Cdx) Transcription Factors1

The Journal of Immunology

Functional Characterization of the CCL25 Promoter in Small Intestinal Epithelial Cells Suggests a Regulatory Role for Caudal-Related Homeobox (Cdx) Transcription Factors1

Anna Ericsson,2* Knut Kotarsky,2,3* Marcus Svensson,* Mikael Sigvardsson,† and William Agace*

The chemokine CCL25 is selectively and constitutively expressed in the small intestinal epithelium and plays an important role in mediating lymphocyte recruitment to this site. In this study, we demonstrate that CCL25 expression in murine small intestinal epithelial cells is independent of signaling through the lymphotoxin ␤ receptor and is not enhanced by inflammatory stimuli, pathways involved in driving the expression of most other chemokines. We define a transcriptional start site in the CCL25 gene and a region ؊141 to ؊5 proximal of exon 1 that is required for minimal promoter activity in the small intestinal epithelial cell lines, MODE-K and mICc12. These cell lines expressed far less CCL25 mRNA than freshly isolated small intestinal epithelial cells indicating that they are missing important factors driving CCL25 expression. The CCL25 promoter contained putative binding sites for the intestinal epithelial-associated Caudal-related homeobox (Cdx) transcription factors Cdx-1 and Cdx-2, and small intestinal epithelial cells but not MODE-K and mICc12 cells expressed Cdx-1 and Cdx-2. EMSA analysis demonstrated that Cdx proteins were present in nuclear extracts from freshly isolated small intestinal epithelial cells but not in MODE-K or mICcl2 cells, and bound to putative Cdx sites within the CCL25 promoter. Finally, cotransfection of MODE-K cells with Cdx transcription factors significantly increased CCL25 promoter activity as well as endogenous CCL25 mRNA levels. Together these results demonstrate a unique pattern of regulation for CCL25 and suggest a role for Cdx proteins in regulating CCL25 transcription. The Journal of Immunology, 2006, 176: 3642–3651.

hemokines are a large family of low m.w. proteins initiation of immune responses, and immune surveillance of primarily recognized for their role as leukocyte che- healthy peripheral tissues (2). The expression of homeostatic C moattractants and in regulating leukocyte trafficking. chemokines in lymphoid organs and the intestine is largely They function through seven transmembrane G protein-coupled dependent on lymphotoxin (LT)4 ␤ receptor signaling (3–5). receptors to induce directed cellular migration and enhanced The division of chemokines into inflammatory and homeostatic integrin-mediated adhesion, which are processes critical for chemokines is, however, not absolute because many homeo- leukocyte extravasation (1). Chemokines can be divided into static chemokines can be up-regulated in response to inflam- two groups, inflammatory and homeostatic chemokines, based matory stimuli (6, 7), and inflammatory chemokines can target on their regulation and function (2). Inflammatory chemokines noneffector leukocytes at sites of leukocyte development (8). control the recruitment of effector leukocytes, including cells The chemokine CCL25 is selectively and constitutively ex- from both the innate and adaptive immune response, to sites of pressed in the small intestine and thymus, primarily by resident infection or inflammation, and can be induced in a wide variety epithelial cells (9–12). Its sole functional receptor, CCR9, is of cells upon exposure to host or pathogen-derived inflamma- expressed on small intestinal lymphocytes, a subset of circulat- tory stimuli (2). Homeostatic chemokines, by contrast, are ing gut tropic lymphocytes, and thymocytes (13, 14). Analysis constitutively expressed in primary and secondary lymphoid Ϫ Ϫ of CCR9 / mice, and in vivo studies using neutralizing organs and in tertiary tissues, such as the skin and intestine, Ϫ Ϫ anti-CCL25 Ab, or CCR9 / TCR transgenic T cells have where they control lymphocyte migration during hemopoiesis, demonstrated a central role for CCL25/CCR9 in the generation of the small intestinal lymphocyte compartment (14–21). De- spite the importance of CCL25/CCR9 in small intestinal immu- *Immunology Section, and †Hematopoietic Stem Cell Laboratory, Stem Cell Center, Lund University, Lund, Sweden nity, the mechanisms underlying the selective and constitutive Received for publication July 22, 2005. Accepted for publication January 6, 2006. expression of CCL25 in the small intestine are not understood. The costs of publication of this article were defrayed in part by the payment of page In the current study, we have examined expression and charges. This article must therefore be hereby marked advertisement in accordance regulation of CCL25 in small intestinal epithelial cells. Our with 18 U.S.C. Section 1734 solely to indicate this fact. results demonstrate that CCL25 displays a unique pattern of 1 This work was supported by grants from the Swedish Medical Research Council; the regulation compared with other inflammatory or homeostatic Crafoordska, O¨ sterlund, Åke Wiberg, Richard and Ruth Julins, Nanna Svartz and Kocks Foundations; the Swedish Medical Society; the Royal Physiographic Society; chemokines and suggest a role for the Caudal-related homeobox the Swedish Foundation for Strategic Research “Microbes and Man” program and transcription factors in enhancing CCL25 promoter activity in INGVAR II program; and a Crohns and Colitis Foundation of America project grant (to W.A.). K.K. is supported by a SWEGENE Postdoctoral Fellowship. the small intestine. 2 A.E. and K.K. contributed equally to this work. 3 Address correspondence and reprint requests to Dr. Knut Kotarsky, Immunology Section, Lund University, Biomedical Centre I-13, S-22184 Lund, Sweden. E-mail address: [email protected] 4 Abbreviations used in this paper: LT, lymphotoxin; IEL, intraepithelial lymphocyte.

Materials and Methods 5Ј-GCGGAATTCACCATGTATGTGGGCTATGTGCTG-3Ј and anti- Ј Ј Mice sense 5 -GCGGAATTCTATGGCAGAAACTCCTCTTTCACA-3 , and for Cdx-2 were sense 5Ј-GCGGAATTCACCATGTACGTGAGCTAC Germfree or conventional Swiss Webster mice were from Taconic Farms, CTTCTG-3Ј and antisense 5Ј-GCGGAATTCACTGGGTGACAGTG and C57BL/6 mice were from the Microbiology, Immunology, and Gly- GAGTTTAAAACC-3Ј. Primers for ␤-actin were sense 5Ј-GGTGGGAAT cobiology animal facility and the Biomedical Centre animal facility (Lund GGGTCAGAAGGACT-3Ј and antisense 5Ј-CCACGCTCGGTCAG University, Lund, Sweden). Small intestinal tissue from athymic mice with GATCTTCAT-3Ј. Primers for CCL25 were sense 5Ј-ATAGGCAATA a truncated common cytokine receptor ␥-chain (CR␥Ϫ/Y nu/nu) and CR␥ϩ CACGCTACAAGC-3Ј and antisense 5Ј-GCGGAATTCGTCTTCAAAG nu/ϩ mice was provided by Dr. H. Ishikawa (Keio University School of GCACCTTGGGCATGG-3Ј. Primers for cytokeratin 18 were sense Medicine, Tokyo, Japan), tissue from TNFR1Ϫ/Ϫ mice was provided by 5Ј-AGATCGACAATGCCCGCCTTG-3Ј and antisense 5Ј-AGACTTG Dr. N. Lycke (University of Gothenburg, Gothenburg, Sweden), and tissue GTGGTGACAACTGT-3Ј. Primers for Madcam-1 were sense 5Ј- from LT␣Ϫ/Ϫ and LT␤Ϫ/Ϫ mice was provided by Dr. D. Finke (University CCTGAGTCTGAGGTAGCTGTGG-3Ј and 5Ј-GAGTGCCTGTGT of Lausanne, Lausanne, Switzerland). All animal studies were approved by GTCTGACAGCAT-3Ј antisense and for intestinal alkaline phosphatase the local ethical committee. sense 5Ј-GCCGTGAAAGTGCTAAGCAGG-3Ј and antisense 5Ј- GGTCAGAGTGTCGCGTTCACTA-3Ј. CCL25 mRNA levels were deter- Epithelial cell isolation and cell line culture mined by real-time RT-PCR as previously described (14). Epithelial cells were removed from the small intestine with EDTA. Briefly, the small intestine was rinsed with ice-cold PBS, inverted, and cut into Rapid amplification of cDNA ends 5-cm fragments. Intestinal fragments were then incubated in PBS contain- Mouse small intestinal cells were isolated by EDTA treatment. Total RNA ing 30 mM EDTA for 30 min at 37°C on a rotating platform and EDTA was prepared using Stratagene RNA Miniprep kit. For 5Ј RACE total RNA was changed every 5 min. Murine small intestinal epithelial crypts were was reverse transcribed onto magnetic beads (Dynal Biotech) using Su- isolated and cultured as described (22). RT-PCR for cytokeratin 18 ex- perscript III (Invitrogen Life Technologies) according to manufacturer’s pression was performed to confirm the epithelial identity of cultured cells instructions. The single-stranded cDNA was tailed with dATP on the 3Ј (data not shown). The murine small intestinal epithelial cell lines MODE-K end using terminal transferase EC 2.7.7.31 (Roche). The second strand was provided by Dr. P. Ernst (University of Texas Medical Branch, Galveston, synthesized with Pfx-polymerase (Invitrogen Life Technologies) and the TX) and S1-H10 provided by Dr. J. I. Gordon (Washington University following adaptor primer: 5Ј-GTCCGCGGCCGCGTAATACGACTCAC School of Medicine, St. Louis, MO), and the murine fibroblast cell line TATAGGGCGTTTTTTTTTTTTTTTTTTT-3Ј. The cDNA was removed BALB/3T3 provided by Dr. C. Owman (Lund University, Lund, Sweden) from the magnetic beads and subjected to PCR using the first gene-specific were maintained in DMEM (Invitrogen Life Technologies) supplemented primer 5Ј-GCGGAATTCTTTGATCCTGTGCTGGTAACCCAGG-3Ј and with 10% FCS (Sigma-Aldrich), nonessential amino acids (Invitrogen Life the adaptor primer. The product was used in a consecutive PCR with a Technologies), 1 mM sodium pyruvate (Invitrogen Life Technologies), and second gene-specific primer 5Ј-GCGGAATTCGTCTTCAAAGGCACCT 50 U/ml penicillin and 50 mg/ml streptomycin (Invitrogen Life Technol- TGGGCATGG-3Ј. The second PCR resulted in a product of 270 bp in ogies). The mouse intestinal mICc12 cells provided by Dr. A. Vandervalle length, which was cloned into pBluescript and seven obtained clones se- (Institut National de la Sante´ et de la Recherche Me´dicale, Faculte´ X, Paris, quenced using BigDye (Applied Biosystems). For the 3Ј RACE, 5 ␮gof France) were grown as described (23). Cells were grown to confluence in cDNA (prepared as previously described) was reverse transcribed using 24-well plates before addition of cytokines. Superscript III (Invitrogen Life Technologies). A PCR using the adaptor Ј Tissue sectioning and laser capture microscopy primer and a gene-specific primer (5 -CCGGCGGCCGCGGAGAAC CCCAACAGTACAAGCG-3Ј) resulted in one gene-specific RACE prod- Small intestinal tissue sections (10 ␮m) were cut from Tissue-Tek OCT uct, which was cloned into pBluescript and eight obtained clones were embedded tissue. The sections were placed on PEN membrane-coated sequenced. slides (P.A.L.M. Microlaser Technologies), and fixed in 70% ethanol for 30 s and acetone for 4 min. Fixed sections were stained with Harris he- Construction of plasmids containing mouse CCL25 promoter matoxylin for 10 s (Sigma-Aldrich), washed, overlaid with 10% DMSO, fragments and placed on dry ice or kept at Ϫ80°C until use. All aqueous solutions were made from diethyl pyrocarbonate-treated water and supplemented The different promoter regions were amplified by PCR using the following with vanadyl-ribonucleoside complex RNase Inhibitor (Sigma-Aldrich). reverse primers: PE1 5Ј-GCGCTCGAGCAATGCCTTTCTGGTCCT Laser capture was performed on a Zeiss microscope equipped with a mi- GAGAGCTGGT-3Ј for PF1 and PF2 or PE2 5Ј-GCGCTCGAGAAGGT crocatapulting laser system (P.A.L.M. Microlaser Technologies). Total TAGATCTCCTCTCCAGATACC-3Ј for PF3, and the following forward RNA was extracted from the catapulted samples using a Stratagene Mi- primers: 5Ј-GCGGAGCTCCGTATCACTCACTGCCCCACTGAAA croprep RNA kit. GTGT-3Ј for PF1 or 5Ј-GCGGAGCTCAAGGCCAGGACAGAGCAA GAGAGCAAGAA-3Ј for PF2 and PF3 on mouse genomic DNA. PCR Immunohistochemistry products were cut with SacI and XhoI, cloned into the pGL3-basic (Pro- mega), and their identity verified by sequencing. Different parts of the Acetone-fixed small intestinal tissue sections (8-␮m thick) were incubated promoter were deleted using restriction endonucleases. The HindIII diges- with 1% H O for 15 min to block endogenous peroxidase activity, followed 2 2 tion gave rise to the plasmids pGL3b-PF4 and pGL3b-PF5, whereas the by an avidin-biotin blocking kit (Vector Laboratories) to block endogenous PstI restriction resulted in the plasmid pGL3b-PF6. The constructs pGL3b- streptavidin-biotin activity. Sections were incubated for 1 h with anti-CCL25 PF7-9 were constructed by ligating double-stranded oligonucleotides in the (at 5 ␮g/ml, clone 89827; R&D Systems) or with an isotype control Ab. pGL3b vector. The following oligonucleotides were used: 5Ј-CTAGCTC Sections were then washed and incubated with biotinylated mouse anti-rat TATCTGAAGGAGAGAGAAGTCCCAAGTCTCACAGAGTGGC-3Ј; IgG2a secondary Ab (2.5 ␮g/ml, clone RG7/1.30; BD Biosciences), and the 5Ј-TCGAGCCACTCTGTGAGACTTGGGACTTCTCTCTCCTTCAG signal was amplified using streptavidin-HRP and tyramid-biotin (NEN/ ATAGAG-3Ј;5Ј-TCGAGAGGAGGAGCAGGAGGAGGGAGGG PerkinElmer) according to the manufacturer’s instructions and visualized AGAGAAGATAGGGGGCAGGTCA-3Ј;5Ј-GATCTGACCTGCCC using streptavidin-Alexa Fluor 488 (2 ␮g/ml; Molecular Probes). Nuclei were CCTATCTTCTCTCCCTCCCTCCTCCTGCTCCTCCTC-3Ј;5Ј-GATCT stained using 4Ј,6Ј-diamidino-2-phenylindole. GCAGGGTGGGGCTCTGACTATAAAGAATGAAGCCAGTTCACT mRNA analysis GA-3Ј; and 5Ј-AGCTTCAGTGAACTGGCTTCATTCTTTATAGT CGGAGCCCCACC CTGCA-3Ј. To construct the positive control plas- RNA isolation, cDNA synthesis, and semiquantitative RT-PCR were mids 3xCdxA and 3xCdxA Mutant, the following oligonucleotides were performed as previously described (17). The following primers were used cloned into pGL3-basic: 5Ј-TCGAGATTTATGCATTTATGATTTAT for RT-PCR: CXCL1 sense 5Ј-ATAGCCACACTCAAGAATGGTCG-3Ј GGGCCCTATAT-3Ј;5Ј-AGCTATATAGGGCCCATAAATCATAAAT and antisense 5Ј-CACCCTTCTACTAGCACAGTGG-3Ј; CCL20 sense GCATAAATC-3Ј;5Ј-TCGAGATCTATGCATCTATGATCTATGG GC 5Ј-ACTGTTGCCTCTCGTACATAC-3Ј and antisense 5Ј-GTGTCCAAT CCTATAT-3Ј; and 5Ј-AGCTATATAGGGCCCATAGATCATAGAT TCCATCCCAA-3Ј; CXCL13 sense 5Ј-CAGAATGAGGCTCAGCA GCATAGATC-3Ј. All constructs were verified by sequencing. CAGC-3Ј and antisense 5Ј-TCTCTTACTCACTGGAGCTT-3Ј; HGPRT (hypoxanthine-guanine phosphoribosyltransferase) sense 5Ј-CACAG Transfections and luciferase assay GACTAGAACACCTGC-3Ј and antisense 5Ј-GCTGGTGAAAAGGAC CTCT-3Ј. Primers for Cdx-1 were sense 5Ј-AGAGCGGCAGGTAAA Cells were transfected using Metafectene (Biontex) according to manufac- GATCT-3Ј and antisense 5Ј-CTACTCTCCAGAGCCAGTCT-3Ј, or sense turer’s recommendations, adding 1.5 ␮g of plasmid DNA complexed with 3644 TRANSCRIPTIONAL REGULATION OF CCL25

5 ␮l of Metafectene to 450,000 cells/well. For analysis of luciferase ac- paris5.fr/genatlas/͘), and transcription factors limited to cells of hemopoietic tivity, cells were harvested 24 h posttransfection by adding passive lysis origin were excluded. The alignment of the human and mouse core promoter buffer and stored at Ϫ80°C until analysis. Luciferase assays were per- sequence was performed using ClustalX (26). formed using the Dual Luciferase Assay kit (Promega) in a BMG LUMI- star Galaxy instrument. For cotransfection experiments using Cdx-1, 0.5 EMSA ␮g of pGL3-PF1 or pGL3-PF7 were cotransfected with 0.9 ␮gof The following dsDNA oligonucleotides were used: Cdx 5Ј-TCTGAC pcDNA3Cdx-1 plasmid (provided by Dr. P. Soubeyran, Institut National de TATAAAGAATGAAGCC-3Ј and mutant Cdx 5Ј-TCTGACTGGG la Sante´ et de la Recherche Me´dicale, Marseille, France) (24) or empty GAGAATGAAGCC-3Ј. The Cdx probe was 32P-labeled using T4 Polynu- pcDNA3 plasmid as control. For cotransfection experiments using Cdx-2, cleotide kinase (Invitrogen Life Technologies) according to the manufac- 0.5 ␮g of pGL3-PF1 or pGL3-PF7 were cotransfected with 0.5 ␮gof turer’s instructions. EMSA was performed in 15 ␮l of binding buffer (20 pTRE Cdx-2 (provided by Dr. T. Uesaka, Hiroshima University, Hiro- tight mM phosphate buffer (pH 6), 10 mM MgCl , 0.1 mM EDTA, 2 mM DTT, shima, Japan) (25) and 0.5 ␮g of pTet-ON (BD Biosciences). Cdx-2 ex- 2 0.01% Nonidet P-40, 0.1 mM NaCl, 100 ␮g/ml BSA, 4% Ficoll) contain- pression was initiated by adding doxycycline (Sigma-Aldrich) to a final ing 5–10 ␮g of nuclear protein extract, 2 ␮g of poly(dI:dC), and 20,000– concentration of 10 ␮M. All transfections performed included 50 ng of the 30,000 cpm of 32P-labeled DNA probe. Reactions were allowed to proceed pTK-RL Renilla plasmid (Promega) for the normalization of transfection for 30 min at room temperature. Complexes were separated on a 6% non- efficiencies. Expression of Cdx-1 and Cdx-2 mRNA was verified by RT- denaturing polyacrylamide gel, and the gels were dried and analyzed by PCR. To assess the effect of Cdx-1 and Cdx-2 on endogenous CCL25 autoradiography. For competition experiments, a 200-fold molar excess of mRNA expression, mICc12 cells were transfected with 1.5 ␮g of pcDNA3 unlabelled Cdx or mutant Cdx oligonucleotide was added before the ad- Cdx-1, 1.5 ␮g of pcDNA3Cdx-2, or 0.75 ␮g of both, or 1.5 ␮gof dition of labeled probe. For Ab blocking experiments, 2 ␮g of rabbit poly- pcDNA3-EGFP or of pcDNA3 as controls. Stable clones were selected in clonal anti-mouse Cdx Ab (CeMines) or 2 ␮g of rabbit serum (DakoCy- the presence of 250 ␮g/ml G-418 (Sigma-Aldrich) for 12 days. The per- tomation) as control was added, and samples were incubated for 30 min at centage of pcDNA3-EGFP transfected cells expressing enhanced GFP after room temperature before the addition of probe. this time ranged from 65 to 85%. CCL25 and GAPDH mRNA copy number were determined as earlier described. Statistical analysis Database analysis of transcription factor binding sites and Statistical analysis was performed using the Student’s unpaired t test with alignment of the mouse and human CCL25 promoters GraphPad InStat software. Putative transcription factor binding sites in the CCL25 promoter were iden- Results tified using Transcription Element Search System (TESS) (͗www.cbil. upenn.edu/tess/͘) and MatInspector (͗www.genomatix.de/cgi-bin/matinspector/ Epithelial cells are the major source of CCL25 mRNA in the matinspector.pl͘). All reported hits were sorted by their matrix similarity (for murine small intestine TESS according to TRANSFAC and for MatInspector according to MatIn- spector matrices). All hits with a threshold over 0.8 were considered relevant. CCL25 mRNA was constitutively and selectively expressed in the Factors predicted to interact with both human and murine sequences were murine small intestine and thymus (Fig. 1A) consistent with previous checked for their predicted tissue distribution using Geneatlas (͗www.dsi.univ- reports (9, 12, 14). In situ hybridization studies have demonstrated

FIGURE 1. Epithelial cells are the major source of CCL25 mRNA in the murine small intestine. A, CCL25 mRNA expression in different tissues as assessed by real-time RT-PCR. Data are mean Ϯ SEM (n ϭ 3–4 mice). B, CCL25 protein expression in small intestinal epithelium. Small intestinal tissue sections were stained with an anti-CCL25 or isotype control Ab, and cell nuclei were counterstained with 4Ј,6Ј-diamidino- 2-phenylindole. C, Hematoxylin staining of murine proximal small intestine (PSI) before and after treatment with EDTA. Location of crypt (c) .are indicated (ء) and villous (v) epithelial cells These cells are not present in EDTA-treated tissue (arrow). D, CCL25 mRNA expression in the intestine. Intestinal epithelial cells (IEC), proximal small intestine (PSI), and proximal small intestine after treatment with EDTA (PSI ϩ EDTA) are shown. CCL25 mRNA expression was determined by real-time RT-PCR. Data are mean (ϮSEM) for n ϭ 3 mice. E, Epithelial cells from crypt and villous region were isolated from small intestinal sections by laser capture microscopy. F, CCL25 mRNA expression in crypt and villous epithelium as assessed by real-time RT-PCR. Re- sults from one representative experiment of two performed. G, CCL25 mRNA expression in laser capture microscopy samples from EDTA-treated small intestinal tissue. cDNA derived from laser capture isolated lamina propria (lane 1); small intestine cDNA (lane 2); dH2O as template (lane 3) are shown. MadCAM-1 is expressed by microvascular endothelial cells, and intestinal alkaline phosphatase (IAP) is selectively expressed by small intestinal epithelial cells. The Journal of Immunology 3645

that epithelial cells constitutively express CCL25 mRNA (10, 12), and immunohistochemical staining of small intestinal sections with anti- CCL25 Ab showed predominant staining in small intestinal epithelial cells (Fig. 1B) (20). To determine the contribution of epithelial cells to total CCL25 mRNA levels in the murine small intestine, epithelial cells were removed from small intestinal tissue with EDTA, and CCL25 mRNA expression assessed by quantitative real-time RT- PCR. Hematoxylin staining of small intestinal sections confirmed that villous and crypt epithelium were effectively removed by this procedure (Fig. 1C). Small intestinal epithelial cells expressed CCL25 mRNA, whereas CCL25 mRNA was barely detected in intestinal tissue devoid of epithelial cells (Fig. 1D). FACS sorted CD8ϩ intraepithelial lymphocytes (IEL) failed to express CCL25 mRNA (data not shown), excluding the possibility that any contaminating IEL in epithelial cell preparations are a significant source of CCL25 mRNA. CCL25 mRNA was expressed at high levels by both crypt and villous epithelium (Fig. 1, E and F). Previous immunohistochemical studies have suggested that lamina propria microvascular endothelial cells are a potential additional source of intestinal CCL25 (11, 20). However, CCL25 mRNA was not detected in laser capture microscopy samples taken from the lamina propria of EDTA-treated small intestinal tissue that contained microvascular endothelial cells (as assessed by a positive signal for MadCAM-1 mRNA) (Fig. 1G), indicating that CCL25 detected on these cells may derive from epithelial cells. Indeed, cellular presentation of exogenous-derived chemokines has been previously described in other systems (27–29). FIGURE 2. Intestinal CCL25 mRNA expression is independent of LT Together, these results demonstrate that small intestinal epithe- and TNFR signaling, inflammatory stimuli, and the presence of intestinal lial cells are the major if not the sole source of the CCL25 mRNA microflora or lymphocytes. A, CCL25 mRNA levels in whole proximal in the murine small intestine. small intestine of LT␣Ϫ/Ϫ and LT␤Ϫ/Ϫ and TNFR1Ϫ/Ϫ mice as assessed by real-time RT-PCR. Results are the mean (ϮSEM) of three mice per group. CCL25 mRNA expression is not enhanced by inducers of B, LPS fails to enhance CCL25 mRNA expression in the murine small homeostatic or inflammatory chemokines intestine. CCL20 and CCL25 mRNA levels were determined by semiquan- Constitutive expression of homeostatic chemokines in secondary titative and real-time RT-PCR, respectively, 3 h after administration of 200 ␮ lymphoid organs and the small intestine is regulated by members g of LPS (Escherichia coli, serotype O55:B5; Sigma-Aldrich) i.v. into Ϯ of the LT/TNF family of cytokines (3–5). To determine whether C57BL/6 mice. For real-time RT-PCR results are the mean ( SEM) of three mice. For semiquantitative PCR, cDNA was serially diluted 1/10, and these factors are required for high CCL25 mRNA in the small results are representative of three mice per group. C, TNF-␣ and IFN-␥ fail intestine, CCL25 mRNA expression was examined in the small to enhance CCL25 mRNA expression in MODE-K cells. MODE-K cells ␣Ϫ/Ϫ ␤Ϫ/Ϫ Ϫ/Ϫ intestine of LT ,LT , or TNFR1 mice. CXCL13 were stimulated with TNF-␣ (100 ng/ml; PeproTech) and IFN-␥ (500 Ϫ/Ϫ mRNA expression was reduced in the small intestine of LT␣ U/ml; PeproTech) for 24 h and CXCL1 and CCL25 mRNA expression Ϫ Ϫ and LT␤ / mice (data not shown), consistent with a previous determined by semiquantitative and real-time RT-PCR, respectively. For report (5). In contrast, CCL25 mRNA was expressed at similar real-time RT-PCR results are the mean (ϮSEM) of triplicate wells and levels in the small intestine of LT␣Ϫ/Ϫ,LT␤Ϫ/Ϫ, and TNFR1Ϫ/Ϫ from one representative experiment of three performed. For semiquantita- mice as in wild-type mice (Fig. 2A). To determine whether intes- tive PCR, cDNA was serially diluted 1/10. Results are from one represen- tinal CCL25 mRNA was enhanced after exposure to inflammatory tative experiment of three performed. D, CCL25 mRNA expression in stimuli, mice were injected i.v. with LPS and intestinal CCL25 whole small intestine of athymic mice with a truncated common cytokine receptor ␥-chain (CR␥Ϫ/Ϫ) and germfree mice as determined by real-time mRNA levels determined 3 h later. This procedure has been shown RT-PCR. Results are representative from three mice per group (germfree to induce expression of multiple chemokine mRNA species in the and conventional (convent.)) or from three intestinal pieces from different small intestine, including CCL20, a chemokine constitutively ex- sites along the small intestine (athymic CR␥Ϫ/Ϫ mice). pressed by small intestinal epithelial cells (30, 31). As expected, injection of LPS i.v. enhanced levels of CCL20 mRNA in the murine small intestine (Fig. 2B); however, intestinal CCL25 failed to induce CCL25 mRNA expression in human HT-29 and mRNA levels remained unchanged (Fig. 2B). In a second set of Caco-2 colonic epithelial lines, and FHs 74 Int cells (CCL-241; experiments, we determined whether proinflammatory cytokines American Type Culture Collection), a morphologically epithelial- could enhance CCL25 mRNA expression in the small intestinal like cell line derived from the human small intestine (data not epithelial cell lines MODE-K, S1-H10, or mICcl2. TNF-␣ and shown). Together, these results indicate that transcriptional regu- IFN-␥ were chosen because these cytokines enhance inflammatory lation of CCL25 mRNA is unique compared with that of other chemokine expression in a wide range of intestinal epithelial lines homeostatic or inflammatory chemokines. including MODE-K cells (32–35). Addition of TNF-␣ (10–100 ng/ml) or IFN-␥ (10–500 U/ml) alone to confluent MODE-K, S1- Small intestinal CCL25 expression is independent of the H10, or mICcl2 epithelial cell layers for 6 and 24 h failed to alter presence of intestinal microflora or IEL CCL25 mRNA levels (data not shown). Furthermore, whereas The expression of CCL25 mRNA in the murine small intestine is TNF-␣ and IFN-␥ enhanced CXCL1 expression (Fig. 2C) as pre- increased between 2 and 3 wk of age (17), a time point correlating viously described (33), this cytokine combination failed to enhance with increased numbers of IEL (17) and bacterial colonization of CCL25 mRNA expression (Fig. 2C). Similarly these cytokines the intestine. Because the small intestinal epithelium is in intimate 3646 TRANSCRIPTIONAL REGULATION OF CCL25 contact with intestinal microflora and IEL, we determined whether full-length CCL25 mRNA transcript previously described in the presence of these components was important in maintaining mouse thymus (GenBank accession number NM_009138). Anal- high constitutive CCL25 expression in the murine small intestine. ysis of expressed sequence tags obtained in the mouse (GenBank) CCL25 mRNA was expressed at similar levels in the small intes- showed good agreement with these sequences (data not shown). tine of athymic mice with a truncated common cytokine receptor Eight clones obtained from 3Ј RACE were sequenced, four of ␥-chain (CR␥Ϫ/Y nu/nu) (Fig. 2D), which contains a negligible which showed identity to clone 1 and four to clone 2 (Fig. 3A). TCRϪ IEL population (36). Furthermore, coincubation of conflu- Thus, murine CCL25 mRNA appears to have two alternative poly- ent MODE-K monolayers for 6 h with 50–500,000 freshly isolated adenylation sites. Having identified mRNA ends of the murine syngeneic CD8ϩ IEL failed to enhance epithelial CCL25 expres- CCL25 message, we determined the genomic organization of the sion (data not shown). Finally, CCL25 mRNA was expressed at murine and human CCL25 gene (Fig. 3B). Both murine and human similar levels in the small intestine of germfree and conventional CCL25 consists of six exons covering 10.5 and 9.9 kb of genomic mice (Fig. 2D). Thus constitutive CCL25 mRNA expression in the DNA, respectively. Exon 1 of human CCL25 that was missing murine small intestine is independent on interactions with mature from the human CCL25 mRNA sequence (NM_005624) was iden- IEL, signaling through the cytokine receptor ␥-chain or the pres- tified from expressed sequence tags derived from the human small ence of intestinal bacteria. intestine (BX415301) (Fig. 3B).

Mapping of murine small intestinal epithelial CCL25 mRNA 5Ј Identification and characterization of the murine CCL25 and 3Ј ends and organization of the murine CCL25 gene promoter The findings suggesting that CCL25 expression is regulated in a To determine whether the region upstream of the predicted CCL25 manner different from other chemokines motivated us to search for transcriptional start site contained promoter activity, we generated regulatory elements involved in the transcriptional regulation of a series of constructs comprising regions surrounding exon 1 (Fig. this gene. To identify the putative CCL25 transcriptional start site, 4A) fused to a luciferase reporter gene. Next we attempted to iden- 5Ј RACE was performed on cDNA prepared from small intestinal tify a cell line that constitutively expressed high levels of CCL25 epithelial cells (Fig. 3A). Seven clones obtained from 5Ј RACE mRNA for use in transfection studies. Screening of a wide range were sequenced, three of which showed identity to clone A, three of murine and human epithelial cell lines including MODE-K, to clone B, and one to clone C (Fig. 3A). These transcripts are a mICc12, S1-H10, HT-29, FHs 74 Int, Caco-2, and T-84 cells failed few base pairs longer at the 5Ј end as compared with the presumed to identify any cell line constitutively expressing CCL25 mRNA levels comparable to that of freshly isolated epithelial cells (Fig. 4B and data not shown). Furthermore freshly isolated epithelial cells showed a dramatic reduction in CCL25 mRNA expression after culture (Fig. 4B). We therefore determined whether any of the constructs displayed basal promoter activity in murine MODE-K and mICc12 cells. The murine fibroblast line BALB/c 3T3 was used as a control cell line because these cells failed to express full-length CCL25 mRNA as assessed by RT-PCR using a 5Ј primer in exon 1 and 3Ј primer in exon 2 of CCL25 (data not shown). Constructs PF1 to PF3 increased luciferase activity ϳ20 times compared with the control in MODE-K and mICcl2 cells (Fig. 4C), but not in BALB/c 3T3 cells. Thus the area upstream of exon 1 contains the minimal CCL25 promoter, and intron 1 does not appear to contribute to promoter activity in these cell lines. Construct PF4 consistently induced greater luciferase expression compared with PF2, indicating that the region Ϫ583 to Ϫ311 contained elements that repressed promoter activity in these cell lines. In addition, PF6 failed to induce luciferase expression, demon- strating that a region between Ϫ45 and Ϫ167 is critical for promoter activity in these cells. Consistent with this finding PF7 covering region Ϫ141 to Ϫ5 had strong promoter activity, whereas PF8, covering region Ϫ99 to Ϫ5, and PF9, covering region Ϫ49 to Ϫ5, showed poor promoter activity (Fig. 4C). Constructs containing region Ϫ141 to Ϫ99 alone or region Ϫ141 to Ϫ99 fused to region Ϫ49 to Ϫ5 showed no promoter activity (data not shown). Together these results demonstrate that the region upstream of exon 1 of the CCL25 gene contains the minimal murine CCL25 promoter FIGURE 3. Genomic structure and transcriptional start site of the including both activating and repressive elements (Fig. 4D). mouse CCL25 gene. A, The transcriptional start site of the murine CCL25 transcript in small intestinal epithelial cells was identified by 5Ј RACE. The The CCL25 promoter contains putative binding sites for Cdx 3Ј RACE identified two different polyadenylation sites separated by 17 bp. transcription factors are indicated. The sequenced CCL25 (ء) The two sites of poly(A) addition To identify potential transcriptional factors that may contribute to cDNA has been deposited (GenBank accession number DQ158256). B, Schematic genomic organization of the mouse (GenBank accession num- the high CCL25 expression in normal small intestinal epithelial ber NT_039455.2) and human (NM_005624 and BX415301) CCL25 gene cells, we examined the CCL25 promoter for putative transcription locus. The exons are presented as boxes and numbered with roman nu- factor binding sites. A consensus TATA box could be identified merals (I-VI) and the introns are numbered using Arabic numerals (1–5). ϳ25 bp upstream of the transcriptional start site (Fig. 5). Addi- Indicates translated (f) and untranslated (Ⅺ) regions. tional binding sites were predicted for a number of transcription The Journal of Immunology 3647

FIGURE 4. Identiﬁcation of the CCL25 minimal promoter. A, Schematic represen- tation of the promoter constructs, their names, length and position in respect to the ﬁrst and second exon of the mouse CCL25 gene. B, CCL25 mRNA expression in freshly isolated murine small intestinal epithelial cells (SIEC), MODE-K and mICcl2 cells, and in primary small intestinal epithelial cells before and after culture as assessed by real-time RT-PCR. Results are representative from one experiment of three performed for the cell lines and from one experiment of two performed for cultured epithelial cells. C, Nine constructs of varying length were transiently transfected into mICc12, MODE-K, and BALB/3T3 cells. The empty plasmid pGL3-basic was used as a negative control and the cotransfected Renilla plasmid was used for stan- dardization. Data are mean Ϯ SD from one representative experiment of three to six p Ͻ 0.01 and ,ءء ,p Ͻ 0.05 ,ء .performed p Ͻ 0.001. D, Summary schematic of ,ءءء the CCL25 promoter depicting positive (ϩ) and negative (Ϫ) regulatory elements.

factors, including Cdx-1 and Cdx-2, Krueppel-like factor, GATA, 5). A single complex was observed when labeled Cdx probe was MAZ, and TFII-I (Fig. 5 and data not shown). Binding sites for incubated with nuclear extracts from primary small intestinal epithe- IRF-3 and classical NF-␬B dimers, such as RelA:p50 implicated in lial cells, and this complex was competed away with unlabeled probe, the induction of inflammatory chemokines, or binding sites for but not unlabeled mutant probe (Fig. 7A). The complex formation was nonclassical NF-␬B dimers RelB:p52, implicated in driving ho- also inhibited by the addition of anti-Cdx, but not control Ab, to the meostatic chemokine gene expression (7, 37–45), were not de- nuclear extracts before addition of labeled probe (Fig. 7A). In contrast, tected within a region covering Ϫ2000 to ϩ25 bp of the CCL25 this complex was not observed using nuclear extracts from MODE-K promoter, supporting the notion that these factors are not directly cells (Fig. 7A), mICcl2 cells, or the thymocyte cell line 2017 (data not involved in the regulation of CCL25 expression. shown) (53). Thus, Cdx is present in freshly isolated epithelial cells, but not MODE-K or mICcl2 cells, and epithelial cell derived Cdx can Cdx interacts with the predicted binding site in the CCL25 bind to the putative Cdx binding motif within the TATA box of the promoter in vitro CCL25 promoter. The presence of putative Cdx-1 and Cdx-2 binding sites within the murine CCL25 promoter was of interest because Cdx-1 and Cdx-2 expression is restricted to the gut epithelium in adult mice (46), Cdx-1 and Cdx-2 enhance activity of the CCL25 promoter in and these transcription factors have been implicated in regulating small intestinal epithelial cell lines expression of intestinal specific genes (47–52). Importantly, al- To determine whether Cdx could enhance the activity of the though freshly isolated small intestinal epithelial cells expressed CCL25 promoter in epithelial cell lines, the PF1 CCL25 promoter Cdx-1 and Cdx-2 mRNA (Fig. 6A), MODE-K and mICcl2 cells, construct or the PF7 CCL25 promoter construct (which contained and cultured primary small intestinal epithelial failed to express a single Cdx binding site corresponding to that predicted within the Cdx mRNA (Fig. 6). Thus expression of Cdx mRNA correlated TATA box) were cotransfected with expression plasmids encoding with the cells ability to express high levels of CCL25 mRNA. Cdx-1 or Cdx-2 into MODE-K cells. In initial experiments func- To determine whether small intestinal Cdx protein could interact tionality of the Cdx expression plasmids in MODE-K cells was with Cdx binding motifs within the CCL25 promoter, nuclear extracts confirmed using a promoter construct containing a TATA box pro- were prepared from primary small intestinal epithelial cells, ceeded by three consensus Cdx binding sites or three mutant Cdx MODE-K and mICcl2 cells, and incubated with a 32P-labeled probe sites, as a negative control, upstream of the luciferase gene. Co- covering the Cdx-binding site predicted within the TATA box (Fig. transfection with the Cdx expression plasmids leads to a 3-fold 3648 TRANSCRIPTIONAL REGULATION OF CCL25

FIGURE 5. Predicted transcription factor binding sites in the mouse CCL25 promoter. Alignment of the core promoter regions of the mouse and human CCL25 gene. Conserved -and mouse exons (un (ء) nucleotides derlined) are indicated, the transcriptional start site is marked (ϩ1), and putative relevant transcription factor binding sites are shown in italic and labeled. Restriction enzyme sites are shown in bold italic and labeled.

increase in luciferase activity from the reporter construct contain- CCL25 mRNA levels expressed Cdx-1 and Cdx-2, whereas epi- ing 3 CdxA consensus sites but not from the mutant reporter (data thelial cell lines and cultured primary epithelial cells that ex- not shown). MODE-K cells transfected with constructs encoding pressed low levels of CCL25 mRNA failed to express Cdx-1 and either Cdx-2 or Cdx-1 significantly enhanced CCL25 promoter Cdx-2. Finally EMSA and transfection studies suggested a role for activity compared with their relevant controls (Fig. 7, B and C). In Cdx transcription factors in contributing to the high CCL25 ex- contrast neither Cdx expression construct enhanced CCL25 pro- pression levels in small intestinal epithelial cells. moter activity in BALB/c 3T3 cells (data not shown). Furthermore, Chemokines have been broadly separated into homeostatic and stable ectopic expression of Cdx-2, or the combination of Cdx-1 inflammatory chemokines with partially distinct mechanisms of ␣ ␤ ␤ and Cdx-2 in mICc12 cells, leads to an increase in endogenous regulation. The LT 1 2 heterotrimer binds to the LT R (3), sig- CCL25 mRNA expression in these cells (Fig. 7D). Together these naling through which is required for maintaining constitutive ex- results suggest a role for Cdx transcription factors in enhancing pression of the homeostatic chemokines CXCL12, CCL21, CCL25 mRNA transcription in small intestinal epithelial cells; CCL19, and CXCL13 in lymph nodes, spleen, and intestine (4, 5) however, because transfection with Cdx failed to enhance CCL25 as well as the epithelial-derived chemokine CCL20 (37). LT␤R mRNA expression to levels observed in freshly isolated small in- mediated expression of homeostatic chemokines functions through testinal epithelial cells, they also suggest that other factors in ad- the alternative NF-␬B pathway, involving translocation of RelB: dition to Cdx are responsible for maintaining the high constitutive p52 dimers into the nucleus (44). In this study we show that in expression of CCL25 in these cells. contrast to homeostatic chemokine promoters (37, 43, 44, 54), putative binding sites for RelB:p52 dimers are not present in the Discussion CCL25 promoter and that small intestinal CCL25 mRNA expres- Despite the importance of CCL25 in small intestinal immunity sion is independent of LT␣ and LT␤ signaling. In addition, tran- (14–20), the mechanisms underlying the tissue selective expres- scriptional binding sites for NF-␬B and IRF-3, both of which are sion and regulation of this chemokine in the small intestine are involved in the induction of inflammatory chemokines (40, 45, 54), unknown. In the present study we demonstrate that epithelial cells were not present in the CCL25 promoter, and intestinal epithelial are the major source of CCL25 mRNA in the small intestine, and CCL25 mRNA levels remained unaltered in two inflammatory that constitutive CCL25 mRNA expression is independent of the models known to induce epithelial expression of inflammatory presence of intestinal bacteria and lymphocytes. CCL25 expres- chemokines (30–35). This inability of inflammatory mediators to sion was not regulated by the LTR␤ and TNFR1 signaling path- induce CCL25 mRNA expression, is in apparent odds with the ways and was not enhanced by inflammatory mediators suggesting enhanced CCL25 expression reported in small intestinal crypts of a unique pattern of regulation compared with homeostatic and in- Crohn’s disease patients in areas of lymphocytic infiltration; how- flammatory chemokines, respectively. The CCL25 promoter con- ever, no quantitative data were presented in this study (55). To- tained several putative binding sites for Cdx transcription factors gether, our results demonstrate that CCL25 is not regulated as and primary small intestinal epithelial cells that expressed high other homeostatic or inflammatory chemokines. The Journal of Immunology 3649

FIGURE 6. Small intestinal epithelial cell lines and cultured primary epithelial cells express reduced levels of Cdx. Cdx-1 and Cdx-2 expression in small intestinal epithelial cells (SIEC) (A) and in mICcl2 and MODE-K cells (B), as determined by RT-PCR. Results are representative of one experiment from two performed for the cultured cells and one from three performed on the cell lines.

We were unable to identify a cell line that could model freshly isolated small intestinal epithelial cells in their levels of CCL25 mRNA expression. Nevertheless constructs encompassing the area immediately upstream of exon 1 of the CCL25 gene showed promoter activity in MODE-K and mICc12 cells. Constructs lacking region Ϫ311 to Ϫ511 showed enhanced promoter activity, indicating that this area contained potential suppressor elements. We were also able to identify important activating elements required for minimal promoter activity in these cells. A region between Ϫ45 to Ϫ167 appeared critical for minimal promoter activity, and constructs containing region Ϫ99 to Ϫ141 had significantly higher promoter activity than constructs lacking this sequence. Cdx-1 and Cdx-2 are clearly not involved in driving this activity because MODE-K and mICc12 cells failed to express these transcription factors and the region Ϫ99 to Ϫ141 contained no Cdx binding sites. However, two binding sites for the TFII family of general transcription factors, which make up part of the initiation complex involved in gene transcription (56), were located within this region and are thus likely to be critical for minimal promoter activity in these cells. FIGURE 7. Cdx proteins bind to the CCL25 promoter and enhance pro- The far higher levels of CCL25 mRNA transcription in primary moter activity in small intestinal epithelial cell lines. A, EMSA is per- small intestinal epithelial cells compared with epithelial cell lines formed using nuclear protein extracts from small intestinal epithelial cells suggested that critical components required for driving CCL25 (SIEC) and MODE-K cells as indicated. Labeled probe and nuclear extract promoter activity were missing or not functional within the cell only (lanes 1 and 6), with excess of unlabeled probe (lanes 2 and 7), with lines. Furthermore, because cultured primary small intestinal epi- excess of mutant probe (lanes 3 and 8), anti-Cdx Ab (lanes 4 and 9), and thelial cells showed reduced CCL25 transcription, the intestinal control Ab (lanes 5 and 10). Results are representative from one experi- environment appears important in maintaining the expression ment of three performed. Cotransfection of Cdx-2 (pTREtightCdx-2) (B)or and/or activity of these components. Several results from the cur- Cdx-1 (C) together with the pGL3b-PF1 or pGL3b-PF7 increased luciferase activity in MODE-K cells. Results are from one representative ex- rent study suggest that Cdx transcription factors are one such com- ;p Ͻ 0.01 ,ءء .periment of three performed. Bars indicate mean Ϯ SEM p Ͻ 0.0001. D, Transfection of Cdx-2, or Cdx-1 and Cdx-2 enhanced ,ءءءء ponent. Firstly analysis of the CCL25 promoter predicted several binding sites for Cdx proteins. Secondly, Cdx was expressed by the endogenous CCL25 transcript level. Stable transfectants of mICc12 primary small intestinal epithelial cells but not cultured primary cells expressing Cdx-2, Cdx-1, or both were generated as described in epithelial cells or epithelial cell lines. Thirdly, Cdx present in nu- Materials and Methods. The endogenous CCL25 mRNA level was deter- clear extracts from freshly isolated small intestinal epithelial cells, mined by quantitative real-time PCR. Data are mean Ϯ SEM (n ϭ 3–4 .p Ͻ 0.01 ,ءء ;p Ͻ 0.05 ,ء .(which has previously been shown to interact with TATA boxes of wells several intestinal specific genes (49–52), bound to the Cdx site predicted in the CCL25 TATA box. Finally, transfection of Cdx expression plasmids caused a significant enhancement CCL25 pro- ditional factors other than Cdx must be involved in maintaining moter construct activity in MODE-K and mICc12 cells and en- high CCL25 transcription levels in these cells. Consistent with hanced CCL25 mRNA expression in mICc12 cells. this, Cdx transcription factors are expressed by small intestinal and Nevertheless, because transfection of epithelial cell lines with colonic epithelium (46), and colonic epithelial cells express ϳ50 Cdx expression plasmids failed to induce CCL25 mRNA expres- times less CCL25 mRNA than small intestinal epithelial cells (Fig. sion to the levels observed in small intestinal epithelial cells, ad- 1A). The mechanisms driving selective gene expression in the 3650 TRANSCRIPTIONAL REGULATION OF CCL25 small intestine vs colon are poorly understood. For the small in- tracts double- and single-positive thymocytes expressing the TECK receptor testinal specific gene sucrase isomaltase, tissue selectivity is CCR9. Eur. J. Immunol. 30: 262–271. 13. Zabel, B. A., W. W. Agace, J. J. Campbell, H. M. Heath, D. Parent, A. I. Roberts, achieved through the combination of HNF-1␣, GATA-4, and E. C. Ebert, N. Kassam, S. Qin, M. Zovko, et al. 1999. Human G protein-coupled Cdx-2 (48), together with active suppression of this gene in the receptor GPR-9-6/CC chemokine receptor 9 is selectively expressed on intestinal homing T lymphocytes, mucosal lymphocytes, and thymocytes and is required colon (57). Promoter analysis using MatInspector predicted a bind- for thymus-expressed chemokine-mediated chemotaxis. J. Exp. Med. 190: ing site for this repressor (CLOX/CDP) within the CCL25 pro- 1241–1256. moter. Similarly, regulation of the chemokine CCL20 in colonic 14. Svensson, M., J. Marsal, A. Ericsson, L. Carramolino, T. Broden, G. Marquez, ␬ and W. W. Agace. 2002. CCL25 mediates the localization of recently activated epithelial cells is obtained through the combined effects of NF- B, CD8␣␤ϩ lymphocytes to the small-intestinal mucosa. J. Clin. Invest. 110: Sp1, and ESE-1, an enterocyte specific transcription factor (42). In 1113–1121. this regard, the CCL25 promoter contains putative binding sites for 15. Wurbel, M. A., M. Malissen, D. Guy-Grand, E. Meffre, M. C. Nussenzweig, M. Richelme, A. Carrier, and B. Malissen. 2001. Mice lacking the CCR9 CC- Krueppel-like factors, a family of transcription factors implicated chemokine receptor show a mild impairment of early T- and B-cell development in driving expression of intestinal genes (58), indicating that and a reduction in T-cell receptor ␥␦ϩ gut intraepithelial lymphocytes. Blood 98: CCL25 may belong to an intestine restricted gene battery. We also 2626–2632. 16. Uehara, S., A. Grinberg, J. M. Farber, and P. E. Love. 2002. A role for CCR9 in detected potential binding sites for broadly expressed transcription T lymphocyte development and migration. J. Immunol. 168: 2811–2819. factors such as GATA, MAZ, Sp1, and TFII-I (56, 59–63) and 17. Marsal, J., M. Svensson, A. Ericsson, A. H. Iranpour, L. Carramolino, although they are unlikely to directly participate in the tissue spe- G. Marquez, and W. W. Agace. 2002. Involvement of CCL25 (TECK) in the generation of the murine small-intestinal CD8␣␣ϩCD3ϩ intraepithelial lympho- cific gene regulation of CCL25, they may play important roles in cyte compartment. Eur. J. Immunol. 32: 3488–3497. overall promoter function. Of note, CCL25 was recently identified 18. Pabst, O., L. Ohl, M. Wendland, M. A. Wurbel, E. Kremmer, B. Malissen, and as an Egr-1 target gene in human endothelial cells after expression R. Forster. 2004. Chemokine receptor CCR9 contributes to the localization of plasma cells to the small intestine. J. Exp. Med. 199: 411–416. of this transcriptional activator in these cells (64). However, we 19. Johansson-Lindbom, B., M. Svensson, M. A. Wurbel, B. Malissen, G. Ma´rquez, failed to identify Egr-1 binding sites within the CCL25 promoter and W. Agace. 2003. Selective generation of gut tropic T cells in gut-associated indicating that this effect may be mediated indirectly through the lymphoid tissue (GALT): requirement for GALT dendritic cells and adjuvant. J. Exp. Med. 198: 963–969. activation of additional transcriptional activators. 20. Hieshima, K., Y. Kawasaki, H. Hanamoto, T. Nakayama, D. Nagakubo, In conclusion, our results demonstrate that CCL25 displays a A. Kanamaru, and O. Yoshie. 2004. CC chemokine ligands 25 and 28 play es- unique regulation pattern for chemokines, and points to a complex sential roles in intestinal extravasation of IgA antibody-secreting cells. J. Immu- nol. 173: 3668–3675. network of transcription factors, involving Cdx, responsible for 21. Onai, N., M. Kitabatake, Y. Y. Zhang, H. Ishikawa, S. Ishikawa, and maintaining the high and tissue selective expression of CCL25 in K. Matsushima. 2002. Pivotal role of CCL25 (TECK)-CCR9 in the formation of gut cryptopatches and consequent appearance of intestinal intraepithelial T lym- the murine small intestinal epithelium. phocytes. Int. Immunol. 14: 687–694. 22. Booth, C., J. A. O’Shea, and C. S. Potten. 1999. Maintenance of functional stem Disclosures cells in isolated and cultured adult intestinal epithelium. Exp. Cell Res. 249: The authors have no financial conflict of interest. 359–366. 23. Bens, M., A. Bogdanova, F. Cluzeaud, L. Miquerol, S. Kerneis, J. P. Kraehenbuhl, A. Kahn, E. Pringault, and A. Vandewalle. 1996. Transim- References mortalized mouse intestinal cells (m-ICc12) that maintain a crypt phenotype. 1. Kunkel, E. J., D. J. Campbell, and E. C. Butcher. 2003. Chemokines in lympho- Am. J. Physiol. 270(6 Pt. 1): C1666–C1674. cyte trafficking and intestinal immunity. Microcirculation 10: 313–323. 24. Soubeyran, P., F. Andre, J. C. Lissitzky, G. V. Mallo, V. Moucadel, 2. Moser, B., M. Wolf, A. Walz, and P. Loetscher. 2004. Chemokines: multiple M. Roccabianca, H. Rechreche, J. Marvaldi, I. Dikic, J. C. Dagorn, and levels of leukocyte migration control. Trends Immunol. 25: 75–84. J. L. Iovanna. 1999. Cdx1 promotes differentiation in a rat intestinal epithelial cell 3. Mebius, R. E. 2003. Organogenesis of lymphoid tissues. Nat. Rev. Immunol. 3: line. Gastroenterology 117: 1326–1338. 292–303. 25. Uesaka, T., and N. Kageyama. 2004. Cdx2 homeodomain protein regulates the 4. Ngo, V. N., H. Korner, M. D. Gunn, K. N. Schmidt, D. S. Riminton, expression of MOK, a member of the mitogen-activated protein kinase super- M. D. Cooper, J. L. Browning, J. D. Sedgwick, and J. G. Cyster. 1999. Lym- family, in the intestinal epithelial cells. FEBS Lett. 573: 147–154. ␣␤ photoxin and tumor necrosis factor are required for stromal cell expression of 26. Jeanmougin, F., J. D. Thompson, M. Gouy, D. G. Higgins, and T. J. Gibson. homing chemokines in B and T cell areas of the spleen. J. Exp. Med. 189: 1998. Multiple sequence alignment with Clustal X. Trends Biochem. Sci. 23: 403–412. 403–405. 5. Kang, H. S., R. K. Chin, Y. Wang, P. Yu, J. Wang, K. A. Newell, and Y. X. Fu. 27. Middleton, J., S. Neil, J. Wintle, I. Clark-Lewis, H. Moore, C. Lam, M. Auer, 2002. Signaling via LT␤R on the lamina propria stromal cells of the gut is re- E. Hub, and A. Rot. 1997. Transcytosis and surface presentation of IL-8 by quired for IgA production. Nat. Immunol. 3: 576–582. venular endothelial cells. Cell 91: 385–395. 6. Ogawa, H., M. Iimura, L. Eckmann, and M. F. Kagnoff. 2004. Regulated pro- 28. Palframan, R. T., S. Jung, G. Cheng, W. Weninger, Y. Luo, M. Dorf, duction of the chemokine CCL28 in human colon epithelium. Am. J. Physiol. D. R. Littman, B. J. Rollins, H. Zweerink, A. Rot, and U. H. von Andrian. 2001. 287: G1062–G1069. Inflammatory chemokine transport and presentation in HEV: a remote control 7. Homey, B., H. Alenius, A. Muller, H. Soto, E. P. Bowman, W. Yuan, L. McEvoy, mechanism for monocyte recruitment to lymph nodes in inflamed tissues. J. Exp. A. I. Lauerma, T. Assmann, E. Bunemann, et al. 2002. CCL27-CCR10 interac- Med. 194: 1361–1373. tions regulate T cell-mediated skin inflammation. Nat. Med. 8: 157–165. 8. Romagnani, P., F. Annunziato, E. Lazzeri, L. Cosmi, C. Beltrame, L. Lasagni, 29. Baekkevold, E. S., T. Yamanaka, R. T. Palframan, H. S. Carlsen, F. P. Reinholt, G. Galli, M. Francalanci, R. Manetti, F. Marra, et al. 2001. Interferon-inducible U. H. von Andrian, P. Brandtzaeg, and G. Haraldsen. 2001. The CCR7 ligand elc protein 10, monokine induced by interferon ␥, and interferon-inducible T-cell ␣ (CCL19) is transcytosed in high endothelial venules and mediates T cell recruit- chemoattractant are produced by thymic epithelial cells and attract T-cell receptor ment. J. Exp. Med. 193: 1105–1112. (TCR) ␣␤ϩ CD8ϩ single-positive T cells, TCR␥␦ϩ T cells, and natural killer- 30. Tanaka, Y., T. Imai, M. Baba, I. Ishikawa, M. Uehira, H. Nomiyama, and type cells in human thymus. Blood 97: 601–607. O. Yoshie. 1999. Selective expression of liver and activation-regulated chemo- 9. Vicari, A. P., D. J. Figueroa, J. A. Hedrick, J. S. Foster, K. P. Singh, S. Menon, kine (LARC) in intestinal epithelium in mice and humans. Eur. J. Immunol. 29: N. G. Copeland, D. J. Gilbert, N. A. Jenkins, K. B. Bacon, and A. Zlotnik. 1997. 633–642. TECK: a novel CC chemokine specifically expressed by thymic dendritic cells 31. Widney, D. P., Y. R. Xia, A. J. Lusis, and J. B. Smith. 2000. The murine che- and potentially involved in T cell development. Immunity 7: 291–301. mokine CXCL11 (IFN-inducible T cell ␣ chemoattractant) is an IFN-␥- and 10. Kunkel, E. J., J. J. Campbell, G. Haraldsen, J. Pan, J. Boisvert, A. I. Roberts, lipopolysaccharide-inducible glucocorticoid-attenuated response gene expressed E. C. Ebert, M. A. Vierra, S. B. Goodman, M. C. Genovese, et al. 2000. Lym- in lung and other tissues during endotoxemia. J. Immunol. 164: 6322–6331. phocyte CC chemokine receptor 9 and epithelial thymus-expressed chemokine 32. Song, F., K. Ito, T. L. Denning, D. Kuninger, J. Papaconstantinou, W. Gourley, (TECK) expression distinguish the small intestinal immune compartment: epi- G. Klimpel, E. Balish, J. Hokanson, and P. B. Ernst. 1999. Expression of the thelial expression of tissue-specific chemokines as an organizing principle in neutrophil chemokine KC in the colon of mice with enterocolitis and by intestinal regional immunity. J. Exp. Med. 192: 761–768. epithelial cell lines: effects of flora and proinflammatory cytokines. J. Immunol. 11. Papadakis, K. A., J. Prehn, V. Nelson, L. Cheng, S. W. Binder, P. D. Ponath, 162: 2275–2280. D. P. Andrew, and S. R. Targan. 2000. The role of thymus-expressed chemokine 33. Izadpanah, A., M. B. Dwinell, L. Eckmann, N. M. Varki, and M. F. Kagnoff. and its receptor CCR9 on lymphocytes in the regional specialization of the mu- 2001. Regulated MIP-3␣/CCL20 production by human intestinal epithelium: cosal immune system. J. Immunol. 165: 5069–5076. mechanism for modulating mucosal immunity. Am. J. Physiol. 280: G710–G719. 12. Wurbel, M. A., J. M. Philippe, C. Nguyen, G. Victorero, T. Freeman, 34. Dwinell, M. B., N. Lugering, L. Eckmann, and M. F. Kagnoff. 2001. Regulated P. Wooding, A. Miazek, M. G. Mattei, M. Malissen, B. R. Jordan, et al. 2000. The production of interferon-inducible T-cell chemoattractants by human intestinal chemokine TECK is expressed by thymic and intestinal epithelial cells and at- epithelial cells. Gastroenterology 120: 49–59. The Journal of Immunology 3651

35. Berin, M. C., M. B. Dwinell, L. Eckmann, and M. F. Kagnoff. 2001. Production box by intestine-specific homeodomain proteins CDX1 and CDX2. Nucleic Acids of MDC/CCL22 by human intestinal epithelial cells. Am. J. Physiol. 280: Res. 31: 5238–5246. G1217–G1226. 50. Lambert, M., S. Colnot, E. Suh, F. L’Horset, C. Blin, M. E. Calliot, 36. Oida, T., K. Suzuki, M. Nanno, Y. Kanamori, H. Saito, E. Kubota, S. Kato, M. Raymondjean, M. Thomasset, P. G. Traber, and C. Perret. 1996. cis-Acting M. Itoh, S. Kaminogawa, and H. Ishikawa. 2000. Role of gut cryptopatches in elements and transcription factors involved in the intestinal specific expression of early extrathymic maturation of intestinal intraepithelial T cells. J. Immunol. 164: the rat calbindin-D9K gene: binding of the intestine-specific transcription factor 3616–3626. Cdx-2 to the TATA box. Eur. J. Biochem. 236: 778–788. 37. Rumbo, M., F. Sierro, N. Debard, J. P. Kraehenbuhl, and D. Finke. 2004. Lym- 51. Barley, N. F., S. R. Prathalingam, P. Zhi, S. Legon, A. Howard, and J. R. Walters. ␤ photoxin receptor signaling induces the chemokine CCL20 in intestinal epi- 1999. Factors involved in the duodenal expression of the human calbindin-D9k thelium. Gastroenterology 127: 213–223. gene. Biochem. J. 341: 491–500. 38. Sugita, S., T. Kohno, K. Yamamoto, Y. Imaizumi, H. Nakajima, T. Ishimaru, and 52. Suh, E., Z. Wang, G. P. Swain, M. Tenniswood, and P. G. Traber. 2001. Clusterin T. Matsuyama. 2002. Induction of macrophage-inflammatory protein-3␣ gene ␬ gene transcription is activated by caudal-related homeobox genes in intestinal expression by TNF-dependent NF- B activation. J. Immunol. 168: 5621–5628. epithelium. Am. J. Physiol. 280: G149–G156. 39. Wickremasinghe, M. I., L. H. Thomas, C. M. O’Kane, J. Uddin, and 53. Petersson, K., F. Ivars, and M. Sigvardsson. 2002. The pT␣ promoter and en- J. S. Friedland. 2004. Transcriptional mechanisms regulating alveolar epithelial hancer are direct targets for transactivation by E box-binding proteins. Eur. J. Im- cell-specific CCL5 secretion in pulmonary tuberculosis. J. Biol. Chem. 279: munol. 32: 911–920. 27199–27210. 40. Ohmori, Y., and T. A. Hamilton. 1993. Cooperative interaction between inter- 54. Weih, F., and J. Caamano. 2003. Regulation of secondary lymphoid organ de- ␬ feron (IFN) stimulus response element and ␬B sequence motifs controls IFN␥- velopment by the nuclear factor- B signal transduction pathway. Immunol. Rev. and lipopolysaccharide-stimulated transcription from the murine IP-10 promoter. 195: 91–105. J. Biol. Chem. 268: 6677–6688. 55. Papadakis, K. A., J. Prehn, S. T. Moreno, L. Cheng, E. A. Kouroumalis, R. Deem, 41. Matsukura, S., F. Kokubu, H. Kuga, M. Kawaguchi, K. Ieki, M. Odaka, T. Breaverman, P. D. Ponath, D. P. Andrew, P. H. Green, et al. 2001. CCR9- S. Suzuki, S. Watanabe, H. Takeuchi, M. Adachi, et al. 2003. Differential regu- positive lymphocytes and thymus-expressed chemokine distinguish small bowel lation of eotaxin expression by IFN-␥ in airway epithelial cells. J. Allergy Clin. from colonic Crohn’s disease. Gastroenterology 121: 246–254. Immunol. 111: 1337–1344. 56. Roy, A. L. 2001. Biochemistry and biology of the inducible multifunctional tran- 42. Kwon, J. H., S. Keates, S. Simeonidis, F. Grall, T. A. Libermann, and scription factor TFII-I. Gene 274: 1–13. A. C. Keates. 2003. ESE-1, an enterocyte-specific Ets transcription factor, reg- 57. Boudreau, F., E. H. Rings, G. P. Swain, A. M. Sinclair, E. R. Suh, D. G. Silberg, ulates MIP-3␣ gene expression in Caco-2 human colonic epithelial cells. J. Biol. R. H. Scheuermann, and P. G. Traber. 2002. A novel colonic repressor element Chem. 278: 875–884. regulates intestinal gene expression by interacting with Cux/CDP. Mol. Cell. Biol. 43. Bonizzi, G., M. Bebien, D. C. Otero, K. E. Johnson-Vroom, Y. Cao, D. Vu, 22: 5467–5478. A. G. Jegga, B. J. Aronow, G. Ghosh, R. C. Rickert, and M. Karin. 2004. Ac- 58. Hinnebusch, B. F., A. Siddique, J. W. Henderson, M. S. Malo, W. Zhang, tivation of IKK␣ target genes depends on recognition of specific ␬B binding sites C. P. Athaide, M. A. Abedrapo, X. Chen, V. W. Yang, and R. A. Hodin. 2004. by RelB:p52 dimers. EMBO J. 23: 4202–4210. Enterocyte differentiation marker intestinal alkaline phosphatase is a target gene 44. Dejardin, E., N. M. Droin, M. Delhase, E. Haas, Y. Cao, C. Makris, Z. W. Li, of the gut-enriched Kruppel-like factor. Am. J. Physiol. 286: G23–G30. ␤ M. Karin, C. F. Ware, and D. R. Green. 2002. The lymphotoxin- receptor 59. Song, J., H. Ugai, H. Nakata-Tsutsui, S. Kishikawa, E. Suzuki, T. Murata, and ␬ induces different patterns of gene expression via two NF- B pathways. Immunity K. K. Yokoyama. 2003. Transcriptional regulation by zinc-finger proteins Sp1 17: 525–535. and MAZ involves interactions with the same cis-elements. Int. J. Mol. Med. 11: 45. Sakaguchi, S., H. Negishi, M. Asagiri, C. Nakajima, T. Mizutani, A. Takaoka, 547–553. K. Honda, and T. Taniguchi. 2003. Essential role of IRF-3 in lipopolysaccharide- 60. Williams, L. J., and A. B. Abou-Samra. 2000. The transcription factors SP1 and induced interferon-␤ gene expression and endotoxin shock. Biochem. Biophys. MAZ regulate expression of the parathyroid hormone/parathyroid hormone-re- Res. Commun. 306: 860–866. lated peptide receptor gene. J. Mol. Endocrinol. 25: 309–319. 46. Duprey, P., K. Chowdhury, G. R. Dressler, R. Balling, D. Simon, J. L. Guenet, and P. Gruss. 1988. A mouse gene homologous to the Drosophila gene caudal is 61. Okamoto, S., K. Sherman, G. Bai, and S. A. Lipton. 2002. Effect of the ubiquitous expressed in epithelial cells from the embryonic intestine. Genes Dev. 2: transcription factors, SP1 and MAZ, on NMDA receptor subunit type 1 (NR1) 1647–1654. expression during neuronal differentiation. Brain Res. Mol. Brain Res. 107: 47. Beck, F. 2004. The role of Cdx genes in the mammalian gut. Gut 53: 1394–1396. 89–96. 48. Boudreau, F., E. H. Rings, H. M. van Wering, R. K. Kim, G. P. Swain, 62. Patient, R. K., and J. D. McGhee. 2002. The GATA family (vertebrates and S. D. Krasinski, J. Moffett, R. J. Grand, E. R. Suh, and P. G. Traber. 2002. invertebrates). Curr. Opin. Genet. Dev. 12: 416–422. Hepatocyte nuclear factor-1␣, GATA-4, and caudal related homeodomain protein 63. Duncan, D. D., A. Stupakoff, S. M. Hedrick, K. B. Marcu, and G. Siu. 1995. A Cdx2 interact functionally to modulate intestinal gene transcription: implication Myc-associated zinc finger protein binding site is one of four important functional for the developmental regulation of the sucrase-isomaltase gene. J. Biol. Chem. regions in the CD4 promoter. Mol. Cell. Biol. 15: 3179–3186. 277: 31909–31917. 64. Fu, M., X. Zhu, J. Zhang, J. Liang, Y. Lin, L. Zhao, M. U. Ehrengruber, and 49. Gautier-Stein, A., C. Domon-Dell, A. Calon, I. Bady, J. N. Freund, G. Mithieux, Y. E. Chen. 2003. Egr-1 target genes in human endothelial cells identified by and F. Rajas. 2003. Differential regulation of the glucose-6-phosphatase TATA microarray analysis. Gene 315: 33–41.