Inflammatory bowel disease (IBD) is characterized by a

Transcription

1 GASTROENTEROLOGY 2012;142: CCL17 Promotes Intestinal Inflammation in Mice and Counteracts Regulatory T Cell Mediated Protection From Colitis ALEXANDER F. HEISEKE,* ANTONIA C. FAUL,* HANS ANTON LEHR, IRMGARD FÖRSTER, ROLAND M. SCHMID,* ANNE B. KRUG,* and WOLFGANG REINDL* *II. Medizinische Klinik, Klinikum rechts der Isar, Technische Universität München, Munich, Germany; Centre Hospitalier Universitaire Vaudois, Institut Universitaire de Pathologie, Lausanne, Switzerland; and Molecular Immunology, Institut für Umweltmedizinische Forschung an der Heinrich-Heine-Universität Düsseldorf, Düsseldorf, Germany BACKGROUND & AIMS: Priming of T cells by dendritic cells (DCs) in the intestinal mucosa and associated lymphoid tissues helps maintain mucosal tolerance but also contributes to the development of chronic intestinal inflammation. Chemokines regulate the intestinal immune response and can contribute to pathogenesis of inflammatory bowel diseases. We investigated the role of the chemokine CCL17, which is expressed by conventional DCs in the intestine and is up-regulated during colitis. METHODS: Colitis was induced by administration of dextran sodium sulfate (DSS) to mice or transfer of T cells to lymphopenic mice. Colitis activity was monitored by body weight assessment, histologic scoring, and cytokine profile analysis. The direct effects of CCL17 on DCs and the indirect effects on differentiation of T helper (Th) cells were determined in vitro and ex vivo. RE- SULTS: Mice that lacked CCL17 (Ccl17 E/E mice) were protected from induction of severe colitis by DSS or T-cell transfer. Colonic mucosa and mesenteric lymph nodes from Ccl17-deficient mice produced lower levels of proinflammatory cytokines. The population of Foxp3 regulatory T cells (Tregs) was expanded in Ccl17 E/E mice and required for long-term protection from colitis. CCR4 expression by transferred T cells was not required for induction of colitis, but CCR4 expression by the recipients was required. CCL17 promoted Toll-like receptor induced secretion of interleukin-12 and interleukin-23 by DCs in an autocrine manner, promoted differentiation of Th1 and Th17 cells, and reduced induction of Foxp3 Treg cells. CONCLUSIONS: The chemokine CCL17 is required for induction of intestinal inflammation in mice. CCL17 has an autocrine effect on DCs that promotes production of inflammatory cytokines and activation of Th1 and Th17 cells and reduces expansion of Treg cells. Keywords: TLR; IBD; Crohn s Disease; Mouse Model. Inflammatory bowel disease (IBD) is characterized by a failure to regulate the immune response against luminal antigens, leading to a chronic relapsing inflammation of the gastrointestinal tract. This dysregulated immune response is sustained by the influx of naïve T cells that have not adapted to the mucosal milieu and are now exposed to luminal antigens. 1 The differentiation of naïve T cells recruited into the mucosa requires their interaction with antigen-presenting cells. This interaction is particularly crucial for the differentiation of CD4 T cells to proinflammatory T-helper (Th) 1 cells, interleukin (IL)- 17 producing Th cells (Th17), and forkhead box P3 (Foxp3) regulatory T cells (Tregs). 2 Dendritic cells (DCs) are professional antigen-presenting cells that attract appropriate T-cell populations by producing chemokines and shape the differentiation of these T cells via the production of cytokines. The chemokine CCL (chemokine [C-C motif] ligand) 17, which is expressed by CD11b DCs outside the spleen and binds to CCR4 (chemokine [C-C motif] receptor 4), is essential for the homing of CD4 T cells to the skin 3,4 and the lung. 5 CCR4, the only known receptor for CCL17 and also for CCL22, is not only expressed by CD4 T cells, but also by CD8 cytotoxic T lymphocytes, natural killer cells, macrophages, and subsets of DCs. 6 9 On CD4 T cells, CCR4 is expressed on Th17 and Th2 effector cells as well as on Tregs. Ccr4-deficient Tregs do not migrate to mesenteric lymph nodes (MLNs) and fail to suppress colitis in the T-cell transfer colitis model. 10 Studies in Ccr4- and Ccl17- deficient mice identified many different and partly conflicting functions of this molecule. Ccr4-deficient mice fail to develop allograft tolerance due to diminished CCR4- and CCL22-dependent recruitment of Tregs to the cardiac allograft. 11 In contrast, cardiac allograft survival is prolonged in Ccl17-deficient mice. 12 In a murine model of atopic dermatitis, the absence of CCL17, but not the absence of CCR4, reduces inflammation and the migration of cutaneous DCs to lymph nodes. 13 These examples show the nonredundant function of the chemokine CCL17, which is not overlapping with that of CCR4. Abbreviations used in this paper: BMDC, bone marrow derived dendritic cell; DC, dendritic cell; DSS, dextran sodium sulfate; egfp, enhanced green fluorescent protein; IFN, interferon; IL, interleukin; LPL, lamina propria leukocyte; LPS, lipopolysaccharide; IEL, intraepithelial leukocyte; MLN, mesenteric lymph node; qrt-pcr, quantitative reverse-transcription polymerase chain reaction; TGF, transforming growth factor; Th, T helper; TLR, Toll-like receptor; Treg, regulatory T cell; WT, wild-type by the AGA Institute /$36.00 doi: /j.gastro

2 336 HEISEKE ET AL GASTROENTEROLOGY Vol. 142, No. 2 Little is known concerning the role of CCL17 in T-cell homing to the intestinal mucosa. CCL17 expression is up-regulated in the inflamed mucosa during experimental mouse models of colitis, 14,15 and one report found that CCL17 expression is up-regulated in the mucosa of patients with Crohn s disease. 16 However, it is not known to what extent and by which mechanism CCL17 influences the course of colitis. We show that Ccl17-deficient mice are protected from colitis induced by naïve CD4 T-cell transfer and that this protection is independent of CCR4 expression on the transferred T cells. In the absence of CCL17, proinflammatory cytokine production by DCs and effector T cells is significantly reduced and Tregs are expanded, conferring long-term protection against colitis. Materials and Methods Mice Mice were on the C57BL/6 background (at least 10 generations backcrossed). Rag (recombination-activating gene)1 /, Ccl17 egfp/egfp (Ccl17 E/E ), 12 Rag1 / Ccl17 E/E, Ccr4 /, Rag1 / Ccr4 /, OTII, and DEREG 23.2 mice 17 were bred in our specific pathogen-free animal facility. Experiments were performed in accordance with the German animal care and ethics legislation and approved by the local government authorities. See Supplementary Materials and Methods for information on media and reagents, bone marrow derived dendritic cells (BMDCs), isolation of colonic lamina propria leukocytes (LPLs) and intraepithelial leukocytes (IELs), flow cytometry, enzymelinked immunosorbent assay, RNA isolation, and quantitative reverse-transcription polymerase chain reaction (qrt-pcr). Isolation of T Cells and DCs T cells (97% CD45RB hi ) were isolated using the CD4 CD62L T Cell Isolation Kit II (Miltenyi Biotec, Bergisch- Gladbach, Germany). DCs were isolated from MLNs after digestion for 30 minutes at 37 C using CD11c-MicroBeads (Miltenyi Biotec). In Vitro Stimulation and Coculture BMDCs were stimulated with lipopolysaccharide (LPS) (Sigma-Aldrich, Munich, Germany), Pam 3 Cys-SKKKK (Pam 3 Cys) (EMC microcollections GmbH, Tübingen, Germany), and/or recombinant murine CCL17 (R&D Systems, Minneapolis, MN) at cells/ml. Cells were harvested for messenger RNA (mrna) isolation after 6 hours. Supernatants were collected for enzyme-linked immunosorbent assay after 24 hours. DCs were pulsed with 5 g/ml OVA-peptide (amino acids , ISQAVHAAHAEINEAGR; GenScript, Piscataway, NJ) and were cocultured 1:10 with CD4 CD62L OTII T cells labeled with carboxyfluorescein succinimidyl ester (1 mol/l) for 4 days in the presence of 100 ng/ml LPS. Induction of Colitis Dextran sodium sulfate (DSS) colitis was induced by adding 4% (wt/vol) DSS to the drinking water for 5 days. T-cell transfer colitis was induced by adoptive transfer of CD4 CD62L splenic T cells intraperitoneally into Rag1 / mice. Mice were killed when they had lost 20% of their initial body weight. Histology Histologic scoring was performed as described. 18 Statistical Analysis Data are shown as mean SD. Student t test or one-way analysis of variance followed by Dunn s test or Student Newman Keuls test were performed using SigmaStat (Systat Software GmbH, Erkrath, Germany). P values less than.05 were considered to indicate statistically significant differences ( * P.05, ** P.01, *** P.001). Results Ccl17-Deficient Mice Are Protected From Severe Experimental Colitis Ccl17 mrna expression in the colon is induced during active colitis. 15 We investigated the expression of CCL17 in DCs during experimental murine colitis using Ccl17 E/E reporter mice expressing enhanced green fluorescent protein (egfp) instead of CCL17. DSS treatment led to an increase in the percentage of Ccl17-eGFP expressing DCs in the MLNs of Ccl17 E/E mice (Figure 1A). Also, a 2- to 3-fold increase in the percentage of Ccl17-eGFP DCs was found in MLNs, LPLs, and IELs of Rag1 / Ccl17 E/E mice, which had received a CD4 CD62L T-cell transfer for induction of colitis 21 days earlier compared with control mice. No induction of CCL17 expression was found before the onset of colitis (Figure 1B). egfp cells expressed high levels of CD11c, CD11b, and major histocompatibility complex class II but very low levels of F4/80 (Supplementary Figure 1). In the DSS model as well as the T-cell transfer model, severe colitis did not develop in the absence of CCL17. The induction of colitis by DSS led to a significantly reduced weight loss in Ccl17 E/E animals ( 6.2% 2.7 %) compared with wild-type animals ( 15.9% 4.3%) at day 7(Figure 1C, right panel). Ccl17 E/E mice had significantly less colonic shortening and a lower histologic score ( [Ccl17 E/E ]vs [wild-type]), representing diminished colonic inflammation (Figure 1D). The induction of colitis by T-cell transfer also led to a significantly blunted reduction in body weight in Rag1 / Ccl17 E/E compared with Rag1 / control mice ( 0.1% 6.3% vs 18.9% 8.1%, day 21) (Figure 1E, right panel), and colonic inflammation was also almost entirely suppressed in Rag1 / Ccl17 E/E mice (histologic score vs , day 21) (Figure 1F). These findings show that CCL17 plays a crucial role in development of colitis and that its absence protects from active colitis. Lower Induction of Proinflammatory Cytokines During Colitis in Ccl17 E/E Mice To gain further insight into the proinflammatory activity of CCL17, we investigated the cytokine expression

3 February 2012 ROLE OF CCL17 IN MURINE COLITIS 337 Figure 1. CCL17 expression in intestinal and MLN DCs after induction of colitis. (A) CCL17- egfp expression was measured by fluorescence-activated cell sorting in CD3 CD11c DCs in the MLNs of CCL17 E/E mice either untreated or treated for 5 days with 4% DSS in the drinking water and analyzed on day 7. (B) CCL17-eGFP expression by DCs was detected by fluorescence-activated cell sorting analysis in MLNs, colonic LPLs, and IELs of Rag1 / Ccl17 E/E mice (gray, negative control; black, egfp). (C) Body weight was measured in mice treated with 4% DSS for 5 days (n 4 [left panel] and n 12 [right panel, day 7]; Student t test). (D) Colon length (n 8) and histologic score (n 12) were determined. H&E-stained colon sections were used to assess severity of colitis (Student t test). (E) Body weight was measured at the indicated time points after transfer of T cells into Rag1 / and Rag1 / Ccl17 E/E mice (n 3[left panel] and n 17 [right panel, day 21]). (F) Histologic scoring was performed on colon sections (n 5; Student t test). Results of one representative of at least 3 independent experiments are shown in panels A, B, and F; mean SD is shown in panels C F; cumulative data are shown in panels C E.**P.01, ***P.001. profile in the colon and the MLNs during colitis. In DSS-induced colitis, no significant difference was observed in tumor necrosis factor, IL-6, and IL-23 mrna expression levels in colonic tissue and MLNs between the 2 mouse strains (Supplementary Figure 2A). However, expression of IL-12p35 mrna in the MLNs and of interferon (IFN)- mrna in the colon was significantly higher in wild-type mice compared with Ccl17 E/E mice treated with DSS (Figure 2A) with no significant differences vice versa (Supplementary Figure 2A). These results suggest that IL-12 expression in MLNs and IFN- expression in the colon are selectively enhanced by DC-derived CCL17 in acute DSS-induced colitis, correlating with higher severity of colitis. In the T-cell transfer model of colitis, Rag1 / and Rag1 / Ccl17 E/E mice did not show differences in proinflammatory cytokine expression in colon tissue as early as 5 to 8 days after T-cell transfer (Supplementary Figure 2B). However, 21 days after transfer, tumor necrosis factor, IL-6, IFN-, IL-17A, and IL-22 expression were significantly higher in colonic tissue of Rag1 / mice compared with Rag1 / Ccl17 E/E mice (Figure 2C). We also found significantly lower IL-23 receptor expression in Ccl17-deficient recipient mice 21 days after T-cell transfer (Figure 2C). In the MLNs of T-cell transfer mice, induction of IL-12 and IFN- was significantly stronger in wild-type mice (Figure 2B). Consistent with these results, increased percentages of IFN- producing, IL-17 producing, and IFN- /IL-17 double-producing T cells were detected in MLNs, colon tissue, and spleen of CCL17-competent versus CCL17- deficient mice (Figure 2D and E). Only a low percentage of CD3-negative LPLs and IELs were found to produce IFN- and/or IL-17 (Supplementary Figure 7). Thus, the reduced induction of colitis in CCL17-deficient mice is also reflected in diminished cytokine induction in the MLNs and the colon as well as in reduced Th1 and Th17 cell generation.

4 338 HEISEKE ET AL GASTROENTEROLOGY Vol. 142, No. 2 Figure 2. Enhanced proinflammatory cytokine expression in the MLNs and colon of WT and Ccl17 E/E mice during colitis. (A) Cytokine mrna expression was quantified by qrt-pcr in MLNs and colon of DSS-treated WT and Ccl17 E/E mice after 5 days of treatment with 4% DSS followed by 2 days of drinking water (fold induction in WT compared with Ccl17 E/E mice, n 4; Student t test). (B) Cytokine mrna expression profile of MLNs 21 days after T-cell transfer (fold induction in Rag1 / compared with Rag1 / Ccl17 E/E mice, n 3; Student t test). (A and B) Results of one representative of at least 3 independent experiments with 3 4 mice per group are displayed as mean SD. (C) Cytokine mrna expression was measured in colon tissues 21 days after T-cell transfer in Rag1 / and Rag1 / Ccl17 E/E mice (fold induction compared with steady-state expression without T-cell transfer) (n 6 10, merged data of 2 3 experiments; Student t test). (D) Representative fluorescence-activated cell sorting plots of intracellular cytokine staining of IFN- and IL-17 in CD3 /CD4 T cells. (E) Cumulative data of intracellular cytokine production in T cells (n 9 15, Student t test). *P.05, **P.01, ***P.001. CCL17 Promotes T-Cell Transfer Colitis Independently of CCR4-Mediated T-Cell Recruitment but Requires CCR4 Expression in Innate Immune Cells The observed differences in the cytokine profile might be due to a reduced recruitment of T cells via the CCR4-CCL17 axis in Rag1 / Ccl17 E/E mice. Upon transfer of Ccr4 / CD4 CD62L T cells, Rag1 / Ccl17 E/E mice lost almost no weight whereas Rag1 / mice lost up to 20% of their body weight (Figure 3A), similar to what was observed after transfer of wild-type (WT) T cells (Figure 1E). Thus, lack of CCR4 expression on transferred T cells did not recapitulate the phenotype of Rag1 / Ccl17 E/E mice after transfer of WT T cells. We also found a similar percentage of CD4 T cells in MLNs 5 days after transfer and in MLN, LPL, and IEL populations 21 days after transfer into Rag1 / versus Rag1 / Ccl17 E/E mice (Figure 3B and C). These results indicate that an altered recruitment of T cells via the CCR4-CCL17 axis is not responsible for the differences observed in development of colitis

5 February 2012 ROLE OF CCL17 IN MURINE COLITIS 339 Figure 3. Transfer of Ccr4 / T cells leads to induction of colitis in Rag1 / but not Rag1 / Ccl17 E/E mice. (A) CD4 CD62L T cells were isolated from Ccr4 / mice and transferred into Rag1 / and Rag1 / Ccl17 E/E mice, and body weight was measured (n 4; Student t test). (B) The percentage of CD4 T cells in MLNs was determined by fluorescence-activated cell sorting 5 days after transfer (n 3). (C) The percentage of CD4 T cells in MLNs, LPLs, and IELs was determined by fluorescence-activated cell sorting 21 days after T-cell transfer (n 3). (D) Body weight of Rag1 / Ccr4 / and Rag1 / mice after T-cell transfer (n 4; Student t test). (A D) Data represent mean SD. *P.05, **P.01, ***P.001. between WT and Ccl17-deficient mice. However, CCR4- deficient Rag1 / recipient mice were protected from induction of colitis by WT T-cell transfer, similar to Rag1 / Ccl17 E/E mice (Figure 3D), suggesting that CCL17 exerts its proinflammatory effects via CCR4 expressed on innate immune cells in Rag1 / recipient mice. WT DCs Produce Higher Amounts of Cytokines Than Ccl17-Deficient DCs and CCL17 Directly Induces IL-12 Production in DCs Because we found no difference in the recruitment of T cells as well as DCs into MLNs and colon in experimental colitis (Figure 3B and C and Supplementary Figure 3), we hypothesized that CCL17 has an autocrine or a paracrine stimulatory effect on DCs. To test this hypothesis, we stimulated Ccl17-deficient and WT BMDCs with LPS in vitro. After 24 hours, no relevant differences were detected in the expression of costimulatory molecules or major histocompatibility complex class II between WT and Ccl17 E/E DCs (Figure 4A). However, stimulation with LPS for 6 hours led to a significantly reduced induction in IL-12p35 and IL-23p19 mrna levels in Ccl17 E/E DCs compared with control DCs, whereas IL-6 expression was not differentially induced and IL-10 only tended to be reduced (Figure 4B). This was also true for stimulation with the Toll-like receptor (TLR) 2 ligand Pam 3 Cys (Figure 4C). Reduced cytokine expression by Ccl17-deficient DCs could be due to a direct, autocrine effect of CCL17 on the DCs. As shown in Figure 4D, CCL17 alone did not induce IL-12p40 production, but when Ccl17 E/E DCs were stimulated simultaneously with a low dose of LPS, IL-12 production was further enhanced by CCL17 in a dose-dependent manner. Furthermore, LPS-induced expression of IL-23p19 and IL-10 mrna and IL-12p40 protein expression tended to be reduced in DCs lacking the CCL17 receptor CCR4 (Figure 4E), suggesting the involvement of CCR4 in the amplification of the DC cytokine response by CCL17. Thus, these results show that CCL17 has a direct autocrine or paracrine effect on the activation of DCs enhancing TLR-induced cytokine expression. WT DCs Are More Potent T-Cell Activators Than Ccl17-Deficient DCs Reduced production of Th1/Th17-inducing cytokines by Ccl17-deficient DCs might affect Th cell differentiation, which is critical for development of colitis. To investigate this, OVA-pulsed BMDCs from WT or Ccl17 E/E mice were cocultured with CD4 CD62L OTII T cells in the presence of LPS for 4 days. Although no differences in the proliferation of T cells were detectable (Supplementary Figure 4A), more IFN- producing T cells were induced (Supplementary Figure 4B) and higher amounts of IFN- were produced (Figure 5A), correlating with higher expression of T-bet mrna (Figure 5B), when the T cells were cultured with WT DCs compared with Ccl17 E/E DCs, whereas the production of IL-17A was not affected (Figure 5C). We therefore cocultured OTII T cells with OVApulsed DCs isolated from the MLNs of WT or Ccl17 E/E mice treated with 100 g LPS intraperitoneally 18 hours earlier. A higher induction of transcription factor RORc mrna was observed in T cells cocultured with DCs isolated from MLNs of WT mice (Figure 5E), correlating with higher levels of IL-17A (Figure 5D) but not IFN- in the

6 340 HEISEKE ET AL GASTROENTEROLOGY Vol. 142, No. 2 Figure 4. CCL17-dependent enhancement of TLR ligand induced cytokine expression by BMDCs. (A) The expression of activation markers by WT and Ccl17 E/E BMDCs stimulated with 100 ng/ml LPS for 24 hours was determined by fluorescence-activated cell sorting (light gray: fluorescence minus one control, mean fluorescence intensity [MFI], and percentage of positive cells are indicated). (B and C) C57BL/6 WT and Ccl17 E/E BMDCs were stimulated with (B) 100 ng/ml LPS or (C) Pam 3 Cys for 6 hours. Cytokine mrna expression was determined by qrt-pcr (fold induction compared with unstimulated control; n 3; Student t test). (D) Ccl17 E/E BMDCs were stimulated with recombinant murine CCL17 and LPS. After 24 hours, IL-12p40 was measured in the supernatants by enzyme-linked immunosorbent assay. (E) WTandCcr4 / BM- DCs were stimulated with 100 ng/ml LPS for 6 hours. Cytokine mrna expression was measured as in panel B. One representative experiment is shown; mean SD of 3 experimental replicates is shown in panels D and E. *P.05, **P.01. coculture supernatants (Figure 5F). In both experimental settings, no difference in Foxp3 expression was observed between T cells cocultured with WT and Ccl17-deficient DCs in the absence of transforming growth factor (TGF)- (Supplementary Figure 4C). Thus, correlating with lower expression of IL-12 and IL-23, despite comparable expression of IL-6, Ccl17-deficient DCs activated by LPS have a lower capacity to induce IFN- producing T cells in vitro and IL-17 producing T cells ex vivo. These results are consistent with the detection of a lower percentage of Th1 and Th17 cells in the colons of CCL17- deficient mice in T-cell transfer colitis. Foxp3 Tregs Mediate Long-term Protection From Colitis in Rag1 / Ccl17 E/E Mice To investigate whether CCL17 deficiency offers long-term protection from colitis, the body weight of Rag1 / Ccl17 E/E mice was monitored for up to 47 days after T-cell transfer. As shown in Figure 6A, these mice were able to sustain their body weight. However, Rag1 / Ccl17 E/E mice failed to gain weight as would have been expected in healthy mice, suggesting that inflammation was markedly reduced but not entirely suppressed in mice lacking CCL17. We therefore examined if Tregs are re-

7 February 2012 ROLE OF CCL17 IN MURINE COLITIS 341 Figure 5. Reduced Th1 and Th17 cell induction by LPS-stimulated DCs lacking CCL17. (A C) OTII T cells were cocultured with WT or Ccl17 E/E BMDCs pulsed with OVA-peptide (5 g/ml) and stimulated with LPS (100 ng/ml) during the coculture for 4 days. T-bet mrna expression was quantified by qrt-pcr (fold induction compared with unstimulated T cells). IFN- and IL-17A concentrations in the coculture supernatants were measured by enzyme-linked immunosorbent assay. (D F) OTII T cells were cocultured for 5 days with OVA-peptide pulsed CD11c DCs, isolated from MLNs of WT and Ccl17 E/E mice injected with 100 g LPS 18 hours earlier. RORc mrna expression was measured by qrt-pcr. IFN- and IL-17 concentrations in the supernatants were measured by enzyme-linked immunosorbent assay. Results of one of 2 experiments with similar results are shown in panels B, C, E, and F. Data are shown as mean SD (n 3, Student t test) in panels A and D. *P.05. sponsible for the long-term immunoregulatory effect seen in this model. The CD4 CD62L T cells that were used for induction of colitis routinely contained about 6% to 8% Foxp3 Tregs expressing CD62L, which do not prevent initiation of colitis, however. Similarly, CD45RB hi T cells used for induction of colitis have been described to contain a small fraction of Foxp3 Tregs. 19 As shown in Figure 6B, the fraction of Tregs of all CD4 T cells progressively declined during the first 3 weeks after T-cell transfer in both mouse strains; however, the loss of Tregs was more pronounced in Ccl17-competent than in Ccl17- deficient Rag1 / mice. Analysis of MLNs revealed no difference in the fractions of Tregs 16 hours as well as 5 days after transfer. However, 21 days after transfer, a significantly higher percentage of Tregs was found in MLNs of Rag1 / Ccl17 E/E mice as compared with Rag1 / mice (3.8% 1.4% vs 1.2% 0.2%, n 3, P.05, Figure 6B). After 47 days, the fraction of Tregs in Rag1 / Ccl17 E/E mice had increased to 14.8% 2.4% (Figure 6B). Similar to the MLNs, a significantly higher percentage of Tregs was found in the LPL ( vs ) and IEL fraction ( vs ) of Rag1 / Ccl17 E/E compared with Rag1 / mice 21 days after transfer (Figure 6D). We also found higher mrna levels of Foxp3 in the colon at days 21 and 47 (Figure 6D). This finding and the sustained levels of TGF- and Aldh1a2 as well as elevated levels of granulocyte-macrophage colony stimulating factor in the colon of Rag1 / Ccl17 E/E mice (Figure 6E) support the interpretation that the observed increase in the fraction of Tregs reflects the selective maintenance and expansion of Tregs in Rag1 / Ccl17 E/E recipient mice rather than reduced expansion or loss of CD4 effector cells. In addition to the high levels of Tregs detected in the colon of Rag1 / Ccl17 E/E mice after 47 days, we found higher numbers of CD103 DCs in the LPLs and the IELs of Rag1 / Ccl17 E/E mice compared with Rag1 / mice 21 days after transfer (Supplementary Figure 5). Analysis of WT compared with Ccl17 E/E mice revealed higher levels of Foxp3 mrna in the MLNs of Ccl17 E/E mice in the steady state (Supplementary Figure 6). In addition, a higher induction of Foxp3 could be detected in T cells cocultivated with DCs isolated from MLNs of LPS-treated Rag1 / Ccl17 E/E mice than Rag1 / mice (Figure 6F). To elucidate if Tregs are required for the observed protection of Rag1 / Ccl17 E/E mice, CD4 CD62L T cells used for transfer were isolated from DEREG 23.2 mice 17 after Tregs had been depleted (Figure 7A). Weight loss was significantly more rapid in Rag1 / compared with

8 342 HEISEKE ET AL GASTROENTEROLOGY Vol. 142, No. 2 Figure 6. Correlation of longterm protection from colitis with maintenance and expansion of Foxp3 Tregs in Rag1 / Ccl17 E/E mice. (A) Colitis was induced by transfer of CD4 CD62L WT T cells, and body weight was measured. (B) The percentages of Foxp3 T cells in MLNs were measured by fluorescence-activated cell sorting (n 3, Student t test). (C) The percentage of Foxp3 cells within CD4 T cells in IELs and LPLs was determined 21 days after T-cell transfer. (D) Foxp3 mrna induction was measured in colon tissues of Rag1 / and Rag1 / Ccl17 E/E mice compared with control mice (n 3, Student t test, P.01, comparison of Rag1 / [day 21] and Rag1 / Ccl17 E/E [day 47]). (E) TGF- (tgf- 1, granulocytemacrophage colony stimulating factor (csf2), and Aldh1a2 (aldh1a2) mrna expression were measured in colon tissues of Rag1 / and Rag1 / Ccl17 E/E mice compared with control mice (n 3; Student t test). Results representative of at least 2 independent experiments are shown (mean SD in panels A, B, D, and E). (F) Foxp3 expression in T cells cocultivated with CD11c MLN DCs from LPSstimulated mice for 4 days in the presence of TGF-, IL-2, and -CD3 (one representative of 4 independent experiments with comparable differences is displayed). *P.05, **P.01, ***P.001. Rag1 / Ccl17 E/E mice after transfer of Treg-depleted T cells, similar to the results obtained with transfer of T cells containing natural Tregs. However, Rag1 / Ccl17 E/E mice were no longer protected from colitis at later time points after transfer of Treg-depleted T cells (Figure 7B) and no de novo generation or expansion of Tregs was observed in LPLs, IELs, and MLNs (Figure 7C). Additionally, cytokine expression was comparable in the colons of Rag1 / Ccl17 E/E and Rag1 / recipient mice reconstituted with Treg-depleted CD4 CD62L T cells when the experiment was terminated, reflecting similar degrees of inflammation (Figure 7D). These results indicate that Tregs are essential for long-term protection of Ccl17-deficient recipient mice from development of severe colitis. Discussion Previous studies have suggested that CCL17 is predominantly expressed in cutaneous DCs, where it is responsible for recruitment of T cells into the skin, whereas CCL25 and its receptor CCR9 are essential for T-cell homing to the intestinal mucosa. 20 However, we and others have shown that CCL17 is expressed in mucosal DCs and up-regulated in Crohn s disease as well as murine experimental colitis ,21 In our study, the absence of CCL17 protected mice from severe colitis in T cell independent and dependent models of colitis. Mice lacking CCL17 produced less proinflammatory cytokines (IL-22, IL-17, IFN-, IL-6, and tumor necrosis factor ) after induction of colitis. To investigate whether the reduced cytokine production in Ccl17-deficient mice was due to a direct CCL17-dependent action, we analyzed the cytokine profile of stimulated DCs and performed coculture experiments to determine T-cell activation and differentiation. On stimulation with LPS, Ccl17-deficient as well as Ccr4- deficient BMDCs produced less IL-12 and IL-23, whereas the production of IL-6 was equal in both groups. This defect in cytokine production in Ccl17-deficient DCs could be overcome by adding recombinant CCL17 to the

9 February 2012 ROLE OF CCL17 IN MURINE COLITIS 343 Figure 7. Long-term protection of Rag1 / Ccl17 E/E mice from colitis is mediated by cotransferred Tregs. (A) CD4 CD62L T cells were isolated from WT mice and DEREG 23.2 mice, which had been injected with 500 ng diphtheria toxin to deplete Tregs. The percentage of Foxp3 CD4 Tregs in the CD4 CD62L T-cell population was determined by fluorescence-activated cell sorting. (B) Body weight of Rag1 / and Rag1 / Ccl17 E/E mice was measured after transfer of Foxp3-depleted CD4 CD62L T cells and compared with the body weight of Rag1 / Ccl17 E/E mice that received WT T cells (n 3; Student t test; *P.05 for Rag1 / Ccl17 E/E vs Rag1 / mice reconstituted with Foxp3- depleted T cells; ## P.01 for Rag1 / Ccl17 E/E mice reconstituted with WT T cells vs Rag1 / Ccl17 E/E mice reconstituted with Foxp3-depleted T cells). (C) The percentages of Foxp3 CD4 Tregs in IELs, LPLs, and MLNs were determined at the indicated time points after T-cell transfer. (D) Cytokine mrna expression was measured by qrt- PCR in colon tissues of mice reconstituted with Foxp3-depleted CD4 CD62L T cells at the indicated time points (fold induction compared with control mice). Results representative of at least 2 independent experiments are displayed (mean SD is shown in panels B and D). culture. IL-23 plays a key role in intestinal inflammation 19,22,23 and, in combination with IL-6, IL-23 is an important player in the differentiation of Th17 cells, which also produce IL The emergence of IFN- / IL-17 double-producing Th cells in colitis depends on IL-23R expression in Th cells. 25 Accordingly, the lower percentage of IFN- /IL-17 producing Th cells in the colon of CCL17-deficient mice during colitis correlated with reduced IL-23R expression. Furthermore, IL-23 not only activates T cells but also has a stimulating effect on innate effector cells. 22 In an ex vivo coculture system, DCs isolated from WT MLNs induced a significantly stronger Th17 response than DCs isolated from MLNs of Ccl17- deficient animals. These results show that CCL17 not only attracts cells to a population of DCs producing proinflammatory cytokines but, in an autocrine or paracrine feedback loop via CCR4 on DCs, also actively enhances the TLR ligand induced production of cytokines, hence promoting Th1/Th17 differentiation by the CCL17-producing DCs. The finding that the chemokine CCL17 can act in such a cytokine-like manner has not been shown before and has only been reported for a few other chemo-

10 344 HEISEKE ET AL GASTROENTEROLOGY Vol. 142, No. 2 kines such as CX3CL1 26 and CCL2. 27 A recent study showed that CCL17 is required for release of intracellular calcium stores and migration of cutaneous DCs in response to CCR7 and CXCR4 ligands. 13 A similar facilitation of TLR signaling could be responsible for the observed enhancing effect of CCL17 on cytokine production by DCs in our study. CCR4 is the only confirmed receptor for CCL17. Therefore, our data obtained with transfer of Ccr4 / T cells for induction of colitis show that the protection of Ccl17- deficient mice was not due to a reduced recruitment of CD4 T cells to the colonic mucosa or MLNs, but rather correlated with a reduced production of proinflammatory cytokines. Accordingly, we were able to show the protective effect of CCL17 deficiency also in the T cell independent acute DSS-induced colitis model. In contrast to Yuan et al, who reported that CCR4-mediated recruitment of Tregs to the MLNs is necessary to reverse the induction of colitis in the T-cell transfer model, 10 we saw no difference in the recruitment of Tregs or effector T cells or DCs during T-cell transfer colitis and lack of CCR4 expression on transferred T cells did not protect Ccl17-competent mice from development of colitis. Thus, CCL17 must act on innate immune cells expressing CCR4, such as DCs. Other CCR4 cells, including innate effector lymphocytes, might also contribute to this effect, 22 although no significant production of IFN- or IL-17 was observed in CD3 cells in the colon during T-cell transfer colitis. Long-term experiments using Ccl17-deficient mice confirmed that these mice do not develop relevant clinical signs of disease even at later time points. These results suggest that inflammation is kept in check by immune regulatory factors in Ccl17-deficient mice. Several publications addressed the issue of expansion and induction of Tregs during intestinal inflammation, 19,28 and Tregs have been shown to possess the capacity to actively suppress colonic inflammation. 29 We therefore analyzed the Treg compartment in more detail and found that the fraction of Tregs transferred with the initial graft of CD4 CD62L T cells is maintained initially and then strongly expanded in mice lacking CCL17, in contrast to CCL17-competent mice, suggesting that these cells are responsible for longterm protection. Among several factors, TGF- 2, Aldh1a2, 30 and granulocyte-macrophage colony stimulating factor have been reported to promote and/or maintain Foxp3 expression in T cells. Granulocyte-macrophage colony stimulating factor has also been shown to be beneficial in treating patients with IBD. 34 Analysis of these 3 factors in the colon of Ccl17-deficient mice clearly showed that they were either sustained or even induced, compared with the levels observed in WT mice, explaining the higher numbers of Tregs in Ccl17-deficient mice. Along this line, the lower levels of IL-23 and IL-6 measured in Ccl17-deficient mice allow expansion of Tregs during intestinal inflammation, whereas the higher levels of IL-23 and IL-6 in WT mice are inhibitory. 19,24,25 By transferring Foxp3-depleted T cells, we show that the cotransferred natural Tregs are essential to establish the long-term protection of Ccl17-deficient mice, whereas protection during the early phase of colitis correlates with the reduced level of proinflammatory cytokines. We also provide clear evidence that the high number of Tregs at day 47 in Ccl17-deficient mice is attributable to expansion of the cotransferred natural Tregs and not to newly generated Tregs. In summary, the data from this study show for the first time that CCL17 has a central role in the orchestration of acute and chronic intestinal inflammation. CCL17 displays a dual function by enhancing expression of proinflammatory cytokines that promote Th1/Th17 differentiation of recruited naïve T cells and by preventing Treg expansion in the MLNs and the colonic lamina propria. The data obtained in our murine models of colitis suggest that investigation into the role of CCL17 in patients could provide further insights into the mechanisms promoting intestinal pathogenesis. The fact that CCL17 has been found to be up-regulated in patients with Crohn s disease would certainly support this line of reasoning. 16 Targeting the CCR4-CCL17 axis might turn out to be a promising new strategy for the treatment of intestinal inflammatory diseases. Supplementary Materials Note: To access the supplementary material accompanying this article, visit the online version of Gastroenterology at and at doi: /j.gastro References 1. Podolsky DK. Inflammatory bowel disease. N Engl J Med 2002; 347: Coombes JL, Siddiqui KR, Arancibia-Carcamo CV, et al. A functionally specialized population of mucosal CD103 DCs induces Foxp3 regulatory T cells via a TGF-beta and retinoic acid-dependent mechanism. J Exp Med 2007;204: Reiss Y, Proudfoot AE, Power CA, et al. CC chemokine receptor (CCR)4 and the CCR10 ligand cutaneous T cell-attracting chemokine (CTACK) in lymphocyte trafficking to inflamed skin. J Exp Med 2001;194: Saeki H, Tamaki K. Thymus and activation regulated chemokine (TARC)/CCL17 and skin diseases. J Dermatol Sci 2006;43: Vijayanand P, Durkin K, Hartmann G, et al. Chemokine receptor 4 plays a key role in T cell recruitment into the airways of asthmatic patients. J Immunol 2010;184: Inngjerdingen M, Damaj B, Maghazachi AA. Human NK cells express CC chemokine receptors 4 and 8 and respond to thymus and activation-regulated chemokine, macrophage-derived chemokine, and I-309. J Immunol 2000;164: Ness TL, Ewing JL, Hogaboam CM, et al. CCR4 is a key modulator of innate immune responses. J Immunol 2006;177: Sallusto F, Lanzavecchia A. Mobilizing dendritic cells for tolerance, priming, and chronic inflammation. J Exp Med 1999;189: Semmling V, Lukacs-Kornek V, Thaiss CA, et al. Alternative crosspriming through CCL17-CCR4-mediated attraction of CTLs toward NKT cell-licensed DCs. Nat Immunol 2010;11: Yuan Q, Bromley SK, Means TK, et al. CCR4-dependent regulatory T cell function in inflammatory bowel disease. J Exp Med 2007; 204:

11 February 2012 ROLE OF CCL17 IN MURINE COLITIS Lee I, Wang L, Wells AD, et al. Recruitment of Foxp3 T regulatory cells mediating allograft tolerance depends on the CCR4 chemokine receptor. J Exp Med 2005;201: Alferink J, Lieberam I, Reindl W, et al. Compartmentalized production of CCL17 in vivo: strong inducibility in peripheral dendritic cells contrasts selective absence from the spleen. J Exp Med 2003;197: Stutte S, Quast T, Gerbitzki N, et al. Requirement of CCL17 for CCR7- and CXCR4-dependent migration of cutaneous dendritic cells. Proc Natl Acad Sci U S A 2010;107: Kristensen NN, Brudzewsky D, Gad M, et al. Chemokines involved in protection from colitis by CD4 CD25 regulatory T cells. Inflamm Bowel Dis 2006;12: Scheerens H, Hessel E, de Waal-Malefyt R, et al. Characterization of chemokines and chemokine receptors in two murine models of inflammatory bowel disease: IL-10-/- mice and Rag-2-/- mice reconstituted with CD4 CD45RBhigh T cells. Eur J Immunol 2001; 31: Jugde F, Alizadeh M, Boissier C, et al. Quantitation of chemokines (MDC, TARC) expression in mucosa from Crohn s disease and ulcerative colitis. Eur Cytokine Netw 2001;12: Lahl K, Loddenkemper C, Drouin C, et al. Selective depletion of Foxp3 regulatory T cells induces a scurfy-like disease. J Exp Med 2007;204: Siegmund B, Rieder F, Albrich S, et al. Adenosine kinase inhibitor GP515 improves experimental colitis in mice. J Pharmacol Exp Ther 2001;296: Izcue A, Hue S, Buonocore S, et al. Interleukin-23 restrains regulatory T cell activity to drive T cell-dependent colitis. Immunity 2008;28: Pabst O, Ohl L, Wendland M, et al. Chemokine receptor CCR9 contributes to the localization of plasma cells to the small intestine. J Exp Med 2004;199: Melgar S, Drmotova M, Rehnstrom E, et al. Local production of chemokines and prostaglandin E2 in the acute, chronic and recovery phase of murine experimental colitis. Cytokine 2006;35: Buonocore S, Ahern PP, Uhlig HH, et al. Innate lymphoid cells drive interleukin-23-dependent innate intestinal pathology. Nature 2010;464: Kullberg MC, Jankovic D, Feng CG, et al. IL-23 plays a key role in Helicobacter hepaticus-induced T cell-dependent colitis. J Exp Med 2006;203: Bettelli E, Carrier Y, Gao W, et al. Reciprocal developmental pathways for the generation of pathogenic effector TH17 and regulatory T cells. Nature 2006;441: Ahern PP, Schiering C, Buonocore S, et al. Interleukin-23 drives intestinal inflammation through direct activity on T cells. Immunity 2010;33: Ishida Y, Hayashi T, Goto T, et al. Essential involvement of CX3CR1-mediated signals in the bactericidal host defense during septic peritonitis. J Immunol 2008;181: Jimenez F, Quinones MP, Martinez HG, et al. CCR2 plays a critical role in dendritic cell maturation: possible role of CCL2 and NFkappa B. J Immunol 2010;184: Park SG, Mathur R, Long M, et al. T regulatory cells maintain intestinal homeostasis by suppressing gammadelta T cells. Immunity 2010;33: Mottet C, Uhlig HH, Powrie F. Cutting edge: cure of colitis by CD4 CD25 regulatory T cells. J Immunol 2003;170: Mucida D, Pino-Lagos K, Kim G, et al. Retinoic acid can directly promote TGF-beta-mediated Foxp3( ) Treg cell conversion of naive T cells. Immunity 2009;30: ; author reply Cheatem D, Ganesh BB, Gangi E, et al. Modulation of dendritic cells using granulocyte-macrophage colony-stimulating factor (GM- CSF) delays type 1 diabetes by enhancing CD4 CD25 regulatory T cell function. Clin Immunol 2009;131: Kared H, Leforban B, Montandon R, et al. Role of GM-CSF in tolerance induction by mobilized hematopoietic progenitors. Blood 2008;112: Yokota A, Takeuchi H, Maeda N, et al. GM-CSF and IL-4 synergistically trigger dendritic cells to acquire retinoic acid-producing capacity. Int Immunol 2009;21: Barahona-Garrido J, Yamamoto-Furusho JK. New treatment options in the management of IBD - focus on colony stimulating factors. Biologics 2008;2: Received April 21, Accepted October 17, Reprint requests Address requests for reprints to: Wolfgang Reindl, MD, and Anne Krug, MD, II. Medizinische Klinik, Klinikum rechts der Isar, Technische Universität München, Ismaninger Str. 22, D Munich, Germany. wolfgang.reindl@lrz.tum.de and anne.krug@lrz.tum.de; fax: (49) Acknowledgments A.K. and W.R. contributed equally to this study. The authors thank A. Ölsner for excellent technical assistance, K. Eisenächer for critically reading the manuscript, and the collaborators of the diagnostic laboratory of the Institute of Pathology (CHUV) for excellent technical assistance. The DEREG 23.2 mice were a kind gift of T. Sparwasser. Conflicts of interest The authors disclose no conflicts. Funding Supported by the German Research Council DFG grant KR2199/3-1 and SFB 571 (to A.K. and W.R.). A.F.H. is supported by the DFG Graduate School GRK 1482 and TUM Graduate School. This work is part of the thesis of A.F.H.