Adipocytes as a Link Between Gut Microbiota-Derived Flagellin and Hepatocyte Fat Accumulation

This is an electronic reprint of the original article. This reprint may differ from the original in pagination and typographic detail. Author(s): Title: Year: Version: Please cite the original version: All material supplied via JYX is protected by copyright and other intellectual property rights, and duplication or sale of all or part of any of the repository collections is not permitted, except that material may be duplicated by you for your research use or educational purposes in electronic or print form. You must obtain permission for any other use. Electronic or print copies may not be offered, whether for sale or otherwise to anyone who is not an authorised user. Adipocytes as a Link Between Gut Microbiota-Derived Flagellin and Hepatocyte Fat Accumulation Munukka, Eveliina; Wiklund, Petri; Partanen, Tiina; Välimäki, Sakari; Laakkonen, Eija; Lehti, Maarit; Fischer-Posovzsky, Pamela; Wabitsch, Martin; Cheng, Sulin; Huovinen, Pentti; Pekkala, Satu Munukka, E., Wiklund, P., Partanen, T., Välimäki, S., Laakkonen, E., Lehti, M., Fischer- Posovzsky, P., Wabitsch, M., Cheng, S., Huovinen, P., & Pekkala, S. (2016). Adipocytes as a Link Between Gut Microbiota-Derived Flagellin and Hepatocyte Fat Accumulation. PLoS ONE, 11(4), Article e0152786. https://doi.org/10.1371/journal.pone.0152786 2016 RESEARCH ARTICLE Adipocytes as a Link Between Gut Microbiota-Derived Flagellin and Hepatocyte Fat Accumulation Eveliina Munukka 1,2 , Petri Wiklund 1 , Tiina Partanen 1 , Sakari Välimäki 1 , Eija K. Laakkonen 1 , Maarit Lehti 3 , Pamela Fischer-Posovzsky 4 , Martin Wabitsch 4 , Sulin Cheng 1,5 , Pentti Huovinen 2 , Satu Pekkala 1,2 * 1Department of Health Sciences, University of Jyväskylä, Jyväskylä, Finland, 2Department of Clinical Microbiology and Immunology, University of Turku, Turku, Finland, 3LIKES Research Center for Sport and Health Sciences, Jyväskylä, Finland, 4Division of Pediatric Endocrinology and Diabetes, University Medical Center Ulm, Ulm, Germany, 5Department of Physical Education, Shanghai Jiao Tong University, Shanghai, China *[email protected] Abstract While the role of both elevated levels of circulating bacterial cell wall components and adipose tissue in hepatic fat accumulation has been recognized, it has not been considered that the bacterial components-recognizing adipose tissue receptors contribute to the hepatic fat content. In this study we found that the expression of adipose tissue bacterial flagellin (FLG)-rec- ognizing Toll-like receptor (TLR) 5associated with liver fat content (r = 0.699, p = 0.003) and insulin sensitivity (r = -0.529, p = 0.016) in humans (n = 23). No such associations were found for lipopolysaccharides (LPS)-recognizing TLR4. To study the underlying molecular mechanisms of these associations, human HepG2 hepatoma cells were exposed in vitro to the conditioned culture media derived from FLG or LPS-challenged human adipocytes. The adipocyte-mediated effects were also compared to the effects of direct HepG2 exposure to FLG and LPS. We found that the media derived from FLG-treated adipocytes stimulated fat accumulation in HepG2 cells, whereas either media derived from LPS-treated adipocytes or direct FLG or LPS exposure did not. This is likely due to that FLG-treatment of adipocytes increased lipolysis and secretion of glycerol, which is known to serve a substrate for triglyceride synthesis in hepatocytes. Similarly, only FLG-media significantly decreased insulin signaling-related Akt phosphorylation, IRS1 expression and mitochondrial respiratory chain ATP5A. In conclusion, our results suggest that the FLG-induced TLR5 activation in adipocytes increases glycerol secretion from adipocytes and decreases insulin signaling and mitochondrial functions, and increases fat accumulation in hepatocytes. These mechanisms could, at least partly, explain the adipose tissue TLR5 expression associated with liver fat content in humans. PLOS ONE | DOI:10.1371/journal.pone.0152786 April 1, 2016 1/16 OPEN ACCESS Citation: Munukka E, Wiklund P, Partanen T, Välimäki S, Laakkonen EK, Lehti M, et al. (2016) Adipocytes as a Link Between Gut Microbiota- Derived Flagellin and Hepatocyte Fat Accumulation. PLoS ONE 11(4): e0152786. doi:10.1371/journal. pone.0152786 Editor: Pratibha V. Nerurkar, College of Tropical Agriculture and Human Resources, University of Hawaii, UNITED STATES Received: November 19, 2015 Accepted: March 18, 2016 Published: April 1, 2016 Copyright: © 2016 Munukka et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability Statement: All relevant data are within the paper. The microarray data used in this study has been deposited in ArrayExpress. The accession number for the data is E-MTAB-2649. Funding: This study was financially supported by The Academy of Finland postdoctoral research fellow for Dr. Satu Pekkala (Grant ID 267719), by The Finnish Diabetes Research Foundation and by the Finnish Culture Foundation (Central and Southwest Finland funds). The study of the human project was Introduction The prevalence of non-alcoholic fatty liver disease (NAFLD) is increasing, currently affecting 14–24% of the general population and up to 80% of morbidly obese individuals [1]. The pathophysiology of NAFLD is complex and has not been fully elucidated. However, growing evidence indicates that inflammation and insulin resistance are important factors in its causation. Previous studies suggest that the inflammatory processes in NAFLD are caused by mitochondrial dysfunction, oxidative stress and subsequent lipid peroxidation [2]. While the metabolic consequences of NAFLD are well documented, the exact underlying mechanisms are not well understood. Recently, the bacterial surface molecules originating from gut microbiota in hepatic fat accumulation has been acknowledged [3,4]. Animal studies have shown that altered gut microbiota composition can cause intestinal inflammation, leading to leakage of gut bacterial components or translocation of entire commensal bacteria to the circulation and peripheral tissues of the host [5–7]. Our recent findings in humans suggest that adipose tissue inflammation could be the key link between the gut microbiota-derived bacterial molecules and NAFLD [3]. Specifically, subjects with high hepatic fat content overexpressed inflammatory Vav1,Lat and MMP-9 in their subcutaneous adipose tissue, which, in addition to function in leukocyte transendothelial migration, are components of the Toll-like receptor (TLR) signaling pathway. TLR family members are key mediators of the innate immune system. They recognize bacterial surface molecules, such as lipopolysaccharides (LPS) from the outer wall of Gram-negative bacteria and flagellin (FLG), which is a structural protein of flagellum, a bacterial locomotive organelle. TLR activation leads to the pleiotropic activation of various inflammatory and metabolic genes [8]. Previous studies have shown that three TLR family members (TLR2, 4, and 9) play a role in the development of NAFLD [9]. For the FLG-recognizing TLR5, contradictory roles have been reported [10–12]. Of the pattern recognition receptors of the innate immune system that sense the exogenic pathogens also Nod-like receptors (NLRs) have been recognized as links between immune function and metabolism [13]. By far, pyrin domain containing 3 (NLRP3) inflammasome is the most studied inflammasome of the NLR family. NRLP3 expression has been found to be increased in the adipose tissue of subjects with metabolic syndrome and insulin resistance [14], and more recently its role in regulating NAFLD progression has been recognized [10]. Our recent results showed that in adipocytes bacterial FLG and LPS increased secretion of glycerol [15], which serves as a substrate for hepatic triglyceride synthesis. In insulin resistant adipose tissue glycerol is produced by increased lipolysis. In this study, therefore, we hypothesized that adipose tissue expression levels of TLR4 and TLR5, and possibly other TLRs and NOD-like receptors would associate with liver fat content and insulin resistance in humans. In order to verify the associations and establish the adipocytes as a link between gut-derived molecules and hepatocytes, cell cultures were selected as a model instead of animals to eliminate any confounding factors. Materials and Methods Human Subjects Participants were recruited from a larger study (the AMB-study The Finnish Academy SKIDKID program), conducted at the University of Jyväskylä in accordance with the Helsinki Declaration and approved by the ethical committee of the Central Finland Health Care district. A written informed consent was obtained from all study participants. To exclude the confounding effect of sex, only women were included in this study. All together 23 women (age 18–57 years old) were included. The health history and the current status were checked by a study Bacterial Flagellin in Hepatocyte Fat Accumulation PLOS ONE | DOI:10.1371/journal.pone.0152786 April 1, 2016 2/16 supported by Academy of Finland SKIDI-KID Project (PI: Sulin Cheng, Grant ID 135038). Competing Interests: The authors have declared that no competing interests exist. Abbreviations: FLG, flagellin; NAFLD, non-alcoholic fatty liver disease; LPS, lipolysaccharides; TLR, Tolllike receptor; DGAT, Diglyceride acyltransferase; ROS, reactive oxygen species; NF-κB, Nuclear factor kappa B; MMP-9, Matrix metalloprotease 9; IRS1, Insulin receptor substrate 1; MRC, mitochondrial respiratory chain; AS160, TBC1 domain family member 4, Akt substrate of 160 kDa; ATP5A, ATP synthase, H+ transporting, mitochondrial F1 complex, alpha subunit; UQCRC2, Cytochrome b-c1 complex subunit 2; MTCO1, Mitochondrially encoded cytochrome c oxidase I; SDHB, Succinate dehydrogenase complex, subunit B, iron sulfur (Ip); NDUFS8, NADH dehydrogenase (ubiquinone) Fe-S protein 8. physician. There was no type 1 or 2 diabetes, cardiac diseases, autoimmune diseases or major liver (cancer, hepatitis) diseases. All subjects were clinically euthyreotic. The subjects underwent a 75-g oral glucose tolerance test (OGTT). Whole-body insulin sensitivity was calculated from glucose and insulin values during the OGTT as proposed by Matsuda and DeFronzo [16]. Glucose was measured using KONELAB 20XTi analyzer (Thermo Fischer Scientific inc.) and serum insulin was determined by immunoassay using the IMMULITE Analyser (Diagnostic Products Corporation, Los Angeles). The subcutaneous adipose tissue biopsies were taken under sterile conditions between 7 and 9 am after overnight fasting. Biopsies were frozen in liquid nitrogen and stored at -80°C until used. Total RNA was isolated using FastPrep systems (MP Biomedicals, France) and RNeasy Lipid Tissue Mini Kit (Qiagen, Valencia, CA, USA), amplified and processed using the Gene- Chip 3´IVT Express Kit (Affymetrix, Santa Clara, CA, USA) and hybridized on Affymetrix Human Genome U219 Array Plates as described previously [3]. From the microarray data adipose tissue Toll-like receptors, NOD-like receptors and several lipolysis-associated gene expression levels were determined. The array data of this study have been submitted to ArrayExpress (E-MTAB-2649). The assessment of hepatic fat content was performed by 1 H MRS (1.5T GE Signa CV/i scanner, GE Medical Systems, Waukesha, WI, USA) as described in Munukka et al. 2014. Cells Lines, Cultures and Reagents SGBS preadipocytes [17] were maintained in DMEM/F12 supplemented with 10% FBS (Invitrogen), 0.17 μM panthotenat plus 0.33 μM biotin (P/B, Sigma Aldrich) and penicillin/streptomycin solution (Invitrogen). Preadipocytes were differentiated into mature adipocytes by incubating them first for 4 days in DMEM/F12 supplemented with P/B, 0.01 mg/ml transferrin, 20 nM insulin, 100 nM cortisol, 0.2 nM triiodothyronine, 25 nM dexamethasone, 250 μM IBMX and 2 μM rosiglitazone (QUICK DIF MEDIUM) (all from Sigma-Aldrich), and afterwards in DMEM/F12 supplemented with P/B, 0.01 mg/ml transferrin, 20 nM insulin, 100 nM cortisol and 0.2 nM triiodothyronine (3FC MEDIUM) for 10 days. When SGBS adipocytes were exposed to FLG for RNA extraction, the 14-days differentiated adipocytes in 3FC MEDIUM were treated for 1 and 4 hours with 10 ng/ml of FLG. When HepG2 cells were exposed to the conditioned SGBS culture media, 14-days differentiated adipocytes in 3FC MEDIUM were treated for 24 hours with 100 ng/ml of LPS or 10 ng/ml of FLG. These concentrations were determined to efficiently affect adipocytes in our former study 13 . Afterwards the culture media was collected and centrifuged for 5 min at 300xg. The freshly collected conditioned culture media was then applied to HepG2 cells that were previously washed twice with PBS. HepG2 hepatoma cells (from ATCC) were maintained in DMEM/Glutamax media supplemented with 10% FBS, 100 U/mL penicillin, 100 μg/mL streptomycin, and 1% sodium pyruvate (all from Invitrogen, Carlsbad, CA, USA). When HepG2 cells were exposed to recombinant Salmonella typhimurium FLG (Invivogen, San Diego, CA, USA) and Escherichia coli LPS (Sigma-Aldrich, St Louis, USA), FBS was omitted from the media and the cells were previously washed twice with PBS (Invitrogen). To mimic the conditions used for adipocytes, for the direct HepG2 exposures LPS was used at a concentration of 100 ng/ml and FLG at 10 ng/ml. Immunofluorescence SGBS adipocytes were grown and allowed to differentiate for 14-days on cover slips. To study the effects of FLG treatment on lipid droplets, FLG at 10ng/ml was added to the culture media for 3 hours. Afterwards the cells were washed twice with PBS and fixed for 15 min with 4% Bacterial Flagellin in Hepatocyte Fat Accumulation PLOS ONE | DOI:10.1371/journal.pone.0152786 April 1, 2016 3/16 PFA-PBS, permeabilized with 0.5% Triton-X for 5 min, blocked 1 hour with 5% donkey serum and thereafter incubated with primary antibody over night at 4°C. Immunolabeling was performed using rabbit polyclonal antibody against TLR5 (Pierce, Appleton, WI, USA, 1:50 in 1% donkey serum) and mouse monoclonal antibody against perilipin (Progen, Heidelberg, Germany, 1:200 in 1% donkey serum). As secondary antibodies donkey anti-mouse Alexa Fluor 555 and donkey anti-rabbit 488 (Invitrogen) were used. The labeled cells were imaged using an inverted wide-field microscope (Carl Zeiss) with a confocal unit and 40× oil/1.4 NA objective (Carl Zeiss). To count the percentage of cells in which lipid droplets were degraded, 200 control and FLG-treated cells were observed in randomly selected fields using 63× oil/1.4 NA objective. RNA Extraction and Real-Time Quantitative PCR HepG2 and SGBS cells were homogenized in Trizol reagent (Invitrogen) and the total RNA was extracted according to the supplier’s protocol. Total RNA was reversely transcribed according to the manufacturer’s instructions using High Capacity cDNA Synthesis Kit (Applied Biosystems, Foster City, CA, USA). Real-time PCR analysis was performed according to MIQE guidelines using in-house designed primers (from Invitrogen), iQ SYBR Supermix and CFX96™Real-time PCR Detection System (Bio-Rad Laboratories, Richmond, CA, USA). The primer sequences were as follows: IRS1: Fwd5’TATGCCAGCATCAGTTTCCA’3 and Rev5’GGATTTGCTGAGGTCATTTAGG’3; DGAT2: Fwd5’GAGGGGTCTGGGAGATGG’3 and Rev 5’CCTGTAGCTGCTTTTCCACC’3; β-actin: Fwd5’AGAGCTAGCTGCCTGAC’3 and Rev5’GGATGCCACAGGACTCCA’3; MMP9: Fwd5’GAGTGGCAGGGGGAAGATGC’3 and Rev 5’CCTCAGGGCACTGCAGGATG’3; NFκB(p65): Fwd5’ATGGCTTCTATGAGG CTGAG’3 and Rev5’CACAGCATTCAGGTCGTAGT’3; GAPDH: Fwd5’CCACCCATGG CAAATTCC’3 and Rev5’TGGGATTTCCATTGATGACAA’3; ATGL: Fwd5’CAACGCCACG CACATCTA’3 and Rev5’ACCTCAATGAACTTGGCACC’3; MGL: Fwd5’TATGAAGAG CTGGCTCGGAT’3 and Rev5’TACCATCCTCTCCCCTTCG’3; FABP3: Fwd5’GCACCTG GAAGCTAGTGGAC’3 and Rev5’GTGGTAGGCTTGGTCATGCT’3, and PLIN1: Fwd5’A CCTTGCTGGATGGAGACC’3 and Rev5’AGGTCTTCTGGAAGCATTCG’3. All results from HepG2 cells were normalized to the expression of β-actin and from the SGBS cells to the expression of GAPDH. The relative expression levels for each gene were calculated with the ΔΔC t method (by setting a fold change value of 1 to the average Ct value of the control samples). The amplification efficiency for each gene was 100±2%. Protein Extraction and Western Blot Proteins from HepG2 cells were extracted at 4°C using an ice-cold lysis buffer (10 mM Tris- HCl, 150 mM NaCl 2 , 2 mM EDTA, 1% Triton X-100, 10% glycerol and 1 mM DTT), supplemented with protease and phosphatase inhibitors (Sigma Aldrich). 20–30 micrograms of protein extracts were separated by SDS-Page using 4–20% Criterion gradient gels (Bio-Rad Laboratories, Richmond, CA) and transferred to nitrocellulose membranes. After blocking, the membranes were probed overnight at 4°C with primary antibodies purchased from Cell Signaling Technology (Danvers, MA, USA) (p-Akt), Sigma-Aldrich (anti-GAPDH) and Abcam (MitoProfile1Total OXPHOS Rodent WB Antibody Cocktail, Abcam, Cambridge, MA, USA). Odyssey anti-rabbit IRDye 800CW and anti-mouse IRDye 680RD (LI-COR Biosciences, Lincoln, NE, USA) were used as secondary antibodies. Finally, the blots were scanned and quantified by using Odyssey CLX Infrared Imager of Li-COR and manufacturer's software. If re-probing was needed, the membranes were incubated for 10 min in 0.2 M NaOH at RT, Bacterial Flagellin in Hepatocyte Fat Accumulation PLOS ONE | DOI:10.1371/journal.pone.0152786 April 1, 2016 4/16 washed with TBS and re-probed with appropriate antibodies. All samples and results were normalized to GAPDH. Oil Red O Staining for Neutral Lipid Measurement in HepG2 Cells Oil Red O stain (Sigma-Aldrich) was used to stain the neutral triglycerides and lipids in HepG2 cells. Briefly, the HepG2 cells on 12-well culture plates were exposed for 24 hours either directly to FLG and LPS, or to the culture media derived from treated adipocytes. Afterwards, the cells were fixed with 10% formalin, and then washed with H 2 O and 60% isopropanol. After washing, the cells were incubated with Oil Red O staining solution at RT for 10 min. Before eluting the staining solution with 100% isopropanol and measuring the absorbance at 550 nm, the cells were washed several times with H 2 O. ROS Measurement from HepG2 Cells by Luminescence For ROS measurement the HepG2 cells were treated with FLG or LPS or with the culture media from treated adipocytes for 24 hours. ROS were detected using non-lytic ROS-Glo™ H2O2 Assay and Glomax Multidetection System (Promega, Madison, WI, USA) according to the manufacturer’s instructions. Glycerol Measurement from the Adipocyte Cell Culture Media SBGS adipocytes were allowed to differentiate for 14 days. For glycerol measurement the cells were treated with FLG for 3 hours. Afterwards the cell culture media was collected and centrifuged for 5 min at 300xg. Glycerol was measured using the KONELAB 20XTi analyzer (Thermo Fischer Scientific inc.). Statistical Analysis All data were checked for normality using the Shapiro-Wilk´s test in PASW 18.0 for Windows. HepG2 cell culture data were analyzed using ANOVA. Bonferroni post hoc tests were performed to localize the effects for the results. Data obtained from FLG-treated SGBS cells were analyzed using Kruskal-Wallis. Spearman correlation analyses were performed to determine the relationship between adipose tissue gene expression and clinical characteristics. The level of significance was set at p<0.05. Results Correlation of Adipose Tissue TLR4,TLR5 as well as Other TLRs and NOD-Like Receptors mRNA with Liver Fat Content and Insulin Sensitivity in Humans The mean age of the study subjects was 34 years ± 15 years. All subjects were normoglycemic and three subjects had non-alcoholic fatty liver based on the clinical diagnostic cut-off criterion (liver fat content >5.6%). Adipose tissue TLR5 expression correlated closely with liver fat content (r = 0.699, p = 0.003) and insulin sensitivity index (r = -0.529, p = 0.016, Fig 1A). TLR4 expression was not associated with liver fat content or insulin sensitivity (Fig 1B). Of the other TLRs and NOD-like receptors TLR1 (r = -0.71, p<0.001), TLR2 (r = -0.599, p = 0.005), TLR7 (r = -0.687, p = 0.001), TLR8 (r = -0.641, p = 0.002), NLRC4 (r = -0.505, p = 0.023) and NLRP3 (r = -0.628, p = 0.003) associated negatively with insulin sensitivity index. In addition to TLR5 only NLRP3 correlated positively with liver fat content (r = 0.545, p = 0.029). Bacterial Flagellin in Hepatocyte Fat Accumulation PLOS ONE | DOI:10.1371/journal.pone.0152786 April 1, 2016 5/16 The Effects of FLG and LPS on Lipid Metabolism and Lipid Content of the HepG2 Cells We further determined the underlying mechanisms linking FLG to hepatic fat accumulation and explaining the lack of association between TLR4 and hepatic fat content. For the purpose, HepG2 cells were either exposed directly to FLG and LPS or to the conditioned culture media derived from FLG- and LPS-challenged adipocytes. The direct exposure of hepatoma cells to LPS increased triglyceride-synthesizing Diglyceride acyltransferase 2, DGAT2 expression after Fig 1. Correlation of liver fat content and Matsuda index (= insulin sensitivity) with (A) adipose tissue TLR5 mRNA and (B) adipose tissue TLR4 mRNA in humans (n = 23). doi:10.1371/journal.pone.0152786.g001 Bacterial Flagellin in Hepatocyte Fat Accumulation PLOS ONE | DOI:10.1371/journal.pone.0152786 April 1, 2016 6/16 24 hours, and to FLG after 4 and 24 hours (p<0.05). However, the direct FLG or LPS exposure did not increase HepG2 intracellular lipid content (Fig 2A). Exposure of HepG2 cells to the conditioned culture media from FLG and LPS-treated adipocytes increased DGAT2 expression at 1 hour, and in addition FLG-treated conditioned media increased DGAT2 expression at 4 hours (p<0.05 for all). The lipid content of the hepatoma cells was moderately increased only by the exposure to the media from FLG-treated adipocytes (p<0.05, Fig 2B). FLG Exposure Causes Glycerol Secretion from Adipocyte Lipid Droplets by Increasing Expression of Lipolytic Genes The exposure of adipocytes to FLG increased glycerol secretion into the culture media (p<0.05, Fig 3A). We further determined the effects of FLG treatment on lipolysis-related gene Fig 2. The (A) direct and (B) adipocyte-mediated effects of flagellin and LPS on the lipid metabolism of HepG2 cells. The figure presents the direct and adipocyte-mediated effects of FLG and LPS on the DGAT2 mRNA fold changes. The values are means + SD (n = 5). The results were normalized to βactin and the fold changes were calculated with the ΔΔC t method. The bars on the right side present graphically the total lipid content of HepG2 cells measured spectrophotometrically using Oil Red O after (A) direct 24 hours exposure to FLG and LPS, and (B) 24 hours exposure to conditioned media derived from FLG and LPS-treated adipocytes. The values are means + SD (n = 12). *indicates a p-value of <0.05. doi:10.1371/journal.pone.0152786.g002 Bacterial Flagellin in Hepatocyte Fat Accumulation PLOS ONE | DOI:10.1371/journal.pone.0152786 April 1, 2016 7/16 expression to reveal the mechanisms behind glycerol release. We found that after 1 hour exposure to FLG the adipocytes expressed more adipose triglyceride lipase (ATGL) and fatty acid binding protein 3 (FABP3), and after 4 hours more monoglyceride lipase (MGL) and less perilipin-1 (PLIN1) (p<0.001 for all, Fig 3B). Accordingly, in human adipose tissue TLR5 correlated positively with FABP3 (R = 0.549, p = 0.007) and negatively with PLIN1 (R = -0.631, p = 0.002). To visualize whether the secreted glycerol was derived from adipocyte lipid droplets, FLG- treated adipocytes were labeled with antibody against lipid droplet membrane protein, perilipin. After 3 hours FLG induced lipid droplet degradation, observed as disintegration of lipid Fig 3. Glycerol is secreted from adipocyte lipid droplets due to increased lipolysis in response to flagellin treatment. (A) Adipocytes were treated with FLG for 3 hours and glycerol was measured from the cell culture media. The values are means + SD (n = 4). *indicates a p-value of <0.05. (B) Adipocytes were treated with FLG for 1 and 4 hours and changes in the expression of ATGL,PLIN1,FABP3 and MGL mRNA were determined. The values are means + SD (n = 4). ** indicates a p-value of <0.001. The results were normalized to GAPDH and the fold changes were calculated with the ΔΔC t method. (C) Adipocytes with and without 3-hours FLG treatment were labeled with TLR5 and perilipin antibodies, and imaged with confocal microscope. Compared to 30% in control cells, in 70% of FLG-treated adipocytes degradation of lipid droplets membranes was observed (perilipin-labeling in red).In mature adipocytes TLR5 is mainly located in intracellular pool near lipid droplets and nuclei without co-localization, and only a minority on cell membranes. doi:10.1371/journal.pone.0152786.g003 Bacterial Flagellin in Hepatocyte Fat Accumulation PLOS ONE | DOI:10.1371/journal.pone.0152786 April 1, 2016 8/16 11. Kanuri G, Ladurner R, Skibovskaya J, Spruss A, Konigsrainer A, Bischoff SC, et al. (2014) Expression of toll-like receptors 1–5 but not TLR 6–10 is elevated in livers of patients with non-alcoholic fatty liver disease. Liver Int. 12. Miura K, Ohnishi H (2014) Role of gut microbiota and Toll-like receptors in nonalcoholic fatty liver disease. World J Gastroenterol 20: 7381–7391. doi: 10.3748/wjg.v20.i23.7381 PMID: 24966608 13. Vandanmagsar B, Youm YH, Ravussin A, Galgani JE, Stadler K, Mynatt RL, et al. (2011) The NLRP3 inflammasome instigates obesity-induced inflammation and insulin resistance. Nat Med 17: 179–188. doi: 10.1038/nm.2279 PMID: 21217695 14. Esser N, L'Homme L, De Roover A, Kohnen L, Scheen AJ, Moutschen M, et al. (2013) Obesity phenotype is related to NLRP3 inflammasome activity and immunological profile of visceral adipose tissue. Diabetologia 56: 2487–2497. doi: 10.1007/s00125-013-3023-9 PMID: 24013717 15. Pekkala S, Munukka E, Kong L, Pollanen E, Autio R, Roos C, et al. (2015) Toll-like receptor 5 in obesity: The role of gut microbiota and adipose tissue inflammation. Obesity (Silver Spring) 23: 581–590. 16. Matsuda M, DeFronzo RA (1999) Insulin sensitivity indices obtained from oral glucose tolerance testing: comparison with the euglycemic insulin clamp. Diabetes Care 22: 1462–1470. PMID: 10480510 17. Fischer-Posovszky P, Newell FS, Wabitsch M, Tornqvist HE (2008) Human SGBS cells—a unique tool for studies of human fat cell biology. Obes Facts 1: 184–189. doi: 10.1159/000145784 PMID: 20054179 18. Tsukumo DM, Carvalho-Filho MA, Carvalheira JB, Prada PO, Hirabara SM, Schenka AA, et al. (2007) Loss-of-function mutation in Toll-like receptor 4 prevents diet-induced obesity and insulin resistance. Diabetes 56: 1986–1998. PMID: 17519423 19. Rivera CA, Adegboyega P, van Rooijen N, Tagalicud A, Allman M, Wallace M (2007) Toll-like receptor- 4 signaling and Kupffer cells play pivotal roles in the pathogenesis of non-alcoholic steatohepatitis. J Hepatol 47: 571–579. PMID: 17644211 20. Spruss A, Kanuri G, Wagnerberger S, Haub S, Bischoff SC, Bergheim I (2009) Toll-like receptor 4 is involved in the development of fructose-induced hepatic steatosis in mice. Hepatology 50: 1094–1104. doi: 10.1002/hep.23122 PMID: 19637282 21. Pinzone MR, Celesia BM, Di Rosa M, Cacopardo B, Nunnari G (2012) Microbial translocation in chronic liver diseases. Int J Microbiol 2012: 694629. doi: 10.1155/2012/694629 PMID: 22848224 22. Gaggini M, Morelli M, Buzzigoli E, DeFronzo RA, Bugianesi E, Gastaldelli A (2013) Non-alcoholic fatty liver disease (NAFLD) and its connection with insulin resistance, dyslipidemia, atherosclerosis and coronary heart disease. Nutrients 5: 1544–1560. doi: 10.3390/nu5051544 PMID: 23666091 23. Konner AC, Bruning JC (2011) Toll-like receptors: linking inflammation to metabolism. Trends Endocrinol Metab 22: 16–23. doi: 10.1016/j.tem.2010.08.007 PMID: 20888253 24. Zhao Y, Yang J, Shi J, Gong YN, Lu Q, Xu H, et al. (2011) The NLRC4 inflammasome receptors for bacterial flagellin and type III secretion apparatus. Nature 477: 596–600. doi: 10.1038/nature10510 PMID: 21918512 25. Leemans JC, Cassel SL, Sutterwala FS (2011) Sensing damage by the NLRP3 inflammasome. Immunol Rev 243: 152–162. doi: 10.1111/j.1600-065X.2011.01043.x PMID: 21884174 26. Fukata M, Vamadevan AS, Abreu MT (2009) Toll-like receptors (TLRs) and Nod-like receptors (NLRs) in inflammatory disorders. Semin Immunol 21: 242–253. doi: 10.1016/j.smim.2009.06.005 PMID: 19748439 27. Nielsen TS, Jessen N, Jorgensen JO, Moller N, Lund S (2014) Dissecting adipose tissue lipolysis: molecular regulation and implications for metabolic disease. J Mol Endocrinol 52: R199–222. doi: 10. 1530/JME-13-0277 PMID: 24577718 28. Boord JB, Fazio S, Linton MF (2002) Cytoplasmic fatty acid-binding proteins: emerging roles in metabolism and atherosclerosis. Curr Opin Lipidol 13: 141–147. PMID: 11891416 29. Franchini M, Monnais E, Seboek D, Radimerski T, Zini E, Kaufmann K, et al. (2010) Insulin resistance and increased lipolysis in bone marrow derived adipocytes stimulated with agonists of Toll-like receptors. Horm Metab Res 42: 703–709. doi: 10.1055/s-0030-1261872 PMID: 20603780 30. Ke B, Shen XD, Ji H, Kamo N, Gao F, Freitas MC, et al. (2011) HO-1-STAT3 axis in mouse liver ischemia/reperfusion injury: regulation of TLR4 innate responses through PI3K/PTEN signaling. J Hepatol 56: 359–366. doi: 10.1016/j.jhep.2011.05.023 PMID: 21756853 31. Perez-Carreras M, Del Hoyo P, Martin MA, Rubio JC, Martin A, Castellano G, et al. (2003) Defective hepatic mitochondrial respiratory chain in patients with nonalcoholic steatohepatitis. Hepatology 38: 999–1007. PMID: 14512887 32. Begriche K, Massart J, Robin MA, Bonnet F, Fromenty B (2013) Mitochondrial adaptations and dysfunctions in nonalcoholic fatty liver disease. Hepatology 58: 1497–1507. doi: 10.1002/hep.26226 PMID: 23299992 Bacterial Flagellin in Hepatocyte Fat Accumulation PLOS ONE | DOI:10.1371/journal.pone.0152786 April 1, 2016 15 / 16 33. Dasu MR, Ramirez S, Isseroff RR (2012) Toll-like receptors and diabetes: a therapeutic perspective. Clin Sci (Lond) 122: 203–214. 34. Brenner C, Galluzzi L, Kepp O, Kroemer G (2013) Decoding cell death signals in liver inflammation. J Hepatol 59: 583–594. doi: 10.1016/j.jhep.2013.03.033 PMID: 23567086 35. Svineng G, Ravuri C, Rikardsen O, Huseby NE, Winberg JO (2008) The role of reactive oxygen species in integrin and matrix metalloproteinase expression and function. Connect Tissue Res 49: 197–202. doi: 10.1080/03008200802143166 PMID: 18661342 36. Stockhammer OW, Zakrzewska A, Hegedus Z, Spaink HP, Meijer AH (2009) Transcriptome profiling and functional analyses of the zebrafish embryonic innate immune response to Salmonella infection. J Immunol 182: 5641–5653. doi: 10.4049/jimmunol.0900082 PMID: 19380811 37. Seki E, Brenner DA (2008) Toll-like receptors and adaptor molecules in liver disease: update. Hepatology 48: 322–335. doi: 10.1002/hep.22306 PMID: 18506843 38. Wobser H, Dorn C, Weiss TS, Amann T, Bollheimer C, Buttner R, et al. (2009) Lipid accumulation in hepatocytes induces fibrogenic activation of hepatic stellate cells. Cell Res 19: 996–1005. doi: 10. 1038/cr.2009.73 PMID: 19546889 Bacterial Flagellin in Hepatocyte Fat Accumulation PLOS ONE | DOI:10.1371/journal.pone.0152786 April 1, 2016 16 / 16