The Maintenance of Intestinal Homeostasis Expression In

Epigenetic Regulation of TLR4 Gene Expression in Intestinal Epithelial Cells for the Maintenance of Intestinal Homeostasis

This information is current as Kyoko Takahashi, Yutaka Sugi, Akira Hosono and Shuichi of October 4, 2021. Kaminogawa J Immunol 2009; 183:6522-6529; Prepublished online 21 October 2009; doi: 10.4049/jimmunol.0901271 http://www.jimmunol.org/content/183/10/6522 Downloaded from

References This article cites 34 articles, 20 of which you can access for free at: http://www.jimmunol.org/content/183/10/6522.full#ref-list-1 http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication by guest on October 4, 2021 *average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2009 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Epigenetic Regulation of TLR4 Gene Expression in Intestinal Epithelial Cells for the Maintenance of Intestinal Homeostasis1

Kyoko Takahashi,2 Yutaka Sugi, Akira Hosono, and Shuichi Kaminogawa

Intestinal epithelial cells (IECs) are continuously exposed to large numbers of commensal bacteria but are relatively insensitive to them, thereby averting an excessive inflammatory reaction. In this study, we show that the low responsiveness of human IEC lines to LPS was mainly brought about by a down-regulation of TLR4 gene transcription. Additionally, the presence of an IEC-specific repressor element in the 5؅ region of the TLR4 gene and binding of a NF to the element was shown. The transcription factor ZNF160, which was expressed more abundantly in a LPS-low responder IEC line than in a LPS-high responder IEC line, repressed TLR4 gene transcription. ZNF160 is known to interact with the scaffold protein KAP1 via its N terminus to recruit histone deacetylase. Histone deacetylation, as well as DNA methylation, at the 5؅ region of the TLR4 gene was significantly higher in LPS-low responder IEC lines than in a monocyte line or a LPS-high responder IEC line. It was demonstrated that TLR4 gene transcription was repressed by these epigenetic regulations, which were, at least in part, dependent on ZNF160. Down-regulaton Downloaded from of TLR4 gene expression by these mechanisms in IECs possibly contributes to the maintainance of homeostasis in the intestinal commensal system. The Journal of Immunology, 2009, 183: 6522–6529.

he intestinal immune system, which is the largest immune cell balances (14, 15). Third, production of IgA against the com- system in the body, plays an essential role in the mainte- mensals is involved in keeping the number of the commensals at

nance of health. In addition to food, pathogenic bacteria an appropriate level (16). Additionally, it is also important that http://www.jimmunol.org/ T 3 and viruses enter the intestinal tract through the oral cavity. On the intestinal epithelial cells (IECs), which form the boundary be- other hand, an enormous number of commensal bacteria inhabit tween the lumen and the internal milieu, acquire hyporesponsive- the intestinal tract. The intestinal immune system accurately rec- ness to the commensals, because the lumenal surface of the epi- ognizes these different organisms, discriminates between safe/ben- thelium is continually exposed to the commensals or their eficial and dangerous components, and attacks only those that are components. hazardous to the host. IECs not only physically separate luminal contents from the Lately, the crucial role of the symbiosis between the commensal internal milieu but also actively participate in the immune reac- bacteria and the intestinal immune system in maintaining good tions as a frontline defense in the mucosal immune system. IECs health has attracted considerable attention. For instance, the de- have been shown to receive stimulation from intestinal commensal by guest on October 4, 2021 velopment of GALTs such as Peyer’s patches and the induction of bacteria through TLRs, the pattern recognition receptors for mi- oral tolerance are known to be impaired or delayed in germ-free crobial components, to maintain their homeostasis (17). However, animals (1–5). Moreover, a correlation between the incidence of to prevent triggering excessive inflammatory reactions, they do not atopic eczema in children and the composition of their intestinal respond in a sensitive manner to the commensals. Specifically, microbiota has been reported (6–11). The following phenomena IECs contribute to the maintenance of intestinal homeostasis by observed specifically in the intestinal mucosa are thought to be partially tolerating the commensals and regulating mucosal inflam- involved in establishment and maintenance of the symbiosis, al- mation. Decreased expression of specific TLRs and accessory mol- though the precise underlying mechanisms have not yet been fully ecules such as MD2 in IECs has been reported as one of the mech- elucidated. First, maturation of a specific population of dendritic anisms controlling the hyporesponsiveness (18, 19). Additionally, cells expressing CD103, which preferentially induce T cells with the expression of Toll-interacting protein, a negative regulator of regulatory properties, is accelerated in the intestine (12, 13). Sec- intracellular signals from TLRs, is known to be increased in IECs ond, the commensal bacteria regulate the differentiation of effector (18). In contrast, excessive responses to the commensals and in- T cells and achieve appropriate Th1/Th2 and Th17/regulatory T creased expression of specific TLRs, including TLR4, are often observed in patients with inflammatory bowel disease (20). There- fore, expression of these molecules is thought to be regulated by Food and Physiological Functions Laboratory, College of Bioresource Sciences, Ni- hon University, Fujisawa-shi, Kanagawa, Japan cell type-specific mechanisms in IECs to maintain the intestinal symbiosis. Here we have analyzed the regulatory mechanisms of Received for publication April 22, 2009. Accepted for publication September 18, 2009. TLR4 gene expression in IECs to elucidate one of the mechanisms The costs of publication of this article were defrayed in part by the payment of page for maintaining the intestinal homeostasis. charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 This work was supported in part by a Grant-in Aid for Young Scientists from the Ministry of Education, Culture, Sports, Science, and Technology of Japan (to K.T.) 3 Abbreviations used in this paper: IEC, intestinal epithelial cell; DN, dominant neg- and a Grant-in Aid for Scientific Research from Japan Society for the Promotion of ative; qPCR, quantitative PCR; ChIP, chromatin immunoprecipitation; 5-aza-dC, Science (to S.K.). 5-aza-2Ј-deoxycytidine; TSA, trichostatin A; siRNA, small interfering RNA; KRAB, 2 Address correspondence and reprint requests to Dr. Kyoko Takahashi, Food and Kruppel-associated box; KAP, KRAB-associated protein; ICSBP, interferon consen- Physiological Functions Laboratory, College of Bioresource Sciences, Nihon Univer- sus sequence-binding protein. sity, 1866 Kameino, Fujisawa-shi, Kanagawa 252-8510, Japan. E-mail address: [email protected] Copyright © 2009 by The American Association of Immunologists, Inc. 0022-1767/09/$2.00 www.jimmunol.org/cgi/doi/10.4049/jimmunol.0901271 The Journal of Immunology 6523

Materials and Methods nt Ϫ675/ϩ188 region, 5Ј-gcggtacCTTCAGTATATTCCATTCTGAAA Ј Ϫ ϩ Ј Cell culture TC-3 ; for amplification of the nt 218/ 188 region, 5 -gcggtacCACC GTCATCCTAGAGAGTTACAAG-3Ј; for amplification of the nt The human epithelial colonic adenocarcinoma cell lines Caco-2 and HT-29 Ϫ6/ϩ188 region, 5Ј-gcggtacCAGAATGCTAAGGTTGCCGCTTTC-3Ј. and the human epithelial colonic carcinoma cell lines HCT 116 and T84, The nucleotides denoted by lowercase letters were inserted to introduce a as well as the human monocyte line THP-1, were purchased from DS KpnI site (underlined). The synthetic oligonucleotide 5Ј-gaagatctCTGG Pharma Biomedical. The human epithelial colonic adenocarcinoma cell CATCATCCTCACTGCTTCTG-3Ј was used as a common reverse primer line SW480 was provided by the Cell Resource Center for Biomedical (nucleotides denoted by small letters were inserted to introduce a BglII site Research Institute of Development, Aging, and Cancer Center at Tohoku (underlined)). University (Miyagi, Japan). Caco-2 cells were cultured in Eagle’s MEM The TLR4 siRNA expression plasmid was constructed as follows: a (Nissui) supplemented with nonessential amino acids (Invitrogen). HCT double-stranded DNA was generated by annealing the following 116 and HT-29 cells were cultured in McCoy’s 5a medium (MP Biomedi- synthetic oligonucleotides: top strand, 5Ј-GATCCGCTCACAATC cals). T84, SW480, and THP-1 cells were cultured in a 1/1 mixture of TTACCAATCTCTGTGAAGCCACGATGGGAATTGGTAAGATTGTG Ham’s F12 (Nissui) and DMEM (Nissui), a 1/1 mixture of L15 (MP Bio- AGCTTTTTTA-3Ј; bottom strand, 5Ј-AGTTAAAAAGCTACAATTTA medicals) and DMEM, and RPMI 1640 (Nissui), respectively. All media TCAATCTCCATCGTGGCTTCACAGAGATTGGTAAGATGTGACG- were supplemented with 10% (v/v) FBS (Biowest), 100 U/ml of penicillin 3Ј. The double-stranded DNA was inserted into pBAsi-hU6 Pur vector (Meiji), 100 ␮g/ml streptomycin (Meiji), and 5 ϫ 10Ϫ5 M 2-ME. Cells (Takara Bio) at the BamHI/HindIII site. After confirming the sequence, the

were cultured at 37°C in a humidified incubator with 5% CO2. resulting plasmid was named pBAsi-TLR4. The control plasmid, pBAsi- cont, carrying the insert with the same A, T, C, and G composition as seen Measurement of IL-8 production in pBAsi-TLR4 was similarly constructed using the following synthetic oligonucleotides: top strand, 5Ј-GACGACATATCTTCACACCATCTT Cells were stimulated for 18 h with 0.1–1000 ng/ml ultra-pure Escherichia GAAGCACAGATGGATGGTGTGAAGATATGTCTTTTT-3Ј; bottom coli K12 LPS (InvivoGen), which is guaranteed to only activate the TLR4 strand, 5Ј-AGCTTAAAAAGACATATCTTCACACCATCCCTCTGTGG pathway. Concentrations of secreted IL-8 in the culture supernatants were TTCACGATGGTGTGAAAGAAATGTCG-3Ј. Downloaded from measured with a human IL-8 ELISA kit (BioSource International) or Quan- To construct an expression plasmid for ZNF160, full-length ZNF160 tikine human CXCL/IL-8 (R&D Systems) according to the manufacturers’ cDNA was reverse-transcribed and amplified from total RNA prepared instructions. from Caco-2 cells. The following synthetic oligonucleotide primers were Ј Measurement of NF-␬B activation by a reporter gene assay used for PCR amplification: forward, 5 -GGATGGCCCTTACTCAGGTA CGGTTGACA-3Ј; reverse, 5Ј-TTTGTAAAAATTAACTGATGTGAAAG Cells were transfected with 1.5 ␮gofNF-␬B reporter plasmid carrying GAGTG-3Ј. The amplified product was cloned into pcDNA3.1 vector (In-

NF-␬B binding sites upstream of the luciferase gene using FuGene HD vitrogen) in the correct and reverse directions to yield pcDNA3.1-ZNF160 http://www.jimmunol.org/ transfection reagent (Roche). Alternatively, cells were cotransfected with sense and pcDNA3.1-ZNF160 antisense, respectively. The control empty 1.5 ␮g of NF-␬B reporter plasmid and 1.5 ␮g of an expression plasmid vector pCDNA3.1-empty was generated by self-ligation of BstXI-digested encoding the dominant-negative (DN) form of MyD88. The plasmid phRL- pcDNA3.1. CMV (1.25 ng/well; Promega) carrying the Renilla luciferase gene under the control of the CMV promoter was introduced to normalize the trans- Knockdown of TLR4 expression fection and cell lysis efficiencies in every experiment. Four hours after transfection, ultra-pure E. coli K12 LPS (InvivoGen) was added to the Cells were transfected with pBAsi-TLR4 or pBAsi-cont, both of which cells. After culturing the cells for an additional 20 h, the cells were har- carry a puromycin resistance gene as a selectable marker, using FuGene vested. Cell lysis and determination of luciferase activity were conducted HD transfection reagent (Roche) and cultured for 24 h. After adding 5 ␮ using a dual-luciferase assay kit (Promega) according to the manufacturer’s g/ml puromycin (Sigma-Aldrich) to select the transfected cells, cells were instructions. Luminescence was measured with a Tristar LB941 luminom- cultured for an additional 48 h. Cells were then employed for measurement by guest on October 4, 2021 eter (Berthold). of IL-8 production upon LPS stimulation as described above. Quantitative RT-PCR (qRT-PCR) FACS Total RNA was prepared from each cell line using an RNeasy Mini Kit After treatment with Fc block (anti-mouse CD16/CD32; BD Biosciences), (Qiagen) or a High Pure RNA isolation kit (Roche), and the first-strand cells were incubated with anti-human TLR4 mAb (clone HTA125; cDNA synthesis was then conducted using SuperScript II reverse transcrip- eBioscience) or mouse IgG2a isotype control (eBioscience) in FACS tase (Invitrogen). Messenger RNA expression was quantified by real-time buffer (DMEM (pH 7.2) containing 0.5% BSA, 1 mM EDTA, 10 mM PCR with a LightCycler 480 (Roche) using SYBR Green I Master reagent HEPES, 2 mM sodium pyruvate, and 0.05% sodium azide) on ice for 30 (Roche). The sequences of the synthetic oligonucleotides used as primers min. After washing, the cells were incubated with biotinylated anti- are as follows: for detection of TLR4 gene expression, forward, 5Ј-AAGC mouse IgG (HϩL) (eBioscience) on ice for 30 min followed by incu- CGAAAGGTGATTGTTG-3Ј; reverse, 5Ј-CTGAGCAGGGTCTTCTCC bation with streptavidin-labeled PE (Invitrogen) and then analyzed on a AC-3Ј; for detection of GAPDH gene expression as a control, forward, 5Ј- FACSCanto (BD Biosciences). TGAACGGGAAGCTCACTGG-3Ј; reverse, 5Ј-TCCACCACCCTGTTGC TGTA-3Ј. Measurement of transcriptional promoter activity by a reporter gene assay Plasmid construction The reporter plasmid pGL-TLR4Ϫ1013/ϩ118 (0.7 ␮g/well) was intro- Ϫ ϩ A DNA fragment corresponding to the nt 1013/ 188 region of the hu- duced into cells using FuGene HD transfection reagent (Roche). For over- man TLR4 gene (nucleotide numbers are counted from the major tran- expression experiments, cells were cotransfected with 0.5 ␮g of the re- ϩ scription start site as 1) was amplified by PCR from a human genomic porter plasmid pGL-TLR4Ϫ1013/ϩ188 and 0.5 ␮g of pcDNA3.1-ZNF160 DNA library (Clontech). The following synthetic oligonucleotides were sense, pcDNA3.1-ZNF160 antisense, or pcDNA3.1-empty. The plasmid Ј used as primers: forward, 5 -CATACTGTATGATTCCACTTCTAAGA phRL-CMV (1.25 ng/well) was introduced to normalize the transfection Ј Ј GCTGCCT-3 ; reverse, 5 -CTGGCATCATCCTCACTGCTTCTGTGAG and cell lysis efficiencies in every experiment. After 20–24 h of culture, Ј CAGCAG-3 . The amplified product was cloned into the pCR2.1 vector cells were harvested and washed with PBS. Cell lysis and determination of (Invitrogen) to confirm the nucleotide sequence and then inserted into the luciferase activity were performed as described above. pGL3-basic vector (Promega) at the KpnI/XhoI sites upstream of the luciferase gene. The resulting plasmid was named pGL3-TLR4Ϫ1013/ϩ188. Immunoprecipitation A series of deletion mutants of pGL3-TLR4Ϫ1013/ϩ188, which carried nt Ϫ675/ϩ188, nt Ϫ427/ϩ188, nt Ϫ218/ϩ188, and nt Ϫ6/ϩ188 regions up- Cells were washed with ice-cold PBS and incubated on ice for 30 min in stream of the luciferase gene, was constructed as follows: pGL3- lysis buffer (20 mM Tris (pH 7.6), 1% Nonidet P-40, 60 mM octyl-␤- TLR4Ϫ427/ϩ188 was generated by self-ligation of blunt-ended products glucoside, 2 mM PMSF, 10 ␮g/ml aprotinin, 2 ␮g/ml leupeptin, 2 ␮g/ml of KpnI/ApaI digested pGL3-TLR4Ϫ1013/ϩ188. The other three frag- pepstatin). Cell lysates were immunoprecipitated with anti-TLR4 mAb (clone ments, nt Ϫ675/ϩ188, nt Ϫ218/ϩ188, and nt Ϫ6/ϩ188, were amplified by HTA125; eBioscience) or anti-TLR2 mAb (clone TL2.1; eBioscience) using PCR, digested with KpnI and BglII, and then inserted into KpnI/BglII- protein A-Sepharose (GE Healthcare). Immunoprecipitated proteins were de- digested-pGL3-basic vector. The sequences of the synthetic oligonucleo- tected by immunoblotting with anti-TLR4 or anti-TLR2 mAb. Part of each tides used for the forward primers are as follows: for amplification of the nonimmunoprecipitated cell lysate was subjected to immunoblotting with 6524 REGULATION OF TLR4 GENE EXPRESSION IN IECs

FIGURE 1. Responsiveness of human IEC lines to LPS correlates with their TLR4 mRNA expression levels. A, Several human IEC lines (Caco-2, HCT Downloaded from 116, T84, HT-29, and SW480) and a human monocyte line (THP-1) were stimulated with 0.1–1000 ng/ml LPS for 18 h. Secretion of IL-8 into the culture supernatant was measured by ELISA. IL-8 production is represented on the graph by subtracting the concentrations in the supernatant of unstimulated cells from those of stimulated cells. Results are represented as means Ϯ SD of two independent experiments. N.D. indicates not detected. B, Activation of NF-␬B by stimulation with LPS was measured by reporter assays. Each cell line was cotransfected with a NF-␬B reporter plasmid and an expression plasmid carrying MyD88 DN or an empty vector control and stimulated with the indicated concentrations of LPS. The fold increase in luciferase activity relative to that of unstimulated cells transfected with the empty vector control is shown. Results are represented as means Ϯ SD of three independent experiments. http://www.jimmunol.org/ C, TLR4 mRNA expression in each cell line was determined by qRT-PCR. Relative values normalized using GAPDH mRNA levels are given. Results are represented as means Ϯ SD of three independent experiments. D, A TLR4 siRNA expression plasmid or a control siRNA expression plasmid was introduced into SW480 cells. After selection of the transfected cells with puromycin, TLR4 mRNA expression was determined by qRT-PCR. Relative values normalized using GAPDH mRNA levels are given. Results are expressed as means Ϯ SD of four independent experiments: —, no transfection; p Ͻ 0.001. E, SW480 cells ,ء ;TLR4, transfected with the TLR4 siRNA expression plasmid; control, transfected with the control siRNA expression plasmid were transfected with the TLR4 siRNA expression plasmid (filled bars) or the control siRNA expression plasmid (open bars). After selection with puromycin, the transfected cells were stimulated with 0.1–1000 ng/ml LPS for 18 h to measure IL-8 secretion into the culture supernatant by ELISA. IL-8 production was calculated by subtracting the concentrations in the culture supernatant of unstimulated cells from those of stimulated cells. Results are p Ͻ 0.005 significantly different from control stimulated with the same ,ءء p Ͻ 0.05 and ,ء .expressed as means Ϯ SD of four independent experiments concentration of LPS. by guest on October 4, 2021 anti-␤-actin mAb (Abcam) to quantify the amounts of proteins present in the cells to analyze the effects of each cDNA clone. The nucleotide se- each lysate before immunoprecipitation. quences of the cDNA clones that repressed TLR4 gene transcription were then determined. Nuclear extract preparation Chromatin immunoprecipitation (ChIP) assay Nuclear extract was prepared from Caco-2 and THP-1 cells as described previously (21). ChIP assays were performed as described previously (22) with slight modifications. Briefly, cells were exposed to 1% formaldehyde to crosslink the EMSA chromatin, solubilized, and then sonicated with a Sonifier 450 (Branson) to shear the genomic DNA to an average length of 500-1000 bp. Precleared Double-stranded DNAs were prepared as probe A, B, and C by annealing lysates were immunoprecipitated with anti-acetyl histone H3 Ab (Upstate Alexa 647-labeled synthetic oligonucleotides. Nucleotide sequences of the Biotechnology) or rabbit IgG as control (Upstate Biotechnology) using probes were as follows: probe A, 5Ј-GACTGTATATGGAGAGAGAGCC salmon sperm DNA/protein A-agarose (Upstate Biotechnology). After TTG-3Ј; probe B, 5Ј-CCTTGAAAGAGGTATGTAAGGTAGA-3Ј; probe washing sequentially with four different wash buffers, the immunoprecipi- C, 5Ј-GTAGAATGAGGTCATTATGGTGGGC-3Ј. tated chromatin was eluted to reverse the cross-links. The 5Ј region of the Fifteen micrograms of nuclear extract and 5 pmol of DNA probe were TLR4 gene corresponding to nt Ϫ529/Ϫ339 was amplified by PCR from incubated at room temperature in 10 mM HEPES buffer (pH 7.9) contain- the recovered DNA using the synthetic oligonucleoides 5Ј-ACATATCG ing 400 ng of poly(dI:dC), 1 mM MgCl2, 30 mM KCl, 1 mM DTT, and 5% AAGTCCTAACCCCTCTAC-3Ј and 5Ј-GAGGCTTCCAGTCTGAACTC glycerol for 20 min. The mixtures were separated by electrophoresis on 4% AAG-3Ј as primers. The GAPDH promoter region was amplified as a con- polyacrylamide gels at 120 V for 2.0–2.5 h in 0.25ϫ TBE buffer (22.5 mM trol using primers with the sequences 5Ј-TACTAGCGGTTTTACGGG Tris, 22.5 mM boric acid, 0.5 mM DTT). Fluorescence was detected using CG-3Ј and 5Ј-TCGAACAGGAGGAGCAGAGAGCGA-3Ј. a Typhoon 9410 (GE Healthcare). For the ZNF160 overexpression experiment, cells were transfected with pcDNA3.1-ZNF160 sense or pcDNA3.1-empty using FuGene HD trans- cDNA subtraction fection reagent. After 8 h, G418 (Sigma-Aldrich) was added to select for cDNAs of the genes that were expressed more abundantly in Caco-2 cells transfected cells. Cells were cultured for an additional 40 h, harvested, and Ј than in SW480 cells were enriched with a PCR-Select cDNA subtraction subjected to ChIP assays employing anti-acetyl histone H3 Ab. The 5 kit (Clontech) according to the manufacturer’s instructions. Obtained region of the TLR4 gene and that of the GAPDH gene in the immunopre- cDNA candidates were inserted into the pcDNA3.1/V5-His-TOPO vector cipitated DNA were quantified by real-time PCR analyses using the same (Invitrogen) to yield their expression plasmids. The expression plasmids primers described above. were introduced into SW480 cells as groups of five clones together with the Bisulfite conversion reaction reporter plasmid pGL3-TLR4Ϫ1013/ϩ188 for a transient expression assay. Expression plasmids, contained in mixtures that gave luciferase ac- Genomic DNA was prepared from cells using a PureLink Genomic DNA tivity less than half of that of the empty vector generated by self-ligation of Mini kit (Invitrogen) according to the manufacturer’s instructions. To an- BstXI-digested pcDNA3.1/V5-His-TOPO, were separately introduced into alyze methylation of CpG motifs, 400 ng of genomic DNA was denatured The Journal of Immunology 6525 at 98°C for 10 min, modified by conversion reagent at 64°C for 2.5 h, and then purified using a MethylCode bisulfite conversion kit (Invitrogen). The 5Ј region of the TLR4 gene was amplified by PCR from the modified genomic DNA. The sequences of the synthetic oligonucleotides used as PCR primers are as follows: forward, 5Ј-GGTAGAGGTTAGATGATTAA TTGGG-3Ј; reverse, 5Ј-CCCCAATAACTACCTCTAATACCCT-3Ј. Following purification of PCR products, they were cloned into the pCR2.1 vector for sequencing. Nucleotide sequences of 14–16 clones for each cell line were analyzed.

Inhibition of histone deacetylase and DNA methyltransferase Caco-2 cells were treated with 10 ␮M 5-aza-2Ј-deoxycytidine (5-Aza-dC; Calbiochem) for 4 days, with 80 nM trichostatin A (TSA; Cayman Chem- ical) for 24 h, or with TSA and 5-Aza-dC for 24 h after treatment with 5-aza-dC alone for 3 days. Cells were then harvested to obtain total RNA. TLR4 mRNA expression was quantiﬁed by qRT-PCR as described above.

Results Responsiveness of human IEC lines to LPS is down-regulated mainly by repressing TLR4 gene transcription Various human IEC lines were ﬁrst analyzed for their ability to Downloaded from respond to stimulation through TLR4. Five human IEC lines (Caco-2, HCT 116, T84, HT-29, SW480) and a control monocyte line (THP-1) were stimulated with LPS to measure IL-8 secretion into the culture supernatant (Fig. 1A). Most IEC lines, with the exception of SW480, showed low responsiveness to LPS com- FIGURE 2. LPS responsiveness of IEC lines is mainly regulated at the

transcriptional level. A, Cell surface expression of TLR4 on each IEC line http://www.jimmunol.org/ pared with the monocyte line THP-1. Caco-2 and HCT 116 cells was determined by FACS using anti-TLR4 Ab. Results are representative barely responded, and HT-29 and T84 cells weakly responded to of three independent experiments. Shadowed areas, unstained; thin lines, LPS stimulation. In contrast, production of IL-8 from SW480 cells stained with isotype control Ab; bold lines, stained with anti-TLR4 Ab. B, was similar to, or even higher than, IL-8 production from the Transcriptional enhancing activity of the 5Ј region of the human TLR4 monocyte line THP-1. The MyD88 dependency of these responses gene (nt Ϫ1013/ϩ188) in each cell line was measured by a reporter gene was further examined by NF-␬B reporter assays (Fig. 1B). Con- assay. Luciferase activities relative to that from a SV40 promoter control sistent with the results for IL-8 production, activation of NF-␬B are shown. Results are represented as means Ϯ SD of three independent was observed in SW480 and THP-1 cells when stimulated with experiments. C, Protein expression of TLR4 and TLR2 in each cell line LPS, while it was hardly detected in Caco-2 cells. Activation of was determined by immunoprecipitation. Amounts of protein included in by guest on October 4, 2021 NF-␬B was inhibited by exogenous expression of the DN form of the cell lysates before immunoprecipitation were determined by immunoblotting with anti-␤-actin Ab. Data are representative of two independent MyD88 in SW480 and THP-1 cells, indicating that the responses experiments. to LPS were dependent on MyD88. The TLR4 mRNA expression in each cell line was determined by qRT-PCR (Fig. 1C). The amounts of TLR4 mRNA almost exactly correlated with the LPS responsiveness of each cell line. Furthermore, TLR4 mRNA ex- shown that the responsiveness of IECs to LPS is down-regulated pression was reduced to about one-third in SW480 cells trans- mainly by suppressing transcription of the TLR4 gene. fected with a TLR4 siRNA expression plasmid when compared Ϫ Ϫ with the cells transfected with a control siRNA expression plasmid Nucleotide 489/ 428 region of the TLR4 gene contains an (Fig. 1D), which led to a significant decrease in IL-8 production IEC-specific repressor element and is recognized by an upon stimulation with LPS (Fig. 1E). These results show that LPS IEC-specific NF sensitivity is regulated by TLR4 gene expression in IECs. Next, To further examine the mechanisms of transcriptional repression the cell surface expression of TLR4 and the transcriptional en- of the TLR4 gene in IECs, cis-acting elements in the 5Ј region of hancing activity of the 5Ј region of the gene upstream of the trans- the TLR4 gene were compared between Caco-2 and THP-1 cells. lation start codon were analyzed in each cell line (Fig. 2, A and B). For this purpose, a series of deletion constructs were employed for The cell surface expression of TLR4 was undetectable on Caco-2 luciferase assays. Deletion of the nt Ϫ675/Ϫ428 region increased and HCT 116 cells. On the other hand, HT-29 and T84 cells ex- the luciferase activity in Caco-2 cells, while it hardly affected the pressed a small amount of cell surface TLR4. SW480 cells ex- activity in THP-1 cells, suggesting that this region contains an pressed a higher level of cell surface TLR4 than did HT-29 and IEC-specific repressor element (Fig. 3A). Moreover, this region T84 cells. The highest cell surface expression of TLR4 was ob- was suggested to contain a responsive element that is repressed by served on the monocyte line THP-1. The transcriptional promoter some microbe, because the transcriptional enhancing activity of nt activity of the 5Ј region of the TLR4 gene was, in order, THP-1, Ϫ675/ϩ188 but not of nt Ϫ427/ϩ188 was decreased by the syn- SW480 Ͼ HT-29, T84 Ͼ Caco-2, HCT 116. Intracellular protein thetic TLR2 ligand, Pam3CSK4 (Fig. 3B). Further mapping of the expression of TLR4 in each cell line was further analyzed by im- repressor element in the nt Ϫ675/Ϫ428 region by luciferase assays munoprecipitation (Fig. 2C). TLR4 protein was barely detected in using deletion constructs in Caco-2 cells revealed that the nt Ϫ489/ Caco-2 cells but was detected at higher levels in SW480 and Ϫ428 region acts as a repressor element (Fig. 3C). We next ex- THP-1 cells, while TLR2 was apparently detected in all cell lines. amined the presence of a NF binding to this region by EMSA using These results indicate that expression of the TLR4 gene is mainly three oligonucleotide probes covering this region and nuclear ex- regulated at the transcriptional level, although additional posttran- tracts prepared from Caco-2 and THP-1 cells. The shifted band scriptional regulation also seems to be present. Collectively, it was indicated by an arrow in Fig. 3D appeared when the nuclear extract 6526 REGULATION OF TLR4 GENE EXPRESSION IN IECs Downloaded from

FIGURE 3. Nucleotide Ϫ489/Ϫ428 region contains an IEC-specific repressor element and is recognized by an IEC-specific NF. A, cis-Acting elements in the 5Ј region of the TLR4 gene were determined by a reporter gene assay using a series of deletion constructs in Caco-2 and THP-1 cells. Luciferase activities relative to that of pGL3-TLR4Ϫ1013/ϩ188 are shown. Results are expressed as means Ϯ SD of three independent experiments. PU.1 and ICSBP binding sites reported by Rehli et al. (31) and enhancer and repressor elements suggested from our study are schematically drawn. B, Caco-2 cells were transfected with the reporter plasmid pGL3-TLR4Ϫ1013/ϩ188, pGL3-TLR4Ϫ675/ϩ188, pGL3-TLR4Ϫ427/ϩ188, or pGL3-TLR4Ϫ218/ϩ188 for a http://www.jimmunol.org/ transient expression assay in the presence or absence of 1 ␮g/ml Pam3CSK4. Luciferase activities relative to those without Pam3CSK4 treatment are shown in each graph. Results are expressed as means Ϯ SD of two independent experiments. C, For further mapping of a repressor element in the nt Ϫ675/Ϫ428 region, Caco-2 cells were transfected with a series of deletion constructs for a transient expression assay. Luciferase activities relative to thoseof pGL3-TLR4Ϫ1013/ϩ188 are shown. Results are expressed as means Ϯ SD of three independent experiments. D, EMSA was performed using three double-stranded oligonucleotide probes corresponding to the regions nt Ϫ492/Ϫ468, nt Ϫ472/Ϫ448, and nt Ϫ452/Ϫ428 of the TLR4 gene and nuclear extracts prepared from Caco-2 and THP-1 cells. Results are representative of three independent experiments. by guest on October 4, 2021 from Caco-2 cells but not from THP-1 cells was added, indicating of nt Ϫ675/ϩ188 but not of nt Ϫ427/ϩ188 in SW480 cells, that a NF that was specifically expressed in Caco-2 cells bound to indicating that ZNF160 acts through the nt Ϫ675/Ϫ428 region the nt Ϫ452/Ϫ428 region. Collectively, these results suggest that (Fig. 4, D and E). a specific NF binds to the nt Ϫ489/Ϫ428 region and represses TLR4 gene transcription in IECs. TLR4 gene transcription is down-regulated by epigenetic modification including histone deacetylation and DNA ZNF160 represses TLR4 gene transcription in IECs methylation in IECs To elucidate the molecules involved in the transcriptional repres- KRAB domains have been reported to recruit KRAB-associated sion of the TLR4 gene in IECs, we screened a human cDNA li- protein (KAP) 1 when tethered to DNA via their zinc finger motifs brary, using the PCR-select subtraction method, for molecules that (24, 25). KAP1 directly binds to KRAB and functions as a scaffold were expressed more abundantly in the LPS-low responder IEC for the formation of a multimolecular complex comprising histone line Caco-2 than in the LPS-high-responder IEC line SW480. The deacetylases, which induces transcriptional repression through the obtained cDNA clones were screened again to identify molecules formation of heterochromatin (26–28). In addition to histone that repressed transcription of the TLR4 gene when overexpressed deacetylation, it was recently reported that the KRAB domain can in SW480 cells by a transient expression assay. Out of the 360 trigger de novo DNA methylation (29). Histone acetylation and cDNA clones selected by subtraction, 7 clones, including 6 clones DNA methylation are known to mediate epigenetic regulation of that encoded 3 different molecules and 1 clone that did not encode gene expression. To investigate if TLR4 gene expression is con- the protein in frame, decreased the TLR4 promoter activity to less trolled by these epigenetic mechanisms in IECs, acetylation of than half. One of the three identified molecules that were encoded histones interacting with the 5Ј region of the TLR4 gene was an- by cDNAs in frame and repressed activation of the TLR4 gene alyzed by ChIP assays using anti-acetyl histone H3 Ab. As shown promoter was ZNF160. As shown in Fig. 4A, it was confirmed that in Fig. 5, acetylated histones interacting with the 5Ј region of the ZNF160 mRNA expression was higher in Caco-2 cells than in TLR4 gene were significantly lower in Caco-2 and HCT 116 cells, SW480 cells by qRT-PCR analyses. ZNF160 is a Kruppel-related both of which barely express TLR4, than in SW480 and THP-1 zinc finger protein characterized by the presence of the Kruppel- cells, indicating that the histones interacting with the 5Ј region of associated box (KRAB), a potent repressor of transcription, at the the TLR4 gene were highly deacetylated in LPS-low responder N terminus (23). Overexpression of ZNF160 decreased the tran- IEC lines. DNA methylation of 11 CpG motifs existing in the 5Ј scriptional promoter activity of the TLR4 gene in SW480 cells region of the TLR4 gene (from nt Ϫ69 to nt ϩ277) was next (Fig. 4B), whereas, interestingly, it up-regulated the promoter analyzed by the bisulfite conversion reaction. As is shown in Fig. activity in the human monocyte line THP-1 (Fig. 4C). Further- 6, the number of methylated CpG motifs was significantly higher more, ZNF160 repressed the transcriptional enhancing activity in Caco-2 and HCT 116 cells than in SW480 and THP-1 cells. The Journal of Immunology 6527

FIGURE 6. The 5Ј region of the TLR4 gene is highly methylated in LPS-low responder IEC lines. DNA methylation of 11 CpG motifs existing in the 5Ј region of the TLR4 gene (nt Ϫ69/ϩ277) was analyzed by the bisulfite conversion reaction. The modified DNA was amplified by PCR, cloned, and sequenced for 14–16 independent clones. A, Filled squares indicate methylated CpG motifs, and open squares indicate unmethylated CpG motifs. B, Mean numbers of methylated CpG motifs in the nt Ϫ69/ ,ءء p Ͻ 1 ϫ 10Ϫ7 and ,ء .ϩ277 region and SD in each cell line are shown Downloaded from p Ͻ 5 ϫ 10Ϫ9.

that DNA methylation and histone deacetylation are cooperatively involved in the repression of TLR4 gene transcription (Fig. 7A).

Collectively, these results show that TLR4 gene transcription is http://www.jimmunol.org/ FIGURE 4. ZNF160, one of the molecules expressed more abundantly down-regulated by epigenetic modification including histone in the LPS-low responder IEC line Caco-2 than in the LPS-high responder deacetylation and DNA methylation in LPS-low responder IEC IEC line SW480, represses transcription of the TLR4 gene. A, ZNF160 lines. mRNA expression in Caco-2 and SW480 cells was determined by qRT- PCR. Relative values normalized using GAPDH mRNA levels are given. The epigenetic modification is, at least in part, dependent on ء Ϯ Results are expressed as means SD of three independent experiments. , ZNF160 p Ͻ 0.05. B and C, SW480 (B), or THP-1 (C) cells were cotransfected with the reporter plasmid pGL3-TLR4Ϫ1013/ϩ188 and a ZNF160 expression To analyze the involvement of ZNF160 in epigenetic regulation, plasmid (pcDNA3.1-ZNF160 sense or pcDNA3.1-ZNF160 antisense) or an ChIP assays of the endogenous TLR4 gene were performed em- empty vector control (pcDNA3.1-empty) for a transient expression assay. ploying SW480 cells transfected with a ZNF160 expression plas- by guest on October 4, 2021 Results are represented as means Ϯ SD of five to six independent exper- mid or an empty vector control. Acetylation of histones that inter- iments. D and E, SW480 cells were cotransfected with a reporter plasmid acted with the 5Ј region of the TLR4 gene was significantly carrying nt Ϫ675/ϩ188 (D)orntϪ427/ϩ188 (E) of the TLR4 gene and reduced by the overexpression of ZNF160, while acetylation of a ZNF-160 expression plasmid or an empty vector control for a transient Ј expression assay. Results are expressed as means Ϯ SD of three indepen- histones that interacted with the 5 region of the GAPDH gene was p Ͻ 0.005, significantly different not affected (Fig. 7, B and C). The results indicate that epigenetic ,ءء p Ͻ 0.05 and ,ء .dent experiments from control transfected with an empty vector.

Furthermore, the effects of the DNA methyltransferase inhibitor 5-aza-dC and/or the histone deacetylase inhibitor TSA were analyzed in Caco-2 cells. The expression of TLR4 increased ϳ10-fold as a result of treatment with both 5-aza-dC and TSA, indicating

FIGURE 7. Epigenetic modification is involved in repression of TLR4 gene transcription and is, at least in part, dependent on ZNF160. A, Caco-2 cells were treated with 10 ␮M 5-aza-dC (DNA methyltransferase inhibitor) for 4 days or with 80 nM TSA (histone deacetylase inhibitor) for 24 h. TLR4 mRNA expression was analyzed by qRT-PCR. Relative values, normalized using GAPDH mRNA levels, are expressed. Results are repre- FIGURE 5. Histones interacting with the 5Ј region of the TLR4 gene sented as means Ϯ SD of three independent experiments. B and C, SW480 are deacetylated in LPS-low responder IEC lines. ChIP assays of endog- cells were transfected with a ZNF160 expression plasmid (pcDNA3.1- enous TLR4 genes in IEC lines were performed employing anti-acetyl ZNF160 sense) or an empty vector control (pcDNA3.1-empty) and cul- histone H3 Ab. Rabbit IgG was used as a control. The 5Ј regions of the tured for 8 h. After adding G418 to select the transfected cells, the cells TLR4 gene, and of the GAPDH gene as a control, were amplified by PCR were cultured for an additional 40 h. ChIP assays were performed using from the immunoprecipitated chromatin. The “input” lanes represent the anti-acetyl histone H3 Ab or control rabbit IgG followed by qPCR for the results of PCR using diluted fractions of nonimmunoprecipitated chromatin 5Ј regions of the TLR4 gene (B) and GAPDH control gene (C). Results are .p Ͻ 0.05 ,ء .as templates. Results are representative of three independent experiments. represented as means Ϯ SD of three independent experiments 6528 REGULATION OF TLR4 GENE EXPRESSION IN IECs modification at the 5Ј region of the TLR4 gene is, at least in part, oligomerization and direct binding to KRAB (24, 25). A reduction dependent on ZNF160. in acetylation of histones that interact with the 5Ј region of the TLR4 gene in IECs is thought to be dependent on histone deacety- Discussion lase, which is recruited by ZNF160 through KAP1. Similarly, an In this report, we show that TLR4 gene transcription is epigeneti- increase in methylation of the 5Ј region of the TLR4 gene in IECs cally suppressed in IECs to prevent excessive inflammatory re- is thought to be dependent on ZNF160. However, the contribution sponses. This means that epigenetic regulation of TLR4 gene ex- of ZNF160 to the regulation of TLR4 gene expression seems to be pression in IECs can act as one mechanism for maintaining rather small compared with the large contribution of epigenetic intestinal homeostasis by suppressing excessive responses to the modifications to the repression of TLR4 gene expression. Actually, commensals and regulating mucosal inflammation in the gut. Epi- overexpression of ZNF160 only suppressed transcriptional pro- genetic information is encoded by differential methylation of DNA moter activity of the TLR4 gene to about a half in SW480 cells on cytosines and by proteins associating with DNA such as his- (Fig. 4B), while inhibitors of histone deacetylase and DNA meth- tones, which may be modified covalently by acetylation, methyl- yltransferase together increased TLR4 gene expression by ϳ10- ation, phosphorylation, and/or ubiquitination. Since this epigenetic fold in Caco-2 cells (Fig. 7A). Additionally, overexpression of information is heritable beyond cell division, the importance of ZNF160 in the monocyte line increased transcriptional promoter epigenetic regulation is established especially in the fields of de- activity, while overexpression in the IEC line suppressed transcrip- velopmental and cancer biology. This is the first report to describe tion (Fig. 4, B and C). These results suggest that ZNF160 and an the involvement of epigenetic regulation of transcription in the additional cell type-specific regulatory factor cooperatively repress maintenance of the intestinal commensal system. Although the in- transcription of the TLR4 gene in IECs. Recently, a single nucle- testinal epithelium is known to be continuously renewed by rapid otide polymorphism with strong association to ileal Crohn’s dis- Downloaded from turnover of IECs, there are many reports supporting that the char- ease was mapped to an intergenic region that is flanked on the acteristics of IECs including their tolerated responses to the com- centromeric side by a gene encoding another zinc finger protein, mensals are inherited by renewed cells, raising the possibility that ZNF365 (32). Although the functions of ZNF365 are little known, epigenetic regulation as seen in our study is involved in such in- specific members of the C2H2 zinc finger proteins, which consti- heritance. These basic mechanisms may work commonly in other tute the largest class of transcription factors in humans, may play commensal systems in specific tissues in our body such as the skin a role in the maintenance of intestinal homeostasis. http://www.jimmunol.org/ epidermis or the mucosa of the oral cavity, in addition to the in- LPS responsiveness of IECs depends on TLR4 expression levels testine, all of which are continuously exposed to nonpathogenic as shown in Fig. 1E. Additionally, expression of MD-2 is thought microbes. Zampetaki et al. recently reported that murine TLR4 to be another important factor that determines the LPS responsive- gene expression is epigenetically repressed in embryonic stem ness because optimal LPS recognition by TLR4 requires MD-2 as cells but not in embryonic stem cell-derived differentiated smooth a coreceptor. It has been reported that MD-2 expression in healthy, muscle cells (30). Epigenetic repression may be released in dif- normal intestinal mucosa is minimal, while it is increased in active ferentiated cells except specific types of cells including IECs. inflammatory bowel disease colitis (33, 34). Moreover, inhibition Rehli et al. reported that transcription factors of PU.1 and IFN of CpG methylation and histone deacetylation was shown to result by guest on October 4, 2021 consensus sequence-binding protein (ICSBP) regulate human in increased mRNA expression of MD-2 gene in IECs just recently TLR4 gene expression in myeloid cells through elements just up- (35). TLR4 and MD-2 expression may be down-regulated in part stream of the transcription start site (31). As shown in Fig. 3A, by common or related mechanisms in IECs. On the other hand, additional enhancer elements seem to be present further upstream expression of TLR4 and TLR2 is thought to be regulated at dif- of the PU.1 and ICSBP binding sites identified by Rehli et al. ferent stages. TLR4 expression has been found to be regulated Moreover, the presence of an IEC-specific repressor element in the mainly at the transcriptional level because its cell surface expres- 5Ј region of TLR4 gene and an IEC-specific NF binding to the sion correlates relatively well with mRNA expression levels in element has been suggested. We tried to identify the transcription each cell line. However, regulation at the posttranslational level is factor forming the Caco-2-specific band shown in Fig. 3D but were also thought to be present, as a difference in the cell surface TLR4 unsuccessful. Since it was found that ZNF160 acts through the nt expression was seen between SW480 and THP1 cells despite these Ϫ675/Ϫ428 region (Fig. 4, D and E), it was thought that ZNF160 cells having almost the same mRNA expression levels (Figs. 1 and binds to this region. However, in vitro translation products of 2). Since IL-8 production in response to LPS stimulation was ZNF160 did not bind to the double-stranded DNA probe contain- rather higher in SW480 cells than in THP-1 cells, a certain extent ing this region in EMSA (data not shown). Some modification of of cell surface TLR4 expression seems to be sufficient to respond the ZNF-160 protein may be required for DNA binding, or, alter- to LPS. On the other hand, expression of TLR2 seemed to be natively, ZNF-160 may indirectly bind to this region through an- regulated mainly at the posttranslational level because its mRNA other DNA binding factor. Stimulation with Pam3CSK4 repressed and intracellular protein expression was not apparently different the transcriptional enhancing activity of the 5Ј region of the TLR4 between the IEC lines and the monocyte line, but its cell surface gene through the nt Ϫ675/Ϫ428 region (Fig. 3B). It is unclear at expression was much lower in the IEC lines than in the monocyte present whether specific microbial components modify transcrip- line (Fig. 2C and our unpublished data). Intracellular transport of tion of the TLR4 gene, although LPS may not affect transcriptional TLR2 protein to the cell surface is thought to be inhibited by a activation because expression of TLR4 recognizing LPS was un- specific mechanism in IECs. detectable on Caco-2 cells (Fig. 2A). Further study should allow elucidation of the entire regulatory The epigenetic regulation of TLR4 gene expression has been mechanism underlying transcription of symbiosis-associated found to be partly dependent on ZNF160, a repressive transcription genes, including TLR4 in IECs and its contribution to regulation of factor possessing the KRAB domain, which has been reported to mucosal inflammation triggered by IECs, and, as a result, to main- recruit histone deacetylase through the scaffold protein KAP1. tenance of the intestinal commensal system. KAP1 is characterized by the presence of a RING finger, B boxes, a coiled-coil region, a PHD finger, and a bromodomain; the first Disclosures three of these motifs are both necessary and sufficient for homo- The authors have no financial conflicts of interest. The Journal of Immunology 6529

References 19. Abreu, M. T., P. Vora, E. Faure, L. S. Thomas, E. T. Arnold, and M. Arditi. 2001. Decreased expression of Toll-like receptor-4 and MD-2 correlates with intestinal 1. Sudo, N., S. Sawamura, K. Tanaka, Y. Aiba, C. Kubo, and Y. Koga. 1997. The epithelial cell protection against dysregulated proinflammatory gene expression requirement of intestinal bacterial flora for the development of an IgE production in response to bacterial lipopolysaccharide. J. Immunol. 167: 1609–1616. system fully susceptible to oral tolerance induction. J. Immunol. 159: 1739–1745. 20. Cario, E., and D. K. Podolsky. 2000. Differential alteration in intestinal epithelial 2. Cebra, J. J., S. B. Periwal, G. Lee, F. Lee, and K. E. Shroff. 1998. Development cell expression of Toll-like receptor 3 (TLR3) and TLR4 in inflammatory bowel and maintenance of the gut-associated lymphoid tissue (GALT): the roles of disease. Infect. Immun. 68: 7010–7017. enteric bacteria and viruses. Dev. Immunol. 6: 13–18. 21. Takahashi, K., N. Hayashi, S. Shimokawa, N. Umehara, S. Kaminogawa, and 3. Umesaki, Y., H. Setoyama, S. Matsumoto, A. Imaoka, and K. Itoh. 1999. Dif- C. Ra. 2008. Cooperative regulation of Fc receptor ␥-chain gene expression by ferential roles of segmented filamentous bacteria and clostridia in development of multiple transcription factors, including Sp1, GABP, and Elf-1. J. Biol. Chem. the intestinal immune system. Infect. Immun. 67: 3504–3511. 283: 15134–15141. 4. Talham, G. L., H. Q. Jiang, N. A. Bos, and J. J. Cebra. 1999. Segmented fila- 22. Takahashi, K., N. Hayashi, S. Kaminogawa, and C. Ra. 2006. Molecular mech- mentous bacteria are potent stimuli of a physiologically normal state of the mu- anisms for transcriptional regulation of human high affinity IgE receptor ␤-chain rine gut mucosal immune system. Infect. Immun. 67: 1992–2000. gene induced by GM-CSF. J. Immunol. 177: 4605–4611. 5. Gaboriau-Routhiau, V., and M. C. Moreau. 1996. Gut flora allows recovery of 23. Halford, S., M. G. Mattei, S. Daw, and P. J. Scambler. 1995. A novel C2H2 oral tolerance to ovalbumin in mice after transient breakdown mediated by chol- zinc-finger protein gene (ZNF160) maps to human chromosome 19q13.3–q13.4. era toxin or Escherichia coli heat-labile enterotoxin. Pediatr. Res. 39: 625–629. Genomics 25: 322–323. 6. Penders, J., C. Thijs, P. A. van den Brandt, I. Kummeling, B. Snijders, F. Stelma, 24. Kim, S. S., Y. M. Chen, E. O’Leary, R. Witzgall, M. Vidal, and J. V. Bonventre. H. Adams, R. van Ree, and E. E. Stobberingh. 2007. Gut microbiota composition 1996. A novel member of the RING finger family, KRIP-1, associates with the and development of atopic manifestations in infancy: the KOALA Birth Cohort KRAB-A transcriptional repressor domain of zinc finger proteins. Proc. Natl. Study. Gut 56: 661–667. Acad. Sci. USA 93: 15299–15304. 7. Mah, K. W., B. Bjo¨rksteń, B. W. Lee, H. P. van Bever, L. P. Shek, T. N. Tan, 25. Friedman, J. R., W. J. Fredericks, D. E. Jensen, D. W. Speicher, X. P. Huang, Y. K. Lee, and K. Y. Chua. 2006. Distinct pattern of commensal gut microbiota E. G. Neilson, and F. J. Rauscher, 3rd. 1996. KAP-1, a novel corepressor for the in toddlers with eczema. Int. Arch. Alelrgy Immunol. 140: 157–163. highly conserved KRAB repression domain. Genes Dev. 10: 2067–2078. 8. Sepp, E., K. Julge, M. Mikelsaar, and B. Bjo¨rksteń. 2005. Intestinal microbiota 26. Jakobsson, J., M. I. Cordero, R. Bisaz, A. C. Groner, V. Busskamp, and immunoglobulin E responses in 5-year-old Estonian children. Clin. Exp. J. C. Bensadoun, F. Cammas, R. Losson, I. M. Mansuy, C. Sandi, and D. Trono. Downloaded from Allergy 35: 1141–1146. 2008. KAP1-mediated epigenetic repression in the forebrain modulates behav- 9. Watanabe, S., Y. Narisawa, S. Arase, H. Okamatsu, T. Ikenaga, Y. Tajiri, and ioral vulnerability to stress. Neuron 60: 818–831. M. Kumemura. 2003. Differences in fecal microflora between patients with atopic 27. Sripathy, S. P., J. Stevens, and D. C. Schultz. 2006. The KAP1 corepressor dermatitis and healthy control subjects. J. Allergy Clin. Immunol. 111: 587–591. functions to coordinate the assembly of de novo HP1-demarcated microenviron- 10. Kallioma¨ki, M., P. Kirjavainen, E. Eerola, P. Kero, S. Salminen, and E. Isolauri. ments of heterochromatin required for KRAB zinc finger protein-mediated tran- 2001. Distinct patterns of neonatal gut microflora in infants in whom atopy was scriptional repression. Mol. Cell. Biol. 26: 8623–8638. and was not developing. J. Allergy Clin. Immunol. 107: 129–134. 28. Ayyanathan, K., M. S. Lechner, P. Bell, G. G. Maul, D. C. Schultz, Y. Yamada, 11. Bjo¨rksteń, B., E. Sepp, K. Julge, T. Voor, and M. Mikelsaar. 2001. Allergy K. Tanaka, K. Torigoe, and F. J. Rauscher, 3rd. 2003. Regulated recruitment of development and the intestinal microflora during the first year of life. HP1 to a euchromatic gene induces mitotically heritable, epigenetic gene silenc- http://www.jimmunol.org/ Genes Dev. J. Allergy Clin. Immunol. 108: 516–520. ing: a mammalian cell culture model of gene variegation. 17: 1855–1869. 12. Jaensson, E., H. Uronen-Hansson, O. Pabst, B. Eksteen, J. Tian, J. L. Coombes, 29. Wiznerowicz, M., J. Jakobsson, J. Szulc, S. Liao, A. Quazzola, F. Beermann, P. L. Berg, T. Davidsson, F. Powrie, B. Johansson-Lindbom, and W. W. Agace. ϩ P. Aebischer, and D. Trono. 2007. The Kruppel-associated box repressor domain 2008. Small intestinal CD103 dendritic cells display unique functional proper- can trigger de novo promoter methylation during mouse early embryogenesis. ties that are conserved between mice and humans. J. Exp. Med. 205: 2139–2149. J. Biol. Chem. 282: 34535–34541. 13. Coombes, J. L., K. R. Siddiqui, C. V. Arancibia-Ca´rcamo, J. Hall, C. M. Sun, 30. Zampetaki, A., Q. Xiao, L. Zeng, Y. Hu, and Q. Xu. 2006. TLR4 expression in Y. Belkaid, and F. Powrie. 2007. A functionally specialized population of mu- ϩ ϩ ␤ mouse embryonic stem cells and in stem cell-derived vascular cells is regulated cosal CD103 DCs induces Foxp3 regulatory T cells via a TGF- and retinoic by epigenetic modifications. Biochem. Biophys. Res. Commun. 347: 89–99. acid-dependent mechanism. J. Exp. Med. 204: 1757–1764. 31. Rehli, M., A. Poltorak, L. Schwarzfischer, S. W. Krause, R. Andreesen, and 14. Meńard, O., M. J. Butel, V. Gaboriau-Routhiau, and A. J. Waligora-Dupriet. B. Beutler. 2000. PU.1 and interferon consensus sequence-binding protein reg-

2008. Gnotobiotic mouse immune response induced by Bifidobacterium sp. ulate the myeloid expression of the human Toll-like receptor 4 gene. J. Biol. by guest on October 4, 2021 strains isolated from infants. Appl. Environ. Microbiol. 74: 660–666. Chem. 275: 9773–9781. 15. Ivanov, I. I., L. Frutos Rde, N. Manel, K. Yoshinaga, D. B. Rifkin, R. B. Sartor, 32. Rioux, J. D., R. J. Xavier, K. D. Taylor, M. S. Silverberg, P. Goyette, A. Huett, B. B. Finlay, and D. R. Littman. 2008. Specific microbiota direct the differenti- T. Green, P. Kuballa, M. M. Barmada, L. W. Datta, et al. 2007. Genome-wide ation of IL-17-producing T-helper cells in the mucosa of the small intestine. Cell association study identifies new susceptibility loci for Crohn disease and impli- Host Microbe 4: 337–349. cates autophagy in disease pathogenesis. Nat. Genet. 39: 596–604. 16. Macpherson, A. J., K. D. McKoy, F. E. Johansen, and P. Brandtzaeg. 2008. The 33. Abreu, M. T., E. T. Arnold, L. S. Thomas, R. Gonsky, Y. Zhou, B. Hu, and immune geography of IgA induction and function. Mucosal Immunol. 1: 11–22. M. Arditi. 2002. TLR4 and MD-2 expression is regulated by immune-mediated 17. Rakoff-Nahoum, S., J. Paglino, F. Eslami-Varzaneh, S. Edberg, and R Medzhi- signals in human intestinal epithelial cells. J. Biol. Chem. 277: 20431–20437. tov. 2004. Recognition of commensal microflora by Toll-like receptors is re- 34. Cario, E., D. T. Golenbock, A. Visintin, M. Ru¨nzi, G. Gerken, and quired for intestinal homeostasis. Cell 118: 229–241. D. K. Podolsky. 2006. Trypsin-sensitive modulation of intestinal epithelial MD-2 18. Melmed, G., L. S. Thomas, N. Lee, S. Y. Tesfay, K. Lukasek, K. S. Michelsen, as mechanism of lipopolysaccharide tolerance. J. Immunol. 176: 4258–4266. Y. Zhou, B. Hu, M. Arditi, and M. T. Abreu. 2003. Human intestinal epithelial 35. Vamadevan, A. S., M. Fukata, E. T. Arnold, L. S. Thomas, D. Hsu, and cells are broadly unresponsive to Toll-like receptor 2-dependent bacterial ligands: M. T. Abreu. 2009. Regulation of Toll-like receptor 4-associated MD-2 in intes- implications for host-microbial interactions in the gut. J. Immunol. 170: tinal epithelial cells: a comprehensive analysis. Innate Immun. [Epub ahead of 1406–1415. print on Aug. 26, 2009].