IL-4 and IL-13 Up-Regulate Intestinal Trefoil Factor Expression: Requirement for STAT6 and De Novo Protein Synthesis¹

Inflammation of the gastrointestinal tract and airways is frequently characterized by CD4⁺ Th cell-mediated mucosal changes that may be either beneficial or detrimental to the epithelial barrier. Th cells differentiate into at least two subsets, Th1 and Th2 cells, based on their cytokine secretion profile (1). Th1 cells, which predominantly produce IL-2, IFN-γ, and lymphotoxin, are important for the eradication of intracellular pathogens, including bacteria, parasites, yeasts, and viruses. Th2 cells produce IL-4, IL-5, IL-9, and IL-13 and are typical of humoral immune reactions, helminth infections, and atopic diseases such as asthma. Up-regulation of IgE, mastocytosis, eosinophilia, and goblet cell hyperplasia are characteristic of Th2 responses induced by helminth infection and allergic trigger. The production of mucus represents an important response for protecting mucosal surfaces. Paradoxically, Th2 cell-mediated mucus production acts positively in the protection of mucosal surfaces from pathogens (2), but it may be detrimental as well by contributing to airway obstruction during pulmonary inflammatory responses (3). The control of goblet cell production in type 2 inflammatory processes has been attributed to IL-4, IL-9, and IL-13 (4, 5). IL-4 and IL-13 are considered the major mediators of goblet cell production in the type 2 cytokine-driven inflammatory process evoked by intestine helminth infection or experimental asthma (6, 7, 8, 9). IL-4 exerts its biological effects by binding to the α-chain of the IL-4R, a component of both type 1 and type 2 IL-4R (10, 11, 12). In type 2 IL-4R, the IL-4R α-chain is paired with the IL-13R α1-chain, which also binds IL-13 (13, 14). Signal transduction may occur by phosphorylation and activation of STAT6 by Janus kinases 1 and 3. Activated phosphorylated STAT6 dimerizes and translocates into the nucleus, where it binds to specific promoter regions to regulate gene transcription (15, 16). IL-13-, IL-4Rα-, and STAT6-deficient mice fail to develop goblet cell production after infection with the parasite Nippostrongylus brasiliensis (17). However, the IL-13-responsive cells causing these effects are unknown. A direct effect of IL-13 on epithelial cells causes mucus overproduction in asthma (18). As epithelial cells of the gut express the IL-4R α-chain (19), and IL-4/IL-13 directly affect several functions of the epithelial barrier (20, 21, 22, 23), it is reasonable to assume that IL-4/IL-13 may increase the expression of goblet cell products. Trefoil factor family (TFF)⁴peptides are typically secreted by mucin (MUC)-producing epithelia, and a broad body of data indicates that these peptides are important agents in preventing mucosal damage and aiding epithelial repair (24, 25). They include TFF3 (formerly hP1.B/intestinal trefoil factor), which is selectively expressed in high concentrations by mucus-producing cells of the small and large intestines (26). Thus, it can provide a vehicle to define the basis of goblet cell activity.

The current studies were undertaken to determine the pathways from IL-4Rα engagement that lead to TFF3 gene expression. We show that IL-4/IL-13 stimulate TFF3 and MUC2 gene expression in intestinal cell lines that harbor a goblet cell phenotype. The cytokine-mediated increase in TFF3 transcription is shown to be STAT6-dependent and to require de novo protein synthesis.

LY294002, cycloheximide (CHX), actinomycin D, and wortmannin were purchased from Sigma-Chemie (St. Quentin Fallavier, France). A 425-bp human TFF3 (hTFF3) cDNA was used as specific probe for Northern blot analysis and was provided by S. Ribieras (Institut de Génétique et de Biologie Moléculaire et Cellulaire, Strasbourg, France). The plasmid LA778 containing STAT6 cDNA was provided by W. J. LaRochelle (National Cancer Institute, Bethesda, MD). The plasmid pCEFL-HA containing the neomycin resistance gene was provided by S. Emami (Institut National de la Santé et de la Recherche Médicale, Unité 482, Paris, France). The plasmid named p800 containing the first 800 bp of TFF3 promoter cloned into HindIII and SacI sites of plasmid pGL3-basic (positions −867 to + 66), the TFF3.d1 plasmid containing the first 495 bp of TFF3 promoter cloned into the HindIII and KpnI sites of pGL3-basic, and the pUC 18 4.10 plasmid containing 10 kb of hTFF3 genomic DNA have been previously described (27). All oligonucleotides used for RT-PCR and EMSA analysis were custom-synthesized by Invitrogen (Cergy-Pontoise, France).

The Ab directed against P85α subunit of phospatidylinositol 3-kinase (PI3-K) was obtained from Upstate Biotechnology (Mundolsheim, France). The rabbit polyclonal IgG anti-STAT6 M200 directed against the N-terminal fragment of the protein was obtained from Santa Cruz Biotechnology (Le Perray en Yvelines, France). The chicken polyclonal IgG anti-STAT6 used in EMSA analysis has been prepared by one of us (M.H.H.). Specific Abs used for flow cytometric analysis were referred to as mAb-hIL-4Rα (R&D Systems, Lille, France), mAb-hIL-13Rα1 (Diaclone, Besançon, France), or mAb-hIL-13Rα2 (Diaclone). The secondary Ab used for flow cytometric analysis was an FITC-labeled goat anti-mouse IgG (H+L) (Chemicon, Mundolsheim, France). Human rIL-4 was obtained from Schering-Plough through the courtesy of F. Brière (Dardilly, France). Human rIL-13 was a gift from A. Minty (Sanofi, France).

The HT-29 CL.16E cells selected by butyrate treatment of parental cells have been described previously (28). HT-29 cells selected by adaptation to methotrexate (referred to as HT-29 MTX) were provided by T. Lesuffleur (Institut National de la Santé et de la Recherche Médicale, Unité 505, Paris, France). Cells were grown in DMEM (Invitrogen) supplemented with 10% FCS and 1% penicillin/streptomycin in an atmosphere of 5% CO₂ at 37°C.

In cytokine stimulation experiments, IL-4 and IL-13 were added to culture media of cells either at 70/80% confluence or at late confluence at a dose of 0.1–100 ng/ml for a time course of 1–48 h. For PI3-K inhibition under cytokine stimulation, LY294002 (20 μM) or wortmannin (1 μM) was added to culture medium of 70/80% confluent cells 0.5 h before cytokine treatment (10 ng/ml) and maintained for 24 h. Due to its instability, wortmannin was added at the same dose every 8 h during the time course of the experiments. For protein synthesis inhibition experiments, CHX (10 μg/ml) was added to culture medium of 70/80% confluent cells 0.5 h before cytokine treatment (10 ng/ml) and maintained for 24 h. mRNA stability measurement was performed using actinomycin D (10 μg/ml), which was added to culture medium of 70/80% confluent cells 0.5 h before cytokine treatment. The cells were either treated with actinomycin D alone or associated with cytokine (10 ng/ml) for 6, 12, and 24 h. For EMSA experiments, 70/80% confluent HT-29 CL.16E cells were treated with cytokines (10–100 ng/ml) for 20 min. Nuclear contents were then extracted as described below. For PI3-K activity assays, early preconfluent HT-29 CL.16E cells were treated with IL-13 (20 ng/ml) for 10–20 min before lysis as described below.

For TFF3 analysis, cells were lysed by sonication in water, and then 50 μg total proteins were boiled for 5 min in the presence of 2-ME, resolved on a 10% polyacrylamide gel, and transferred to a 0.2-μm pore size nitrocellulose membrane. Membranes were incubated overnight at 4°C with antiserum 8A directed against the C-terminal part of TFF3 (29). For STAT6 detection, cells were lysed in cold solubilization buffer containing 1% Triton X-100, 50 mM HEPES, 150 mM NaCl, 2 mM Na₃VO₄, 100 mM NaF, 100 IU aprotinin/ml, 20 μM leupeptin, and 0.2 mg/ml PMSF (pH 7.5). Membranes were incubated overnight at 4°C with anti-STAT6 M200. The RIA for hTFF3 was performed as previously described (29).

Total cellular RNA was extracted with RNAble (Eurobio, Les Ulis, France) according to the manufacturer’s instructions. Ten micrograms of RNA per assay was separated on a 1% agarose-formaldehyde denaturing gel and then blotted onto nylon-positive membranes. Blots were hybridized overnight with 10⁶ cpm/ml of an [α-³²P]dCTP-labeled probe at 60°C and exposed to Kodak X-OMAT AR films (Eastman Kodak, Rochester, NY) with intensifying screens.

Plasmid denoted BB, XX, and KX contained different lengths of hTFF3 intron I DNA sequence, subcloned upstream of the SV40 promoter of the pGl3 promoter vector (designated P0; Promega, Charbonnières, France). BB plasmid contained the complete 1724 bp of intron I. It was constructed using the BamHI sites (+122/+1845; pUC 18 4.10) and the BglII site of P0. An additional construct XX containing 353 bp of intron I was obtained using PCR product amplified by forward primer ctgctcgaggatctctatggcgtgtg and reverse ctgctcgagcaagtttgcatcctccc (containing the XhoI restriction site), subcloned into the XhoI site of P0 (underlined sequences indicate the restriction sites). The KX plasmid was obtained from the precedent PCR product that was truncated using the KpnI site (+1448) and subcloned into the KpnI/XhoI sites of P0.

The first 1989 bp of human STAT6 cDNA were amplified by sticky-end PCR (30 cycles) using LA778 plasmid as template. Primers used to obtain digested BamHI and EcoRI restriction sites were: for BglII: forward 1, gatctACCATGTCTCTGTGGGGTCTGGTC; and forward 2, tACCATGTCTCTGTGGGGTCTGGTC; and for EcoRI: reverse 1, cTTCATGGGGTAGGAAGTGGTTG; and reverse 2, aattcTTCATGGGGTAGGAAGTGGTTG, containing a chain termination codon. The 1989-bp PCR product encoded the first 659 N-terminal aa of STAT6, corresponding to a dominant-negative form. The PCR product was purified and subcloned with T4 DNA ligase into BglII/EcoRI-digested pCEFL-HA vector. This plasmid was named DNSTAT6.

For transient transfections, ∼2 × 10⁵ HT-29 CL.16E cells/well (six-well plate) were plated on day 1 and transfected by the Lipofectamine reagent (Invitrogen) on day 2. Briefly, for each well, 1 μg of reporter plasmids were mixed with 4 μl of Lipofectamine reagent in 200 μl of serum-free DMEM. In all transfection experiments, a plasmid with a Renilla luciferase reporter gene under the control of a thymidine kinase promoter (pHRL-TK; 50 ng/well; Promega) was used as an internal control. Eight hundred microliters of complete medium was added, and cells were incubated for 48 h. Transfection mixture was then replaced with fresh complete medium for an additional 12-h period before cytokine treatment. On day 4 cells were left untreated or were stimulated with 10 ng/ml IL-4 or IL-13 for 24 h. Cells were then lysed, and the firefly and Renilla luciferase activities were measured in a luminometer (Microlite; Dynatech Laboratories, Chantilly, VA) using the Dual Luciferase Reporter assay system (Promega) in accordance with the manufacturer’s instructions.

Stable transfections were performed as follows: ∼10⁴ HT-29 CL.16E cells/well (9.6-cm²) were plated on day 1 and transfected by the SuperFect transfection reagent (Qiagen, Courtaboeuf, France) on day 2 with the pCEFL-HA or the DNSTAT6 plasmid for 3 h, according to the manufacturer’s protocol. After 4 days, transfected cells were selected for 3 wk of culture in 800 μg/ml G418 (Geneticin; Invitrogen). Resistant colonies were ring-cloned as individual colonies. Cells were then left under selection and were subjected to analysis of ectopic overexpression of the truncated STAT6 form by immunoblot analysis.

First-strand cDNA was synthesized using Superscript II (Invitrogen). cDNA were amplified using the following primers (5′-3′): hGAPDH (983 bp): forward, tgaaggtcggagtcaacggatttggt; reverse, catgtgggccatgaggtccaccac; hTFF3 (92 bp): forward, aaccggggctgctgctttg; reverse, gaggtgcctcagaaggtgc; hMUC2 (401 bp): forward, ctgcaccaagaccgtcctcatg; reverse, gcaaggactgaacaaagactcaga; human β-actin (586 bp): forward, caaggccaaccgcgagaag; reverse, cagggtacatggtggtgcc; IL-13Rα1 (509 bp): forward, ggagccagctcaatttgtag; reverse, cacacgggaagttaaaggca; IL-13Rα2 (1088 bp): forward, ggagagaggcaatatcaagg; reverse, ggccatgactggaaactgt; and IL-4Rα (335 bp): forward, gacctggagcaacccgtatc; reverse, catagcacaacaggcagacg.

Amplifications were performed in an automated thermal cycler (denaturation: 96°C, 30 s; annealing: 60°C, 30 s; extension: 72°C, 1 min). Amplifications of 23 and 25 cycles were performed for GAPDH and TFF3, respectively, and amplifications of 35 cycles were performed for MUC2 and IL receptor chains. In addition to conventional PCR, real-time PCR was conducted by rapid cycling using the LightCycler instrument and Fast Start DNA Master SYBR Green I (Roche, Meylan, France) according to the manufacturer’s instructions. Results normalized to β-actin amplified from the same cDNA mix are expressed as attomoles per microgram of RNA.

Single-cell suspensions of HT-29 CL.16E cells obtained after trypsinization were cultured in petri dishes with constant rocking at 37°C in 5% CO₂ for 48 h to allow re-expression of receptors lost during the treatment with trypsin. Approximately 5 × 10⁵ cells were incubated in PBS/10 mM EDTA for 15 min at 4°C. After washing with PBS, 0.1% BSA, and 0.05% NaN₃, the cells were incubated with 4 μg of specific Abs anti-hIL-4Rα, anti-hIL-13Rα1, or anti-hIL-13Rα2 in PBS/5% BSA for 1 h at room temperature. After washing, cells were incubated with 2 μg of FITC-labeled secondary Ab. Cells were then fixed in PBS/0.8% paraformaldehyde, and 5 × 10³ to 10⁴ cells/ondition were analyzed for fluorescence by single-color flow cytometry using a FACScan and FloMax software (DAKO, Trappes, France).

After treatment with cytokines, cells were lysed in low salt buffer (20 mM HEPES (pH 7.9), 10 mM KCl, 1 mM EDTA, 1 mM EGTA, 0.2% Nonidet P-40, 10% glycerol, 0.1 mM Na₃VO₄, 1 mM PMSF, 1 mM DTT, 2 μg/ml aprotinin, 1 μg/ml leupeptin, and 1 μg/ml pepstatin) at 4°C for 10 min. After centrifugation for 5 min at 3,000 × g, pellets were extracted with high salt buffer (low salt buffer supplemented with 420 mM NaCl and 20% glycerol) for 30 min at 4°C. Samples were cleared by centrifugation at 12,000 × g at 4°C. Nuclear extracts (10 μg of total proteins) were then incubated for 20 min at room temperature in a mixture containing 20 mM HEPES (pH 7.9), 4% Ficoll, 1 mM MgCl₂, 40 mM KCl, 0.1 mM EGTA, 0.5 mM DTT, and 160 μg/ml poly(dI-dC) with 1 ng of ³²P-labeled oligonucleotides. Samples were separated on a 4.5% nondenaturing polyacrylamide gel at 400 V for 3 h at 4°C. Gels were then dried and exposed to BioMax film (Kodak) overnight. The following oligonucleotides corresponding to STAT6 response element sequences were used: Cε, 5′-CACTTCCCAAGAACAGA-3′; and TFF3 (intron 1), 5′-ATCACCTTCTCGGGAAGTCACA-3′. For supershift experiments, 2 μl of anti-STAT6 was added to the samples before or after the addition of labeled oligonucleotides. Competition experiments were performed in the presence of 100-fold molar excesses of the cold STAT6 consensus oligonucleotides.

Cells were lysed at 4°C in 20 mM Tris-HCl (pH 7.4), 140 mM NaCl, 10 mM EDTA, 4 mM NaVO₄, 100 mM NaF, 10 mM pyrophosphate, and 1% Nonidet P-40. After centrifugation, 100 μg of supernatant proteins were used for immunoprecipitation with a specific Ab directed against human P85α PI3-K. PI3-K activity was measured in the immunoprecipitates as previously described (30, 31). Labeled phosphoinositides were visualized and quantified using a PhosphorImager SI and ImageQuant software (Molecular Dynamics, Sunnyvale, CA).

Statistical analysis was conducted using Student’s t test, and significance was assigned at p ≤ 0.05.

The hallmarks of the differentiation of intestinal goblet cells are the early expression of TFF3, followed by the expression of MUCs (26). The subclone 16E of the human colonic epithelial cell line HT29 undergoes differentiation to the goblet cell-like phenotype after reaching confluence (28). We examined the relative expressions of TFF3 and MUC2 in HT-29 CL.16E cells during their differentiation into the MUC-secreting phenotype. The appearance of TFF3 and MUC2 transcripts was analyzed by RT-PCR in the course of the culture from a subconfluent (day 3) up to a late postconfluent (day 34) state. Transcripts for these genes became strongly expressed after confluence (Fig. 1,A). Real-time PCR indicated a dramatic increase in TFF3 expression between day 3 (56 ± 7 attomol/μg total RNA; n = 5) and day 34 (1756 ± 136 attomol/μg total RNA; n = 5). Northern blot analysis revealed that the TFF3 transcripts extracted from the HT-29 CL.16E cells were similar in size to those found in intestinal goblet cells (0.4 kb; Fig. 1,B). In a Western blot analysis, a single species (∼7 kDa), corresponding closely to the molecular mass of the TFF3 peptide, was revealed in cell extracts, and the intensity of the band strongly increased when cells were committed to differentiation (Fig. 1 C).

As illustrated in Fig. 2, the specific transcripts for IL-4Rα (335 bp), IL-13Rα1 (509 bp), and IL-13Rα2 (1088 bp) were detected in HT-29 CL.16E cells by RT-PCR. The presence of these chains was further evidenced by flow cytometry, as the fluorescence intensity was shifted on the right when cells were incubated with the antiserum raised against the IL-4Rα, IL-13Rα1, or IL-13Rα2 chain (Fig. 2). IL-2Rγ chain transcripts were also detected (data not shown). We could not demonstrate a differential expression pattern for the IL-4 and IL-13 receptors in the course of differentiation.

The expression of TFF3 in HT-29 CL.16E cells was first investigated by Northern blot analysis of total RNA after 24 h of stimulation with graded concentrations of IL-13 ranging from 0.1–100 ng/ml. As shown in Fig. 3, HT-29 CL.16E cells constitutively expressed low levels of TFF3 mRNA, and IL-13 induced a concentration-dependent increase in TFF3 mRNA expression, first detectable at 1 ng/ml of IL-13 and maximal at 10 ng/ml. The real-time PCR revealed 8- and 13-fold increases in TFF3 mRNA levels upon treatment with IL-4 (10 ng/ml) and IL-13 (10 ng/ml), respectively. Cell extracts from similar experiments were also quantitated by RIA for TFF3 protein contents. Both cytokines similarly increased the synthesis of TFF3 protein in the HT-29 CL.16E cell line in a dose-dependent fashion (Fig. 3). Comparable results were obtained in HT-29 MTX cells, which harbor a goblet cell-like phenotype (data not shown). Fig. 4 illustrates the kinetics of up-regulation of TFF3 mRNA expression by IL-13, with a 2-fold increase at 6 h and a 7-fold increase at 24 h of treatment, as measured by real-time PCR. Changes in TFF3 mRNA levels were paralleled by an increase in TFF3 cellular content, as assessed by RIA. As illustrated in Fig. 4, both IL-4 and IL-13 markedly enhanced TFF3 peptide synthesis. The effect was detectable after 24 h and was maximal after 48 h of stimulation with the cytokines. Incubation of the cells with IL-13 (10 ng/ml) also resulted in an increase in MUC2 mRNA levels, as evaluated by semiquantitative RT-PCR (data not shown).

In a variety of cell types, activation of gene expression by Th2 cytokines often relies on phosphorylation and activation of the transcription factor STAT6. To identify DNA elements that could be involved in IL-4/IL-13-induced up-regulation of TFF3, we constructed luciferase reporters containing putative regulatory sequences of the human TFF3 gene. The proximal part of the TFF3 promoter does not contain canonical STAT6 binding elements (TTCnnnnGAA), but five noncanonical binding sites (TTnnnnnnAA) were identified. Luciferase reporter constructs containing up to 800 bp of TFF3 promoter have been described previously by one of us (27). They were transiently transfected into intestinal cells. No apparent modification of the reporter activity was observed upon IL-4/IL-13 treatment (data not shown). Intron 1 of TFF3 contains one canonical STAT6 binding site (Fig. 5). Thus, the construct BB containing the 1.7-kb intron 1 was subcloned upstream of the SV40 promoter of the pGL3 promoter vector. Other constructs containing shorter segments of intron 1, including the putative STAT6 DNA binding element, were also prepared (Fig. 5). Transient transfections of these constructs were performed on HT-29 CL16E cells treated with IL-13 for 24 h. The activity of the intron 1-containing reporter was not modified by IL-13 treatment (Fig. 5). Similarly to the TFF3 gene, the polymeric IgR (pIgR) gene contains a unique STAT6 binding site localized in intron 1. This element was shown to be involved in the IL-4-induced up-regulation of pIgR (32). We therefore used the construct p12 containing a 554-bp segment of pIgR intron 1, including the STAT6 recognition sequence, upstream of the SV40 promoter of the pGL3 promoter vector as an internal control (32). We observed a 3-fold increase in luciferase enzymatic activity upon stimulation with IL-13 (Fig. 5). In HT-29 cells, the IL-4-mediated up-regulation of endogenous pIgR has been shown to depend on de novo protein synthesis that cooperates with STAT6 itself bound to its cognate DNA element in the noncoding part of the gene (32). We tested whether it was true for IL-4/IL-13-mediated up-regulation of TFF3 in HT-29 CL.16E cells. Cells were left untreated or were stimulated with IL-13 for 24 h in either the presence or the absence of CHX. Total RNA was isolated, and the levels of TFF3 transcripts were measured by semiquantitative RT-PCR. For each sample, we analyzed mRNA for the housekeeping gene GAPDH. We found that the mRNA level for TFF3 was increased ∼5-fold after IL-13 treatment, whereas simultaneous treatment with CHX abolished this up-regulation (Fig. 6). Otherwise, experiments with actinomycin D revealed that the stability of TFF3 mRNA was not modified in IL-4/IL-13-treated cells (Fig. 7).

Up-regulation of TFF3 by IL-4/IL-13 does not seem to be mediated through the STAT6 binding element located in intron 1. To determine whether IL-13 stimulation affected protein-DNA interactions, we isolated nuclear extracts from HT-29 CL.16E cells treated with IL-4 or IL-13 for 20 min and performed in vitro EMSA experiments with a probe spanning the canonical STAT6 site (Fig. 8). We identified one major IL-4- or IL-13-inducible nucleoprotein complex. Competitor experiments and treatment with Ab to STAT6 demonstrated that this factor was present in the complex. These data indicate that type 2 cytokines may induce activation of STAT6 and its binding to this DNA element from intron 1.

To evaluate further the function of STAT6 in IL-4/IL-13-mediated TFF3 gene induction, we established stable HT-29 CL.16E cells expressing a dominant-negative STAT6 protein. As shown in Fig. 9,A, clone M expressed high amounts of the truncated STAT6 protein. We then asked whether TFF3 gene induction was affected by dominant-negative STAT6. Cells were treated with or without IL-13 at 10 ng/ml for 24 h. Total RNA was prepared, and TFF3 induction was analyzed by real-time PCR. As shown in Fig. 9,B, TFF3 mRNA was profoundly induced by IL-13 in the parental HT-29 CL.16E cells and in the control clone D, which was obtained after transfection with an empty vector. In contrast, TFF3 gene expression was not activated in clone M by IL-13 treatment. In the same set of experiments, quantification of TFF3 in cell extracts revealed 4- and 2-fold increases in immunoreactive material in IL-13-treated parental cells or clone D cells, respectively, whereas TFF3 immunoreactivity was not increased in clone M (Fig. 9 C). Similar results were observed using transfected HT-29 MTX cells (data not shown). Taken together, these results indicate that STAT6 plays a crucial role in TFF3 expression.

Several investigations showed that IL-4 and IL-13 also signal along the insulin receptor substrate-1/-2, PI3-K pathway; for example, to increase paracellular permeability in the intestinal wall (22). We therefore tested the hypothesis that PI3-K could be involved in the IL-4/IL-13-mediated activation of TFF3 gene expression in HT-29 CL.16E cells. As shown in Fig. 10,A, IL-13 increased PI3-K activity in HT-29 CL.16E cells. To block PI3-K signaling, we used the specific inhibitor wortmannin at a concentration of 1 μM, which has been reported to be optimal for PI3-K inhibition in several intestinal cell lines (33). Wortmannin treatment of the cells did not modify the IL-13-induced TFF3 gene expression (Fig. 10, B and C). Similar results were obtained with LY294002, another inhibitor of PI3-K (data not shown).

Infection with nematode parasites typically induces a Th2-type response, which contributes to the expulsion of worms (4, 34). IL-4 has been considered to play a key role in the host’s protective immunity, and recently, the importance of IL-13, a related cytokine, has been highlighted in Th2 development and host protection against nematode infection (11). The Th2-type response is characterized by mucosal changes, including goblet cell hyperplasia (35, 36). Goblet cells contained in the intestinal epithelium are the main source of MUCs and TFF3 in the gut. TFF3 is thought to associate with MUC to increase the viscoelasticity of the mucus gel (37). These high molecular mass structures may play an important role in the trapping and removal of intestinal nematodes from the gut (38). Recently, it has been reported that IL-13- or STAT6-deficient mice infected with N. brasiliensis or Trichinella spiralis, respectively, did not exhibit goblet cell hyperplasia (6, 39). Additionally, a study performed with SCID or TCR^−/− mice showed that dysregulated Th2 cells cause goblet cell metaplasia in the intestine through IL-4 release (40). Nevertheless, little is known about the mechanism that regulates goblet cell proliferation and differentiation in the gut. In the present study we hypothesized that IL-4 or IL-13 candirectly induce intestinal epithelial cells to engage in the goblet cell differentiation program and express TFF3 and MUCs.

Because IL-4Rs are known to be present on intestinal epithelial cells (19), we first investigated the presence of IL-4Rs on cultured HT-29 CL.16E cells. These cells harbor a goblet-cell like phenotype in the early postconfluence (28), and we show a marked increase in TFF3 and MUC2 expression in cells committed to differentiation. We also demonstrated the presence of IL-4R α- and IL-13R α-chains, which are subunits of the functional IL-4Rs. These cells can therefore respond to IL-4 or IL-13 and constitute a valid system to study the direct effects of IL-4 and IL-13 on epithelial cell differentiation. As one hallmark of the differentiation of intestinal goblet cells is the early expression of TFF3, we first investigated the effects of IL-4/IL-13 on TFF3 expression, as evaluated by Northern blot, real-time PCR, and RIA. These two prototypic Th2 cytokines up-regulate the expression of TFF3 in a goblet cell line. This up-regulation presumably involves indirect trans-activation of the TFF3 gene and is associated with enhanced TFF3 peptide synthesis.

We found that IL-13 induced TFF3 gene expression from 6 h onward and that this gene induction was dependent on de novo protein synthesis. It is likely that de novo synthesis of a transcription factor(s) is required to elicit TFF3 gene expression. Alternatively, it might be an enzyme or some other protein that modulates transcription through its effect on DNA-bound factors. Whatever the precise underlying mechanism, STAT6 is an essential component of the TFF3 gene induction by IL-13, because in our study the expression of a dominant-negative STAT6 in HT-29 CL.16E cells abolished the cytokine-mediated TFF3 expression.

The downstream molecular events associated with the STAT6 pathway upon IL-4/IL-13 activation and leading to goblet cell metaplasia are difficult to assess because the identification of differentiation factors for goblet cell lineage has remained elusive. Kruppel-like transcription factor 4 (formerly gut-enriched Kruppel-like factor) is a zinc finger transcription factor expressed in the epithelia of skin, lungs, gastrointestinal tract, and several other organs (41). A recent study provided the first evidence that Kruppel-like transcription factor 4 is a goblet cell-specific differentiation factor in the intestine (42). Otherwise, a goblet cell silencer inhibitor element interacting with a goblet cell silencer inhibitor element binding protein has been shown to be essential for goblet cell-specific expression of TFF3 upon HT-29 subclone exposure to keratinocyte growth factor (43). An attractive hypothesis would be that transcription factors are up-regulated in the course of IL-4/IL-13-induced goblet cell metaplasia according to a STAT6-dependent pathway.

Previous studies using airway hypersensitivity models showed that IL-4/IL-13 were key elements in allergen-induced goblet cell metaplasia (7, 8, 18, 44). Inhalation of IL-4 led to up-regulation of the MUC gel and to goblet cell metaplasia and IL-4-induced MUC2 gene expression in cultured airway epithelial cells (45). Although the expression of MUC2 apoprotein was not specifically evaluated in our experiments, our data show that TFF3 and MUC2 genes are coexpressed upon IL-4/IL-13 treatment of the cells.

PI3-K, a ubiquitous lipid kinase involved in receptor signal transduction by tyrosine kinase receptors, has been shown to stimulate adipocyte and myogenic differentiation (46, 47). Two distinct studies performed with different cell lines show that PI3-K may either stimulate or inhibit enterocyte differentiation (33, 48). We observed that PI3-K inhibition in HT-29 CL.16E cells significantly reduced the spontaneous goblet cell-like differentiation,⁵As several investigations showed that IL-4/IL-13 may also signal along the insulin receptor substrate-1/-2, PI3-K pathway in intestinal cells (22), we tested the hypothesis that the PI3-K and STAT6 pathways could both be activated upon IL-4/IL-13 treatment of cells, leading to TFF3 gene induction. As wortmannin or LY294002 did not modify cytokine-evoked TFF3 gene induction, it seems likely that distinct pathways govern TFF3 expression in the course of spontaneous differentiation of goblet-like cells and in type 2 cytokine-induced goblet cell metaplasia.

In summary, our study is the first to show that IL-4 or IL-13 induces TFF3 expression via a direct effect on intestinal epithelial cells, through a STAT6- and de novo protein synthesis-dependent pathway. These results do not rule out other important effects of IL-4/IL-13 on goblet cell growth, for example, that may occur concurrently to the activation of goblet cell-specific transcriptional machinery.

We thank Dr. Hilde Schjerven for the gift of the construct p12, and Drs. Françoise Brière and Adrian Minty for the gifts of human IL-4 and IL-13, respectively.