Potential conflict of interest: Nothing to report.
For many years, adipose tissue has been mainly considered as an inert reservoir for storing triglycerides. Since the discovery that adipocytes may secrete a variety of bioactive molecules (hormones, chemokines, and cytokines), an endocrine and paracrine role for white adipose tissue (WAT) in the regulation of energy balance and other physiological processes has been established, particularly with regard to brain and muscle. In contrast, little is known about the interactions of WAT with liver. Hence, we examined the effect of the secretory products of WAT on hepatocytes. Conditioned medium of human WAT explants induced significant steatosis in hepatocyte cell lines. Factor(s) responsible for the conditioned medium-induced steatosis were screened by a battery of blocking antibodies against different cytokines/chemokines shown to be secreted by WAT. In contrast to interleukin-8 and interleukin-6, the monocyte chemoattractant protein-1 was capable of inducing steatosis in hepatocytes in a time-dependent manner at concentrations similar to those found in conditioned medium. Incubation of conditioned medium with antimonocyte chemoattractant protein-1 antibodies prevented triglyceride accumulation. Investigation of the mechanism leading to the triglyceride accumulation showed that both a diminution of apolipoprotein B secretion and an increase in phosphoenolpyruvate carboxykinase messenger RNA may be involved. Conclusion: The monocyte chemoattractant protein-1 secreted by adipose tissue may induce steatosis not only recruiting macrophages but also acting directly on hepatocytes. (HEPATOLOGY 2008.)
Recent lifestyle changes have resulted in a significant increase in the prevalence of obesity, and the distribution of excess adipose tissue in different anatomical locations plays a major role in the development of obesity-associated comorbidities. Epidemiological studies have reported an association between central obesity and insulin resistance,1–4 two components of the metabolic syndrome.5 An important feature of the metabolic syndrome is liver steatosis, that is, the accumulation of triglycerides in hepatocytes. The molecular mechanisms that link the metabolic syndrome and steatosis are partially understood, but are thought to be mainly mediated by insulin resistance.6 In adipose tissue, insulin resistance leads to increased lipase activity, resulting in elevated rates of lipolysis and enhanced free fatty acid flux to the liver. Here, the synergism between sterol regulatory element binding protein-1c and carbohydrate responsive element-binding protein, activated by hyperinsulinemia and hyperglycemia, respectively, leads to increased lipogenic enzymes and the conversion of excess glucose into fatty acids. Thus, in patients with insulin resistance, the net effect of free fatty acid overflow into the liver and increased de novo lipogenesis is an enhanced esterification to triglycerides, hence steatosis.
Since the discovery that adipocytes may secrete a variety of bioactive molecules, such as interleukin (IL)-6, tumor necrosis factor-α, IL-1β, IL-87, 8, and monocyte chemoattractant protein-1 (MCP-1),8, 9 an endocrine role for the adipose tissue in the regulation of several physiological processes has been proposed. These cytokines may exert their effects also on the liver, and their imbalance, as observed in pathological conditions, may lead to chronic, progressive liver damage characterized by inflammation and fibrosis.10
To investigate whether adipose tissue could directly induce liver steatosis, we analyzed the effect of secretory products from human white adipose tissue (WAT) on hepatocytes in vitro. We show that, among the cytokines secreted by WAT, MCP-1 induced a direct accumulation of lipids in hepatocytes.
Rabbit polyclonal anti-phospho cyclic adenosine monophosphate response element binding protein (CREB) (Ser133) and rabbit anti-phospho-p44/42 mitogen-activated protein kinase (MAPK) (extracellular signal-regulated kinase, ERK) (Thr202/Tyr204) antibodies were purchased from Cell Signaling Technology (Danvers, MA). Blocking anti-IL-6, anti-IL-8, anti-MCP-1, anti-growth regulated oncogene (GRO) antibodies, human and mouse recombinant MCP-1 and IL-6 were from R&D Systems Europe (Abingdon, UK). ERK inhibitor (PD98059), p38 MAPK inhibitor (SB203580), and Ca2+ chelator 1,2-bis-(o-Aminophenoxy)-ethane-N,N,-N′,N′-tetraacetic acid tetraacetoxy-methyl ester (BAPTA//AM) were purchased from Calbiochem (San Diego, CA). The chemokine (C-C motif) receptor 2 (CCR2) antagonist (RS102895) and 2-naphtol-AS-E-phosphate (KG501) were from Sigma Aldrich (Buchs, Switzerland).
Human Adipose Tissue and Adipose Tissue Culture.
Samples of subcutaneous WAT (scWAT) were collected from patients undergoing abdominal plastic surgery, and conditioned media (CM) was prepared from scWAT as described.11 The study was approved by the Ethical Committee of Geneva University Hospitals and all patients gave written informed consent to participate.
Cell Culture and Treatments.
Mouse primary hepatocytes were isolated from adult male C57BL/6 mice as described.12 Human hepatoma cells (Huh-7) and mouse primary hepatocytes were cultured in low glucose Dulbecco's modified Eagle's medium (Invitrogen, Basel, Switzerland) supplemented with 10% fetal bovine serum, 100 U/mL penicillin, and 100 μg/mL streptomycin (Invitrogen). Cells were treated with CM or recombinant MCP-1. For MCP-1 (50-200 ng/mL for 8-120 hours) treatment, growth medium was replaced with serum-free medium consisting of Dulbecco's modified Eagle's medium/Ham's nutrient F12 (1:1) supplemented with insulin-transferrin-selenium (Invitrogen) and vitamin C (0.2 mM).
RNA Isolation, Reverse Transcription, and Real-Time Quantitative Polymerase Chain Reaction.
Total RNA was extracted using the RNeasy Mini Kit (Qiagen, Hombrechtikon, Switzerland) and digested by DNase I. complementary DNA was synthesized with SuperScript II RNase H(−) reverse transcriptase (Invitrogen) and random hexanucleotides. For real-time reverse-transcription polymerase chain reaction (PCR), we used the following SYBR Green QuantiTect Primers from Qiagen: human fatty acid synthase (FAS, QT00014588); human sterol regulatory element binding protein (SREBP-1, QT00036897), human apolipoprotein B (ApoB, QT00020139), human microsomal triglyceride transfer protein (MTP, QT00043050), human peroxisome proliferator-activated receptor alpha (PPARα, QT00017451), human peroxisome proliferator-activated receptor gamma (PPARγ, QT00029841), human CCR2 (QT00000224), human ACC-α (QT00033761), human ACC-β (QT00996352), human camptothecin 1 (carnitine palmitoyltransferase 1 [CPT1], QT00082236), and human phosphoenolpyruvate carboxykinase (PEPCK, QT00001197). Relative quantification was performed by real-time PCR using an iCycler iQ Detection System (BioRad Laboratories, Reinach, Switzerland). Mouse CCR2 primers forward 5′-agagagctgcagcaaaaagg-3′ and reverse 5′-ggaaagaggcagttgcaaag-3′ were from Invitrogen.
Raw fluorescence threshold (Ct) values were obtained using SDS 2.0 software (Applied Biosystems). Relative expression level of target genes was normalized using human eukaryotic translation elongation factor A-1 (eEF1A1) forward 5′-agcaaaaatgacccaccaatg-3′ and reverse 5′-ggcctggatggttcaggata-3′, human glucuronidase beta (Gus B) forward 5′-ccaccagggaccatccaat-3′ and reverse 5′-agtcaaaatatgtgttctggacaaagtaa-3′, mouse hypoxanthine-guanine phosphoribosyltransferase (HPRT) forward 5′-gctcgagatgtcatgaaggagat-3′ and reverse 5′-aaagaacttatagccccccttga-3′, and mouse β-cytoplasmic actin forward 5′-ctaaggccaaccgtgaaaagat-3′ and reverse 5′-cacagcctggatggctacgt-3′ (Invitrogen) as references.
Measurement of Cytokines.
The secretion profile of scWAT was determined using human cytokine antibody arrays covering 120 different cytokines (C series 1000; Raybiotech, Inc., Norcross, GA). Quantitative measurements of MCP-1 secretion in CM were performed using duoset enzyme-linked immunosorbent assay (ELISA) development systems (R&D Systems).
Lipid Accumulation Evaluation.
Cells were fixed in 3% paraformaldehyde for 10 minutes. After three washes in phosphate-buffered saline solution (PBS), lipids were stained with Oil-red-O in 40% isopropanol for 2 minutes. After rinsing, coverslips were mounted in Moviol. Cells were observed using an Axiophot photomicroscope (Zeiss, Oberkochen, Germany), and images were acquired with an Axiocam color camera (Zeiss) and analyzed using the Metamorph software (Molecular Devices Corp., Sunnyvale, CA).
Indirect Immunofluorescence and Confocal Microscopy.
Cells were fixed in 3% paraformaldehyde for 10 minutes at room temperature (RT) and permeabilized with 0.3% Triton X-100 for 10 minutes. Cells were first incubated with anti-phospho CREB or anti-phospho ERK diluted in PBS-Tween 0.1% for 30 minutes at RT and then with rhodamine-conjugated anti-rabbit antibodies (Jackson Immunoresearch Laboratories, West Grove, PA) and [4′,6-diamidino-2-phenylindole] for nuclear staining, for 30 minutes at RT. After washing in PBS, cells were mounted in Moviol. Images were acquired using a confocal microscope (LSM510 Meta; Zeiss).
Confocal Video Microscopy and Cytosolic Ca2+ Signal Recording.
Functional expression of MCP-1 receptors in Huh-7 was determined to examine changes in cytosolic Ca2+.13 Huh-7 were loaded with the Ca2+-sensitive fluorophore fluo-4 AM (Molecular Probes, Invitrogen), and changes in fluo-4 fluorescence were monitored using a laser confocal (Nipkow disc; Visitech, Sunderland, UK). Images were acquired using a cooled, 16-bit CCD camera (CoolSnap HQ; Roper Scientific, Trenton, NJ) operated by Metamorph software. Changes in fluorescence over time were expressed as peak fluorescence (f) divided by initial fluorescence (f0).
Cytotoxicity and ApoB ELISA.
The release of lactate dehydrogenase into the culture media was used as a measure of cell death due to damaged membrane and was measured fluorimetrically using the CytoTox-ONE Assay kit (Promega, Madison, WI). ApoB levels in the culture media were quantified using the ApoB Microwell ELISA Assay Kit (AlerCHEK, Portland, ME).
Oleic Acid Uptake.
Huh-7 were treated for 48 hours with 50 ng/mL MCP-1 and incubated with 50 μM oleic acid (4:1 mix of cold and [14C]-oleic acid; 1 μCi per condition) for various times. Cells were extensively washed with ice-cold PBS, scraped, and oleic acid uptake was measured by counting [14C]-associated radioactivity in cell lysates per μg of protein (β-scintillation counter Wallac 1409; Perkin Elmer).
Mitochondrial β-oxidation of oleic acid over 4 hours was assessed by measuring collected 14CO2, as described.14
Results are expressed as mean ± standard deviation of three independent experiments. Results were analyzed either by Student t test or by one-way analysis of variance followed by Tukey test in case of comparison between the means of three or more groups, using GraphPad Prism 4. Values of P < 0.001 (***), P < 0.01 (**) and P < 0.05 (*) were considered statistically significant.
CM of Human scWAT Explants Induces Significant Steatosis in Huh-7 Cells.
To investigate the effect of scWAT on the development of liver steatosis, we cultured Huh-7 in presence of CM (Fig. 1A). Huh-7 cultured in presence of CM for 48 hours displayed a significant accumulation of lipids within the cytoplasm, visualized by Oil-red-O staining (Fig. 1B,C). This lipid droplets accumulation was not dependent on the presence of triglycerides in the CM, as it was caused by a protein factor contained in trichloroacetic acid precipitates (Fig. 1B,C).
MCP-1 Secreted by scWAT Induces Lipid Droplet Accumulation in Both Huh-7 Cells and Mouse Primary Hepatocytes.
To characterize the cytokine profile of human scWAT CM, we used a cytokine antibody array allowing the qualitative assessment of 120 cytokines. The cytokine pattern of one patient is shown in Fig. 2A. Similar patterns of cytokines were observed with CM obtained from all other patients (not shown). IL-6 (Fig. 2A, 1), leptin (Fig. 2A, 2), MCP-1 (Fig. 2A, 3), PARC (Fig. 2A, 4), PDGF-BB (Fig. 2A, 5), adiponectin (Fig. 2A, 6), GRO (Fig. 2A, 7), GROα (Fig. 2A, 8), hepatocyte growth factor (Fig. 2A, 9), and IL-8 (Fig. 2A, 10) were detected in CM. As hepatocyte growth factor, leptin, and adiponectin have been shown to act as protective factors against steatosis,15–17 we investigated the effect of blocking antibodies against IL-6, GRO, IL-8, and MCP-1. Only anti-MCP-1 blocked CM-induced lipid accumulation in Huh-7 cells, albeit not completely (Fig. 2B). Experiments using recombinant cytokines showed that IL-6 had no effect on lipid accumulation, excluding putative, poor antibody affinity in the above experiments (not shown). However, recombinant MCP-1 induced lipid accumulation in ∼45% of Huh-7 cells (Fig. 3 Ab, arrows) in a time-dependent manner (Fig. 3B), and at a concentration similar to that found in CM, estimated by ELISA (50.6 ± 4.9 ng/mL). Removal of MCP-1 after 48 hours could partially reverse the effect. The mean diameter of droplets was significantly increased by MCP-1 (Fig. 3C). As hepatoma cells may differentially respond to chemokines compared to normal cells, we tested the effect of MCP-1 on mouse primary hepatocytes. Mouse MCP-1 induced lipid accumulation in normal primary hepatocytes in a dose dependent manner (Fig. 3D).
MCP-1 responses in cells are mediated by cytosolic Ca2+ signals.18 By using confocal video microscopy (Fig. 4), MCP-1 induced Ca2+ waves in most cells examined, suggesting that Huh-7 cells express a functional MCP-1 receptor whose engagement is linked to increases in intracellular free Ca2+ concentrations. To confirm these results, pretreatment with BAPTA/AM to chelate intracellular Ca2+ was performed. However, accumulation of lipids was observed in control cells treated only with BAPTA/AM, in keeping with previous reports showing that depleting the cells of calcium results in a sharp decline in very low density lipoprotein secretion, leading to lipid accumulation.19
Expression of the MCP-1 receptor CCR2, described in monocytes and other cells,20 was determined in Huh-7 cells and mouse primary hepatocytes by real-time PCR (Fig. 5A). Although Huh-7 cells expressed CCR2 messenger RNA (mRNA), this was markedly lower than in peripheral blood mononuclear cells used as control. In mouse primary hepatocytes, CCR2 mRNA was below the level of detection. Moreover, blockade of CCR2 with the antagonist RS102895 did not have any impact on lipid accumulation in Huh-7 (Fig. 5B). These results seem to demonstrate that MCP-1-induced fat accumulation in Huh-7 and mouse primary hepatocytes is probably not mediated by CCR2.
Inhibition of ERK, But Not p38-MAPK, Abrogates MCP-1-Induced Lipid Accumulation.
Recently, the involvement of ERK in lipid secretion has been reported in HepG2 cells.21 In addition, MCP-1 has been shown to activate MAPK pathways.22, 23 To determine whether MCP-1-induced lipid accumulation was dependent on MAPK pathway activation, even in the absence of large amounts of CCR2, Huh-7 cells were treated with MCP-1 and PD98059, an ERK inhibitor, or SB203580, a p38 MAPK inhibitor (Fig. 6A). PD98059 inhibited the increase of lipids, whereas SB203580 did not, indicating that MCP-1-mediated lipid accumulation involves activation of ERK. The ability of MCP-1 to increase ERK activation was further shown by immunofluorescence (32.7 ± 3.5% of cells displayed nuclear localization of phospho-ERK compared to 14.9 ± 2.4% in control; P < 0.001) (Fig. 6B).
MCP-1 Increases PEPCK mRNA Level and Decreases ApoB Secretion.
Steatosis can be explained by increased neosynthesis, increased uptake, impaired oxidation, or decreased secretion of lipids. We therefore determined which of these processes was affected by MCP-1. We first evaluated the expression of key actors involved in the above processes by real-time PCR (Fig. 7A). The level of PEPCK mRNA, an enzyme involved in gluconeogenesis but also controlling fatty acid esterification,24 was increased by more than three-fold by MCP-1 (Fig. 7A). As PEPCK can be regulated by CREB,25 we analyzed the ability of MCP-1 to induce CREB nuclear activation. MCP-1 stimulated CREB phosphorylation (20.9±4.2% of cells displayed nuclear localization of phospho-CREB compared to 2.2±0.77% in control) (Fig. 7B). PEPCK mRNA level was also measured in presence of KG501, a disruptor of the interaction between phospho-CREB and its cofactor CREB binding protein/p300. As expected, treatment with KG501 decreased the level of PEPCK mRNA in control conditions. Interestingly, the interference with the CREB pathway was associated with reduction of the PEPCK upregulation induced by MCP-1 (approximately four-fold; not shown).
Results indicate that oleic acid uptake was not significantly modified at any time point by MCP-1 (Fig. 8A). Mitochondrial β-oxidation of free fatty acids was also not affected by MCP-1, as compared to control cells (not shown). However, the amount of ApoB in cultured medium was decreased by approximately 20% after MCP-1 treatment for 48 hours (Fig. 8B; black columns). This was due neither to an increased cell death nor to a decreased cell proliferation (not shown), suggesting that MCP-1 inhibited ApoB secretion. Cell treatment with MCP-1 together with PD98059 resulted in a recovery of ApoB secretion, reaching a level comparable to control conditions (Fig. 8B; gray columns).
MCP-1 is a member of the CC chemokine family and was first identified as a monocyte chemoattractant released by blood cells.26, 27 MCP-1 has also been implicated in several chronic inflammatory diseases20, 28, 29 and is increased in human obesity.9, 30 Recently, it has been shown that the expression of the MCP-1 transgene in adipose tissue was sufficient to induce macrophage infiltration into adipose tissue, insulin resistance, and increased hepatic triglyceride levels in mice, indicating that MCP-1 contributes to the development of insulin resistance and hepatic steatosis associated with obesity.31 In that work, liver steatosis was attributed to an infiltration of macrophages in adipose tissue induced by increased MCP-1 expression. Here we show that MCP-1 can directly induce lipid accumulation in hepatoma cells and cultured primary hepatocytes. This provides an additional mechanism of induction of liver steatosis, that is, MCP-1 secreted by adipose tissue not only leads to macrophage recruitment, but has also direct effects on hepatocytes.
MCP-1 binds the chemokine receptor CCR2 with high affinity.32 However, Huh-7 do not express significant levels of CCR2 mRNA, in agreement with previous studies showing that the effects of MCP-1 in other cell types may proceed independently of CCR218, 22, 33 and indicating the existence of an alternative, hitherto unknown receptor.
We also provided some insights on the intracellular signaling mediated by MCP-1 in Huh-7 cells. Here, MCP-1-induced fat accumulation depends on the ERK signaling pathway. ERK has already been shown to be induced by MCP-1 in Chinese hamster ovary cells and in smooth muscle cells22, and to be involved in MCP-1-mediated monocyte chemotaxis.34 Interestingly, Tsai et al.21 have recently demonstrated that inhibition of ERK pathway in HepG2 cells led to an increase in ApoB secretion. Given that, in our model, ApoB secretion is diminished upon MCP-1 treatment in Huh-7 cells and that this phenomenon can be rescued by an ERK inhibitor, we hypothesize that lipid accumulation is partly due to a blockade of lipid secretion via the ERK pathway. However, MCP-1 treatment did not decrease ApoB mRNA levels, suggesting that other factors such as translational regulation and/or alterations in microsomal triglyceride transfer protein activity may be involved.
The increased lipid accumulation in Huh-7 may also be partly explained by the changes in PEPCK mRNA levels. This enzyme is not only involved in gluconeogenesis, but it plays an important role also in lipid metabolism via glyceroneogenesis (for review see Reshef et al.35), that is, the de novo synthesis of glycerol-3-phosphate from pyruvate. A previous report stressed the role of PEPCK in the control of fatty acid re-esterification in adipose tissue and, thus, the contribution of glyceroneogenesis to fat accumulation.24 A deregulation of this pathway may have profound pathophysiological effects.36 In the liver, there is strong evidence that glyceroneogenesis occurs to a significant degree.37, 38 PEPCK expression is regulated almost entirely at the level of gene transcription by several hormones and extracellular signals.25, 39 Changes in PEPCK gene transcription are controlled by many tissue-specific factors, including CCAAT/enhancer binding protein α (C/EBPa), C/EBPβ40, 41, and hepatic nuclear factors 1, 3, and 4,42 as well as multiple hormone-responsive factors such as CREB,43, 44 the glucocorticoid retinoic acid, retinoid X, and thyroid hormone receptors, and a poorly defined insulin-responsive system (for review see Chakravarty et al.45). Our result that MCP-1 activates CREB in Huh-7 cells, similar to that reported for HEK-cells,46 allows us to propose a new mechanism of MCP-1-induced fat accumulation in hepatocytes via glyceroneogenesis.
In conclusion, we show that MCP-1 induces lipid accumulation in hepatocytes in vitro via multiple mechanisms including: (1) increased expression of PEPCK, involved in fatty acid esterification; and (2) decreased lipids secretion. These data suggest a direct role of MCP-1 in the pathogenesis of obesity-associated liver steatosis.
We thank Drs. Françoise Assimacopoulos, Manlio Vinciguerra, and Michelangelo Foti for their valuable advice throughout the project; Dr. Serge Arnaudeau (Bioimaging Core Facility, Faculty of Medicine, Geneva) for his help and advice in using the Nipkow system; Dr. Badwi Elias (Plastic and Reconstructive Surgery, Geneva University Hospital) and Dr. Pierre Quinodoz (Plastic Reconstructive Surgery Unit, La Tour Hospital, CH-1217 Meyrin, Switzerland) for providing human WAT samples from patients undergoing abdominal plastic surgery; Dr. Francesca Frigerio for help and advice in the β-oxidation experiments; Agnès Pernin-Chollet, Lionel Nancoz, and Kévin Guilloux for technical help; and the Genomics Platform of the NCCR program “Frontiers in Genetics” (Geneva) for performing and analyzing the real-time PCR experiments.