These authors contributed equally to this work.
In vitro-induced Th17 cells fail to induce inflammation in vivo and show an impaired migration into inflamed sites
Article first published online: 25 JAN 2010
Copyright © 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
European Journal of Immunology
Volume 40, Issue 4, pages 1089–1098, April 2010
How to Cite
Janke, M., Peine, M., Nass, A., Morawietz, L., Hamann, A. and Scheffold, A. (2010), In vitro-induced Th17 cells fail to induce inflammation in vivo and show an impaired migration into inflamed sites. Eur. J. Immunol., 40: 1089–1098. doi: 10.1002/eji.200939487
- Issue published online: 14 APR 2010
- Article first published online: 25 JAN 2010
- Manuscript Accepted: 12 JAN 2010
- Manuscript Revised: 27 NOV 2009
- Manuscript Received: 2 APR 2009
- Deutsche Forschungsgemeinschaft ENDO-Stiftung–Stiftung des Gemeinnützigen Vereins ENDO-Klinik e.V.. Grant Number: sfb650
- European Community. Grant Numbers: OIGU40719-22, LSHG-CT-2007-037291, LSHG-CT-2005-00523
- Cell migration;
- Th17 cells
Recently, IL-17 produced by Th17 cells was described as pro-inflammatory cytokine with an eminent role in autoimmune diseases, e.g. rheumatoid arthritis. A lack of IL-17 leads to amelioration of collagen-induced arthritis. IL-17 induction in naïve CD4+ T cells depends on IL-6 and TGF-β and is enhanced by IL-23. The in vivo inflammatory potential of in vitro-primed Th17 cells however, remains unclear. Here, we show that, although IL-17 neutralisation results in amelioration of murine OVA-induced arthritis, in vitro-primed Th17 cells cannot exacerbate arthritic symptoms after adoptive transfer. Furthermore, Th17 cells cannot induce an inflammatory delayed type hypersensitivity reaction because they fail to migrate into inflamed sites, possibly due to the lack of CXCR3 expression. Also, re-isolated Th17 cells acquired IFN-γ expression, indicating instability of the Th17 phenotype. Taken together, the data show that IL-6, TGF-β and IL-23 might not provide sufficient signals to induce “fully qualified” Th17 cells.
In the last years, the pro-inflammatory cytokines IL-23 and IL-17 came into focus to play a major role in inflammatory autoimmune diseases. IL-23, a heterodimeric protein consisting of p40 (shared subunit with IL-12) and p19, was identified as an important factor to induce or maintain inflammatory diseases, for p19-deficient mice are resistant to EAE and colitis 1, 2. IL-17 acts on multiple cell types to evoke the release of other pro-inflammatory cytokines like TNF-α, IL-1β or IL-6 as well as chemokines like IL-8 to attract neutrophils or macrophages to the inflamed site 3, 4. This may lead to a rising inflammatory response. In addition, IL-17 directly causes bone and cartilage destruction via upregulation of RANK-L and therefore, an eminent role of IL-17 in autoimmune diseases, like rheumatoid arthritis, was reported 5, 6. In 2006 Veldhoen et al. and others showed that the combination of IL-6 and TGF-β is sufficient to induce IL-17 expression by CD4 T cells in vitro 7. Furthermore, by blocking Th1 and Th2 differentiation pathways, the efficiency of Th17 differentiation can be enhanced. Therefore, IL-17-producing Th cells were claimed to represent a distinct Th-cell lineage, termed Th17 cells 7–9. However, IL-1β was reported to be necessary for Th17 induction in vivo and IL-6-independent pathways for IL-17 induction may exist, for IL-6-deficient mice show diminished but detectable levels of Th17 cells 10, 11. RORγT was claimed to be the Th17 lineage-specific transcription factor and sufficient to induce the Th17 phenotype in vitro in naïve T cells 12. The role of IL-23 is currently under discussion. It might be important for the stabilisation or proliferation of Th17 cells, because in vitro the adding of IL-23 leads to increased numbers of IL-17-expressing cells 13. However, whether Th17 cells induced in vitro by IL-6, TGF-β and IL-23 represent fully functional cells relevant for in vivo inflammatory reactions remains unclear. Although the role of IL-17 for autoimmune diseases, e.g. experimental arthritis or EAE, has been established, it is not clear whether IL-17 production by Th17 cells is sufficient for their pathologic role or whether additional functional properties are required. In addition to the cytokine profile, particularly the migratory potential of T cells has a major impact on the functional properties of T-cell subsets in vivo 14. The chemokine receptors CCR5 and CXCR3, which are absent on naïve T cells, are important for guiding effector T cells to inflamed sites 15, 16. In addition, P- and E-selectin are important molecules for guiding T cells to inflamed sites 14. Recently, it was shown that human Th17 cells were identified as CCR2+CCR5− T cells and in addition express CCR6 17, 18. However, the chemokine receptor expression of in vitro Th17 cells as well as their migratory behaviour remains unclear.
To directly address the question on the role of joint-specific Th17 cells for arthritis pathogenesis we employed the OVA-induced arthritis (OIA) model. This arthritis model allows the direct analysis of OVA-specific T cells upon adoptive transfer in vivo. Using OIA and delayed type hypersensitivity (DTH) reaction, we show that in vitro-primed Th17 cells lack pro-inflammatory potential in vivo, although neutralisation of endogenous IL-17 in both models reduces disease severity. Characterisation of in vitro-generated Th17 cells indicates that they lack the expression of the pro-inflammatory chemokine receptor CXCR3 and fail to migrate into inflamed sites in vivo. These data show that the current protocols for in vitro Th17 differentiation, TGF-β, IL-6 and IL-23, as well as the markers typically used for Th17 characterisation, e.g. RORγT and IL-17, are not sufficient to define fully functional Th17 cells.
The role of IL-17 and in vitro-generated Th17 cells in inflammatory arthritis
IL-17 has been described as highly pro-inflammatory in collagen-induced arthritis as well as in antigen-induced arthritis 19, 20, implicating an important role of Th17 cells in arthritis pathology. We confirmed an important role for IL-17 also in the OIA model. This model is based on the classical AIA model comprising two immunisations with OVA in CFA followed by an intra-articular (i.a.) injection of cationic OVA (catOVA) to induce arthritis development. Here, in vivo neutralisation of IL-17 led to significantly reduced arthritis symptoms (Fig. 1). Interestingly, IL-17 neutralisation was effective only in the late phase of acute joint swelling leading to a faster recovery of oedema manifestation (Fig. 1A). For chronic arthritis at day +21 after arthritis induction, IL-17 neutralisation led to a 50% reduction of the overall histological arthritis score (Fig. 1B). The histological analysis revealed a clear reduction of mononuclear cell infiltrates as well as synovial hyperplasia and bone and cartilage destruction (Fig. 1C).
Although the role of IL-17 is firmly established, the additional parameters qualifying Th17 cells for promoting inflammatory reactions in vivo are less clear. To directly analyse the contribution of joint-specific Th17 cells we used the OIA model allowing the transfer of OVA-specific Th17 cells for functional analysis in an arthritis model. Th1 cells were used for comparison of the pro-inflammatory in vivo potential of Th17 cells.
Th1 and Th17 cells were generated in vitro from naïve OVA TCR-transgenic T cells according to standard protocols with IL-12 or IL-6, TGF-β, IL-23, respectively. Th17 cells revealed all typical characteristics as described before, i.e. high expression of the transcription factor RORγT and IL-17 and lack of IFN-γ expression, whereas Th1 cells produced IFN-γ and the transcription factor T-bet but no IL-17 or RORγT (Fig. 2A and B).
First, we tested the pro-inflammatory capacity of in vitro-generated Th17 cells in the OIA model using adoptive cell transfer. Transfer of Th1 or Th17 cells as well as unpolarised or naïve T cells alone without prior OVA immunisation did not result in detectable levels of chronic arthritis (Fig. 6D and data not shown), indicating that T cells alone are not sufficient for disease induction. Therefore, we generated Th1 and Th17 cells in vitro, and at day 6 of the culture 2×106 in vitro-generated Th17 or Th1 cells were transferred into OVA-immunised mice 1 day before arthritis induction. Unexpectedly, transfer of in vitro-primed Th17 cells did neither increase acute joint swelling nor chronic arthritis symptoms (Fig. 3A and B). In contrast, for comparison transferred Th1 cells exacerbated acute joint swelling although chronic symptoms were equal to non-transferred animals.
In vitro-primed Th17 cells fail to induce a DTH reaction
To test the pro-inflammatory potential of in vitro-generated Th17 cells in a different model of inflammation, we used DTH, which is impaired in IL-17-deficient mice 21. To further analyse the role of IL-17 in an acute inflammatory DTH reaction, we neutralised IL-17 in DTH-induced mice. Our results show a significantly reduced footpad swelling after DTH induction in animals treated with anti-IL-17 neutralising antibody (Fig. 4A) claiming an important role for IL-17 in a DTH response. To test the acute inflammatory potential of Th17 cells, 5×105 in vitro-generated Th17 cells (day 6 of culture) were transferred into naïve recipient mice followed by footpad injection of OVA to induce a DTH reaction. Interestingly, Th17 cells failed to induce a strong DTH reaction, while Th1 cells induced a robust footpad swelling for several days, as expected (Fig. 4B). Only in the late phase of inflammation (>day 5) an increased foot swelling in Th17-transferred animals was observed. Combined transfer of Th1 and Th17 cells resulted in the same response as transfer of Th1 cells alone. However, in the late phase double-transferred animals remained on a high level of foot swelling while animals with Th1 cells only showed a declining swelling as expected (Fig. 4B). However, when re-analysing the adoptively transferred Th cells for cytokine production we found that 2 wk after cell transfer Th1 cells still produced IFN-γ but no IL-17. Surprisingly, Th17 cells showed a decreased amount of IL-17 at the single-cell level but a rising number of IL-17/IFN-γ double producers (Fig. 5A). Indeed, more “Th17” cells now produced IFN-γ.
It was reported that in vitro Th17 cultures might be contaminated with Foxp3+ Treg that may interfere with Th17-cell activation in vivo. Also, recently McGeachy et al. and others demonstrated IL-10 expression by in vitro-generated Th17 cells as well as by in vivo-primed cells 22–24. However, in our Th17-cell cultures neither significant Foxp3 expression was observed nor could we confirm IL-10 production by in vitro- or in vivo-primed Th17 cells (Fig. 5B and C, data not shown). IL-10 staining was positively controlled by staining IL-10+ Th2 cells (data not shown).
IL-17/IFN-γ double-producing Th cells have reduced pathogenic potential in vivo
In the DTH model transferred Th17 cells showed a weak pathogenic potential in the late phase of the immune reaction (Fig. 4B). Also, some of the transferred Th17 cells acquired IFN-γ expression resulting in IL-17/IFN-γ double and IFN-γ single producers (Fig. 5A). Recently, it was reported that IL-17/IFN-γ double-producing “Th17” cells are more pathogenic than IL-17 single producers 25. Therefore, we tested double-producing Th17 cells in our models for their pathogenic potential in vivo. Th17 cells that are 1 wk old were further cultivated for a second week with IL-6/TGF-β/IL-23/anti-IL-4 and in addition with IL-12, resulting in about 10% of IL-17/IFN-γ double-producing Th cells (Th17-IL-12, Fig. 6A). We compared the double-producing cells with 2 wk-cultured Th17 cells (Fig. 6A). Double-producing Th17-IL-12 cells showed the same reduced inflammatory potential as “pure” Th17 cells in the DTH model (Fig. 6B). In the OIA model we transferred 2×106 Th17-IL-12 or “pure” Th17 cells into naïve non-immunised mice to determine the pathogenic potential of these cells in inducing an arthritic disease. However, in the same way as Th1 and Th17 cells, double-producing Th17-IL-12 cells failed to induce chronic arthritis (Fig. 6D), although Th17-IL-12 cells provoked a slightly higher acute knee swelling compared with Th17 cell-induced joint swelling (Fig. 6C).
Th17 cells migrate poorly into inflamed sites
A possible explanation for the failure of in vitro-primed Th17 cells to mediate pro-inflammatory functions in vivo might be a disturbed migratory behaviour. To address this question, in vivo cell homing experiments were performed, using with 51chromium radioactively labelled Th cells. The use of chromium to analyse the in vivo fate of transferred lymphocytes is a well established and carefully controlled method excluding the influence of non-viable cells 26, 27. As shown in Fig. 7A, Th17 cells migrated poorly into the inflamed foot in comparison to Th1 cells. This was also true when testing migratory behaviour of Th cells into the inflamed knee joint in the OIA model (data not shown). Instead, Th17 cells migrated better into the spleen. The migration pattern of Th1 and Th17 cells into draining and peripheral lymph nodes as well as lung and liver was comparable, demonstrating that Th17 cells were not simply deleted under the experimental conditions in vivo.
The different migratory behaviour prompted us to analyse the chemokine receptor expression of Th17 cells. As shown in Fig. 7B in vitro-primed Th17 cells lack surface expression of both CCR5 and CXCR3, which might explain the impaired migratory behaviour of these cells. CCR7, a marker for naïve and central memory T cells, is expressed on both cell types at almost the same level, which proves an equal activation status of Th1 and Th17 cells (Fig. 7B). In contrast, Th17 and Th1 cells showed a comparable level of P- and E-selectin-ligand expression. The surface expression data was verified at the mRNA level (Fig. 7C), showing a higher CCR5 expression in Th1 cells and lack of CXCR3 mRNA expression in Th17 cells. However, in contrast to Th1 cells CCR5 protein was not expressed on the cell surface of Th17 cells, indicating a differential regulation of this gene at the posttranscriptional level. Because Th17 cells were shown to express CCR4 and CCR6, we also analysed the expression of these two chemokine receptors on Th17 cells. Figure 7D clearly shows that in vitro-generated Th17 cells lack the expression of CCR4 and CCR6 protein. To validate the expression of CCR6 and CCR4 on in vivo-generated Th17 cells, we isolated CD4 cells of old C57BL/6 mice and sorted these cells for CCR4+CCR6−, CCR4−CCR6+, CCR6+CCR4+ and CCR4−CCR6− cells, respectively. In common with other reports, after in vitro re-stimulation with PMA/ionomycin, IL-17+ Th cells were only detected in the CCR6+ fraction (with either CCR4 expression in addition or not). CCR4+CCR6− as well as double-negative CD4+ T cells showed no IL-17 expression (Fig. 7E).
In the last years, many publications claimed a pivotal role for the pro-inflammatory cytokine IL-17 in autoimmune diseases, e.g. rheumatoid arthritis and multiple sclerosis 6, 28, 29. We confirmed in our murine arthritis model of OIA an important role for IL-17 in the progression of chronic arthritis. Neutralising IL-17 showed the strongest ameliorating effect on disease development late in the acute phase and in the chronic arthritis phase. Therefore, we concluded that IL-17 might play a role in the late phase of inflammation only but not in the induction phase. This is in line with recently published data where Hoyer et al. claimed that IL-17 plays a minor role in the induction phase but a major role in chronic tissue inflammation in the murine model of autoimmune haemolytic anaemia 30. In 2006 several reports described the in vitro differentiation of Th17 cells from naïve T cells using the combination of IL-6 and TGF-β 7, 9, 31. However, whether in vitro-primed Th17 cells are comparable to in vivo-induced pathogenic Th17 cells is unclear. Our results demonstrate that Th17 cells induced in vitro by IL-6 and TGF-β are in vivo neither able to induce or exacerbate chronic arthritis nor to induce an acute inflammation in the OIA or the DTH model in comparison to pro-inflammatory Th1 cells. Th cells are thought to be the main producers of IL-17 32, 33. However, we cannot exclude that other cell types than CD4+ T cells contribute to pathogenic IL-17 production. CD8+ T cells, macrophages and neutrophils are reported to be able to express IL-17 34–36. However, CD8+ T cells are of minor importance in AIA but macrophages and neutrophils are involved in AIA development 37, 38. This might explain the discrepancy between IL-17 neutralisation and the lacking pro-inflammatory potential of in vitro-generated Th17 cells. On the other hand, it was shown that Th17 cells clearly contribute to arthritis development in different mouse models 39, 40. In the OIA model, Th1 cells were able to exacerbate the acute knee swelling. This might be attributed to the much higher potential of transferred Th1 cells to migrate into inflammatory site, and the high expression of TNF-α in Th1 cells. TNF-α mainly mediates the acute inflammation or rather the acute knee swelling in the OIA model (own unpublished data). Therefore, increased levels of TNF-α in the inflamed knee joint may lead to a more severe joint swelling of Th1-transferred mice compared with Th17-transferred or non-transferred mice. The failure of Th17 cells to induce inflammation might be also due to the unstable phenotype concerning cytokine production. Lexberg et al. already demonstrated that in vivo-induced Th17 cells show a more stable phenotype than in vitro-primed Th17 cells regarding IL-17 and RORγ T expression 41. In line with results from other groups we could show that in vitro-generated Th17 cells switched in vivo to the production of IFN-γ after ex vivo re-stimulation 42, 43. However, other factors, e.g. cytokine or chemokine receptor expression, may distinguish in vivo “converted” Th17 cells from pathogenic Th1 cells. Therefore, IFN-γ expression by “converted” Th17 cells might be sufficient to suppress (late) pathogenic effects of IL-17-producing cells but is not sufficient to drive pathogenicity of these cells. On the other hand, Th17 cells could mediate a DTH reaction in the late phase and this may be driven by IFN-γ single producers, which likely arise from non-cytokine-producing “Th17 cells” (Fig. 5A, lower right panel). However, recent publications claim that IFN-γ/IL-17 double-producing Th cells are more pathogenic than IL-17 single producers 25, 44. Our results do not support this idea, because in vitro-generated IL-17/IFN-γ double producers also failed to show pathogenic potential in vivo in the OIA model as well as in the DTH model. Recent data also suggest that pathogenicity of IL-17-producing Th cells depends on the cytokine stimulus 25. IL-6 was found as inducer of inflammatory potential of Th17 cells while the combination of IL-6 and TGF-β resulted in less pathogenic Th17 cells in the EAE model. However, this has to be analysed and confirmed in other disease models but might be a reason for reduced inflammatory potential of our in vitro-generated Th17 cells. The failure of Th17 cells to induce an acute inflammation in vivo was not due to contamination of in vitro-generated cells with Foxp3-expressing Treg or IL-10 production by T cells, which has been reported for in vitro-generated Th17 cells 22. Furthermore, the co-transfer of Th1 together with Th17 cells did not reduce the Th1-mediated DTH induction. Thus, the unresponsiveness of in vitro-generated Th17 cells cannot be attributed to contaminating Treg or intrinsic IL-10 production by Th17 cells.
Homing assays revealed the impaired migratory behaviour of in vitro-primed Th17 cells compared with Th1 cells. Th17 cells migrated two to five times less efficiently into inflamed tissue, which might be another reason for the poor in vivo inflammatory potential of in vitro-primed Th17 cells. CCR5 and CXCR3 are important receptors for the migration into inflamed sites and both receptors are expressed on Th1 cells 16. Indeed, CCR5 and CXCR3 expression was only detectable on the surface of Th1 but not Th17 cells. This suggests that in vitro-generated Th17 cells differ from their in vivo-primed counterparts in their migratory potential as well as in the stability of the Th17 phenotype, which has recently been demonstrated 41, 42. Only a few reports showed direct inflammatory ability of in vitro-primed Th17 in vivo, yet 43, 45. Langrish et al. as well as others showed that Th17 cells may induce passive EAE. However, these authors used in vivo-primed Th cells, which were isolated from draining organs and further cultured in vitro under Th17 differentiation conditions to enrich IL-17-producing T cells 46. Stummvoll et al. demonstrated that in vitro-primed antigen-specific Th17 cells may induce a stronger autoimmune gastritis than Th1 cells after i.p. transfer 43. However, due to the i.p. injection the migratory behaviour of the transferred Th cells was of minor importance. Moreover, whether Th17-derived IL-17 has an impact on gastritis development at all is unclear. All analysed Th-cell subsets, Th1, Th17 and Th2, induced disease symptoms and produced IFN-γ 6 wk after transfer. Thus, the stronger proliferation of Th17 cells in vivo together with the production of IFN-γ might be responsible for the more severe gastritis symptoms induced by Th17 cells in this study 43. Cox et al. showed that in vitro-induced Th1 as well as Th17 cells induce eye inflammation in vivo 45. Although Th17 cells were less efficient than Th1 cells, this indicates again that homing into the eyes requires different migratory receptors than homing into the inflamed knee joint or the footpad. It was reported that ex vivo Th17 cells may express CCR6 and CCR4 chemokine receptors. We confirmed this expression on in vivo-induced Th17 cells but could detect neither significant CCR6 nor CCR4 expression on the surface of 1 wk old in vitro-generated Th17 cells. In contrast, other groups reported CCR6 surface expression on in vitro-generated Th17 cells 45, 47, 48. However, Hirota et al. demonstrated that stimulation with IL-1β is mandatory for optimal CCR6 expression of in vitro-primed Th17 cells 48. This may explain the observed differences in CCR6 expression. Whether the stimulus with a different cytokine cocktail, e.g. IL-6 without TGF-β, IL-1β or IL-21, induces CCR4 on the cell surface of in vitro-generated Th17 cells remains to be elucidated.
Thus, we conclude that Th17 cells differentiated in vitro from naïve T cells by the standard combination IL-6/TGF-β/IL-23 do not exert full pro-inflammatory function in vivo due to impaired migration into inflamed sites.
Taken together, our results show that in addition to IL-6 and TFG-β further signals are required to induce fully functional Th17 cells in vivo including the capacity to migrate into inflamed sites. Together with recent results concerning the instability of the Th17 phenotype of in vitro-primed Th17 cells, this indicates that the in vivo development of functional Th17 cells cannot be mimicked by the current protocols for in vitro Th17 differentiation but is the result of a more complex process requiring the integration of multiple signalling components within the in vivo microenvironment.
Materials and methods
BALB/c wt, C57BL/6 wt, lymphopenic SCID and OVA-TCRtg/tg DO11.10 mice were purchased from the BgVV, Berlin and were housed under specific pathogen free conditions. Animals were used at the age of 6–10 wk. All experiments were done following national guidelines and as approved by our national animal ethics committee.
MACS-sorted CD4+CD62L+ naïve T cells from D011.10 mice were cultured with irradiated MHCII+APC (30 gy, 1:5 ratio) in RPMI medium supplemented with 10% FCS and stimulated with 0.5 μg/mL OVA323–339 peptide. 10 ng/mL rmIL-23, 20 ng/mL rmIL-6, 1 ng/mL rhTGF-β, 10 μg/mL anti-IL-4 (11B11), 10 μg/mL anti-IFN-γ and 10 ng/mL rmIL-12, 10 μg/mL anti-IL-4 were used for Th17 and Th1 conditions, respectively (all cytokines R&D Systems). Cells were harvested at day 6 and used for experiments or re-stimulated for 6 h with 1 ng/mL PMA/1 μg/mL ionomycin in the presence of 5 μg/mL BrefeldinA (all Sigma). In some experiments 1 wk old Th17 cells were further cultured with irradiated APC for another week under the above-mentioned conditions. To induce IL-17/IFN-γ double-producing Th17-IL-12 cells, 1 wk old Th17 cells were cultured for another week with IL-6, TGF-β, IL-23 and anti-IL-4 and 10 ng/mL rmIL-12.
Mice were immunised subcutaneously twice, at day −21 and day −14 with 100 μg catOVA in CFA. In some experiments 2×106 OVA-specific Th1 or Th17 cells were transferred i.v. either into naïve non-immunised mice or into catOVA immunised BALB/c wt mice at day −1. At day 0, arthritis was induced by i.a. injection of 60 μg catOVA in PBS in one knee joint. When indicated, 200 μg of neutralising anti-IL-17 (TC11-18H10 49, kind gift from A. Richter, Miltenyi-Biotec) was injected i.p. twice, at day −1 and day +2. Joint swelling was monitored at indicated time points with an Oditest micrometer gauge (Kroeplin) as difference between injected and non-injected joint. 3 wk after arthritis induction in the chronic arthritis phase, mice were sacrificed; the knee joints were fixed in 10% formalin, decalcified in EDTA-solution and embedded in paraffin. Sections were stained with H&E and were analysed in a blinded manner for the following histological parameters: granulocyte exudates, granulocyte infiltration, hyperplasia of synovial lining, mononuclear cell infiltration/fibroblast activation (each scoring 0–3 points); peri-articular infiltration and bone and cartilage destruction (each 0–4 points) with one additional point for fibrin exudates adding up to a maximum score of 21.
A total of 5×105 OVA-specific Th1 or/and Th17 cells were transferred i.v. into congenic BALB/c or SCID mice. After 1 day, DTH was induced by s.c. injection of 250 ng OVA323–339 peptide emulsified 1:1 in incomplete Freund's adjuvant (IFA) in one footpad. PBS/IFA was given to the contralateral footpad. Swelling of the foot was measured in a blinded manner with an Oditest micrometer gauge (Kroeplin). For analysing the role of IL-17 in a DTH reaction, BALB/c mice were immunised at days −14 and −7 with 100 μg OVA in CFA. At day 0, DTH was induced by s.c. injection of 50 μg OVA protein emulsified 1:1 in IFA in one footpad. When indicated, 200 μg of neutralising anti-IL-17 (TC11-18H10) was injected i.p. twice, at day +2 and day +5. For homing experiments, DTH was induced with 5×105 OVA-specific Th1 cells in BALB/c wt mice and 5×105 OVA-specific radioactively labelled Th1 or Th17 cells were transferred i.v. at day 1 after DTH induction. After 24 h, mice were sacrificed and indicated organs were analysed for radioactivity with a γ-counter (Wallac).
FACS analysis and FACS-sorting
For surface chemokine receptor expression, living Th cells were stained with anti-P-selectin-ligand, anti-E-selectin-ligand (kind gift from A. Hamann, Berlin), anti-CXCR3-APC (R&D Systems), anti-CCR7-Bio (e-Biosciences), anti-CCR5-Bio, anti-CD4-FITC, SA-PE (all BD Biosciences), anti-CCR4-PE and anti-CCR6-Alexa647 (both Biolegend). For intracellular staining, PMA/iono re-stimulated Th cells were fixed in 2% formaldehyde and stained in 0.5% saponin with anti-IL-17-FITC (Biozol), anti-IFN-γ-APC, anti-IL-10-APC and anti-TNF-α-PE (all BD Biosciences). Foxp3 staining was performed with staining kit (e-Biosciences) according to the manufacturer's instructions.
For FACS-Sort, spleen and lymph node cells of C57BL/6 wt mice were enriched for CD4+ T cells with anti-CD4-FITC (BD Biosciences) and anti-FITC-Beads (Miltenyi-Biotec) according to the manufacturer's instructions. Enriched CD4+ T cells were stained with anti-CCR4-PE and anti-CCR6-Alexa647 and were sorted using FACS-ARIA (BD Biosciences) for living (Propidium iodide negative) CD4+ T cells as CCR4+CCR6−, CCR6+CCR4−, CCR4+CCR6+ and CCR4−CCR6− fractions. Sorted fractions were re-stimulated with PMA/iono and stained intracellularly for IL-17 expression as mentioned above.
mRNA was isolated from Th cells with invitrogen RNA isolation kit according to the manufacturer's instructions and transcribed into cDNA (Taq Man® Kit, Roche Diagnostics). Real-time PCR was performed with the LightCycler instrument and the FastStart DNA Master SYBR Green I kit (Roche Diagnostics). Cycling program: 10 min at 95°C followed by 40 cycles of 15 s at 95°C, 15 s at 65°C and 15 s at 72°C; primer 5′–3′ sequences: CCR5 for CAAgACAATCCTgATCgTgCAA, rev TCCTACTCCCAAgCTgCATAgAA; CXCR3 for TgCTAgATgCCTCggACTTT, rev CgCTgACTCAgTAgCACAgC; CCR7 for CATCAgCATTgACCgCTACgT, rev ggTACggATgATAATgAggTAgCA; T-bet for TCCTGCAGTCTCTCCACAAGT, rev CAGCTGAGTGATCTCTGCGT; RORγt for TgCAAgACTCATCgACAAgg, rev AggggATTCAACATCAgTgC; UBC (housekeeping gene) for TCTTgACAATTCATTTCCCAACAg, rev TCAggCACTAAAggATCATCTgg. Data were evaluated with LightCycler software version 3.5.28 (Roche Diagnostics) and the second derivative maximum algorithm, mRNA expression was calculated relative to UBC gene expression.
The authors thank Thordis Hohnstein, Gabriele Fernahl and Uta Lauer for excellent technical support. This study was supported by the sfb650, by grants from the Deutsche Forschungsgemeinschaft ENDO-Stiftung–Stiftung des Gemeinnützigen Vereins ENDO-Klinik e.V., the European Community OIGU40719-22, LSHG-CT-2007-037291 and LSHG-CT-2005-00523. A. Nass was supported by the Grako 1121.
Conflict of interest: Alexander Scheffold is an employee of Miltenyi-Biotec.
- 26Following cellular traffic: methods of labelling lymphocytes and other cells to trace their migration in vivo, in Weir, D. M. (Ed.) Handbook of Experimental Immunology, Blackwell Scientific Publications, Oxford 1986, pp 57.51–57.23.and ,
- 27Use of labelled lymphocytes to analyze trafficking in vivo, in Hamann, A. a. E., B. (Ed.) Leukocyte Trafficking, Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 2005, pp. 497–508.,