage of total newly made triglyceride glycerol is calculated using the following equation, % total newly made triglyceride-glycerol = ((^{H-labeled triglyceride-glycerol × (H-labeled water ×} n)) − 1) × 100, where ^{H-labeled triglyceride-glycerol is the M1 isotopomer, H-labeled water is the average amount labeled in a given mouse, and} n is the exchange factor (experimentally determined from the M2/M1 ratio of triglyceride glycerol). We calculated the total triglyceride pool size (μmol/g of tissue) in the tissues using the following equation, total pool size of triglyceride = ((^{H-labeled triglyceride-glycerol × (H-labeled water × n)) − 1) × 100.}

In Vivo Cholesterol Absorption

The fecal dual isotope ratio method (33) was used, in which WT and AEG-1KO mice were gavaged with 0.5 μCi of [^{C]cholesterol and 1 μCi of [H]sitostanol in 100 μl of soy bean oil. Mice were fed standard chow diet for 3 days, and the feces were collected daily. The radioactivity of the pooled fecal samples was determined by a scintillation counter. The percentage of cholesterol absorption was calculated as follows, ((C/H dosing mixture − C/H feces)/(C/H dosing mixture) × 100.}

In Vivo Fatty Acid Absorption

WT and AEG-1KO mice were fasted overnight and intravenously injected with 500 mg/kg tyloxapol to block serum lipase activity. After 30 min, mice were gavaged with fat emulsion containing 10 μCi of [^{H]triolein. Blood samples were collected from the tail vein at time 0 and hourly for 4 h. Serum radioactivity was determined by a scintillation counter (}34).

Measurement of Fatty Acid β-Oxidation

Products of fatty acid β-oxidation was measured in fresh liver and small intestine homogenates using [^{C]palmitate as described (}35).

Glucose Tolerance Test

Mice were fasted overnight (16 h), and glucose was injected intraperitoneally (1.5 g/kg). Blood samples were collected from the tail tip before and at 10, 30, 60, 90, and 120 min after the injection for the measurement of glucose with a glucometer.

Primary Hepatocyte Culture and Transfection Assays

Primary mouse hepatocytes were isolated as described (9) and were plated on collagen-coated dishes (BD BioCoat collagen type I, BD Biosciences) and cultured in Williams E medium (Sigma) containing NaHCO₃, l-glutamine, insulin (1.5 μm), and dexamethasone (0.1 μm). For transfections, 1 × 10 ^{hepatocytes were plated in 24-well collagen-coated plates and transfected the next day using PromoFectin-Hepatocyte transfection reagent in a 9:1 ratio with firefly luciferase reporter plasmid containing responsive elements (pGL3.luc, pGL3.}LXRE.luc, and pGL3.PPRE.luc) and Renilla luciferase reporter plasmid (pGL3.renilla.luc). After 48 h, cells were incubated in the presence or absence of the respective ligands (T0901317 (5 μm) and CP775146 (2.5 μm)) for another 24 h. Luciferase assays were measured using the Dual-Luciferase Reporter Assay kit (Promega) following the manufacturer's protocol, and firefly luciferase activity was normalized by Renilla luciferase activity. Each experiment was performed in triplicates and three times to calculate means and S.D. values.

Total RNA Extraction, cDNA Preparation, and Real-time PCR

Total RNA was extracted from hepatocytes or mouse tissues using the Qiagen miRNAeasy minikit (Qiagen, Hilden, Germany). cDNA preparation was done using the ABI cDNA synthesis kit (Applied Biosystems, Foster City, CA). RT-PCR was performed using an ABI ViiA7 fast real-time PCR system and Taqman gene expression assays according to the manufacturer's protocol (Applied Biosystems). The expression level of each gene was determined by the ΔΔCT method and was normalized by the level of GAPDH. The level of untreated control for each gene was considered as 1 to determine the -fold difference.

Chromatin Immunoprecipitation (ChIP) Assay

12-week-old WT and AEG-1KO littermates (n = 3) were treated with the LXR agonist T0901317 (30 μg/g) by a single oral gavage. 24 h later, the animals were euthanized, and livers were immediately removed. A ChIP assay was performed using a SimpleChIP Plus kit (Cell Signaling Technology, Danvers, MA) according to the manufacturer's instructions (25). The antibodies used were rabbit anti-LXRα (Santa Cruz Biotechnology, Inc.), anti-acetylated histone H3 (AHH3), and anti-SRC-1 (Cell Signaling). Normal rabbit IgG was used as control. DNA was purified, and quantitative PCR was performed using Power SYBR Green PCR Mastermix (Applied Biosystems) on the ABI ViiA7 fast real-time PCR system according to the manufacturer's protocol (Applied Biosystems). PCR efficiencies of primers were examined by a standard curve of serial-diluted liver extract input and melting curve functionality. The primers used were Abca1 forward (5′-GAACGAGCTTTTCCCCTTTC-3′) and Abca1 reverse (5′-CCCGGCCTCTGTTTATGTAG-3′).

Western Blot Analysis

Western blot analysis were performed as described previously (5). The following primary antibodies were used: anti-ABCA1 (mouse; Santa Cruz Biotechnology; catalogue no. sc-58219; 1:2000), anti-CPT1A (rabbit; Proteintech; catalogue no. 15184-1-AP; 1:1000), anti-NPC1L1 (rabbit; Cell Signaling Technology; catalogue no. 5058; 1:500), anti-PPARα (rabbit; Santa Cruz Biotechnology; catalogue no. sc-9000; 1:500), anti-LXR (rabbit; Santa Cruz Biotechnology; catalogue no. sc-13068; 1:1000), anti-AEG-1 (chicken; in-house; 1:3000) and anti-GAPDH (mouse; Santa Cruz Biotechnology; catalogue no. sc-166545; 1:2500).

Statistical Analysis

Data were represented as the mean ± S.E. and analyzed for statistical significance using one-way analysis of variance followed by Newman-Keuls test as a post hoc test. A p value of <0.05 was considered as significant.

Results

AEG-1KO Mice Are Lean and Survive Longer

We generated an AEG-1KO mouse in which the promoter region, exon 1, and part of intron 1 of the AEG-1 gene was deleted using a Cre-loxP system (25, 29). AEG-1KO mice were viable and fertile, although the litter size was smaller than expected. Both WT and KO mice developed normally to full adulthood, indicating that AEG-1 is probably not involved in developmental regulation. Histological analysis of the internal organs of the adult mice also did not reveal any significant and noticeable difference between WT and AEG-1KO mice (29).

However, a marked difference in body weight gain was observed between age-matched WT and AEG-1KO mice. We performed weekly monitoring of body weight starting at 4 weeks of age when the mice were weaned. Over the next 12 weeks, the average body weight gain by each WT mouse was ∼14 g, whereas that in AEG-1KO mice was only ∼4 g (Fig. 1A). During this period, the food consumption (g of food/g of body weight) was similar in both groups (Fig. 1B). The substantial difference in body weight gain continued with age (Fig. 1, C and D) and was more pronounced in male mice compared with female mice (Fig. 1, C and D). MRI analysis of the mice demonstrated a significant decrease in visceral, subcutaneous, and intramuscular fat in the AEG-1KO mice compared with WT mice (Fig. 2, A and B). Histologically, adipocytes in white and brown adipose tissues were larger and more abundant in WT mice versus AEG-1KO mice (Fig. 2C). Aged (15 months old) WT mice demonstrated steatotic changes in the liver, whereas no such change was observed in AEG-1KO mice (Fig. 2D). The mice were monitored until 24 months, by which time point 19 of 21 WT mice had died, whereas a significant survival benefit was observed in the leaner AEG-1KO mice, of which only 3 of 20 mice had died (Fig. 2E). No significant difference was observed in glucose tolerance, indicating that insulin signaling is functional in AEG-1KO mice (Fig. 2F). Glucose uptake in peripheral tissues by an [^{F]FDG uptake assay also did not show a difference between WT and AEG-1KO mice (data not shown).}

FIGURE 1.

AEG-1KO mice are smaller than WT mice.A, weekly body weight gain (n = 8/group). B, biweekly food consumption (n = 8/group). C, photograph of mice at 15 months of age. D, body weight of male (M) and female (F) mice at 6 and 15 months of age (n = 12/group). For graphs, the data represent mean ± S.E. (error bars). *, p < 0.05.

FIGURE 2.

AEG-1KO mice have less body fat compared with WT mice.A, representative MRI image of 5-month-old mice. B, quantification of body fat distribution determined by MRI (n = 5/group). C, H&amp;E staining of white adipose tissue (WAT) and brown adipose tissue (BAT) at 15 months. D, H&amp;E staining of liver at 15 months. E, Kaplan-Meier survival curve at 24 months (n = 21 for WT and n = 20 for AEG-1KO). In the graphs, the data represent mean ± S.E. (error bars). *, p < 0.05. F, blood glucose was measured at the indicated time points after intraperitoneal injection of a bolus of glucose in overnight fasted mice. The data represent mean ± S.E. (n = 5/group).

Intestinal Fat Absorption Is Inhibited in AEG-1KO Mice

The profound lack of fat in AEG-1KO mice may be due to 1) defects in fat synthesis, 2) increased energy expenditure (fat burning), and/or 3) decreased fat absorption from the intestine. To address this issue, we first measured serum total TG and cholesterol levels, both of which were significantly lower in AEG-1KO compared with WT mice (Fig. 3A). Levels of newly synthesized TG and palmitate in liver and muscle of WT and AEG-1KO mice showed no significant differences (Fig. 3, B and C). An indirect calorimetry analysis measuring O₂ consumption and CO₂ production did not reveal any increased expenditure in AEG-1KO mice (Fig. 3D). Rather, AEG-1KO mice demonstrated a small but statistically significant decrease in energy expenditure compared with WT mice (Fig. 3, D and E). No difference was observed in UCP1 (uncoupling protein 1) levels in white and brown adipose tissue of WT and AEG-1KO mice (data not shown). Both cholesterol and fatty acid absorption by the intestine were markedly inhibited in AEG-1KO versus WT mice (Fig. 3, F and G). Cholesterol absorption was measured by gavaging [^{C]cholesterol and then measuring radioactivity in the feces, whereas fatty acid absorption was quantified by gavaging [H]triolein and then temporally measuring plasma radioactivity. Collectively, the lack of body fat and lean phenotype in AEG-1KO mice may be explained by reduced cholesterol and TG absorption from the intestine.}

FIGURE 3.

Intestinal fat absorption is inhibited in AEG-1KO mouse.A, serum total cholesterol and TG levels (n = 5/group). B and C, measurement of newly synthesized triglyceride (B) and palmitate (C). D and E, indirect calorimetry analysis measuring O₂ consumption and CO₂ production (n = 6/group). Area under the curve (D) and respiration quotient (E) plots are shown. F, measurement of cholesterol absorption from the intestine after gavaging [^{C]cholesterol and then measuring radioactivity in the feces.} G, measurement of fatty acid absorption by gavaging [^{H]triolein and then temporally measuring plasma radioactivity.} H, weekly body weight gain upon feeding HFD (n = 8). I, H&amp;E staining of liver sections of mice fed HFD. In the graphs, the data represent mean ± S.E. (error bars). *, p < 0.05.

We next evaluated the effect of feeding an HFD on the lean phenotype of AEG-1KO mice, starting at 8 weeks of age. Over a period of 5 weeks, each WT mouse gained ∼11 g of body weight on average, whereas no weight gain was observed in AEG-1KO mice (Fig. 3H). Post-mortem evaluation revealed evidence of steatosis in the liver of WT mice but not in AEG-1KO mice (Fig. 3I). These findings support the hypothesis that inhibition of intestinal fat absorption results in lean phenotype in AEG-1KO mice so that even feeding HFD could not induce a gain of body weight.

LXR and PPARα Activities Are Augmented in the Intestine of AEG-1KO Mice

We next searched for the molecular mechanism conferring the phenotypes of AEG-1KO mice. We have identified a novel interaction between AEG-1 and RXR that interferes with co-activator recruitment and inhibits RXR-mediated transcriptional regulation by its heterodimer partners (25). The phenotypes of the AEG-1KO mice may be explained by activation of LXR and PPARα, heterodimer partners of RXR, in the intestines. In enterocytes, activation of LXR inhibits cholesterol absorption by down-regulating cholesterol transporter Npc1l1 and up-regulating cholesterol efflux proteins, Abca1, Abcg5, and Abcg8 (15). Activation of PPARα in the enterocytes promotes β-oxidation of absorbed fatty acids, thereby down-regulating fatty acid absorption into the circulation (23, 24). Products of cholesterol metabolism, such as oxysterols, function as ligands for LXR, whereas fatty acids and fatty acid derivatives are endogenous ligands for PPARα (17, 36). HFD provides ligands to both LXR and PPARα, thereby accentuating this effect with resultant almost complete inhibition of fat absorption.

To confirm our hypothesis, we first checked LXR and PPARα activity by luciferase reporter assays in WT and AEG-1KO hepatocytes upon treatment with synthetic ligands T0901317 (5 μm) and CP775146 (2.5 μm), respectively. Both basal and ligand-dependent LXR and PPARα activities were significantly higher in AEG-1KO hepatocytes compared with WT hepatocytes (Fig. 4, A and B). Analysis of LXR expression showed an increased LXR level in the intestines of AEG-1KO mice compared with WT (Fig. 4C) We treated WT and AEG-1KO mice with T0901317 for 24 h and analyzed recruitment of LXR, AHH3, and co-activator SRC-1 on the promoter region of LXR target gene Abca1 by a ChIP assay using liver samples. A significant increase in recruitment of LXR, AHH3, and SRC-1 was observed in AEG-1KO livers versus WT livers, confirming that our previous in vitro observation (25) holds true in an in vivo situation (Fig. 4D). We checked the expression of LXR target genes in the small intestine of WT and AEG-1KO mice fed regular chow or HFD. The basal expression level of Abca1 was markedly higher in AEG-1KO mice compared with WT mice (Fig. 4E). With regular chow, the heterodimeric efflux pump Abcg5 and Abcg8 showed small but significant increase in AEG-1KO versus WT mice. Upon HFD treatment, the expression of all three genes was significantly induced in both WT and AEG-1KO mice. However, the magnitude of induction was significantly robust, especially for Abca1, in AEG-1KO mice compared with WT mice (Fig. 4F). Conversely, the basal expression level of the cholesterol transporter Npc1l1 was significantly lower in AEG-1KO mice versus WT mice, and upon HFD treatment, this decrease was further accentuated (Fig. 4E). The increased levels of the cholesterol efflux pumps and decrease in the cholesterol transporter explain the inhibition of intestinal cholesterol absorption in AEG-1KO mice, which is further accentuated upon HFD treatment that provides ligands to LXR, augmenting its effects.

FIGURE 4.

LXR is activated in the intestine of AEG-1KO mice.A and B, LXRE (A) and PPARE (B) luciferase activity was determined in WT and AEG-1KO hepatocytes upon treatment with LXR ligand, T0901317 (5 μm), and PPARα ligand, CP775146 (2.5 μm), respectively. R.L.U., relative luciferase units. C, the expression of the indicated proteins was detected by Western blot in the intestines of WT and AEG-1KO mice. Two independent mice of each group are shown. D, ChIP assays were performed using anti-LXRα, anti-AHH3, and anti-SRC-1 antibodies using T0901317-treated WT and AEG-1KO livers and PCR primers amplifying promoter regions of the Abca1 gene. Normal rabbit IgG was used as control, and the data were normalized by results obtained using normal IgG. E and F, expression of the indicated genes by Taqman quantitative RT-PCR in chow- and HFD-fed small intestine. The data represent mean ± S.E. (error bars). *, p < 0.05 versus the corresponding data point in WT mice; #, p < 0.05 chow versus HFD in WT mice; [caret], p < 0.05 chow versus HFD in AEG-1KO mice. A.U., arbitrary units.

We next analyzed the expression levels of PPARα target genes that regulate β-oxidation of fatty acids in the small intestine and liver. For the mitochondrial β-oxidation pathway, we analyzed Cpt1a, the rate-limiting enzyme in the carnitine-dependent transport of fatty acids across the mitochondrial membrane, and Acadl and Acadm, which catalyze the initial steps of fatty acid β-oxidation (37). For peroxisomal β-oxidation, we analyzed Acox1, the first enzyme in this pathway (37). In standard chow-fed mice, Cpt1a and Acadl expression levels were significantly higher in the small intestine, and Cpt1a, Acadl, and Acadm levels were significantly higher in the liver of AEG-1KO mice versus WT mice (Fig. 5, A and B). There were no significant differences in the Acox1 levels between AEG-1KO and WT mice in either organ. With HFD treatment, a significant induction of all four genes was observed in both the small intestine and liver of WT mice. In the small intestine of AEG-1KO mice, HFD treatment resulted in a significant induction of all four genes, which was significantly more pronounced than the induction seen in WT mice (Fig. 5A). However, HFD treatment did not result in further induction of these genes in the liver of AEG-1KO mice (Fig. 5B). The basal Ppara level was significantly higher in the small intestine and liver of AEG-1KO mice versus WT mice (Fig. 5, C and D). HFD treatment resulted in induction of Ppara in both WT and AEG-1KO liver and small intestine. However, the magnitude of this induction was much more pronounced in the small intestine of AEG-1KO mice versus WT mice (Fig. 5C). Additionally, in AEG-1KO mice, the HFD-mediated induction of Ppara was more robust in the small intestine than in the liver (Fig. 5, C and D). No significant difference was observed in the level of Pparg and Ppargc1, the co-activator for PPAR (Fig. 5, C and D). The changes in LXR and PPARα target genes in the small intestines of chow- and HFD-fed mice were confirmed at the protein level by Western blot analysis (Fig. 5E). We measured products of fatty acid β-oxidation in fresh liver and small intestine homogenates using [^{C]palmitate (}35). Acetyl-CoA, generated by β-oxidation, enters the tricarboxylic acid cycle to generate CO₂. [^{C]Palmitate that does not get oxidized forms acid-soluble metabolites. Fatty acid β-oxidation was significantly increased in the liver and small intestine of AEG-1KO mouse} versus WT (Fig. 5F). Collectively, these findings suggest that under basal conditions, higher levels of PPARα and its target genes in the small intestine of AEG-1KO mice confer increased fatty acid β-oxidation in the enterocytes, resulting in decreased fatty acid absorption. This effect is further accentuated upon HFD treatment that provides ligands to PPARα, thus augmenting levels of both Ppara and its target genes. The lack of absorption of fatty acids from the intestines upon HFD treatment results in decreased delivery of PPARα ligands to the liver, so that there was no further induction by HFD compared with the chow diet. No significant difference was observed in the expression levels of enterocyte fatty acid transporters, FABPpm/Got2, FAT/Cd36, and FATP4/Slc27a4 in WT and AEG-1KO mice (Fig. 5G) with regular chow or HFD, further indicating that fatty acid transport from the lumen to the enterocyte is normal, and it is enhanced β-oxidation that abrogates absorption to the plasma. We also did not observe a significant difference in bile acid composition between WT and AEG-1KO mice (Fig. 5H), indicating that anomalies in bile composition may not be a potential mechanism of decreased fat absorption.

FIGURE 5.

PPARα is activated in the intestine of AEG-1KO mice.A–D, expression of the indicated genes by Taqman quantitative RT-PCR in chow- and HFD-fed small intestine (A and C) and chow- and HFD-fed liver (B and D). The data represent mean ± S.E. (error bars). *, p < 0.05 versus corresponding data point in WT mice; #, p < 0.05 chow versus HFD in WT mice; [caret], p < 0.05 chow versus HFD in AEG-1KO mice. White bars, WT; black bars, AEG-1KO. E, expression of the indicated proteins was checked in the small intestines of chow- or HFD-fed WT and AEG-1KO mice by Western blot analysis. F, fatty acid β-oxidation is increased in AEG-1KO mice. Fresh liver and small intestine homogenates were used with [^{C]palmitate as substrate. Oxidation rate is nmol/mg/min. The data represent mean ± S.E. *,} p < 0.01. G, expression of the indicated genes was measured in the small intestines of the chow- and HFD-fed WT and AEG-1KO mice. H, bile acid levels in the gallbladder of adult mice. MCA, muricholic acid; TUDCA, tauroursodeoxycholic acid; TCA, taurocholic acid. For G and H, the data represent mean ± S.E. For A–D and F–H, n = 4/group. A.U., arbitrary units.

AEG-1KO Mice Are Lean and Survive Longer

We generated an AEG-1KO mouse in which the promoter region, exon 1, and part of intron 1 of the AEG-1 gene was deleted using a Cre-loxP system (25, 29). AEG-1KO mice were viable and fertile, although the litter size was smaller than expected. Both WT and KO mice developed normally to full adulthood, indicating that AEG-1 is probably not involved in developmental regulation. Histological analysis of the internal organs of the adult mice also did not reveal any significant and noticeable difference between WT and AEG-1KO mice (29).

However, a marked difference in body weight gain was observed between age-matched WT and AEG-1KO mice. We performed weekly monitoring of body weight starting at 4 weeks of age when the mice were weaned. Over the next 12 weeks, the average body weight gain by each WT mouse was ∼14 g, whereas that in AEG-1KO mice was only ∼4 g (Fig. 1A). During this period, the food consumption (g of food/g of body weight) was similar in both groups (Fig. 1B). The substantial difference in body weight gain continued with age (Fig. 1, C and D) and was more pronounced in male mice compared with female mice (Fig. 1, C and D). MRI analysis of the mice demonstrated a significant decrease in visceral, subcutaneous, and intramuscular fat in the AEG-1KO mice compared with WT mice (Fig. 2, A and B). Histologically, adipocytes in white and brown adipose tissues were larger and more abundant in WT mice versus AEG-1KO mice (Fig. 2C). Aged (15 months old) WT mice demonstrated steatotic changes in the liver, whereas no such change was observed in AEG-1KO mice (Fig. 2D). The mice were monitored until 24 months, by which time point 19 of 21 WT mice had died, whereas a significant survival benefit was observed in the leaner AEG-1KO mice, of which only 3 of 20 mice had died (Fig. 2E). No significant difference was observed in glucose tolerance, indicating that insulin signaling is functional in AEG-1KO mice (Fig. 2F). Glucose uptake in peripheral tissues by an [^{F]FDG uptake assay also did not show a difference between WT and AEG-1KO mice (data not shown).}

FIGURE 1.

AEG-1KO mice are smaller than WT mice.A, weekly body weight gain (n = 8/group). B, biweekly food consumption (n = 8/group). C, photograph of mice at 15 months of age. D, body weight of male (M) and female (F) mice at 6 and 15 months of age (n = 12/group). For graphs, the data represent mean ± S.E. (error bars). *, p < 0.05.

FIGURE 2.

AEG-1KO mice have less body fat compared with WT mice.A, representative MRI image of 5-month-old mice. B, quantification of body fat distribution determined by MRI (n = 5/group). C, H&amp;E staining of white adipose tissue (WAT) and brown adipose tissue (BAT) at 15 months. D, H&amp;E staining of liver at 15 months. E, Kaplan-Meier survival curve at 24 months (n = 21 for WT and n = 20 for AEG-1KO). In the graphs, the data represent mean ± S.E. (error bars). *, p < 0.05. F, blood glucose was measured at the indicated time points after intraperitoneal injection of a bolus of glucose in overnight fasted mice. The data represent mean ± S.E. (n = 5/group).

Intestinal Fat Absorption Is Inhibited in AEG-1KO Mice

The profound lack of fat in AEG-1KO mice may be due to 1) defects in fat synthesis, 2) increased energy expenditure (fat burning), and/or 3) decreased fat absorption from the intestine. To address this issue, we first measured serum total TG and cholesterol levels, both of which were significantly lower in AEG-1KO compared with WT mice (Fig. 3A). Levels of newly synthesized TG and palmitate in liver and muscle of WT and AEG-1KO mice showed no significant differences (Fig. 3, B and C). An indirect calorimetry analysis measuring O₂ consumption and CO₂ production did not reveal any increased expenditure in AEG-1KO mice (Fig. 3D). Rather, AEG-1KO mice demonstrated a small but statistically significant decrease in energy expenditure compared with WT mice (Fig. 3, D and E). No difference was observed in UCP1 (uncoupling protein 1) levels in white and brown adipose tissue of WT and AEG-1KO mice (data not shown). Both cholesterol and fatty acid absorption by the intestine were markedly inhibited in AEG-1KO versus WT mice (Fig. 3, F and G). Cholesterol absorption was measured by gavaging [^{C]cholesterol and then measuring radioactivity in the feces, whereas fatty acid absorption was quantified by gavaging [H]triolein and then temporally measuring plasma radioactivity. Collectively, the lack of body fat and lean phenotype in AEG-1KO mice may be explained by reduced cholesterol and TG absorption from the intestine.}

FIGURE 3.

Intestinal fat absorption is inhibited in AEG-1KO mouse.A, serum total cholesterol and TG levels (n = 5/group). B and C, measurement of newly synthesized triglyceride (B) and palmitate (C). D and E, indirect calorimetry analysis measuring O₂ consumption and CO₂ production (n = 6/group). Area under the curve (D) and respiration quotient (E) plots are shown. F, measurement of cholesterol absorption from the intestine after gavaging [^{C]cholesterol and then measuring radioactivity in the feces.} G, measurement of fatty acid absorption by gavaging [^{H]triolein and then temporally measuring plasma radioactivity.} H, weekly body weight gain upon feeding HFD (n = 8). I, H&amp;E staining of liver sections of mice fed HFD. In the graphs, the data represent mean ± S.E. (error bars). *, p < 0.05.

We next evaluated the effect of feeding an HFD on the lean phenotype of AEG-1KO mice, starting at 8 weeks of age. Over a period of 5 weeks, each WT mouse gained ∼11 g of body weight on average, whereas no weight gain was observed in AEG-1KO mice (Fig. 3H). Post-mortem evaluation revealed evidence of steatosis in the liver of WT mice but not in AEG-1KO mice (Fig. 3I). These findings support the hypothesis that inhibition of intestinal fat absorption results in lean phenotype in AEG-1KO mice so that even feeding HFD could not induce a gain of body weight.

LXR and PPARα Activities Are Augmented in the Intestine of AEG-1KO Mice

We next searched for the molecular mechanism conferring the phenotypes of AEG-1KO mice. We have identified a novel interaction between AEG-1 and RXR that interferes with co-activator recruitment and inhibits RXR-mediated transcriptional regulation by its heterodimer partners (25). The phenotypes of the AEG-1KO mice may be explained by activation of LXR and PPARα, heterodimer partners of RXR, in the intestines. In enterocytes, activation of LXR inhibits cholesterol absorption by down-regulating cholesterol transporter Npc1l1 and up-regulating cholesterol efflux proteins, Abca1, Abcg5, and Abcg8 (15). Activation of PPARα in the enterocytes promotes β-oxidation of absorbed fatty acids, thereby down-regulating fatty acid absorption into the circulation (23, 24). Products of cholesterol metabolism, such as oxysterols, function as ligands for LXR, whereas fatty acids and fatty acid derivatives are endogenous ligands for PPARα (17, 36). HFD provides ligands to both LXR and PPARα, thereby accentuating this effect with resultant almost complete inhibition of fat absorption.

To confirm our hypothesis, we first checked LXR and PPARα activity by luciferase reporter assays in WT and AEG-1KO hepatocytes upon treatment with synthetic ligands T0901317 (5 μm) and CP775146 (2.5 μm), respectively. Both basal and ligand-dependent LXR and PPARα activities were significantly higher in AEG-1KO hepatocytes compared with WT hepatocytes (Fig. 4, A and B). Analysis of LXR expression showed an increased LXR level in the intestines of AEG-1KO mice compared with WT (Fig. 4C) We treated WT and AEG-1KO mice with T0901317 for 24 h and analyzed recruitment of LXR, AHH3, and co-activator SRC-1 on the promoter region of LXR target gene Abca1 by a ChIP assay using liver samples. A significant increase in recruitment of LXR, AHH3, and SRC-1 was observed in AEG-1KO livers versus WT livers, confirming that our previous in vitro observation (25) holds true in an in vivo situation (Fig. 4D). We checked the expression of LXR target genes in the small intestine of WT and AEG-1KO mice fed regular chow or HFD. The basal expression level of Abca1 was markedly higher in AEG-1KO mice compared with WT mice (Fig. 4E). With regular chow, the heterodimeric efflux pump Abcg5 and Abcg8 showed small but significant increase in AEG-1KO versus WT mice. Upon HFD treatment, the expression of all three genes was significantly induced in both WT and AEG-1KO mice. However, the magnitude of induction was significantly robust, especially for Abca1, in AEG-1KO mice compared with WT mice (Fig. 4F). Conversely, the basal expression level of the cholesterol transporter Npc1l1 was significantly lower in AEG-1KO mice versus WT mice, and upon HFD treatment, this decrease was further accentuated (Fig. 4E). The increased levels of the cholesterol efflux pumps and decrease in the cholesterol transporter explain the inhibition of intestinal cholesterol absorption in AEG-1KO mice, which is further accentuated upon HFD treatment that provides ligands to LXR, augmenting its effects.

FIGURE 4.

LXR is activated in the intestine of AEG-1KO mice.A and B, LXRE (A) and PPARE (B) luciferase activity was determined in WT and AEG-1KO hepatocytes upon treatment with LXR ligand, T0901317 (5 μm), and PPARα ligand, CP775146 (2.5 μm), respectively. R.L.U., relative luciferase units. C, the expression of the indicated proteins was detected by Western blot in the intestines of WT and AEG-1KO mice. Two independent mice of each group are shown. D, ChIP assays were performed using anti-LXRα, anti-AHH3, and anti-SRC-1 antibodies using T0901317-treated WT and AEG-1KO livers and PCR primers amplifying promoter regions of the Abca1 gene. Normal rabbit IgG was used as control, and the data were normalized by results obtained using normal IgG. E and F, expression of the indicated genes by Taqman quantitative RT-PCR in chow- and HFD-fed small intestine. The data represent mean ± S.E. (error bars). *, p < 0.05 versus the corresponding data point in WT mice; #, p < 0.05 chow versus HFD in WT mice; [caret], p < 0.05 chow versus HFD in AEG-1KO mice. A.U., arbitrary units.

We next analyzed the expression levels of PPARα target genes that regulate β-oxidation of fatty acids in the small intestine and liver. For the mitochondrial β-oxidation pathway, we analyzed Cpt1a, the rate-limiting enzyme in the carnitine-dependent transport of fatty acids across the mitochondrial membrane, and Acadl and Acadm, which catalyze the initial steps of fatty acid β-oxidation (37). For peroxisomal β-oxidation, we analyzed Acox1, the first enzyme in this pathway (37). In standard chow-fed mice, Cpt1a and Acadl expression levels were significantly higher in the small intestine, and Cpt1a, Acadl, and Acadm levels were significantly higher in the liver of AEG-1KO mice versus WT mice (Fig. 5, A and B). There were no significant differences in the Acox1 levels between AEG-1KO and WT mice in either organ. With HFD treatment, a significant induction of all four genes was observed in both the small intestine and liver of WT mice. In the small intestine of AEG-1KO mice, HFD treatment resulted in a significant induction of all four genes, which was significantly more pronounced than the induction seen in WT mice (Fig. 5A). However, HFD treatment did not result in further induction of these genes in the liver of AEG-1KO mice (Fig. 5B). The basal Ppara level was significantly higher in the small intestine and liver of AEG-1KO mice versus WT mice (Fig. 5, C and D). HFD treatment resulted in induction of Ppara in both WT and AEG-1KO liver and small intestine. However, the magnitude of this induction was much more pronounced in the small intestine of AEG-1KO mice versus WT mice (Fig. 5C). Additionally, in AEG-1KO mice, the HFD-mediated induction of Ppara was more robust in the small intestine than in the liver (Fig. 5, C and D). No significant difference was observed in the level of Pparg and Ppargc1, the co-activator for PPAR (Fig. 5, C and D). The changes in LXR and PPARα target genes in the small intestines of chow- and HFD-fed mice were confirmed at the protein level by Western blot analysis (Fig. 5E). We measured products of fatty acid β-oxidation in fresh liver and small intestine homogenates using [^{C]palmitate (}35). Acetyl-CoA, generated by β-oxidation, enters the tricarboxylic acid cycle to generate CO₂. [^{C]Palmitate that does not get oxidized forms acid-soluble metabolites. Fatty acid β-oxidation was significantly increased in the liver and small intestine of AEG-1KO mouse} versus WT (Fig. 5F). Collectively, these findings suggest that under basal conditions, higher levels of PPARα and its target genes in the small intestine of AEG-1KO mice confer increased fatty acid β-oxidation in the enterocytes, resulting in decreased fatty acid absorption. This effect is further accentuated upon HFD treatment that provides ligands to PPARα, thus augmenting levels of both Ppara and its target genes. The lack of absorption of fatty acids from the intestines upon HFD treatment results in decreased delivery of PPARα ligands to the liver, so that there was no further induction by HFD compared with the chow diet. No significant difference was observed in the expression levels of enterocyte fatty acid transporters, FABPpm/Got2, FAT/Cd36, and FATP4/Slc27a4 in WT and AEG-1KO mice (Fig. 5G) with regular chow or HFD, further indicating that fatty acid transport from the lumen to the enterocyte is normal, and it is enhanced β-oxidation that abrogates absorption to the plasma. We also did not observe a significant difference in bile acid composition between WT and AEG-1KO mice (Fig. 5H), indicating that anomalies in bile composition may not be a potential mechanism of decreased fat absorption.

FIGURE 5.

PPARα is activated in the intestine of AEG-1KO mice.A–D, expression of the indicated genes by Taqman quantitative RT-PCR in chow- and HFD-fed small intestine (A and C) and chow- and HFD-fed liver (B and D). The data represent mean ± S.E. (error bars). *, p < 0.05 versus corresponding data point in WT mice; #, p < 0.05 chow versus HFD in WT mice; [caret], p < 0.05 chow versus HFD in AEG-1KO mice. White bars, WT; black bars, AEG-1KO. E, expression of the indicated proteins was checked in the small intestines of chow- or HFD-fed WT and AEG-1KO mice by Western blot analysis. F, fatty acid β-oxidation is increased in AEG-1KO mice. Fresh liver and small intestine homogenates were used with [^{C]palmitate as substrate. Oxidation rate is nmol/mg/min. The data represent mean ± S.E. *,} p < 0.01. G, expression of the indicated genes was measured in the small intestines of the chow- and HFD-fed WT and AEG-1KO mice. H, bile acid levels in the gallbladder of adult mice. MCA, muricholic acid; TUDCA, tauroursodeoxycholic acid; TCA, taurocholic acid. For G and H, the data represent mean ± S.E. For A–D and F–H, n = 4/group. A.U., arbitrary units.

AEG-1KO Mice Are Lean and Survive Longer

We generated an AEG-1KO mouse in which the promoter region, exon 1, and part of intron 1 of the AEG-1 gene was deleted using a Cre-loxP system (25, 29). AEG-1KO mice were viable and fertile, although the litter size was smaller than expected. Both WT and KO mice developed normally to full adulthood, indicating that AEG-1 is probably not involved in developmental regulation. Histological analysis of the internal organs of the adult mice also did not reveal any significant and noticeable difference between WT and AEG-1KO mice (29).

However, a marked difference in body weight gain was observed between age-matched WT and AEG-1KO mice. We performed weekly monitoring of body weight starting at 4 weeks of age when the mice were weaned. Over the next 12 weeks, the average body weight gain by each WT mouse was ∼14 g, whereas that in AEG-1KO mice was only ∼4 g (Fig. 1A). During this period, the food consumption (g of food/g of body weight) was similar in both groups (Fig. 1B). The substantial difference in body weight gain continued with age (Fig. 1, C and D) and was more pronounced in male mice compared with female mice (Fig. 1, C and D). MRI analysis of the mice demonstrated a significant decrease in visceral, subcutaneous, and intramuscular fat in the AEG-1KO mice compared with WT mice (Fig. 2, A and B). Histologically, adipocytes in white and brown adipose tissues were larger and more abundant in WT mice versus AEG-1KO mice (Fig. 2C). Aged (15 months old) WT mice demonstrated steatotic changes in the liver, whereas no such change was observed in AEG-1KO mice (Fig. 2D). The mice were monitored until 24 months, by which time point 19 of 21 WT mice had died, whereas a significant survival benefit was observed in the leaner AEG-1KO mice, of which only 3 of 20 mice had died (Fig. 2E). No significant difference was observed in glucose tolerance, indicating that insulin signaling is functional in AEG-1KO mice (Fig. 2F). Glucose uptake in peripheral tissues by an [^{F]FDG uptake assay also did not show a difference between WT and AEG-1KO mice (data not shown).}

FIGURE 1.

AEG-1KO mice are smaller than WT mice.A, weekly body weight gain (n = 8/group). B, biweekly food consumption (n = 8/group). C, photograph of mice at 15 months of age. D, body weight of male (M) and female (F) mice at 6 and 15 months of age (n = 12/group). For graphs, the data represent mean ± S.E. (error bars). *, p < 0.05.

FIGURE 2.

AEG-1KO mice have less body fat compared with WT mice.A, representative MRI image of 5-month-old mice. B, quantification of body fat distribution determined by MRI (n = 5/group). C, H&amp;E staining of white adipose tissue (WAT) and brown adipose tissue (BAT) at 15 months. D, H&amp;E staining of liver at 15 months. E, Kaplan-Meier survival curve at 24 months (n = 21 for WT and n = 20 for AEG-1KO). In the graphs, the data represent mean ± S.E. (error bars). *, p < 0.05. F, blood glucose was measured at the indicated time points after intraperitoneal injection of a bolus of glucose in overnight fasted mice. The data represent mean ± S.E. (n = 5/group).

Intestinal Fat Absorption Is Inhibited in AEG-1KO Mice

The profound lack of fat in AEG-1KO mice may be due to 1) defects in fat synthesis, 2) increased energy expenditure (fat burning), and/or 3) decreased fat absorption from the intestine. To address this issue, we first measured serum total TG and cholesterol levels, both of which were significantly lower in AEG-1KO compared with WT mice (Fig. 3A). Levels of newly synthesized TG and palmitate in liver and muscle of WT and AEG-1KO mice showed no significant differences (Fig. 3, B and C). An indirect calorimetry analysis measuring O₂ consumption and CO₂ production did not reveal any increased expenditure in AEG-1KO mice (Fig. 3D). Rather, AEG-1KO mice demonstrated a small but statistically significant decrease in energy expenditure compared with WT mice (Fig. 3, D and E). No difference was observed in UCP1 (uncoupling protein 1) levels in white and brown adipose tissue of WT and AEG-1KO mice (data not shown). Both cholesterol and fatty acid absorption by the intestine were markedly inhibited in AEG-1KO versus WT mice (Fig. 3, F and G). Cholesterol absorption was measured by gavaging [^{C]cholesterol and then measuring radioactivity in the feces, whereas fatty acid absorption was quantified by gavaging [H]triolein and then temporally measuring plasma radioactivity. Collectively, the lack of body fat and lean phenotype in AEG-1KO mice may be explained by reduced cholesterol and TG absorption from the intestine.}

FIGURE 3.

Intestinal fat absorption is inhibited in AEG-1KO mouse.A, serum total cholesterol and TG levels (n = 5/group). B and C, measurement of newly synthesized triglyceride (B) and palmitate (C). D and E, indirect calorimetry analysis measuring O₂ consumption and CO₂ production (n = 6/group). Area under the curve (D) and respiration quotient (E) plots are shown. F, measurement of cholesterol absorption from the intestine after gavaging [^{C]cholesterol and then measuring radioactivity in the feces.} G, measurement of fatty acid absorption by gavaging [^{H]triolein and then temporally measuring plasma radioactivity.} H, weekly body weight gain upon feeding HFD (n = 8). I, H&amp;E staining of liver sections of mice fed HFD. In the graphs, the data represent mean ± S.E. (error bars). *, p < 0.05.

We next evaluated the effect of feeding an HFD on the lean phenotype of AEG-1KO mice, starting at 8 weeks of age. Over a period of 5 weeks, each WT mouse gained ∼11 g of body weight on average, whereas no weight gain was observed in AEG-1KO mice (Fig. 3H). Post-mortem evaluation revealed evidence of steatosis in the liver of WT mice but not in AEG-1KO mice (Fig. 3I). These findings support the hypothesis that inhibition of intestinal fat absorption results in lean phenotype in AEG-1KO mice so that even feeding HFD could not induce a gain of body weight.

LXR and PPARα Activities Are Augmented in the Intestine of AEG-1KO Mice

We next searched for the molecular mechanism conferring the phenotypes of AEG-1KO mice. We have identified a novel interaction between AEG-1 and RXR that interferes with co-activator recruitment and inhibits RXR-mediated transcriptional regulation by its heterodimer partners (25). The phenotypes of the AEG-1KO mice may be explained by activation of LXR and PPARα, heterodimer partners of RXR, in the intestines. In enterocytes, activation of LXR inhibits cholesterol absorption by down-regulating cholesterol transporter Npc1l1 and up-regulating cholesterol efflux proteins, Abca1, Abcg5, and Abcg8 (15). Activation of PPARα in the enterocytes promotes β-oxidation of absorbed fatty acids, thereby down-regulating fatty acid absorption into the circulation (23, 24). Products of cholesterol metabolism, such as oxysterols, function as ligands for LXR, whereas fatty acids and fatty acid derivatives are endogenous ligands for PPARα (17, 36). HFD provides ligands to both LXR and PPARα, thereby accentuating this effect with resultant almost complete inhibition of fat absorption.

To confirm our hypothesis, we first checked LXR and PPARα activity by luciferase reporter assays in WT and AEG-1KO hepatocytes upon treatment with synthetic ligands T0901317 (5 μm) and CP775146 (2.5 μm), respectively. Both basal and ligand-dependent LXR and PPARα activities were significantly higher in AEG-1KO hepatocytes compared with WT hepatocytes (Fig. 4, A and B). Analysis of LXR expression showed an increased LXR level in the intestines of AEG-1KO mice compared with WT (Fig. 4C) We treated WT and AEG-1KO mice with T0901317 for 24 h and analyzed recruitment of LXR, AHH3, and co-activator SRC-1 on the promoter region of LXR target gene Abca1 by a ChIP assay using liver samples. A significant increase in recruitment of LXR, AHH3, and SRC-1 was observed in AEG-1KO livers versus WT livers, confirming that our previous in vitro observation (25) holds true in an in vivo situation (Fig. 4D). We checked the expression of LXR target genes in the small intestine of WT and AEG-1KO mice fed regular chow or HFD. The basal expression level of Abca1 was markedly higher in AEG-1KO mice compared with WT mice (Fig. 4E). With regular chow, the heterodimeric efflux pump Abcg5 and Abcg8 showed small but significant increase in AEG-1KO versus WT mice. Upon HFD treatment, the expression of all three genes was significantly induced in both WT and AEG-1KO mice. However, the magnitude of induction was significantly robust, especially for Abca1, in AEG-1KO mice compared with WT mice (Fig. 4F). Conversely, the basal expression level of the cholesterol transporter Npc1l1 was significantly lower in AEG-1KO mice versus WT mice, and upon HFD treatment, this decrease was further accentuated (Fig. 4E). The increased levels of the cholesterol efflux pumps and decrease in the cholesterol transporter explain the inhibition of intestinal cholesterol absorption in AEG-1KO mice, which is further accentuated upon HFD treatment that provides ligands to LXR, augmenting its effects.

FIGURE 4.

LXR is activated in the intestine of AEG-1KO mice.A and B, LXRE (A) and PPARE (B) luciferase activity was determined in WT and AEG-1KO hepatocytes upon treatment with LXR ligand, T0901317 (5 μm), and PPARα ligand, CP775146 (2.5 μm), respectively. R.L.U., relative luciferase units. C, the expression of the indicated proteins was detected by Western blot in the intestines of WT and AEG-1KO mice. Two independent mice of each group are shown. D, ChIP assays were performed using anti-LXRα, anti-AHH3, and anti-SRC-1 antibodies using T0901317-treated WT and AEG-1KO livers and PCR primers amplifying promoter regions of the Abca1 gene. Normal rabbit IgG was used as control, and the data were normalized by results obtained using normal IgG. E and F, expression of the indicated genes by Taqman quantitative RT-PCR in chow- and HFD-fed small intestine. The data represent mean ± S.E. (error bars). *, p < 0.05 versus the corresponding data point in WT mice; #, p < 0.05 chow versus HFD in WT mice; [caret], p < 0.05 chow versus HFD in AEG-1KO mice. A.U., arbitrary units.

We next analyzed the expression levels of PPARα target genes that regulate β-oxidation of fatty acids in the small intestine and liver. For the mitochondrial β-oxidation pathway, we analyzed Cpt1a, the rate-limiting enzyme in the carnitine-dependent transport of fatty acids across the mitochondrial membrane, and Acadl and Acadm, which catalyze the initial steps of fatty acid β-oxidation (37). For peroxisomal β-oxidation, we analyzed Acox1, the first enzyme in this pathway (37). In standard chow-fed mice, Cpt1a and Acadl expression levels were significantly higher in the small intestine, and Cpt1a, Acadl, and Acadm levels were significantly higher in the liver of AEG-1KO mice versus WT mice (Fig. 5, A and B). There were no significant differences in the Acox1 levels between AEG-1KO and WT mice in either organ. With HFD treatment, a significant induction of all four genes was observed in both the small intestine and liver of WT mice. In the small intestine of AEG-1KO mice, HFD treatment resulted in a significant induction of all four genes, which was significantly more pronounced than the induction seen in WT mice (Fig. 5A). However, HFD treatment did not result in further induction of these genes in the liver of AEG-1KO mice (Fig. 5B). The basal Ppara level was significantly higher in the small intestine and liver of AEG-1KO mice versus WT mice (Fig. 5, C and D). HFD treatment resulted in induction of Ppara in both WT and AEG-1KO liver and small intestine. However, the magnitude of this induction was much more pronounced in the small intestine of AEG-1KO mice versus WT mice (Fig. 5C). Additionally, in AEG-1KO mice, the HFD-mediated induction of Ppara was more robust in the small intestine than in the liver (Fig. 5, C and D). No significant difference was observed in the level of Pparg and Ppargc1, the co-activator for PPAR (Fig. 5, C and D). The changes in LXR and PPARα target genes in the small intestines of chow- and HFD-fed mice were confirmed at the protein level by Western blot analysis (Fig. 5E). We measured products of fatty acid β-oxidation in fresh liver and small intestine homogenates using [^{C]palmitate (}35). Acetyl-CoA, generated by β-oxidation, enters the tricarboxylic acid cycle to generate CO₂. [^{C]Palmitate that does not get oxidized forms acid-soluble metabolites. Fatty acid β-oxidation was significantly increased in the liver and small intestine of AEG-1KO mouse} versus WT (Fig. 5F). Collectively, these findings suggest that under basal conditions, higher levels of PPARα and its target genes in the small intestine of AEG-1KO mice confer increased fatty acid β-oxidation in the enterocytes, resulting in decreased fatty acid absorption. This effect is further accentuated upon HFD treatment that provides ligands to PPARα, thus augmenting levels of both Ppara and its target genes. The lack of absorption of fatty acids from the intestines upon HFD treatment results in decreased delivery of PPARα ligands to the liver, so that there was no further induction by HFD compared with the chow diet. No significant difference was observed in the expression levels of enterocyte fatty acid transporters, FABPpm/Got2, FAT/Cd36, and FATP4/Slc27a4 in WT and AEG-1KO mice (Fig. 5G) with regular chow or HFD, further indicating that fatty acid transport from the lumen to the enterocyte is normal, and it is enhanced β-oxidation that abrogates absorption to the plasma. We also did not observe a significant difference in bile acid composition between WT and AEG-1KO mice (Fig. 5H), indicating that anomalies in bile composition may not be a potential mechanism of decreased fat absorption.

FIGURE 5.

PPARα is activated in the intestine of AEG-1KO mice.A–D, expression of the indicated genes by Taqman quantitative RT-PCR in chow- and HFD-fed small intestine (A and C) and chow- and HFD-fed liver (B and D). The data represent mean ± S.E. (error bars). *, p < 0.05 versus corresponding data point in WT mice; #, p < 0.05 chow versus HFD in WT mice; [caret], p < 0.05 chow versus HFD in AEG-1KO mice. White bars, WT; black bars, AEG-1KO. E, expression of the indicated proteins was checked in the small intestines of chow- or HFD-fed WT and AEG-1KO mice by Western blot analysis. F, fatty acid β-oxidation is increased in AEG-1KO mice. Fresh liver and small intestine homogenates were used with [^{C]palmitate as substrate. Oxidation rate is nmol/mg/min. The data represent mean ± S.E. *,} p < 0.01. G, expression of the indicated genes was measured in the small intestines of the chow- and HFD-fed WT and AEG-1KO mice. H, bile acid levels in the gallbladder of adult mice. MCA, muricholic acid; TUDCA, tauroursodeoxycholic acid; TCA, taurocholic acid. For G and H, the data represent mean ± S.E. For A–D and F–H, n = 4/group. A.U., arbitrary units.

AEG-1KO Mice Are Lean and Survive Longer

We generated an AEG-1KO mouse in which the promoter region, exon 1, and part of intron 1 of the AEG-1 gene was deleted using a Cre-loxP system (25, 29). AEG-1KO mice were viable and fertile, although the litter size was smaller than expected. Both WT and KO mice developed normally to full adulthood, indicating that AEG-1 is probably not involved in developmental regulation. Histological analysis of the internal organs of the adult mice also did not reveal any significant and noticeable difference between WT and AEG-1KO mice (29).

However, a marked difference in body weight gain was observed between age-matched WT and AEG-1KO mice. We performed weekly monitoring of body weight starting at 4 weeks of age when the mice were weaned. Over the next 12 weeks, the average body weight gain by each WT mouse was ∼14 g, whereas that in AEG-1KO mice was only ∼4 g (Fig. 1A). During this period, the food consumption (g of food/g of body weight) was similar in both groups (Fig. 1B). The substantial difference in body weight gain continued with age (Fig. 1, C and D) and was more pronounced in male mice compared with female mice (Fig. 1, C and D). MRI analysis of the mice demonstrated a significant decrease in visceral, subcutaneous, and intramuscular fat in the AEG-1KO mice compared with WT mice (Fig. 2, A and B). Histologically, adipocytes in white and brown adipose tissues were larger and more abundant in WT mice versus AEG-1KO mice (Fig. 2C). Aged (15 months old) WT mice demonstrated steatotic changes in the liver, whereas no such change was observed in AEG-1KO mice (Fig. 2D). The mice were monitored until 24 months, by which time point 19 of 21 WT mice had died, whereas a significant survival benefit was observed in the leaner AEG-1KO mice, of which only 3 of 20 mice had died (Fig. 2E). No significant difference was observed in glucose tolerance, indicating that insulin signaling is functional in AEG-1KO mice (Fig. 2F). Glucose uptake in peripheral tissues by an [^{F]FDG uptake assay also did not show a difference between WT and AEG-1KO mice (data not shown).}

FIGURE 1.

AEG-1KO mice are smaller than WT mice.A, weekly body weight gain (n = 8/group). B, biweekly food consumption (n = 8/group). C, photograph of mice at 15 months of age. D, body weight of male (M) and female (F) mice at 6 and 15 months of age (n = 12/group). For graphs, the data represent mean ± S.E. (error bars). *, p < 0.05.

FIGURE 2.

AEG-1KO mice have less body fat compared with WT mice.A, representative MRI image of 5-month-old mice. B, quantification of body fat distribution determined by MRI (n = 5/group). C, H&amp;E staining of white adipose tissue (WAT) and brown adipose tissue (BAT) at 15 months. D, H&amp;E staining of liver at 15 months. E, Kaplan-Meier survival curve at 24 months (n = 21 for WT and n = 20 for AEG-1KO). In the graphs, the data represent mean ± S.E. (error bars). *, p < 0.05. F, blood glucose was measured at the indicated time points after intraperitoneal injection of a bolus of glucose in overnight fasted mice. The data represent mean ± S.E. (n = 5/group).

Intestinal Fat Absorption Is Inhibited in AEG-1KO Mice

The profound lack of fat in AEG-1KO mice may be due to 1) defects in fat synthesis, 2) increased energy expenditure (fat burning), and/or 3) decreased fat absorption from the intestine. To address this issue, we first measured serum total TG and cholesterol levels, both of which were significantly lower in AEG-1KO compared with WT mice (Fig. 3A). Levels of newly synthesized TG and palmitate in liver and muscle of WT and AEG-1KO mice showed no significant differences (Fig. 3, B and C). An indirect calorimetry analysis measuring O₂ consumption and CO₂ production did not reveal any increased expenditure in AEG-1KO mice (Fig. 3D). Rather, AEG-1KO mice demonstrated a small but statistically significant decrease in energy expenditure compared with WT mice (Fig. 3, D and E). No difference was observed in UCP1 (uncoupling protein 1) levels in white and brown adipose tissue of WT and AEG-1KO mice (data not shown). Both cholesterol and fatty acid absorption by the intestine were markedly inhibited in AEG-1KO versus WT mice (Fig. 3, F and G). Cholesterol absorption was measured by gavaging [^{C]cholesterol and then measuring radioactivity in the feces, whereas fatty acid absorption was quantified by gavaging [H]triolein and then temporally measuring plasma radioactivity. Collectively, the lack of body fat and lean phenotype in AEG-1KO mice may be explained by reduced cholesterol and TG absorption from the intestine.}

FIGURE 3.

Intestinal fat absorption is inhibited in AEG-1KO mouse.A, serum total cholesterol and TG levels (n = 5/group). B and C, measurement of newly synthesized triglyceride (B) and palmitate (C). D and E, indirect calorimetry analysis measuring O₂ consumption and CO₂ production (n = 6/group). Area under the curve (D) and respiration quotient (E) plots are shown. F, measurement of cholesterol absorption from the intestine after gavaging [^{C]cholesterol and then measuring radioactivity in the feces.} G, measurement of fatty acid absorption by gavaging [^{H]triolein and then temporally measuring plasma radioactivity.} H, weekly body weight gain upon feeding HFD (n = 8). I, H&amp;E staining of liver sections of mice fed HFD. In the graphs, the data represent mean ± S.E. (error bars). *, p < 0.05.

We next evaluated the effect of feeding an HFD on the lean phenotype of AEG-1KO mice, starting at 8 weeks of age. Over a period of 5 weeks, each WT mouse gained ∼11 g of body weight on average, whereas no weight gain was observed in AEG-1KO mice (Fig. 3H). Post-mortem evaluation revealed evidence of steatosis in the liver of WT mice but not in AEG-1KO mice (Fig. 3I). These findings support the hypothesis that inhibition of intestinal fat absorption results in lean phenotype in AEG-1KO mice so that even feeding HFD could not induce a gain of body weight.

LXR and PPARα Activities Are Augmented in the Intestine of AEG-1KO Mice

We next searched for the molecular mechanism conferring the phenotypes of AEG-1KO mice. We have identified a novel interaction between AEG-1 and RXR that interferes with co-activator recruitment and inhibits RXR-mediated transcriptional regulation by its heterodimer partners (25). The phenotypes of the AEG-1KO mice may be explained by activation of LXR and PPARα, heterodimer partners of RXR, in the intestines. In enterocytes, activation of LXR inhibits cholesterol absorption by down-regulating cholesterol transporter Npc1l1 and up-regulating cholesterol efflux proteins, Abca1, Abcg5, and Abcg8 (15). Activation of PPARα in the enterocytes promotes β-oxidation of absorbed fatty acids, thereby down-regulating fatty acid absorption into the circulation (23, 24). Products of cholesterol metabolism, such as oxysterols, function as ligands for LXR, whereas fatty acids and fatty acid derivatives are endogenous ligands for PPARα (17, 36). HFD provides ligands to both LXR and PPARα, thereby accentuating this effect with resultant almost complete inhibition of fat absorption.

To confirm our hypothesis, we first checked LXR and PPARα activity by luciferase reporter assays in WT and AEG-1KO hepatocytes upon treatment with synthetic ligands T0901317 (5 μm) and CP775146 (2.5 μm), respectively. Both basal and ligand-dependent LXR and PPARα activities were significantly higher in AEG-1KO hepatocytes compared with WT hepatocytes (Fig. 4, A and B). Analysis of LXR expression showed an increased LXR level in the intestines of AEG-1KO mice compared with WT (Fig. 4C) We treated WT and AEG-1KO mice with T0901317 for 24 h and analyzed recruitment of LXR, AHH3, and co-activator SRC-1 on the promoter region of LXR target gene Abca1 by a ChIP assay using liver samples. A significant increase in recruitment of LXR, AHH3, and SRC-1 was observed in AEG-1KO livers versus WT livers, confirming that our previous in vitro observation (25) holds true in an in vivo situation (Fig. 4D). We checked the expression of LXR target genes in the small intestine of WT and AEG-1KO mice fed regular chow or HFD. The basal expression level of Abca1 was markedly higher in AEG-1KO mice compared with WT mice (Fig. 4E). With regular chow, the heterodimeric efflux pump Abcg5 and Abcg8 showed small but significant increase in AEG-1KO versus WT mice. Upon HFD treatment, the expression of all three genes was significantly induced in both WT and AEG-1KO mice. However, the magnitude of induction was significantly robust, especially for Abca1, in AEG-1KO mice compared with WT mice (Fig. 4F). Conversely, the basal expression level of the cholesterol transporter Npc1l1 was significantly lower in AEG-1KO mice versus WT mice, and upon HFD treatment, this decrease was further accentuated (Fig. 4E). The increased levels of the cholesterol efflux pumps and decrease in the cholesterol transporter explain the inhibition of intestinal cholesterol absorption in AEG-1KO mice, which is further accentuated upon HFD treatment that provides ligands to LXR, augmenting its effects.

FIGURE 4.

LXR is activated in the intestine of AEG-1KO mice.A and B, LXRE (A) and PPARE (B) luciferase activity was determined in WT and AEG-1KO hepatocytes upon treatment with LXR ligand, T0901317 (5 μm), and PPARα ligand, CP775146 (2.5 μm), respectively. R.L.U., relative luciferase units. C, the expression of the indicated proteins was detected by Western blot in the intestines of WT and AEG-1KO mice. Two independent mice of each group are shown. D, ChIP assays were performed using anti-LXRα, anti-AHH3, and anti-SRC-1 antibodies using T0901317-treated WT and AEG-1KO livers and PCR primers amplifying promoter regions of the Abca1 gene. Normal rabbit IgG was used as control, and the data were normalized by results obtained using normal IgG. E and F, expression of the indicated genes by Taqman quantitative RT-PCR in chow- and HFD-fed small intestine. The data represent mean ± S.E. (error bars). *, p < 0.05 versus the corresponding data point in WT mice; #, p < 0.05 chow versus HFD in WT mice; [caret], p < 0.05 chow versus HFD in AEG-1KO mice. A.U., arbitrary units.

We next analyzed the expression levels of PPARα target genes that regulate β-oxidation of fatty acids in the small intestine and liver. For the mitochondrial β-oxidation pathway, we analyzed Cpt1a, the rate-limiting enzyme in the carnitine-dependent transport of fatty acids across the mitochondrial membrane, and Acadl and Acadm, which catalyze the initial steps of fatty acid β-oxidation (37). For peroxisomal β-oxidation, we analyzed Acox1, the first enzyme in this pathway (37). In standard chow-fed mice, Cpt1a and Acadl expression levels were significantly higher in the small intestine, and Cpt1a, Acadl, and Acadm levels were significantly higher in the liver of AEG-1KO mice versus WT mice (Fig. 5, A and B). There were no significant differences in the Acox1 levels between AEG-1KO and WT mice in either organ. With HFD treatment, a significant induction of all four genes was observed in both the small intestine and liver of WT mice. In the small intestine of AEG-1KO mice, HFD treatment resulted in a significant induction of all four genes, which was significantly more pronounced than the induction seen in WT mice (Fig. 5A). However, HFD treatment did not result in further induction of these genes in the liver of AEG-1KO mice (Fig. 5B). The basal Ppara level was significantly higher in the small intestine and liver of AEG-1KO mice versus WT mice (Fig. 5, C and D). HFD treatment resulted in induction of Ppara in both WT and AEG-1KO liver and small intestine. However, the magnitude of this induction was much more pronounced in the small intestine of AEG-1KO mice versus WT mice (Fig. 5C). Additionally, in AEG-1KO mice, the HFD-mediated induction of Ppara was more robust in the small intestine than in the liver (Fig. 5, C and D). No significant difference was observed in the level of Pparg and Ppargc1, the co-activator for PPAR (Fig. 5, C and D). The changes in LXR and PPARα target genes in the small intestines of chow- and HFD-fed mice were confirmed at the protein level by Western blot analysis (Fig. 5E). We measured products of fatty acid β-oxidation in fresh liver and small intestine homogenates using [^{C]palmitate (}35). Acetyl-CoA, generated by β-oxidation, enters the tricarboxylic acid cycle to generate CO₂. [^{C]Palmitate that does not get oxidized forms acid-soluble metabolites. Fatty acid β-oxidation was significantly increased in the liver and small intestine of AEG-1KO mouse} versus WT (Fig. 5F). Collectively, these findings suggest that under basal conditions, higher levels of PPARα and its target genes in the small intestine of AEG-1KO mice confer increased fatty acid β-oxidation in the enterocytes, resulting in decreased fatty acid absorption. This effect is further accentuated upon HFD treatment that provides ligands to PPARα, thus augmenting levels of both Ppara and its target genes. The lack of absorption of fatty acids from the intestines upon HFD treatment results in decreased delivery of PPARα ligands to the liver, so that there was no further induction by HFD compared with the chow diet. No significant difference was observed in the expression levels of enterocyte fatty acid transporters, FABPpm/Got2, FAT/Cd36, and FATP4/Slc27a4 in WT and AEG-1KO mice (Fig. 5G) with regular chow or HFD, further indicating that fatty acid transport from the lumen to the enterocyte is normal, and it is enhanced β-oxidation that abrogates absorption to the plasma. We also did not observe a significant difference in bile acid composition between WT and AEG-1KO mice (Fig. 5H), indicating that anomalies in bile composition may not be a potential mechanism of decreased fat absorption.

FIGURE 5.

PPARα is activated in the intestine of AEG-1KO mice.A–D, expression of the indicated genes by Taqman quantitative RT-PCR in chow- and HFD-fed small intestine (A and C) and chow- and HFD-fed liver (B and D). The data represent mean ± S.E. (error bars). *, p < 0.05 versus corresponding data point in WT mice; #, p < 0.05 chow versus HFD in WT mice; [caret], p < 0.05 chow versus HFD in AEG-1KO mice. White bars, WT; black bars, AEG-1KO. E, expression of the indicated proteins was checked in the small intestines of chow- or HFD-fed WT and AEG-1KO mice by Western blot analysis. F, fatty acid β-oxidation is increased in AEG-1KO mice. Fresh liver and small intestine homogenates were used with [^{C]palmitate as substrate. Oxidation rate is nmol/mg/min. The data represent mean ± S.E. *,} p < 0.01. G, expression of the indicated genes was measured in the small intestines of the chow- and HFD-fed WT and AEG-1KO mice. H, bile acid levels in the gallbladder of adult mice. MCA, muricholic acid; TUDCA, tauroursodeoxycholic acid; TCA, taurocholic acid. For G and H, the data represent mean ± S.E. For A–D and F–H, n = 4/group. A.U., arbitrary units.

Discussion

This work provides a conceptual framework for understanding the role of AEG-1 in normal physiology. Up until now, AEG-1 has been known only for its role in tumorigenesis. Here for the first time, we show that the lack of AEG-1 in mice augments LXR and PPARα activity, thereby preventing fat absorption from the intestine. We observed an increase in LXR and PPARα levels in the intestines of AEG-1KO versus WT mice. It has been documented that LXR and PPARα may indirectly regulate each other (38). As such, relief of AEG-1-induced suppression of their activity may lead to increased expression of these transcription factors. One limitation of our study is that we performed ChIP analysis using liver tissue from mice that were treated with the LXR ligand T0901317. Ideally, this study should have been performed using intestinal tissue. However, we encountered several technical hurdles. The presence of only a single layer of enterocytic cells in the intestine and that of tough muscular and peritoneal tissue resulted in extraction of a very limited amount of DNA that precluded effective use for ChIP analysis. To overcome this limitation, we performed the assay using liver tissue as a “proof of principle” using the same LXR target gene that is relevant for the intestine. We documented increased recruitment of SRC-1 to the promoter of the Abca1 gene in the AEG-1KO mouse that provided support to our hypothesis. Thus, it may be argued that AEG-1 regulates LXR and PPARα both at a functional level and at the level of expression. We also did not observe difference in bile acid composition between WT and AEG-1KO mouse. Bile acid metabolism is primarily regulated by FXR, another RXR heterodimer partner. As yet, why AEG-1 knockdown did not affect bile acid composition is not clear. Studies are being performed to obtain more in depth insights into the molecular mechanism of AEG-1-mediated nuclear factor regulation.

It should be noted that abrogation of intestinal fat absorption may be one of several mechanisms that contribute to the lean phenotype in the AEG-1KO mouse. AEG-1 is expressed in the neurons, and knocking out AEG-1 in the brain may also cause a central effect eliciting the observed phenotypes. We did not observe differences in food consumption between WT and AEG-1KO mouse, suggesting that AEG-1 may not regulate feeding behavior. However, the possibility of dysregulation of a plethora of RXR heterodimer partners in the brain contributing to the lean phenotype should not be ignored. Further studies need to be carried out to elucidate a potential role of AEG-1 in regulation of brain function.

In addition to the lean phenotype, we observed that although AEG-1KO mice were viable and fertile, the litter sizes were very small (1–2 pups/litter). Further, even litters generated by crossing AEG-1 ^{breeding pairs were very small (2–3 pups/litter), which precluded generating large numbers of WT and AEG-1KO mice as littermates. A recent study has demonstrated a role of AEG-1 in male infertility, thereby explaining the phenotype (}39). To obtain a sufficient number of AEG-1KO mice for statistical analysis, some of the experiments were carried out with age-matched mice generated by breeding WT and AEG-1KO mice separately. However, it should be noted that the same phenotypes were observed in AEG-1KO mice generated from AEG-1 ^{× AEG-1 matings as from AEG-1KO × AEG-1KO matings. Thus, our findings are not restricted to strains or littermates. AEG-1KO mice live significantly longer than WT. Caloric restriction is the most effective way of increasing longevity (}40). The inherent inability of AEG-1KO mice to accumulate fat may provide them with the beneficial effect of longer life. We previously documented that AEG-1 is fundamentally required for NF-κB activation and, compared with WT mice AEG-1KO mice, showed profound inhibition in aging-associated inflammation (29). This factor may also contribute to increased survival in AEG-1KO mice.

Taken together, these findings provide evidence that AEG-1 plays an essential role in maintaining homeostasis in lipid metabolism. These findings have important implications for a number of different pathological conditions, including obesity and non-alcoholic fatty liver disease, where AEG-1 overexpression may facilitate increased fat availability and may contribute to fat accumulation and ultimately cirrhosis of the liver and obesity-associated cancers. Our findings highlight AEG-1 as a new potential target to therapeutic intervention of obesity-associated illnesses.

Author Contributions

C. L. R. designed and performed experiments, analyzed the data, and wrote the manuscript. J. S., A. S., R. G., L. E., D. R., M. A., X. S., F. C., G. S., J. Z., C. C., X. G., M. A. S., and J. J. W. designed and performed experiments and analyzed data. S. G., P. B. H., and P. B. F. interpreted data and edited the manuscript. D. S. designed and performed experiments, analyzed data, and edited the manuscript.