• Open Access

Endoplasmic reticulum stress induces hepatic steatosis via increased expression of the hepatic very low-density lipoprotein receptor

Authors

  • Hyunsun Jo,

    1. School of Biological Sciences, Institute of Molecular Biology and Genetics, University of Ulsan, Ulsan, Korea
    Search for more papers by this author
  • Sung Sik Choe,

    1. School of Biological Sciences, Institute of Molecular Biology and Genetics, University of Ulsan, Ulsan, Korea
    Search for more papers by this author
  • Kyung Cheul Shin,

    1. School of Biological Sciences, Institute of Molecular Biology and Genetics, University of Ulsan, Ulsan, Korea
    Search for more papers by this author
  • Hagoon Jang,

    1. School of Biological Sciences, Institute of Molecular Biology and Genetics, University of Ulsan, Ulsan, Korea
    Search for more papers by this author
  • Jae Ho Lee,

    1. School of Biological Sciences, Institute of Molecular Biology and Genetics, University of Ulsan, Ulsan, Korea
    Search for more papers by this author
  • Je Kyung Seong,

    1. Department of Anatomy and Cell Biology, College of Veterinary Medicine and Research Institute for Veterinary Science, University of Ulsan, Ulsan, Korea
    Search for more papers by this author
  • Sung Hoon Back,

    1. School of Biological Sciences, World Class University, University of Ulsan, Ulsan, Korea
    Search for more papers by this author
  • Jae Bum Kim

    Corresponding author
    1. School of Biological Sciences, Institute of Molecular Biology and Genetics, University of Ulsan, Ulsan, Korea
    2. Department of Biophysics and Chemical Biology, Seoul National University, Seoul, Korea
    • School of Biological Sciences, Institute of Molecular Biology & Genetics, Center for Adipose Tissue Remodeling, Seoul National University, San 56-1, Sillim-dong, Gwanak-gu, Seoul, Korea===

    Search for more papers by this author
    • fax: (82)-2-878-5852


  • Potential conflict of interest: Nothing to report.

  • Supported by the National Creative Research Initiative Program (2012-0001241) and the World Class University Program (R31-10032) funded by the Ministry of Education, Science, and Technology. H. Jo, S. S. Choe, K. C. Shin, H. Jang, and J. H. Lee were supported by the BK21 program, Korea

Abstract

Recent evidence suggests that obese animals exhibit increased endoplasmic reticulum (ER) stress in the liver and adipose tissue. Although ER stress is closely associated with lipid homeostasis, it is largely unknown how ER stress contributes to hepatic steatosis. In this study, we demonstrate that the induction of ER stress stimulates hepatic steatosis through increased expression of the hepatic very low-density lipoprotein receptor (VLDLR). Among the unfolded protein response sensors, the protein kinase RNA-like ER kinase–activating transcription factor 4 signaling pathway was required for hepatic VLDLR up-regulation. In primary hepatocytes, ER stress–dependent VLDLR expression induced intracellular triglyceride accumulation in the presence of very low-density lipoprotein. Moreover, ER stress–dependent hepatic steatosis was diminished in the livers of VLDLR-deficient and apolipoprotein E–deficient mice compared with wild-type mice. In addition, the VLDLR-deficient mice exhibited decreased hepatic steatosis upon high-fat diet feeding. Conclusion: These data suggest that ER stress–dependent expression of hepatic VLDLR leads to hepatic steatosis by increasing lipoprotein delivery to the liver, which might be a novel mechanism explaining ER stress–induced hepatic steatosis. (HEPATOLOGY 2013;57:1366–1377)

The endoplasmic reticulum (ER) regulates protein folding, calcium storage, and the biosynthesis of macromolecules such as steroids, lipids, and carbohydrates. Various stimuli increase the accumulation of unfolded proteins in the lumen of ER, leading to ER stress.1 To adapt to ER stress, several pathways, including those of inositol-requiring protein 1α (IRE1α), activating transcription factor 6 (ATF6), and protein kinase RNA-like ER kinase (PERK), are activated to transmit their downstream signaling cascades, which are collectively referred to as the unfolded protein response (UPR).1, 2 Although it has been suggested that ER stress is associated with metabolic diseases such as obesity, atherosclerosis, hyperlipidemia, and type 2 diabetes,3-6 the underlying mechanisms linking ER stress to metabolic dysregulation are not thoroughly understood.

Hepatic steatosis, which is defined by the excessive accumulation of triglycerides in the liver, is one of the hallmarks of obesity-related pathologies.7 Hepatic triglyceride levels are regulated by multiple mechanisms such as de novo synthesis, lipolysis, dietary lipid uptake, and delivery/secretion of lipoprotein particles.7-11 Recently, it has been suggested that ER stress is involved in hepatic lipid metabolism.12-17 For example, hepatic XBP1 deficiency exhibits decreased expression of lipogenic genes.12 Furthermore, attenuation of eukaryotic initiation factor 2α (eIF2α) in GADD34 transgenic mice protects against hepatic steatosis induced by a high-fat diet (HFD).15 Additionally, liver-specific IRE1α and whole-body ATF6 knockout mice exacerbate hepatic steatosis upon pharmacological ER stress.16, 17 Despite these findings, it is not completely understood how elevated ER stress in the liver contributes to hepatic steatosis.

Very low-density lipoprotein receptor (VLDLR) is highly expressed in the brain, heart, muscle, and adipose tissue.18 VLDLR binds to reelin and plays an important role in neuronal development.19 As a member of the LDLR superfamily, VLDLR acts as a receptor for apolipoprotein E (APOE)-containing several lipoproteins such as triglyceride-rich chylomicrons, very low-density lipoprotein (VLDL), and intermediate density lipoproteins.20-22 The VLDLR also mediates lipid uptake through lipoprotein lipase (LPL)-dependent lipolysis or receptor-mediated endocytosis.23 VLDLR-deficient mice are protected from obesity induced by HFD feeding or leptin deficiency.24 In addition, elevated VLDLR levels stimulate lipid accumulation in cardiomyocytes under hypoxic conditions,25 and peroxisome proliferator-activated receptor γ–mediated VLDLR expression facilitates lipid accumulation in adipose tissue.26, 27 However, the pathophysiological roles of hepatic VLDLR have not been thoroughly elucidated due to its low expression in liver.28

In this study, we demonstrate that hepatic VLDLR expression is increased in response to ER stress and that elevated levels of hepatic VLDLR contribute to ER stress–dependent hepatic steatosis. Among the UPR genes, PERK–activating transcription factor 4 (ATF4) signaling mediates hepatic VLDLR expression in the presence of ER stress. Moreover, ER stress–dependent hepatic steatosis is greatly attenuated in VLDLR-deficient and APOE-deficient mice. These findings suggest a novel role for the hepatic VLDLR in ER stress–dependent hepatic steatosis.

Abbreviations

APOE, apolipoprotein E; ATF4, activating transcription factor 4; ATF6, activating transcription factor 6; BIP/GRP78, binding immunoglobulin protein/glucose-regulated protein 78 kDa; CHOP, CCAAT/enhancer-binding protein homologous protein; eIF2α, eukaryotic initiation factor 2α; ER, endoplasmic reticulum; ERDJ4, endoplasmic reticulum-resident DNAj homolog 4; HFD, high-fat diet; IRE1α, inositol-requiring protein 1α; MEF, mouse embryonic fibroblast; mRNA, messenger RNA; PERK, protein kinase RNA-like ER kinase; qRT-PCR, quantitative reverse-transcription polymerase chain reaction; siRNA, small interfering RNA; UPR, unfolded protein response; VLDL, very low-density lipoprotein; VLDLR, very low-density lipoprotein receptor; WT, wild-type.

Materials and Methods

Animals.

All animal experiments were approved by the Seoul National University Animal Experiment Ethics Committee. VLDLR-deficient mice were purchased from the The Jackson Laboratory (Bar Harbor, ME; strain 002529 B6;127S7-Vldlr <tm1Her>/J) and bred in isolated cages. APOE-deficient mice were purchased from Central Lab Animal Inc. (Seoul, Korea). All of the animals were housed at 22 ± 2°C in 55 ± 5% relative humidity with a 12-hour light/dark cycle. Normal chow (Purina Mills) or an HFD (Research diet) was given ad libitum. For the in vivo study of ER stress, tunicamycin (1 μg/g body weight) was injected intraperitoneally into 10-week-old mice. Thereafter, all tissue samples were dissected.

Reagents.

Tunicamycin was purchased from Calbiochem (Germany). Thapsigargin, brefeldin A, and actinomycin D were purchased from Sigma Aldrich (St. Louis, MO). Small interfering RNAs (siRNAs) designed for knockdown were purchased from Genepharma (Shanghai, China). Human VLDL was purchased from Academy Bio-Medical Company, Inc (Houston, TX).

VLDL Incubation and Triglyceride Measurement.

Primary hepatocytes (2.5 × 106 cells) were cultured in 100-mm plates. All of the experiments were performed the day after seeding. Primary hepatocytes were cultured in M199 medium supplemented with 10% fetal bovine serum. Primary hepatocytes were incubated with serum-depleted medium for 3 hours, administrated with ER stress–inducing chemicals for 6 hours, and then incubated with human VLDL (20 μg/mL) for 1 hour. All of the biochemical assays were performed with solution extracted from 2.5 × 106 hepatocytes. Intracellular triglyceride was measured according to the manufacturer's protocol (Bioassay Systems, Hayward, CA). For the loading controls, the values were normalized by total protein content, which was analyzed with a bicinchoninic acid protein assay kit that was compatible with reducing agents (Thermo Scientific, Rockford, IL). The total volume of the lipid-containing solution was also used for the normalization of total triglyceride.

Liver Triglyceride Measurement.

Liver triglyceride measurement was performed according to the manufacturer's protocol (Bioassay systems, Hayward, CA). Liver tissue samples (30-60 mg) were homogenized in 500 μL of 5% triton-X100 with a tissue homogenizer. Total tissue extracts were incubated in a water bath up to 80°C and cooled down to room temperature, and this step was repeated. After further centrifugation of the total tissue extract, the supernatant was collected and used in further biochemical assays. The measurement of hepatic triglycerides was performed using the spectrometric method.

Hepatic VLDL Secretion.

Wild-type (WT) and VLDLR-deficient mice were injected with poloxamer-407 (Sigma Aldrich, catalog #16758) in saline. Poloxamer-407 was administrated intraperitoneally at 1000 mg/kg according to previous studies.29 Tunicamycin (1 μg/g body weight) was pretreated before detergent injection. At 1, 3, and 7 hours after injection, blood samples were collected, serum was prepared, and triglyceride concentrations were determined.

Statistical Comparisons.

Data were compared using a paired Student t test or analysis of variance as appropriate and are represented as the mean ± SD. P < 0.05 was considered statistically significant.

Results

Tunicamycin Rapidly Induces Hepatic Steatosis.

ER stress–inducing chemicals including tunicamycin provoke hepatic steatosis in mice.30-32 Although several mechanisms have been proposed, it is largely unknown how ER stress causes lipid dysregulation in the liver. To monitor tunicamycin-induced ER stress responses, the expression of binding immunoglobulin protein/glucose-regulated protein 78 kDa (BIP/GRP78) messenger RNA (mRNA), as a representative ER chaperone, was examined in several tissues. As shown in Fig. 1A, BIP/GRP78 mRNA was selectively elevated in the liver compared with other tissues. Simultaneously, the expression of ER stress–response genes, including CCAAT/enhancer-binding protein homologous protein (CHOP), endoplasmic reticulum-resident DNAj homolog 4 (ERDJ4), and spliced XBP1, was significantly stimulated in the liver (Supporting Fig. 1A,D). Consistent with previous reports,30-32 tunicamycin induced hepatic steatosis (Fig. 1B). Interestingly, the level of accumulated hepatic triglyceride was rapidly elevated upon tunicamycin treatment (Fig. 1C), whereas the level of serum triglyceride was gradually decreased (Fig. 1D). These data imply that systemic lipid delivery or the secretory system might be involved in ER stress-dependent hepatic steatosis.

Figure 1.

Tunicamycin induces hepatic steatosis. Mice were intraperitoneally injected with tunicamycin (1 μg/g body weight). (A) BIP/GRP78 mRNA expression was analyzed in several tissues. The TATA-binding protein was used as the internal control for qRT-PCR analysis. (B) Oil Red O staining in the liver after cryosection. (C) The hepatic triglyceride level was measured at the indicated time points. (D) Serum triglyceride level was measured at the indicated time points. Data are presented as the mean ± SD (n = 4-5 for each group). Each experiment was performed independently at least three times. *P < 0.05, **P < 0.01 (Student t test). con, control; DMSO, dimethyl sulfoxide; Tunica, tunicamycin.

Hepatic VLDLR Is Up-regulated by ER Stress.

Hepatic steatosis is stimulated by increased lipogenesis, decreased fatty acid oxidation, increased fatty acid/lipoprotein uptake, or decreased lipid secretion. To determine which genes are associated with ER stress–dependent hepatic steatosis, we carefully investigated the expression of genes involved in triglyceride metabolism (e.g., lipogenesis, fatty acid oxidation, and lipid delivery) by quantitative reverse-transcription polymerase chain reaction (qRT-PCR) analysis. In the presence of tunicamycin, most lipogenic genes such as SREBP1c, FAS, and SCD1 were decreased in the liver, whereas there was no consistent change in genes related to fatty acid oxidation (Supporting Fig. 2A-B). Among the lipid delivery genes, the expression of LDLR, LRP1, CD36, and SRBI were decreased or not significantly changed by tunicamycin, whereas the expression of hepatic VLDLR was greatly elevated (Supporting Fig. 2C). To rule out nonspecific chemical effects of tunicamycin on gene expression in vivo, we recapitulated the gene expression analysis in primary hepatocytes. In accordance with the data from the tunicamycin-treated animals, the expression patterns of most genes related to lipid metabolism were not consistently changed by various ER stress–inducing chemicals (Supporting Table 1). In contrast, expression of hepatic VLDLR was significantly elevated by ER stress–inducing chemicals within several hours in liver and primary hepatocytes (Fig. 2A,B). Furthermore, hepatic VLDLR protein was increased by tunicamycin (Fig. 2C). We also examined the level of ER stress marker gene expression in other tissues with or without tunicamycin. Unlike liver, expression of ER stress marker genes including CHOP, ERDJ4, and spliced XBP1 was not significantly altered in other tissues such as adipose tissue and heart (Supporting Fig. 1). Although the basal level of hepatic VLDLR mRNA was quite low,18 the level of tunicamycin-induced hepatic VLDLR mRNA was comparable with that of other tissues (Supporting Fig. 3).

Figure 2.

Hepatic VLDLR is up-regulated by ER stress–inducing chemicals. (A) qRT-PCR analysis of VLDLR mRNA. (B) Primary hepatocytes were treated with several ER stress–inducing chemicals. (C) VLDLR protein level was monitored via western blotting. (D) Actinomycin D (1 μg/mL) was used to block de novo transcription. Primary hepatocytes were incubated with actinomycin D for 3 hours prior to the activation of ER stress. (E,F) Mice were injected intraperitoneally with actinomycin D (2 μg/g body weight) before tunicamycin injection. Total RNA was isolated from these liver tissue samples. (E) qRT-PCR analysis of VLDLR mRNA expression. (F) Hepatic triglyceride level was measured. TATA-binding protein was used as the internal control for qRT-PCR analysis. Data are presented as the mean ± SD (independent plates for n = 3). Each experiment was performed independently more than three times. *P < 0.05, **P < 0.01 (Student t test). ActD, actinomycin D; Bre A, brefeldin A; DMSO, dimethyl sulfoxide; Thapsi, thapsigargin; Tunica, tunicamycin.

To examine whether ER stress–mediated VLDLR expression is mediated by de novo transcription, actinomycin D was used to inhibit RNA polymerase II–dependent transcription. In primary hepatocytes, the expression of VLDLR and ER stress–responding genes such as BIP/GRP78 and ERDJ4 was completely abolished by actinomycin D (Fig. 2D). Next, to test whether ER stress–induced gene expression is essential for hepatic steatosis, actinomycin D was injected into mice with or without tunicamycin. As shown in Fig. 2E, the expression of VLDLR mRNA was dramatically repressed by actinomycin D even in the presence of tunicamycin. Moreover, tunicamycin-induced hepatic triglyceride accumulation was mitigated by actinomycin D (Fig. 2F), implying that transcriptional up-regulation of the hepatic VLDLR is required for tunicamycin-induced hepatic steatosis.

VLDLR Plays a Role in ER Stress–Dependent Hepatic Triglyceride Accumulation.

Although VLDLR has been suggested to be a peripheral receptor for triglyceride-enriched lipoproteins,20-22 its hepatic function is poorly understood due to its low expression in the liver. To examine the role of the hepatic VLDLR, the modified VLDL incubation assay has been adopted in primary hepatocytes.33 In the presence of exogenous VLDL, the level of intracellular triglyceride accumulation was substantially increased by various ER stress–inducing chemicals such as tunicamycin, thapsigargin, and brefeldin A (Fig. 3A,E). Further, the level of intracellular cholesterol was also increased by tunicamycin (Supporting Fig. 4), implying that cholesterol uptake might be mediated via VLDLR from exogenous VLDL particles. To test whether ER stress–mediated intracellular triglyceride accumulation requires newly synthesized proteins, cycloheximide was cotreated with tunicamycin. As shown in Fig. 3B, increased intracellular triglyceride upon tunicamycin treatment was suppressed by cycloheximide, indicating that the newly translated proteins would be necessary for ER stress–dependent hepatic triglyceride accumulation. To determine whether hepatic VLDLR would play a role in ER stress–dependent hepatic triglyceride accumulation, expression of hepatic VLDLR was suppressed via siRNA (Fig. 3C). In primary hepatocytes, VLDLR knockdown alleviated tunicamycin-mediated intracellular triglyceride accumulation (Fig. 3D). Moreover, when we examined the effects of various ER stress–inducing chemicals on intracellular triglyceride accumulation, primary hepatocytes from VLDLR-deficient mice failed to increase intracellular triglyceride level (Fig. 3E). Therefore, these data explicitly indicate that the hepatic VLDLR is necessary for ER stress–dependent intracellular triglyceride accumulation.

Figure 3.

Hepatic VLDLR is required for ER stress–dependent intracellular triglyceride accumulation. (A) Measurement of intracellular triglyceride content with tunicamycin in primary hepatocytes. VLDL(−) means no incubation of exogenous human VLDL as a negative control. (B) Intracellular triglyceride content was measured with the preincubation of cycloheximide. (C) qRT-PCR was performed to analyze the level of VLDLR mRNA with or without VLDLR siRNA. (D) Intracellular triglyceride content was measured in VLDLR-suppressed primary hepatocytes. (E) Intracellular triglyceride content was measured in primary hepatocytes from WT or VLDLR-deficient mice with or without tunicamycin, thapsigargin or brefeldin A. Data are presented as the mean ± SD (independent plates for n = 3). Each experiment was performed independently more than three times. *P < 0.05 (Student t test). Bre A, brefeldin A; CHX, cycloheximide; DMSO, dimethyl sulfoxide; N.S., not significant; Thapsi, thapsigargin; Tunica, tunicamycin.

PERK-ATF4 Regulates Hepatic VLDLR Expression upon ER Stress.

Three UPR sensors—ATF6, PERK, and IRE1α—are key players in ER stress responses.1 To address which UPR pathway is involved in ER stress–induced hepatic VLDLR expression, the three UPR sensors were suppressed via siRNAs in primary hepatocytes (Supporting Fig. 5). Compared with the suppression of IRE1α and ATF6, the suppression of PERK significantly decreased ER stress–mediated hepatic VLDLR expression (Fig. 4A). As a positive control, the expression level of CHOP mRNA was examined (Fig. 4A). Because ATF4 is a well-known transcription factor that mediates PERK downstream signaling,1 we decided to investigate the involvement of ATF4 in hepatic VLDLR expression. As shown in Fig. 4B, the expression of VLDLR and CHOP mRNA was significantly attenuated in ATF4 knockdown hepatocytes in the presence of tunicamycin. To confirm that the increase in hepatic VLDLR expression following ER stress requires ATF4, ATF4-deficient mouse embryonic fibroblasts (MEFs) were challenged with tunicamycin. As shown in Fig. 4C, ER stress–dependent expression of the VLDLR and CHOP was attenuated in ATF4-deficient MEF. In primary hepatocytes, overexpression of ATF4 stimulated hepatic VLDLR expression (Fig. 4D). Because two well-conserved binding motifs of ATF4 (−200 and −1383 bp upstream from the transcription start site) are found in the promoter of the mouse VLDLR gene, the effect of ATF4 overexpression on VLDLR promoter activity was tested. In accordance with the above results, ATF4 overexpression transactivated the promoter activity of the mouse VLDLR gene (Fig. 4E). Next, in order to examine whether VLDLR expression is regulated by PERK-ATF4 axis in other cell types, we investigated the level of VLDLR mRNA in hepa1c1c and 3T3-L1 adipocytes. In the presence of tunicamycin, there was no significant increase of VLDLR mRNA expression in these cells even though they up-regulated the level of BIP/GRP78 mRNA (Supporting Fig. 6). Taken together, these data indicate that the PERK-ATF4 pathway is involved in, at least, hepatic VLDLR expression following ER stress.

Figure 4.

PERK-ATF4 pathway regulates the expression of hepatic VLDLR in response to ER stress. (A) qRT-PCR analysis of mRNA expression for VLDLR and CHOP. (B) qRT-PCR analysis for mRNA expression of the VLDLR and CHOP. Two different ATF4 siRNAs were used. (C) MEFs were cultured with tunicamycin (1 μg/mL) for 6 hours after incubation in serum-free medium for 3 hours. qRT-PCR was performed for analysis of mRNA expression of VLDLR and CHOP genes. (D) qRT-PCR analysis for mRNA expression of VLDLR. Hepatocytes were transfected with the ATF4 expression vector. (E) Luciferase reporter vector containing the promoter region (∼2 kb) for the murine VLDLR gene was cotransfected with the ATF4 expression vector. TATA-binding protein was used as an internal control for qRT-PCR analysis. Data are presented as the mean ± SD (independent plates for n = 3). Each experiment was performed independently more than three times. *P < 0.05 (Student t test). DMSO, dimethyl sulfoxide; Tunica, tunicamycin.

VLDLR-Deficient Mice Are Protected from Tunicamycin-Induced Hepatic Steatosis.

To directly address whether the increase of hepatic VLDLR expression following ER stress is associated with hepatic steatosis, VLDLR-deficient mice were challenged with tunicamycin. Interestingly, tunicamycin-dependent hepatic triglyceride accumulation was significantly decreased in VLDLR-deficient mice compared with WT mice (Fig. 5A,B). Because tunicamycin-treated VLDLR-deficient mice exhibited a marginal but substantial increase in hepatic triglyceride content, it is likely that tunicamycin-induced hepatic steatosis would not be solely mediated by elevated hepatic VLDLR expression. Rather, it is plausible that an increase in hepatic VLDLR expression following ER stress attributes to the stimulation of hepatic lipid accumulation. When the expression of ER stress–response genes and triglyceride metabolism–related genes was compared in WT and VLDLR-deficient mice, there were few differences between the two groups (Fig. 5C-E). Previously, ER stress decreased the secretion of hepatic VLDL, which led to overt hepatic steatosis.34-36 These data prompted us to test whether the alleviation of tunicamycin-mediated hepatic steatosis in VLDLR-deficient mice results from dysregulated hepatic VLDL secretion. To address this question, we performed an in vivo VLDL secretion assay by detergent injection.29 Expectedly, in WT mice, tunicamycin repressed hepatic VLDL secretion in the presence of the detergent, resulting in a smaller increase in the serum triglyceride levels (Fig. 5F). Similarly, tunicamycin reduced hepatic VLDL secretion in VLDLR-deficient mice (Fig. 5F). These results suggest that the hepatic VLDLR participates in ER stress–dependent hepatic steatosis without a significant change in hepatic VLDL secretion.

Figure 5.

VLDLR-deficient mice alleviate ER stress–dependent hepatic steatosis. Ten-week-old VLDLR−/− and WT littermate mice were injected with tunicamycin (1 μg/g body weight). Liver tissue was isolated for the measurement of triglyceride and gene expression analysis. (A) Oil Red O staining in the liver of VLDLR-deficient and WT mice with or without tunicamycin. (B) Measurement of hepatic triglycerides from VLDLR−/− and WT mice with or without tunicamycin. (C-E) qRT-PCR analysis of genes involved in (C) lipogenesis, (D) fatty acid oxidation, and (E) lipid delivery. TATA-binding protein (TBP) was used as the internal control for qRT-PCR analysis. (F) Hepatic VLDL secretion. Poloxamer 407 was used as a detergent. Data are presented as the mean ± SD (n = 5-6 for each group). Each experiment was performed independently three times. *P < 0.05, **P < 0.01, Student t test. DMSO, dimethyl sulfoxide; Tunica, tunicamycin.

APOE Is Necessary for ER Stress–Induced Hepatic Triglyceride Accumulation.

The VLDLR mediates triglyceride delivery from APOE-containing lipoproteins.18, 37 To examine whether ER stress–mediated hepatic steatosis is associated with triglyceride delivery from APOE-containing lipoproteins, APOE-deficient mice were treated with or without tunicamycin. Compared with WT mice, the basal level of hepatic triglyceride was elevated in APOE-deficient mice as determined by Oil Red O staining (Fig. 6A) and biochemical assay (Fig. 6B). However, APOE-deficient mice did not show a further increase in hepatic triglyceride levels following tunicamycin treatment (Fig. 6A,B). In APOE-deficient mice, the expression level of lipid metabolism-related genes was not significantly different than that of WT mice (Fig. 6C-E). These results suggest that APOE-containing lipoproteins including VLDL might be important for ER stress–dependent hepatic steatosis, probably, via the hepatic VLDLR.

Figure 6.

APOE-deficient mice relieve ER stress–induced hepatic triglyceride accumulation. Ten-week-old APOE−/− and WT (C57BL6/J) mice were injected with tunicamycin (1μg/g body weight). Liver tissue was isolated for the measurement of triglyceride and gene expression analysis. (A) Oil Red O staining of lipid droplets in the liver of APOE-deficient and WT mice with tunicamycin. (B) The measurement of hepatic triglycerides from APOE+/+ and APOE−/− mice with tunicamycin. (C-E) qRT-PCR analysis of genes involved in (C) lipogenesis, (D) fatty acid oxidation, and (E) lipid delivery. TATA-binding protein was used as an internal control for qRT-PCR analysis. The values are the mean ± SD (n = 3-5 for each group). Each experiment was performed independently three times. *P < 0.05 (Student t test). DMSO, dimethyl sulfoxide; Tunica, tunicamycin.

Hepatic VLDLR Expression Is Increased in Obese Animals.

The PERK pathway is activated in the livers of obese animals.38 Interestingly, the level of hepatic VLDLR mRNA was increased in several obese animal models such as leptin-deficient (ob/ob), leptin receptor–deficient (db/db), and diet-induced obese mice (Fig. 7A). The level of hepatic VLDLR protein was also increased in the liver of HFD-fed mice (Fig. 7B). To affirm the potential role of the hepatic VLDLR in obesity-associated hepatic steatosis, VLDLR-deficient mice were fed an HFD. As shown in Fig. 7C,D, the level of hepatic triglyceride was lower in VLDLR-deficient mice than in WT mice following the HFD. In addition, there was no significant difference of serum triglyceride and cholesterol between WT and VLDLR-deficient mice (Supporting Fig. 7). These data suggest that elevated hepatic VLDLR levels in obese animals might be linked with hepatic steatosis.

Figure 7.

Hepatic VLDLR is increased in the liver of obese animals showing hepatic steatosis. (A) qRT-PCR analysis of the mRNA expression of hepatic VLDLR in obese animal models including leptin receptor–deficient (db/db), leptin-deficient (ob/ob), and HFD-fed obese mice. (B) Immunoblotting for the protein level of VLDLR in the liver tissue of HFD-fed mice compared with control mice. Heath shock protein 70 (HSP70) antibody was used as a loading control. (C) VLDLR-deficient mice were fed with a 60% HFD for 8 weeks and compared with their littermate WT mice. Liver tissue was prepared for Oil Red O staining. (D) Measurement of hepatic triglyceride content from WT or VLDLR-deficient mice following an HFD. TATA-binding protein was used as an internal control for qRT-PCR analysis. Data are presented as the mean ± SD (n = 4-5 for each group). Each experiment was performed independently two times. *P < 0.05, **P < 0.01 (Student t test). NC, normal chow.

Discussion

In this study, we found a novel role for hepatic VLDLR in ER stress–dependent hepatic steatosis. Our results show that hepatic VLDLR is up-regulated by the activation of PERK-ATF4 under ER stress. Additionally, increased hepatic VLDLR levels stimulate intracellular triglyceride accumulation with VLDL uptake following ER stress. Given that VLDLR-deficient and APOE-deficient mice are protected from tunicamycin-induced hepatic steatosis, these data indicate that elevated hepatic VLDLR levels during ER stress would be one of the mechanisms to induce hepatic steatosis through the accumulation of lipids in the liver (Supporting Fig. 8).

Consistent with previous reports,30, 31 we observed that tunicamycin promptly induced hepatic steatosis. Notably, the expression of most lipogenic genes was somewhat repressed in tunicamycin-treated animals. Although thapsigargin induced a slight increase in expression of certain lipogenic genes in primary hepatocytes, other ER stress–inducing chemicals (i.e., tunicamycin and brefeldin A) did not cause a similar change in most lipogenic genes. Similarly, it has been reported that ER stress suppresses the expression of lipogenic genes through repression of SREBP1, a key lipogenic transcription factor.14, 39-41 In addition, we also observed that, in VLDLR-deficient and APOE-deficient mice, tunicamycin decreased the expression of lipogenic genes, indicating that the regulation of lipogenic gene expression might not be a causal factor for hepatic triglyceride accumulation in VLDLR-deficient and APOE-deficient mice. In contrast to these results, it has been reported that ER stress potently activates SREBP1c, which subsequently leads to elevated hepatic lipogenesis.30, 42 Because gene expression patterns do not always reflect their enzymatic activities, further studies are required to explore which ER stresses mediate hepatic lipogenesis.

Apolipoprotein B100 and microsomal transfer protein have well-known roles in hepatic VLDL production.43 The levels of microsomal transfer protein mRNA and protein or secretion of apolipoprotein B100 were greatly mitigated by tunicamycin.17, 43-45 These findings led us to test the capacity of VLDL secretion in VLDLR-deficient mice. However, we observed that there was no significant difference in hepatic VLDL secretion between WT and VLDLR-deficient mice, implying that decreased hepatic triglyceride accumulation in tunicamycin-challenged VLDLR-deficient mice would not result from the dysregulation of hepatic VLDL secretion.

Due to the low level of the VLDLR in the liver, pathophysiological roles for the hepatic VLDLR have not been completely elucidated.18 Our study showed that decreased hepatic triglyceride accumulation in tunicamycin-challenged VLDLR-deficient mice did not seem to be primarily mediated by changes in lipogenesis or VLDL secretion. Thus, we proposed the hypothesis that hepatic VLDLR might directly regulate lipid delivery from lipoproteins. Interestingly, we observed that mRNA and protein expression levels of hepatic VLDLR were promoted by ER stress. In primary hepatocytes, various ER stress–inducing chemicals increased the intracellular triglyceride content through hepatic VLDLR in the presence of exogenous VLDL. In contrast, VLDLR deletion or suppression with siRNA impaired ER stress–dependent hepatic lipid accumulation, indicating that hepatic VLDLR plays a key role in APOE-containing VLDL uptake, which is then converted into triglyceride. We observed that the inhibition of protein synthesis or transcription decreased ER stress–induced hepatic triglyceride accumulation. Consistent with the roles for the VLDLR in adipose and heart tissue,25-27 the hepatic VLDLR might contribute to tunicamycin-dependent hepatic triglyceride accumulation. Although the role of the hepatic VLDLR needs to be further investigated, our data indicate that hepatic VLDLR plays a role in lipid delivery into the liver under various ER stress conditions.

Recent studies have suggested that ER stress and UPR signaling are tightly associated with hepatic lipid metabolism.12-17 For instance, hepatic XBP1 regulates the expression of lipogenic enzymes such as SCD1, ACC2, and DGAT2, which are crucial for de novo lipogenesis,12 whereas the UPR branches through IRE1α and/or ATF6 play a role in preventing ER stress–dependent hepatic steatosis.16, 17 In addition, the UPR pathway via PERK/eIF2α is required for the expression of lipogenic genes.15, 46 Here, we discovered that the PERK branch plays a role in ER stress–induced VLDLR expression. Our data revealed that VLDLR expression was regulated by ATF4, a downstream factor of the PERK branch, following ER stress. We observed that ER stress–mediated VLDLR induction was substantially attenuated in ATF4-deficient MEFs or ATF4-suppressed hepatocytes, whereas the overexpression of ATF4 induced the up-regulation of hepatic VLDLR expression in primary hepatocytes. Unlike primary hepatocytes, MEFs failed to increase intracellular triglyceride accumulation upon ER stress, which might be a result of insufficient activity for lipid metabolism in MEFs (data not shown). In accordance with our findings, ATF4-deficient mice are protected from age-related and diet-induced obesity and diet-induced hepatic steatosis.47 Because ATF4 is a key transcription factor for integrating stress responses by eIF2α, it appears that VLDLR might be regulated by amino acid depletion or infection of a double-stranded RNA virus. The role of VLDLR expression in other conditions related to the eIF2α signaling pathway needs to be investigated.

Unlike primary hepatocytes, the expression of VLDLR mRNA was not increased upon ER stress in 3T3-L1 adipocytes and hepa1c1c-7, implying that ER stress would selectively induce VLDLR expression in certain cell types. Recently, it has been reported that ER stress increases intracellular triglyceride levels in Huh-7 cell,48 in which VLDLR expression was not induced by ER stress (data not shown). Similarly, in yeast cells, ER stress elevates lipid droplet formation.49 These data indicate that ER stress–mediated VLDLR expression would be a selective mechanism, probably, limited in certain tissues and/or cell types.

Our data suggest that up-regulation of hepatic VLDLR is one of the potential links between ER stress and hepatic steatosis. Based on its expression profile, the hepatic VLDLR may not actively deliver lipids from lipoproteins under normal conditions. However, increased hepatic VLDLR levels during severe or prolonged ER stress might result in the transport of lipids into the liver to prepare for extreme conditions. Although the roles of hepatic VLDLR in the liver and in hepatic steatosis need to be investigated further, our data suggest that hepatic VLDLR might potentially be a novel therapeutic target against hepatic steatosis induced by ER stress and obesity.

Acknowledgements

We thank Randal Kaufman for sharing ATF4 MEFs, Ki-Up Lee for critical comments on the manuscript, and Esther Park and J. Y. Jang for proofreading the manuscript.

Ancillary