Lymphotoxin β receptor signaling induces the chemokine CCL20 in intestinal epithelium

GASTROENTEROLOGY 2004;127:213–223

Lymphotoxin ␤ Receptor Signaling Induces the Chemokine CCL20 in Intestinal Epithelium MARTIN RUMBO,* FRE´DE´RIC SIERRO,*,‡ NATHALIE DEBARD,* JEAN–PIERRE KRAEHENBUHL,*,‡ and DANIELA FINKE*,‡ *Swiss Institute for Experimental Cancer Research, Lausanne Branch, Epalinges; and ‡Institute of Biochemistry, University of Lausanne, Epalinges, Switzerland

Background & Aims: The follicle-associated epithelium (FAE) that overlies Peyer’s patches (PPs) exhibits distinct features compared with the adjacent villus epithelium. Besides the presence of antigen-sampling membranous M cells and the down-regulation of digestive functions, it constitutively expresses the chemokine CCL20. The mechanisms that induce FAE differentiation and CCL20 expression are poorly understood. The aim of this work was to test whether lymphotoxin ␤ receptor signaling (LT␤R), which plays a central role in PPs’ organogenesis, mediates CCL20 gene expression in intestinal epithelial cells. Methods: CCL20, lymphotoxin ␤ (LT␤) and LT␤R expression were monitored during embryonic development by in situ hybridization of mouse intestine. The human intestinal epithelial cell line T84 was used to study CCL20 expression following LT␣1/␤2 stimulation. In vivo CCL20 expression following agonistic anti-LT␤R antibody treatment was studied by laser microdissection and quantitative RT-PCR. Results: CCL20 was expressed in the FAE before birth at the time when the first hematopoietic CD4ⴙCD3ⴚ appeared in the PP anlage. LT␤R was expressed in the epithelium during PP organogenesis, making it a putative target for LT␣1␤2signals. In vitro, CCL20 was induced in T84 cells upon LT␤R signaling, either using an agonistic ligand or anti-LT␤ receptor agonistic antibody. LT␣1␤2-induced CCL20 expression was found to be NF-␬B dependent. LT␤R signaling upregulated CCL20 expression in the small intestinal epithelium in vivo. Conclusions: Our results show that LT␤R signaling induces CCL20 expression in intestinal epithelial cells, suggesting that this pathway triggers constitutive production of CCL20 in the FAE.

he mucosal surface of the intestine is constantly being exposed to antigens and microorganisms. Peyer’s patches (PP) are organized lymphoid tissues found in the small intestine that act as inductive sites for adaptive immune responses. PPs are covered by a specialized epithelium, the follicle-associated epithelium (FAE) that lacks goblet and enterochromaffin cells but harbors M cells that efficiently transport material from the intestinal lumen to the underlying lymphoid compartment. In the FAE, the brush border hydrolytic enzymes such as

T

sucrase isomaltase1 and lactase2 or transporters such as the dipeptide transporter PepT13 and the polymeric Ig receptor4 are down-regulated when compared with the epithelium of adjacent villi, facilitating access of luminal antigens to the epithelium. In addition, the chemokine CCL20 (MIP-3␣, LARC, Exodus) is constitutively expressed in the FAE but not in the absorptive epithelium.5,6 CCL20 was identified as the chemokine that recruits immature CD11b⫹ myeloid dendritic cells into PPs.7,8 CCR6 is the only known chemokine receptor that binds to CCL20.9 –11 CCR6-deficient mice fail to recruit myeloid dendritic cells into the subepithelial dome of PP,12;13 indicating that the FAE controls the cellular composition of the subepithelial microdomain of PPs. CCL20 expression in epithelial cells can be induced by proinflammatory cytokines such as tumor necrosis factor ␣ (TNF-␣) and interleukin (IL)-1 through the nuclear factor ␬B (NF-␬B)-signaling pathway.14,15 How CCL20 expression in the FAE is regulated is as yet unknown. It has been proposed that B cells control FAE and M-cell formation.16,17 However, in mice lacking B cells, small PPs with FAE containing M cells have been observed, indicating that additional cell types may contribute to the development of both FAE and PPs.18 The earliest hematopoietic cells that migrate into the developing gut around embryonic (E) day E15.5 to 16.5 are CD4⫹CD3⫺IL-7R⫹ cells. These cells express the membrane-bound heterotrimer LT␣1␤2 and activated ␣4␤1integrin that interacts with intestinal VCAM-1⫹ mesenchymal cells.19 LT␤R signaling is crucial for PP development.20 LT␤R engagement by LT␣1␤2⫹ hematopoietic cells are known to induce the secretion of chemokines CXCL13, CCL21, and CCL19 by mesenchymal cells.21–23 Whether LT␣1␤2/LT␤R interactions are Abbreviations used in this paper: FAE, follicle-associated epithelium; IL7R, interleukin 7 receptor; LT␣1␤2 , lymphotoxin ␤ heterotrimer; LT␤R, lymphotoxin ␤ receptor; PP, Peyer’s patch. © 2004 by the American Gastroenterological Association 0016-5085/04/$30.00 doi:10.1053/j.gastro.2004.04.018

214

RUMBO ET AL.

also important for FAE development and CCL20 expression has remained elusive. To address this question, we have monitored CCL20 expression during embryonic development and found that CCL20 mRNA was detected in the FAE as soon as LT␣1␤2⫹ CD4⫹CD3⫺ hematopoietic cells accumulated in PP anlage. LT␤R signaling efficiently triggered CCL20 expression in a human colon epithelial cell line grown as monolayers. Moreover, CCL20 mRNA accumulated in the small intestine epithelium of mice treated with agonistic LT␤R-specific Ab. Our data indicate that CCL20 expression in intestinal epithelial cells is regulated by LT␤R signaling and suggest a role of hematopoietic LT␣1␤2⫹ CD4⫹CD3⫺ cells in the regulation of FAE gene expression during PP ontogeny.

GASTROENTEROLOGY Vol. 127, No. 1

nologicals, Camarillo, CA) and hamster IgG antibodies with biotinylated goat anti-hamster IgG antibodies (Pierce, Rockford, IL), followed by a streptavidin-alkaline phosphatase (Amersham Pharmacia Biotech, Du¨ bendorf, Switzerland). Peroxidase activity was revealed by incubating the sections with diaminobenzidine (Sigma, St Louis, MO) in 0.05 mol/L Tris buffer saline (TBS) (pH 7.6) containing 0.06% H2O2. Alkaline phosphatase activity was visualized with AP substrate conjugated to fast blue (Sigma) in 0.05 mol/L TBS (pH 9.2).

Cell Culture The human colon adenocarcinoma T-84 cells (ATCC No. CCL-248) were grown to confluency for 10 days on Transwell filters (6-mm diameter, 3-␮m pore, Corning Inc., Acton) at 37°C under 5% CO2 in 50% DMEM, 50% Ham’s F12 medium supplemented with 10% FCS.

Anti-LT␤-R or LT␣1␤2 Stimulation

Materials and Methods Antibodies and Reagents Mouse monoclonal antibody specific for human LT␤R (CBE 11), hamster monoclonal antibody specific for mouse LT␤R (ACH6), control antibodies, and TNFR1-Ig fusion protein were kindly provided by Dr. Browning, Biogen, Cambridge, MA. Recombinant human lymphotoxin LT␣1␤2 was from R&D systems. Rat monoclonal antibody specific for mouse CD4 molecule was produced in the laboratory as described.18 Mouse anti-human NF-␬B p52 monoclonal antibody was from Upstate (Charlottesville, VA) and rabbit polyclonal anti-human NFkB p65 antiserum from Santa Cruz Biotechnology (Santa Cruz, CA). Biotin-conjugated hamster anti-mouse CD3⑀ was from PharMingen (145-2C11, San Diego, CA).

Mice C57BL/6J, Rag-1⫺/⫺, and C3H/HeJ mice were obtained from the Jackson Laboratory (Harlan, Horst, The Netherlands). MyD88⫺/⫺ mice were a kind gift from Dr. Akira (Osaka University, Osaka, Japan). LT␣⫺/⫺ mice were previously described.24 Mice were maintained in conventional animal facilities, and all animal experiments complied with the Swiss animal experimentation regulations. All mice used in this study were between 2 and 4 months of age, unless otherwise stated.

Immunoenzymatic Procedures Specimens were frozen in OCT embedding compound (Sakura Finetek Europe, Zoeterwoude, The Netherlands) and stored at ⫺80°C. Sections (7-␮m thick) were fixed in 4% paraformaldehyde in phosphate-buffered saline (PBS), pH 7.4. Sections were then preincubated for 30 minutes in saturation buffer (PBS with 1% bovine serum albumin and 3% goat serum) to block nonspecific binding sites. Next, tissue sections were incubated for 2 hours with primary antibodies. After 3 washes with PBS, rat IgG antibodies were revealed with horseradish peroxidase-conjugated goat anti-rat IgG (Tago Immu-

Anti-LT␤R antibody or LT␣1␤2 recombinant protein was diluted at the indicated concentrations in cell culture medium and added to the epithelial cells in the lower chamber. Total RNA for qRT-PCR analysis was extracted from the monolayers as described below. Mice were injected intraperitoneally with 50 ␮g of either mouse anti-LT␤R agonist antibody or control antibody and killed as indicated. In the case of LT␣⫺/⫺ mice, newborns were injected intraperitoneally with 5 ␮g of either mouse anti-LT␤R agonist antibody or control antibody and killed after 2 hours.

Cloning of CCL20 Promoter and Luciferase Reporter Constructs The BAC clone RP11-90L9 (AQ284493 GENBank accession number) (Life Technologies Gibco-BRL, Novex and Research Genetics divisions) containing the CCL20 gene was digested using Xho1-HindIII. The 1.6-kb promoter region was cloned in pBluescript creating the pCCL20.16 and subcloned in the firefly luciferase reporter pGL-3 basic vector (Promega, Madison, WI). Mutations into the putative CCL20 NF-␬Bbinding site were introduced by site directed mutagenesis using the proof start DNA polymerase (Qiagen, Basel, CH) and primer 5⬘ GGGCCAGTTGATCAATgatgagaattCCATGTGGCAACACGC 3⬘ (mutated nucleotides are written in small letters). This primer was used in combination with the anti-sense primer for amplification and KpnI-NheI fragments were subcloned into pGL3-basic vector. The constructs were verified by sequencing.

Cell Transfections and Dual Luciferase Assay Epithelial cells were transfected with the CCL20 reporter plasmid and the normalizing TK-Renilla luciferase construct (Promega) using Lipofectin (Invitrogen, Carlsbad, CA) according to manufacturer’s conditions. After 12-hour incubation, the transfection mix was replaced by fresh cultured medium. After 48 hours, the cells were stimulated with the different reagents at the indicated concentrations, and, 8 hours

July 2004

after stimulation, the cells were lysed, and firefly luciferase activity was measured and normalized to Renilla luciferase activity using the Dual luciferase assay (Promega).

Nuclear Protein Extraction and Western Blot Analysis After different stimulation times, cells were washed with cold PBS containing phosphatase inhibitors, and nuclear proteins were obtained using the Nuclear Extract kit from Active Motif (Carlsbad, CA), following manufacturer’s instructions. Protein concentration of the extracts was measured using the Bradford Assay (Bio-Rad, Hercules, CA). Equal amounts of proteins were loaded in a 12% PAGE gel and, after electrophoresis, transferred to nitrocellulose using the Miniprotean II system (Bio-Rad). After transfer, nonspecific binding sites were blocked by incubation in PBS containing 1% dry skimmed milk. Subsequent incubations were performed using this solution as diluent. Primary antibodies were incubated overnight at 4°C, and secondary HRP-conjugated antibody were incubated for 30 minutes at 37°C. Blots were developed using chemiluminescent ECL system (Amersham Pharmacia Biotech) on BioMax films (Eastman Kodak Co, Rochester, NY). Semiquantitative analysis was done using ImageJ software (http://rsb.info.nih.gov/ij).

Tissue Processing The gut was removed from dead mice and washed extensively with ice-chilled PBS, pH 7.5. Intestinal fragments containing PP, or total intestine in the case of embryonic tissue, were snap frozen in OCT compound (Sakura). For RNA isolation, 5-mm-long, PP-free intestinal segments were homogenized using a Potter homogenizer (see below). For laser microdissection, the gut was rinsed with icechilled PBS to remove the intestinal content. One-cm-long gut segments containing a visible PP were cut, incubated overnight in zinc fixative/sucrose 30% solution (5 g ZnCl2 , 6 g ZnAc2X2H2O, 0.1 g CaAc2 per L of 0.1 mol/L Tris pH 7.4), finally embedded in OCT, and frozen by immersion in liquid nitrogen.

Laser Dissection Microscopy Twenty-␮m frozen sections were cut and mounted on Leica membranes for dissecting microscopy (code number 11505151, Leica Microsystems, Wezlar, Germany), fixed for 5 minutes with methanol, and stained for 30 seconds with Mayer’s Hemalun solution. Afterwards, they were rinsed in water for 2 minutes and air-dried for 30 minutes. Samples were processed using a laser dissection microscopy (LMD) microscope (Leica Microsystems), coupled to a CCD camera. The selected region of the sample was cut by laser shot, and the samples were collected on a tube cap containing 20 ␮L RNA lysis buffer placed below the sample holders.

RNA Isolation RNA was isolated using Nucleospin RNA II system (Macherey-Nagel, Duren, Germany). Endogenous DNA was eliminated by DNAse treatment on the purification column.

CCL20 INDUCTION BY LYMPHOTOXIN ␣1␤2

215

The quality of RNA was assessed using the Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA) with picoRNA6000 chips according to the manufacturer’s indications.

Reverse Transcription and Semiquantitative Real-Time PCR Reverse transcription (RT) was performed with 100 ng of RNA using Superscript II (Gibco BRL). Resulting cDNA was amplified in triplicate using the SYBR-Green PCR assay, and products were detected on a Prism 5700 detection system (SDS; ABI/Perkin-Elmer, Foster City, CA). PCR reactions were performed for 2 minutes at 50°C and for 10 minutes at 95°C, followed by 40 amplification cycles with 1 minute annealing/extension at 60°C and 15 seconds denaturation at 95°C. Specific primers for human and mouse CCL20 (human forward primer: CCAAGAGTTTGCTCCTGGCT, human reverse primer: TGCTTGCTGCTTCTGATTCG; mouse forward primer: TTTTGGGATGGAATTGGACAC, mouse reverse primer: TGCAGGTGAAGCCTTCAACC), human histone H3.3 (forward primer: AGACTGCCCGAAATCGAC, reverse primer: CTTGCGAGCGGCTTTTGTA), and mouse ␤-actin (forward primer: CGTCATCCATGGCGAACTG, reverse primer: GCTTCTTTGCAGCTCCTTCGT) were designed using the software Primer Express (Applied Biosystem). Because of differences in sample size, normalization was done using ␤-actin for mouse samples and histone H3.3 for the human cell line samples. The identity of PCR amplicons was checked by melting curve analysis and sequencing. Relative mRNA levels (2⌬C) were determined by comparing (1) the PCR cycle threshold (ct) between cDNA of the gene of interest and the normalizer (H3.3 or ␤-actin) (⌬C) and (2) ⌬C values between treated and untreated conditions (⌬⌬C). SD of relative mRNA levels was 2 2 calculated as follows: 2(⌬⌬C⫾√{SD(⌬Ctreated) ⫹SD(⌬Cuntreated) }). Increase of RNA levels lower than 2-fold was not considered significant.

In Situ Hybridization CCL20-specific 35S-labeled riboprobes were synthesized from a plasmid vector pKS⫹ containing full coding region for CCL20 (kindly provided by Christoph Mueller, University of Bern, Switzerland). Sense and anti-sense probes were produced using T7 and T3 RNA polymerases, respectively (Roche, Mannheim, Germany). To produce LT␤ and LT␤R specific riboprobes, the coding sequences were cloned in pGEM plasmid vector (Promega, Madison, WI) using blunt cDNA obtained with gene-specific primers (LT␤ forward primer: ATGGGGACACGGGACT, reverse primer: ACACATTCGTACCGTCAGTC; LT␤R forward primer: ACTCCACCCGGTCTGGTCA, reverse primer: GGCAAATAGGTTCCTTGGCCCTG) from RNA recovered from adult PPs. Sense and anti-sense probes were produced using T7 and SP6 RNA polymerases respectively (Roche). For ISH, 4% p-formaldehyde-fixed frozen sections were treated with 1 ␮g/mL proteinase K at 37°C for 30 minutes.

216

RUMBO ET AL.

Nonspecific binding of probe was reduced by treatment with 0.1 mol/L triethanolamine and acetylation with acetic anhydride. The slides were hybridized at 45°C overnight with a labeled probe at 2 ⫻ 106 cpm/10 ␮L. The sections were then washed and digested with RNase at 37°C for 30 minutes. Finally, slides were washed with SSC and dehydrated in graded ethanol and air dried. The radioactive RNA probe bound to tissue section was detected by 2– 4 weeks exposure to emulsion autoradiography using NTB-2 emulsion (Eastman Kodak Co.) at 4°C in the dark.

Results CCL20 Expression in PP Anlage Is Concomitant to LT␤ⴙ Cell Clustering To determine when the FAE marker CCL20 is induced in fetal mice and to identify when LT␣1␤2⫹ hematopoietic cells appear in the gut during development, we performed in situ hybridization (ISH) in embryonic mice at different stages of development. We monitored the expression of CCL20, LT␤, and LT␤R from day E15.5 to day E18.5 by ISH. The presence of CD4⫹CD3⫺ hematopoietic cells was assessed by immunohistochemistry. At day E16.5, some scattered CD4⫹CD3⫺ cells could be found in the lamina propria (Figure 1A). From day E17.5 onward, CD4⫹CD3⫺ cell clusters were clearly observed (Figure 1B and C). LT␤ expression was found where aggregates of CD4⫹CD3⫺ cells were located (Figure 1D–F). LT␤R was expressed by the epithelium at all stages tested (Figure 1G–I). CCL20 expression, although weak, could be clearly detected from day E17.5 onward. Interestingly, only the epithelium associated with follicles harboring LT␤⫹ lymphoid cells expressed CCL20 (Figure 1J–L). At day E18.5, CCL20 expression became stronger. Taken together, our data indicate that CCL20 was expressed in the FAE once the PP anlage became colonized by LT␤⫹ CD4⫹CD3⫺ cells. CCL20 Expression in FAE Is Not Dependent on Mature B or T Cells or on Toll-Like Receptor Signaling Different cell types from hematopoietic origin may mediate LT␣1␤2 signaling to the FAE. We studied CCL20 and LT␤ expression in RAG-1⫺/⫺ mice that lack mature B- and T-cell populations. These mice present rudimentary PP follicles in which LT␤⫹ cells can be detected (Figure 2A). Notably, CCL20 expression was detected in RAG-1⫺/⫺ FAE, indicating that mature B and T cells are dispensable for CCL20 expression in the FAE (Figure 2B). As assessed by immunostaining, the CD3⫺CD4⫹ cells in the intestinal follicles of RAG1⫺/⫺ deficient mice were found together with LT␤ expressing cells (data not shown).

GASTROENTEROLOGY Vol. 127, No. 1

CCL20 is known to be induced in intestinal epithelium by bacterial flagellin triggering toll-like receptors (TLR).25 To investigate the role of CCL20 expression in the absence of TLR signaling, we studied the expression of CCL20 in MyD88⫺/⫺ mice that have a defect in signaling through TLR ligation.26 The MyD88-deficient mice have only minor alterations in PP.27 By laser dissection microscopy and RT-qPCR, we compared the expression of CCL20 in FAE of MyD88⫺/⫺ and wildtype mice (Figure 3). Similar levels of CCL20 (Figure 3F) were observed in both mouse strains. In addition, similar LT␤ mRNA expression pattern (Figure 3G) was found in both wild-type and MyD88-deficient mice, suggesting that LT␤ signaling was sufficient to induce CCL20 in the FAE. LT␤R Signaling Induces CCL20 Expression in Intestinal Epithelial T84 Cells To test whether LT␣1␤2 induces expression of CCL20 in intestinal epithelial cells, human colon carcinoma-derived T84 cells were treated with agonistic LT␤R antibody or recombinant LT␣1␤2. CCL20-specific mRNA was induced in a dose-dependent manner in T84 cells on exposure to LT␤ ligand (Figure 4A). Recombinant LT␣1␤2 preparations may be contaminated with minute amounts of LT␣3 or LT␣2␤1 that signal through the tumor necrosis factor receptors, TNFRI and TNFRII, instead of the LT␤R.28,29 To rule out any signaling via TNFRI or TNFRII, recombinant TNFRI-Ig fusion protein was used to quench LT␣3 or LT␣2␤1 Under these conditions, LT␣1␤2-mediated CCL20 induction was not altered, whereas TNF␣-induced activity was completely abrogated (Figure 4B). The contribution of LT␤R signaling to CCL20 induction was confirmed using an agonistic LT␤R-specific antibody (Figure 4C). Stimulation with an isotype-matched control antibody failed to induce CCL20. We next analyzed the kinetics of LT␣1␤2 stimulation on T84 cells. After 2 hours of incubation with the agonistic antibody, CCL20-specific mRNA levels were increased 20-fold compared with cells incubated with isotype-matched control antibodies and remained significantly elevated for at least 24 hours (Figure 5A). Flagellin stimulation, known to induce CCL20 in intestinal epithelial cells,25 was used here as a positive control. Interestingly, the kinetics of flagellin-induced CCL20 expression is different, with a transient up-regulation and a decrease almost to basal levels after 12 hours of stimulation. To monitor CCL20 expression after LT␣1␤2 stimulation, a 1.6-kb fragment upstream of the transcription start of human CCL20 and containing the promoter region was inserted in front of a luciferase reporter gene.

July 2004

CCL20 INDUCTION BY LYMPHOTOXIN ␣1␤2

217

Figure 1. Identification of hematopoietic precursor cells, lymphotoxin ␣1␤2 and its receptor, and CCL20 in PP anlage during embryonic development. Serial sections of the small intestine at different embryonic days were analyzed. (A–C) Double staining with anti-CD4 (peroxidase, brown) and anti-CD3 antibodies (alkaline phosphatase, blue). Clusters of CD4⫹ cells were detected at day E17.5 and E18.5. CD3⫹ cells were rare in the follicles. Epithelium strongly express endogenous alkaline phosphatase and shows a strong blue staining. (D–F) In situ hybridization with LT␤ anti-sense probe. At day E17.5 and E18.5, there is a strong labeling of follicles in PP anlage. (G–I) In situ hybridization with LT␤R anti-sense probe. Receptor is expressed in the epithelium at all stages. (J–L) In situ hybridization with CCL20 anti-sense probe. A weak epithelial CCL20 signal over the PP anlage was detected as early as day E17.5 and increased at day E18.5. Scale bars represent 50 ␮m.

Following transient transfection of T84 cells with the construct, luciferase activity was measured following LT␣1␤2 stimulation. As shown in Figure 5B, the CCL20 promoter was responsive to LT␤R signaling. The promoter with a mutated NF-␬B binding site failed to respond to 8 hours of LT␣1␤2 stimulation, showing that

NF-␬B activity is required for LT␣1␤2-mediated CCL20 gene expression. The same result was obtained with 48-hour LT␣1␤2 stimulation (data not shown). LT␤R signaling activates different NF-␬B pathways.30 We followed the kinetics of nuclear translocation of the NF-␬B p65 subunit mediating the canonical

218

RUMBO ET AL.

Figure 2. LT␤ and CCL20 expression in RAG2 mice. (A) In situ hybridization with LT␤ anti-sense probe. A signal is clearly seen in association with the follicle of the rudimentary PPs. (B) In situ hybridization with CCL20 anti-sense probe. CCL20 signal is present on the epithelium over the rudimentary PP’s follicles. Scale bars represent 50 ␮m.

GASTROENTEROLOGY Vol. 127, No. 1

of LPS or other bacterial components contaminating the injected antibody preparation, C3H/HeJ and MyD88⫺/⫺ mice were used. C3H/HeJ mice are hyporesponsive to LPS because of a point mutation in the LPS-sensing receptor TLR4.33 CCL20 induction upon LT␤R agonistic antibody treatment was not significantly altered in these mouse strains (Figure 6A), confirming that the effect observed is due to LT␤R signaling. The effect of LT␤R signaling at different times postinjection was assessed. CCL20 was induced in intestine 2 hours after treatment, and mRNA accumulation remained stable over an 8-hour period (Figure 6C). To investigate whether the gut epithelium was producing CCL20 after LT␤R stimulation, epithelial villi

NF-␬B pathway and the processing of NF-␬B p100 to the p52 subunit to assess the alternative NF-␬B pathway. LT␣1␤2 induced a rapid and transient translocation of NF-␬B p65 into the nuclear fraction (Figure 5C and E as observed with flagellin (Figure 5D and E). Processing of NF-␬B p100 started after 6 hours of LT␣1␤2 stimulation and further increased over 24 hours, in contrast to flagellin where very little change was observed. LT␤R Signaling Induced CCL20 Gene Expression in Intestinal Epithelium In Vivo To test whether LT␤R signaling was sufficient to induce CCL20 gene expression in the small intestine, mice were treated intraperitoneally with 50 ␮g of agonistic anti-LT␤R antibodies. RNA from intestine was recovered and tested for CCL20-specific mRNA accumulation by RT-qPCR. CCL20 mRNA levels were 5- to 7-fold higher in intestine of mice treated with the agonistic anti-LT␤R antibodies when compared with PBStreated or an isotype-matched antibody control (Figure 6A). Trypsin digestion of the anti-LT␤R antibody completely abolished CCL20 induction (data not shown). Similar results were obtained using proximal or distal small intestinal segments. In the distal part of the intestine, however, more interindividual variation was observed, probably reflecting the presence of tiny isolated lymphoid follicles not detected macroscopically that contribute to CCL20 expression.31 In LT␣⫺/⫺ mice, which lack PPs32 or isolated lymphoid follicles,31 treatment with anti-LT␤R antibody elicited CCL20 production in the small intestine (Figure 6B), indicating that induction of CCL20 expression is not due to the presence of organized lymphoid structures. CCL20 was also induced in the colon of mice treated with anti-LT␤R antibody (data not shown). CCL20 is known to be induced either by proinflammatory signals (IL-1␤ and TNF-␣) produced upon tolllike receptor (TLR) signaling or by direct triggering of some TLRs.25 To rule out possible biased results because

Figure 3. LT␤ and CCL20 expression in MyD88⫺/⫺ mice. Tissue before and after laser microdissection of crypt epithelium (parts A and B, respectively) and epithelial villi (parts C and D, respectively) is shown. (E) Laser dissection microscopy was used to prepare samples highly enriched in FAE (1), villus epithelium (2), crypt epithelium (3), or follicular material (4). (F ) CCL20 relative expression in the different tissue types of MyD88⫺/⫺ and C57BL/6 mice. Tissue from 3 different mice from each strain was processed and qPCR performed on the corresponding cDNA. Results show the average and SD among the 3 different mice processed. (G) LT␤ relative expression in the different tissue types in MyD88⫺/⫺ and C57BL/6 mice. Tissue from 3 different mice from each strain was processed and qPCR performed on the corresponding cDNA. Results show the average and SD among the 3 different mice processed. Scale bars represent 50 ␮m.

July 2004

CCL20 INDUCTION BY LYMPHOTOXIN ␣1␤2

219

Figure 4. LT␤R signaling induces CCL20 expression in human intestinal T84 cells. Results are from at least 3 independent experiments. (A) Dose-response curve of LT␣1␤2 stimulation. T84 cells were stimulated for 3 hours with different concentrations of recombinant LT␣1␤2. CCL20-specific mRNA was measured by RT-PCR. (B) TNFR55-Ig fusion protein does not block the effect of recombinant LT␣1␤2. TNF-␣ or LT␤ was preincubated for 1 hour with a 100-fold molar excess of TNFR55-Ig. The cells were incubated as in part A. The fusion protein abolished TNF-␣-induced, but not LT␣1␤2-induced, CCL20 mRNA accumulation. (C) Agonist LT␤R-specific antibodies induce CCL20 expression by T84 cells. Agonist antibodies directed against the LT␤R were used to stimulate cell culture as in part A. CCL20 induction was observed.

were laser dissected and analyzed. CCL20 levels were consistently higher in the epithelium of LT␤R agonistic antibody-treated mice (Figure 6D) either 2 or 8 hours posttreatment when compared with PBS- or isotype antibody-matched controls.

Discussion Formation of Peyer’s patches requires complex interactions between the gut epithelium, the mesenchyme, and the hematopoietic cells (for review see Finke and Kraehenbuhl34). Although attention has focused mainly on the role of early LT␣1␤2⫹ hematopoietic cell and mesenchymal stromal cell cross talk in the formation of PP anlage,35 much less is known about the role of the epithelium in PP organization and the direct or indirect contribution of hematopoietic cells to FAE gene regulation. To analyze how the FAE develops during PP organogenesis, we selected, as a FAE-specific marker, CCL20, a chemokine known to be constitutively expressed in the FAE over PPs.6 We observed that CCL20 mRNA was detectable as early as day E17.5, and significantly increased at E18.5, corresponding to the accumulation of CD4⫹CD3⫺ hematopoietic cells at distinct sites along the small intestine.22,36 This prenatal induction of CCL20 indicates that microbial products that are able to induce CCL20 expression by intestinal epithelium25 are dispensable for FAE-specific CCL20 production. The fact that CCL20 is highly expressed on FAE of MyD88⫺/⫺ mice supports this hypothesis, although we cannot formally exclude that microflora may contribute to CCL20specific expression in adulthood by a MyD88⫺/⫺-independent signaling pathway.

Because LT␣1␤2 and LT␤R are critical for PP and FAE formation, we investigated when and where LT␤ and its receptor were expressed in the gut. In agreement with a previous report,37 LT␤ mRNA starts to accumulate in PP anlage around day E17.5, whereas LT␤R was already detected in the gut epithelium as early as day E15.5. Interestingly, LT␤R expression was much higher in the gut epithelium than in stromal cells, making this tissue a putative target for LT␤ signaling. Indeed, T84 intestinal epithelial cells respond to LT␤ stimulation and accumulate CCL20 transcripts. Although CCL20 mRNA accumulation is transient when epithelial cells are stimulated via TLR5 with bacterial flagellin,25 the accumulation is sustained up to 24 hours upon LT␣1␤2 stimulation. In Caco-2 cells, NF-␬B-binding sites present in the CCL20 promoter are required for CCL20 induction by TNF-␣ or IL-1.14 In this study, we demonstrate that LT␣1␤2-induced CCL20 expression is also NF-␬B dependent. Signal transduction by LT␤R in fibroblasts and in the intestinal epithelia HT29 cell line is mediated by at least 2 different NF-␬B pathways.30,38 LT␤R signals initially through the canonical NF-␬B pathway (p50/ p65), but, with time (12–24 hours), p65 nuclear translocation drops to basal levels, and the alternative NF-␬B pathway leading to p100 processing into active p52 takes over, resulting in sustained expression of several target genes. Activation of this alternative pathway by other receptors has also been shown to produce sustained induction of target genes.39 – 41 In concordance with findings of Dejardin et al.,30 for HT29 cells, we have found a similar dual kinetics of activation of canonical NF-␬B pathway in T84 cells after LT␤R triggering. p65 Is

220

RUMBO ET AL.

GASTROENTEROLOGY Vol. 127, No. 1

Figure 5. (A) Kinetics of CCL20 induction. T84 cells were stimulated for the indicated times with LT␣1␤2 recombinant protein (1 ␮g/mL). CCL20 expression after stimulation was measured as described in the Materials and Methods section. A sustained induction of CCL20 was observed at least up to 24 hours of stimulation. (B) Analysis of CCL20 promoter activity on LT␤ stimulation. T84 cells were transfected with a normalizer plasmid and a luciferase reporter construct containing the 1.6-kb sequence upstream of the CCL20 putative transcription start. In parallel, construct with mutated NF-␬B-binding site sequences was used. Reporter activity was measured 8 hours after LT␣1␤2 stimulation and normalized to Renilla luciferase. (C) Kinetics of NF-␬B activation on LT␣1␤2 stimulation. T84 cells were stimulated for different times using LT␣1␤2 recombinant protein (1 ␮g/mL) preincubated 1 hour with an excess of TNFR55-Ig fusion protein as in Figure 3C. Nuclear fractions from treated cells were analyzed by Western blots using either an anti-NF-␬B p65 or anti-NF-␬B p52 antibody. Time of treatment (hours) is indicated on top of each lane. As negative control (⫺), nuclear extract from untreated T84 cells was used. Total protein lysate from Raji B cells was used as positive control. (D) Kinetics of NF-␬B activation on flagellin stimulation. T84 cells were stimulated during different times with flagellin (1 ␮g/mL). Nuclear fractions from treated cells were analyzed by Western blots as in part C. Time of treatment (h) is indicated on top of each lane. Similar controls as in C were used. (E) Quantitative analysis of p52 and p65 induction in nuclear extracts upon flagellin and LT␣1␤2 stimulation. Blots showed in sections C and D were scanned using the ImageJ software. Relative quantification was done by assigning an arbitrary value of 1 to the lowest signal in each group of experiments.

rapidly translocated to the nucleus upon stimulation, and a subsequent sustained activation of p100 processing and p52 nuclear localization is observed. LT␣1␤2-induced CCL20 gene expression in T84 cells followed a similar kinetics to NF-␬B activation, suggesting that different NF-␬B-signaling pathways can trigger CCL20 induction in intestinal epithelial cells. This is further supported by the correlation found in CCL20 expression and NF-␬B activation upon flagellin stimulation. In vivo, it is likely

that the alternative pathway is used for FAE formation because RelB, which associates with the NF-␬B/p52 complex, is detected in the nuclei of mouse FAE anlage at E18.5,42 and CCL20 expression is reduced in the FAE of p52-deficient mice.43 The injection of agonistic LT␤R-specific antibodies induced CCL20 expression in the small intestinal epithelium, with kinetics similar to that observed in vitro, suggesting that LT␤R signaling is sufficient for the

July 2004

CCL20 INDUCTION BY LYMPHOTOXIN ␣1␤2

221

Figure 6. CCL20 expression is induced in vivo after anti-LT␤R agonist antibody treatment. (A) CCL20 expression on treatment in different mice strains. CCL20-specific expression was determined by real-time PCR using RNA recovered from duodenal segments of mice injected 2 hours before with 50 ␮g of agonist antibody or control antibody. Wild-type C57BL/6, Myd88⫺/⫺ mice, and LPS hyporesponsive C3H/HeJ strains were tested. The results shown are the average of 3 independent experiments. (B) CCL20 expression in treated LT␣⫺/⫺ mice. CCL20 specific expression was determined as described in part A from duodenal segments of newborn mice injected 2 hours before with either 5 ␮g of agonist antibody or control antibody. The results shown are the average of 3 independent experiments. (C) CCL20 mRNA up-regulation is induced at 2-hours postinjection and sustained for up to 8-hours postinjection. Treatment as described in part A was performed. Mice were killed at different times posttreatment, and CCL20 specific expression was monitored. The results shown are the average of 3 independent experiments. (D) CCL20 expression in small intestine epithelial tissue from treated mice. CCL20 specific expression was determined by real-time PCR on RNA from laser-dissected epithelial tissue from anti-LT␤R-treated mice or from PBS-treated or isotype matched antibody controls. CCL20 is up-regulated in the epithelium of mice treated with the anti-LT␤R antibody.

regulation of CCL20 gene expression in the gut epithelium. LT␣-deficient mice lack both PP and FAE. We were unable to detect CCL20 expression when consecutive sections of total intestine of these mice were analyzed by in situ hybridzation (data not shown). Treatment of newborn LT␣⫺/⫺ mice with agonistic anti-LT␤R significantly induced intestinal CCL20 expression (Figure 5B). This indicates that LT␣1␤2/LT␤R cross-linking is required for CCL20 expression. Agonistic anti-LT␤R antibody triggered CCL20 expression along the entire intestinal epithelium, although, under normal conditions, CCL20 is restricted to the FAE. Because LT␤R is expressed along the entire gut epithelium, all enterocytes are responsive to LT␣1␤2 signaling and can potentially express CCL20 as shown in this study. What restricts CCL20 expression to the FAE is the presence of the LT␣1␤2 ligand, which is restricted

to the lymphoid follicles rich in LT␣1␤2 bearing cells. The observation that PPs develop normally in CCR6⫺/⫺ mice supports the idea that CCL20 is not a PP-inducing factor.12,13 Constitutive CCL20 expression in the FAE, at distinct sites along the gut epithelium, may result from the interaction with LT␣1␤2-bearing cells that are recruited during PP ontogeny by a yet undefined mechanism. The intestine is first seeded around day 16.5 with IL-7R␣⫹ CD4⫹CD3⫺CD45⫹ cells. Signaling through IL-7R␣ had been shown to induce expression of LT␣1␤2 , which is essential for lymphoid organogenesis.22,36,44 We have previously shown that fetal wild-type CD4⫹CD3⫺CD45⫹ cells can induce PP formation in PP-deficient mice.19 Here, a second function of CD4⫹CD3⫺CD45⫹ PP-inducing cells during fetal development, namely, the induction of CCL20 in FAE via LT␣1␤2 , is suggested. At the present, we

222

RUMBO ET AL.

cannot rule out that other cell types, i.e., B, T, and NK cells, which are known to express LT␣1␤2 ,45 could also play a role in inducing FAE-specific gene expression. In RAG-1-deficient mice, which lack both mature B and T cells, CCL20 is expressed in the FAE over rudimentary PPs that harbor LT␤⫹ cells and CD4⫹CD3⫺CD45⫹ cells. Thus, B and T cells probably are dispensable for CCL20 expression in the FAE. We cannot exclude the possibility, however, that mature lymphocytes participate in potentiating and maintaining CCL20 expression. The communication of hematopoietic cells with epithelial cells via the LT␤R-signaling pathway seems to be important not only in the gut but also in the thymus. A normal configuration of medullar epithelial cells is dependent on the presence of LT␣1␤2-expressing thymocytes.46 Taken together, LT␤R signaling has a central role in cellular homeostasis and function of primary and secondary lymphoid organs, including epithelial cells within these organs. In conclusion, we provide in vitro and in vivo evidence that LT␤R signaling induces CCL20 in the gut epithelium. We described a concomitant expression of LT␤ and CCL20 in embryonic intestines. Altogether, these results indicate that LT␤R signaling may be responsible for the production of CCL20 in FAE and subsequent recruitment of dendritic cells to this anatomical site.

References 1. Sierro F, Pringault E, Simon Assman P, Kraehenbuhl JP, Debard N. Transient expression of M-cell phenotype by enterocyte-like cells of the follicle-associated epithelium of mouse Peyer’s patches. Gastroenterology 2000;119:734 –743. 2. Smith MW. Selective expression of brush border hydrolases by mouse Peyer’s patch and jejunal villus enterocytes. J Cell Physiol 1985;124:219 –225. 3. Freeman TC, Bentsen BS, Thwaites DT, Simmons NL. H⫹/ditripeptide transporter (PepT1) expression in the rabbit intestine. Pflu¨ gers Arch 1995;430:394 – 400. 4. Pappo J, Owen RL. Absence of secretory component expression by epithelial cells overlying rabbit gut-associated lymphoid tissue. Gastroenterology 1988;95:1173–1177. 5. Rossi DL, Vicari AP, Franz-Bacon K, McClanahan TK, Zlotnik A. Identification through bioinformatics of two new macrophage proinflammatory human chemokines: MIP-3␣ and MIP-3␤. J Immunol 1997;158:1033–1036. 6. Tanaka Y, Imai T, Baba M, Ishikawa I, Uehira M, Nomiyama H, Yoshie O. Selective expression of liver and activation-regulated chemokine (LARC) in intestinal epithelium in mice and humans. Eur J Immunol 1999;29:633– 642. 7. Iwasaki A, Kelsall BL. Localization of distinct Peyer’s patch dendritic cell subsets and their recruitment by chemokines macrophage inflammatory protein (MIP)-3 ␣, MIP-3 ␤, and secondary lymphoid organ chemokine. J Exp Med 2000;191:1381–1393. 8. Dieu-Nosjean MC, Massacrier C, Homey B, Vanbervliet B, Pin JJ, Vicari A, Lebecque S, Dezutter-Dambuyant C, Schmitt D, Zlotnik A, Caux C. Macrophage inflammatory protein 3 ␣ is expressed at inflamed epithelial surfaces and is the most potent chemokine known in attracting Langerhans cell precursors. J Exp Med 2000; 192:705–717. 9. Zaballos A, Varona R, Gutierrez J, Lind P, Marquez G. Molecular

GASTROENTEROLOGY Vol. 127, No. 1

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

cloning and RNA expression of two new human chemokine receptor-like genes. Biochem Biophys Res Commun 1996;227:846 – 853. Greaves DR, Wang W, Dairaghi DJ, Dieu MC, Saint-Vis B, FranzBacon K, Rossi D, Caux C, McClanahan T, Gordon S, Zlotnik A, Schall TJ. CCR6, a CC chemokine receptor that interacts with macrophage inflammatory protein 3␣ and is highly expressed in human dendritic cells. J Exp Med 1997;186:837– 844. Baba M, Imai T, Nishimura M, Kakizaki M, Takagi S, Hieshima K, Nomiyama H, Yoshie O. Identification of CCR6, the specific receptor for a novel lymphocyte-directed CC chemokine LARC. J Biol Chem 1997;272:14893–14898. Varona R, Villares R, Carramolino L, Goya F, Zaballos A, Gutieerrez J, Torres M, Martinez AC, Marquez G. CCR6-deficient mice have impaired leukocyte homeostasis and altered contact hypersensitivity and delayed-type hypersensitivity responses. J Clin Invest 2001;107:R37–R45. Cook DN, Prosser DM, Forster R, Zhang J, Kuklin NA, Abbondanzo SJ, Niu XD, Chen SC, Manfra DJ, Wiekowski MT, Sullivan LM, Smith SR, Greenberg HB, Narula SK, Lipp M, Lira SA. CCR6 mediates dendritic cell localization, lymphocyte homeostasis, and immune responses in mucosal tissue. Immunity 2000;12: 495–503. Fujiie S, Hieshima K, Izawa D, Nakayama T, Fujisawa R, Ohyanagi H, Yoshie O. Proinflammatory cytokines induce liver and activation-regulated chemokine/macrophage inflammatory protein-3␣/ CCL20 in mucosal epithelial cells through NK-␬B. Int Immunol 2001;13:1255–1263. Sugita S, Kohno T, Yamamoto K, Imaizumi Y, Nakajima H, Ishimaru T, Matsuyama T. Induction of macrophage-inflammatory protein-3␣ gene expression by TNF-dependent NF-␬B activation. J Immunol 2002;168:5621–5628. Golovkina TV, Shlomchik M, Hannum L, Chervonsky A. Organogenic role of B lymphocytes in mucosal immunity. Science 1999; 286:1965–1968. Kerne´ is S, Bogdanova A, Kraehenbuhl JP, Pringault E. Conversion by Peyer’s patch lymphocytes of human enterocytes into M cells that transport bacteria. Science 1997;277:948 –952. Debard N, Sierro F, Browning J, Kraehenbuhl JP. Effect of mature lymphocytes and lymphotoxin on the development of the follicleassociated epithelium and M cells in mouse Peyer’s patches. Gastroenterology 2001;120:1173–1182. Finke D, Acha-Orbea H, Mattis A, Lipp M, Kraehenbuhl JP. CD4⫹ CD3⫺ cells induce Peyer’s patch development: role of ␣4␤1integrin-activation by CXCR5. Immunity 2002;17:367–373. Rennert PD, Browning JL, Mebius R, Mackay F, Hochman PS. Surface lymphotoxin ␣/␤ complex is required for the development of peripheral lymphoid organs. J Exp Med 1996;184:1999 – 2006. Ngo VN, Korner H, Gunn MD, Schmidt KN, Riminton DS, Cooper MD, Browning JL, Sedgwick JD, Cyster JG. Lymphotoxin ␣/␤ and tumor necrosis factor are required for stromal cell expression of homing chemokines in B and T cell areas of the spleen. J Exp Med 1999;189:403– 412. Honda K, Nakano H, Yoshida H, Nishikawa S, Rennert P, Ikuta K, Tamechika M, Yamaguchi K, Fukumoto T, Chiba T, Nishikawa SI. Molecular basis for hematopoietic/mesenchymal interaction during initiation of Peyer’s patch organogenesis. J Exp Med 2001; 193:621– 630. Cuff CA, Sacca R, Ruddle NH. Differential induction of adhesion molecule and chemokine expression by LT␣3 and LT␣␤ in inflammation elucidates potential mechanisms of mesenteric and peripheral lymph node development. J Immunol 1999;162:5965– 5972. DeTogni P, Goellner J, Ruddle NH, Streeter PR, Fick A, Mariathasan S, Smith SC, Carlson R, Shornick LP, Strauss-Schoenberger J, Russell JH, Karr R, Chaplin DD. Abnormal development

CCL20 INDUCTION BY LYMPHOTOXIN ␣1␤2

July 2004

25.

26.

27.

28.

29.

30.

31.

32.

33.

34. 35.

36.

37.

38.

of peripheral lymphoid organs in mice deficient in lymphotoxin. Science 1994;264:703–706. Sierro F, Dubois B, Coste A, Kaiserlian D, Kraehenbuhl JP, Sirard JC. Flagellin stimulation of intestinal epithelial cells triggers CCL20-mediated migration of dendritic cells. Proc Natl Acad Sci U S A 2001;98:13722–13727. Akira S, Hoshino K. Myeloid differentiation factor 88-dependent and -independent pathways in toll-like receptor signaling. J Infect Dis 2003;187(suppl 2):S356 –S363. Iiyama R, Kanai T, Uraushihara K, Ishikura T, Makita S, Totsuka T, Yamazaki M, Nakamura T, Miyata T, Yoshida H, Takeuchi O, Hoshino K, Takeda K, Ishikawa H, Akira S, Watanabe M. Normal development of the gut-associated lymphoid tissue except Peyer’s patch in MyD88-deficient mice. Scand J Immunol 2003;58: 620 – 627. Browning JL, Dougas I, Ngam-ek A, Bourdon PR, Ehrenfels BN, Miatkowski K, Zafari M, Yampaglia AM, Lawton P, Meier W. Characterization of surface lymphotoxin forms. Use of specific monoclonal antibodies and soluble receptors. J Immunol 1995; 154:33– 46. Browning JL, Miatkowski K, Griffiths DA, Bourdon PR, Hession C, Ambrose CM, Meier W. Preparation and characterization of soluble recombinant heterotrimeric complexes of human lymphotoxins ␣ and ␤. J Biol Chem 1996;271:8618 – 826. Dejardin E, Droin NM, Delhase M, Haas E, Cao Y, Makris C, Li ZW, Karin M, Ware CF, Green DR. The lymphotoxin-␤ receptor induces different patterns of gene expression via two NF-␬B pathways. Immunity 2002;17:525. Lorenz RG, Chaplin DD, McDonald KG, McDonough JS, Newberry RD. Isolated lymphoid follicle formation is inducible and dependent upon lymphotoxin-sufficient B lymphocytes, lymphotoxin ␤ receptor, and TNF receptor I function. J Immunol 2003;170: 5475–5482. Banks TA, Rouse BT, Kerley MK, Blair PJ, Godfrey VL, Kuklin NA, Bouley DM, Thomas J, Kanangat S, Mucenski ML. Lymphotoxin␣-deficient mice: effects on secondary lymphoid organ development and humoral immune responsiveness. J Immunol 1995; 155:1685–1693. Qureshi ST, Lariviere L, Leveque G, Clermont S, Moore KJ, Gros P, Malo D. Endotoxin-tolerant mice have mutations in toll-like receptor 4 (Tlr4). J Exp Med 1999;189:615– 625. Finke D, Kraehenbuhl JP. Formation of Peyer’s patches. Curr Opin Gen Dev 2001;11:562–569. Adachi S, Yoshida H, Kataoka H, Nishikawa S. Three distinctive steps in Peyer patch formation of murine embryo. Int Immunol 1997;9:507–514. Yoshida H, Honda K, Shinkura R, Adachi S, Nishikawa S, Maki K, Ikuta K. IL-7 receptor ␣⫹CD3⫺ cells in the embryonic intestine induces the organizing center of Peyer’s patches. Int Immunol 1999;11:643– 655. Browning JL, French LE. Visualization of lymphotoxin-␤ and lymphotoxin-␤ receptor expression in mouse embryos. J Immunol 2002;168:5079 –5087. Derudder E, Dejardin E, Pritchard LL, Green DR, Korner M, Baud V. RelB/p50 dimers are differentially regulated by TNF-␣ and

39.

40.

41.

42.

43.

44.

45.

46.

223

lymphotoxin-␤ receptor activation: critical roles for p100. J Biol Chem 2003;278:23278 –23284. Saitoh T, Nakayama M, Nakano H, Yagita H, Yamamoto N, Yamaoka S. TWEAK induces NF-␬B2 p100 processing and longlasting NF-␬B activation. J Biol Chem 2003;1:1. Coope HJ, Atkinson PG, Huhse B, Belich M, Janzen J, Holman MJ, Klaus GG, Johnston LH, Ley SC. CD40 regulates the processing of NF-␬B2 p100 to p52. EMBO J 2002;21:5375–5385. Claudio E, Brown K, Park S, Wang H, Siebenlist U. BAFF-induced NEMO-independent processing of NF-␬ B2 in maturing B cells. Nat Immunol 2002;3:958 –965. Yilmaz ZB, Weih DS, Sivakumar V, Weih F. RelB is required for Peyer’s patch development: differential regulation of p52-RelB by lymphotoxin and TNF. EMBO J 2003;22:121–130. Paxian S, Merkle H, Riemann M, Wilda M, Adler G, Hameister H, Liptay S, Pfeffer K, Schmid RM. Abnormal organogenesis of Peyer’s patches in mice deficient for NF-␬B1, NF-␬B2, and Bcl-3. Gastroenterology 2002;122:1853–1868. Adachi S, Yoshida H, Honda K, Maki K, Saijo K, Ikuta K, Saito T, Nishikawa S. Essential role of IL-7 receptor ␣ in the formation of Peyer’s patch anlage. Int Immunol 1998;10:1– 6. Ware CF, Crowe PD, Grayson MH, Androlewicz MJ, Browning JL. Expression of surface lymphotoxin and tumor necrosis factor on activated T, B, and natural killer cells. J Immunol 1992;149: 3881–3888. Boehm T, Scheu S, Pfeffer K, Bleul CC. Thymic medullary epithelial cell differentiation, thymocyte emigration, and the control of autoimmunity require lympho-epithelial cross talk via LT␤R. J Exp Med 2003;198:757–769.

Received October 28, 2003. Accepted March 25, 2004. Address requests for reprints to: Jean-Pierre Kraehenbuhl, ISREC, Chemin des Boveresses 155, CH1066 Epalinges, Switzerland. e-mail: [email protected]; fax: (41) 21 652 69 33. Supported by scholarships from the Roche Research Foundation, the Consejo Nacional de Investigaciones Cientificas y Tecnicas (CONICET), and the Antorchas Foundation (to M.R.) and by grants from the Swiss National Science Foundation (3100-067926.02 to J.-P.K.) and (PPOOA-68855/1 to D.F.) and Oncosuisse (OCS 1135-02-2001 to D.F.). The authors thank Corinne Tallichet-Blanc, Monique Reinhardt, Isabelle Surbek, and Catherine Roger for their skillful technical assistance; Dr. Christoph Mueller for providing CCL20 plasmid and protocols for in situ hybridization; Dr. S. Akira for the MyD88ⴚ/ⴚ mice; Jean-Claude Sirard and Arnaud Didierlaurent for helpful discussions and contribution to the CCL20 promoter cloning; Dr. J. Browning for providing several reagents and helpful comments; and Rod Ceredig for critical reading of the manuscript. M.R. and F.S. contributed equally to this article. D.F.’s present address is: Department of Clinical and Biological Sciences (DKBW), Division of Developmental Immunology, Pharmacenter/University of Basel, CH-4056 Basel, Switzerland. F.S.’s present address is: Arthritis and Inflammation Research Department, The Garvan Institute of Medical Research, Darlinghurst NSW 2010 Australia.