Th memory for interleukin-17 expression is stable in vivo

Authors


Abstract

Based on the memory for the re-expression of certain cytokine genes, different subsets of Th cells have been defined. In Th type 1 (Th1) and Th2 memory lymphocytes, the genes for the cytokines interferon-γ and interleukin (IL)-4 are imprinted for expression upon restimulation by the expression of the transcription factors T-bet and GATA-3, respectively, and epigenetic modification of the cytokine genes. In Th17 cells, IL-17 expression is dependent on the transcription factors RORγt and RORα. Here, we analyze the stability and plasticity of IL-17 memory in Th17 cells. We have developed a cytometric IL-17 secretion assay for the isolation of viable Th cells secreting IL-17. For Th17 cells generated in vitro, IL-17 expression itself is dependent on continued TGF-β/IL-6 or IL-23 signaling and is blocked by interferon-γ and IL-4 signaling. In response to IL-12 and IL-4, in vitro generated Th17 cells are converted into Th1 or Th2 cells, respectively. Th17 cells isolated ex vivo, however, maintain their IL-17 memory upon subsequent in vitro culture, even in the absence of IL-23. Their cytokine memory is not regulated by IL-12 or IL-4. Th17 cells generated in vivo are a stable and distinct lineage of Th cell differentiation.

Introduction

Th memory lymphocytes are imprinted for the re-expression of distinct cytokine genes upon restimulation. Originally, two types of Th effector memory cells had been defined: T helper type 1 (Th1) cells re-expressing interferon-γ (IFN-γ), and Th2 cells, re-expressing interleukin (IL)-4, -5 and -13 1. Recently, a third lineage of Th effector memory cells has been described, characterized by the re-expression of IL-17A, IL-17F and IL-22 (reviewed in 2). Th17 cells can induce autoimmune inflammation 3 and are protective in response to fungal infection 4. In vitro, naïve murine Th cells can be induced to differentiate into Th17 cells by combined TGF-β and IL-6 signalling 5, 6. IL-23 promotes survival and proliferation of Th17 cells 6. IL-21 can induce IL-17 independent of IL-6 and is expressed by Th17 cells themselves, as part of a positive regulatory feedback loop for IL-17 re-expression 7, 8. In human Th cells, similar signals are required for the differentiation of IL-17 re-expressing Th memory cells 9–11. STAT3 is involved as a signal transducer and IRF-4 12 and the retinoic acid receptor-related orphan receptors RORγt 13 and RORα 14 as transcription factors controlling lineage development. Ectopic over-expression of RORγt and RORα in naïve Th cells is sufficient to induce IL-17 expression 14.

As part of their functional memory, the capacity of effector memory Th cells to stably re-express particular cytokines has been demonstrated for Th1 cells and IFN-γ expression and for Th2 cells and their IL-4 and IL-10 expression 15, 16. This memory cytokine expression depends on TcR signals, but does not require the original instructive signals. It even occurs in the presence of adverse instructive signals. Cytokine memory for Il4 and Ifnγ is based on epigenetic modification of the cytokine genes and expression of the transcription factors T-bet and GATA-3, for Th1 and Th2 differentiation, respectively (reviewed in 17). A more complex situation has been described for the cytokine IL-10, which can be expressed by Th1 and Th2 cells. In Th2 cells, the Il10 gene is imprinted by GATA-3, but this imprinting requires multiple restimulations of the Th2 cells, while the Il4 gene is imprinted in the primary activation 18–20. Th1 cells are imprinted to re-express Il10 by Notch 21, but re-expression requires continued IL-12 signaling 18.

Chromatin of the Il17a and Il17f genes is modified in Th17 cells, as compared with Th1 and Th2 cells 22, and their re-expression depends on RORγt and RORα 14. It is not clear, however, whether the chromatin-modifications and the ROR transcription factors imprint the Il17 genes for re-expression, and whether Th17 cells induced with the signals identified so far are resistant to reprogramming by IL-12 or IL-4? In vitro, IL-4 and IFN-γ inhibit the induction of Th17 differentiation 23. Interestingly, Th cells co-expressing IL-17 and IFN-γ are frequent in vivo, indicating that in vivo IFN-γ and IL-17 inductions are not mutually exclusive 4, 24–26.

To analyze the cytokine memory of individual Th cells expressing IL-17, we here have isolated Th cells according to the secretion of IL-17 and analyzed their cytokine memory upon further stimulation in vitro. IL-17-secreting cells generated in vitro, by TGF-β, IL-6 and IL-23 in the absence of IFN-γ and IL-4, upregulate the lineage-specific transcription factors RORγt and RORα but fail to re-express IL-17 in the absence of the original inducing signals, even after repeated instructive restimulation. They are converted to express IFN-γ when stimulated with IL-12, or IL-4 when stimulated in the presence of IL-4. In contrast, IL-17 expression could not be induced in differentiated Th1 and Th2 cells by TGF-β, IL-6 and IL-23, blocking IL-4 and IFN-γ. IL-17-expressing memory Th cells isolated ex vivo faithfully re-express IL-17, even when restimulated in the absence of IL-17-inducing signals or in the presence of IL-4 or IL-12.

Results

IL-17 is not induced in Th1 and Th2 memory cells

Naïve CD4+CD62L+ T cells from TCR transgenic DO11.10 mice were activated with their cognate antigen and differentiated into Th1 cells with IL-12 and anti-IL-4 antibody, or into Th2 cells with IL-4, anti-IL-12 and anti-IFNγ. After 6 days the Th1 and Th2 memory cells were restimulated with antigen, but this time under Th17 inducing conditions, i.e. in the presence of TGF-β, IL-6, IL-23, anti-IL-4 and anti-IFN-γ. Six days later, the cells were restimulated with PMA/ionomycin, fixed after 4 h, permeabilized and stained intracellularly for cytokine expression. Induction of IL-17 by TGF-β and IL-6 was not effective in either Th1 (1.6% IL-17+ cells) or Th2 (4.7%) memory cells (Figs. 1A and B). When naïve Th cells were differentiated into Th17 cells for 6 days and then restimulated under Th17 polarizing conditions for an additional 6 days, IL-17 expression increased from 12.9 to 29% (Fig. 1C). To exclude inhibition of Th17 differentiation by IFN-γ in established Th1 cells, we also analyzed cells deficient for the IFN-γ receptor. Also in these cells, once the cells had been polarized into Th1 cells, IL-17 expression was not effectively induced (4%) (Supporting Information Fig. 1), as compared with naïve cells (25%) (Supporting Information Fig. 2). In Th1 cells, T-bet was upregulated under Th17-inducing conditions (Fig. 1D). The transcription factors RORα and RORγt were upregulated 2- and 6-fold, respectively, in Th1 cells under Th17-inducing conditions. In Th2 cells, GATA-3 was downregulated 2–3-fold when they had been restimulated under Th17-inducing conditions, and RORγt expression was upregulated 5-fold. However, RORγt levels remained well below the expression level in cells stimulated twice under Th17 polarizing conditions. The expression of IL23R and RORα remained unchanged.

Figure 1.

Th1 and Th2 cells are refractory to Th17 polarization. Naïve CD4+CD62Lhigh cells from DO11.10 mice were differentiated in the presence of rIL-12, anti-IL-4, irradiated APC and equation image to Th1 (A), in the presence of IL-4, anti-IL-12, anti-IFN-γ, irradiated APC and equation image to Th2 (B), or in the presence of TGF-β, IL-6, IL-23, anti-IFN-γ, anti-IL-4, irradiated APC and equation image to Th17 (C) for 6 days. The cells were cultured for another 6 days in the presence of IL-17-inducing conditions (TGF-β, IL-6, IL-23, anti-IL-4 and anti-IFN-γ). Cytokine expression was analyzed after 5 h of restimulation with PMA/ionomycin by intracellular cytokine staining. Data are representative of four experiments. (D) RNA was extracted from unstimulated (unst) cells and stimulated cells (stim) for 3 h and quantitative real-time PCR was performed for T-bet, GATA-3, IL-23R, RORα and RORγt. Data are representative of two experiments.

Figure 2.

Isolation of viable IL-17 producing and non-producing cells with the IL-17 secretion assay. (A) Naïve CD4+CD62Lhigh cells were differentiated under IL-17-inducing conditions for 6 days, and the IL-17 secretion assay was performed. IL-17 secreting and IL-17 non-secreting cells were separated by FACS sorting. (B) After labeling with the IL-17 capture matrix, cells were either kept on ice immediately to prevent secretion (low control) or incubated with recombinant murine IL-17 at RT for 15 min (high control). (C) The staining of secreted IL-17 correlates with the staining of intracellular IL-17. After the IL-17 secretion assay, the cells were fixed and stained for intracellular IL-17. To confirm the staining of intracellular IL-17, the staining was performed either in the absence or in the presence of saponin. Data are representative of two experiments.

Direct isolation of IL-17-expressing Th cells

To analyze the stability of Th17 memory cells on the single cell level, we developed a cytometric cytokine secretion assay 27, 28 for murine IL-17. Upon PMA/ionomycin restimulation the maximal frequency of IL-17-expressing Th cells is reached already after 1–2 h (Supporting Information Fig. 3). Accordingly, Th17 cells were restimulated for 1 h to induce cytokine expression, labeled with the IL-17 capture matrix, and allowed to secrete IL-17 for 30 min. The secreted IL-17 bound to the capture matrix was then detected by a fluorochrome-conjugated anti-IL-17 antibody. The cells were analyzed by flow cytometry and separated by fluorescent-activated cell sorting (FACS) (Fig. 2A) or magnetic cell sorting (data not shown). Cells placed on ice for the secretion period, thus blocking secretion, were used as control (Fig. 2B, left plot). The capacity of the capture matrix was determined by adding recombinant IL-17 (Fig. 2B, right plot). To control for false-positive cells due to cross-feeding of IL-17 from secreting cells to the capture matrix of non-secreting cells, cells of the IL-17 secretion assay were fixed and stained intracellularly for IL-17 in the presence or absence of the membrane-permeabilizing agent saponin. All IL-17-secreting but none of the IL-17-non-secreting Th cells expressed intracellular IL-17 (Fig. 2C).

Figure 3.

In vitro-generated Th17 cells fail to re-express IL-17. Naïve CD4+CD62Lhigh cells from DO11.10 mice were differentiated under IL-17 inducing conditions for 6 days. (A) IL-17 producing and non-producing cells were separated and recultured for 6 days in the presence of equation image and irradiated APC only (stim), under neutral conditions (anti-IL-4, anti-IL-12, anti-IFN-γ) (Ab) or under different Th17 favoring conditions (IL-23, IL-23 with anti-IL-4 and anti-IFN-γ (IL23+Ab) or TGF-β, IL-6, IL-23, anti-IL-4 and anti-IFN-γ (Th17)). Cytokine expression was analyzed by intracellular staining after PMA/ionomycin for 5 h. Data are representative of three experiments. (B) mRNA was extracted from IL-17+ and IL-17 cells directly after isolation (post-sort) and 6 days later from cells restimulated for 2 h, reverse transcribed and quantified by quantitative real-time PCR for RORγt, RORα, T-bet, GATA-3, IL-17F, IL-22, IL-23R and IL-21. Data are representative of two experiments.

IL-17 re-expression is blocked by IL-4 and IFN-γ

We stimulated naïve CD4+CD62L+ cells from DO11.10 mice with antigen in the presence of TGF-β, IL-6, IL-23, anti-IL-4 and anti-IFN-γ to induce Th17 differentiation. After 6 days, such cells expressed IL-17, IL-22, IL-17F, IL-23 receptor, RORγt and RORα (Supporting Information Fig. 2). Cells from such cultures were separated into IL-17 expressing and non-expressing cells, with a purity of >97 and >99%, respectively (Fig. 3A). In either population, we could not detect IFN-γ- or IL-4-expressing cells (data not shown). IL-17+ and IL-17 cells were restimulated with the cognate antigen, cultured for another 6 days under various conditions, and analyzed for IL-17 re-expression. We did not observe selective outgrowth of contaminating cells in either purified population, using CFSE to track proliferation of the cells (Supporting Information Fig. 4). Cell numbers were comparable and viability was above 90% throughout the culture period, as monitored microscopically (data not shown). When cultured in the absence of exogenous cytokines and blocking antibodies, only 13% of the IL-17+ Th cells re-expressed IL-17, 49% now expressed IFN-γ (Fig. 3A). In the presence of IL-23, 32% of the IL-17+ cells re-expressed IL-17, and 16% of them expressed IFN-γ. In the presence of blocking antibodies to IL-4 and IFN-γ, more than 60% of the IL-17+ cells re-expressed IL-17, irrespective of whether IL-23 was blocked by anti-IL-12p40, or added. This frequency was also not influenced by addition of recombinant TGF-β and IL-6. IL-17 non-expressing cells expressed IFN-γ (35%) but no IL-17 (<1%), if no cytokines or antibodies were added. In the presence of IL-23, 2% of these cells expressed IL-17. Blocking of IFN-γ and IL-4 resulted in the expression of IL-17 in 9% of the cells in the absence of IL-23, 12% in the presence of IL-23 and 18% in the presence of TGF-β, IL-6 and IL-23 (Fig. 3A). In IL-17+ and IL-17 cells, RORγt and RORα were expressed at similar levels (Fig. 3B) and were downregulated when the cells were cultured without the addition of exogenous antibodies or cytokines. Regulation of RORγt expression corresponded with the expression of IL-17, being high under conditions when IL-17 was expressed (anti-IL-4 and anti-IFNγ with anti-IL-12p40, IL-23 or TGF-β/IL-6). RORα was generally downregulated upon reculture, except for a 4-fold upregulation in the presence of TGF-β/IL-6 compared with cells cultured just in the presence of anti-IL-4, anti-IFN-γ and IL-23. T-bet expression in IL-17+ cells was 3-fold enhanced in the presence of (endogenous) IFN-γ. The expression of T-bet was higher in IL-17 than in IL-17+ cells. IL-17F, IL-22, IL-23R and IL-21 were only highly expressed by IL-17+ sorted cells, and their expression was downregulated in the absence of added cytokines or antibodies. IL-23 receptor expression was maintained in the presence of IL-23, but downregulated in the presence of TGF-β and IL-6. IL-17F expression was only maintained in the presence of TGF-β, IL-6 and IL-23. IL-22 and IL-21 expression was downregulated under all conditions analyzed (Fig. 3B).

Figure 4.

In vitro generated-Th17 cells can be converted to Th1 and Th2 cells. Naïve CD4+CD62Lhigh cells from DO11.10 mice were differentiated under IL-17 inducing conditions. (A) After 6 days, the cells were restimulated for IL-17 secretion assay. The IL-17+ and IL-17 cells were cultured under Th1 (IL-12, anti-IL-4) and Th2 (IL-4, anti-IL-12, anti-IFN-γ) polarizing conditions for 6 days. Cytokine expression was analyzed by intracellular staining after PMA/ionomycin restimulation. (B) Cells were cultured for 18 days under Th17 polarizing conditions. IL-17+ and IL-17 cells were separated and cultured under Th1 and Th2 polarizing conditions. Data are representative of three experiments. (C) mRNA of 1-week-old IL-17+ and IL-17 cells directly after isolation and after 6 day culture under Th1 and Th2 polarizing conditions was isolated after 2 h restimulation with PMA/ionomycin, reverse transcribed and quantified by quantitative real-time PCR for RORγt, RORα, T-bet, GATA-3, IL-17F, IL-22, IL-23R and IL-21. Data are representative of two experiments.

Th17 cells generated in vitro can cross-differentiate into Th1 and Th2 cells

We stimulated naïve CD4+ Th cells with TGF-β, IL-6, IL-23, anti-IL-4 and anti-IFN-γ for 6 days and isolated IL-17 expressing and non-expressing cells. IL-17+ and IL-17 cells were restimulated and cultured for an additional 6 days either under Th1 polarizing conditions (IL-12 and anti-IL-4) or under Th2 polarizing conditions (IL-4, anti-IFNγ and anti-IL-12). As shown in Fig. 4A, IL-17+ cells re-expressed IL-17 with a frequency of 8% under Th1 polarizing conditions, and 26% under Th2 polarizing conditions. IL-17+ and IL-17 cells started to express IFN-γ (>70% of CD4+ cells) or IL-4 (19% in IL-17+ and 28% in IL-17 cultures), respectively. RORγt, RORα, IL-17F, IL-22, IL-23R and IL-21 expression was downregulated when the cells were cultured under Th1 or Th2 conditions. Under Th1 conditions the cells upregulated the expression of T-bet 3–8-fold. Under Th2 conditions GATA-3 expression was upregulated 16–20-fold (Fig. 4C). The memory for IL-17 re-expression was also not stabilized in cells polarized toward Th17 differentiation for 3 wk, with weekly restimulations (Fig. 4B). After 3 wk, IL-17+ cells were isolated and further cultured for 6 days under Th1 or Th2 polarizing conditions. Only 9% re-expressed IL-17 under such conditions. About 64% of the IL-17+ or IL-17 cells expressed IFN-γ and 12–13% expressed IL-4 under Th1 or Th2 conditions, respectively.

In vivo-generated Th17 cells maintain IL-17 expression in vitro

IL-17-expressing Th memory cells generated in vivo were isolated directly ex vivo from unmanipulated DO11.10 or BALB/c mice (Fig. 5, Supporting Information Fig. 5). CD4+CD62Llow splenocytes were stimulated polyclonally with PMA/ionomycin and IL-17-expressing cells isolated with the IL-17 secretion assay. Of the PMA-ionomycin-stimulated cells, 8.2% expressed IFN-γ, 0.2% IL-4, and 3.6% expressed IL-17, of which approx. 25% co-expressed IFN-γ (Fig. 5A). Purified IL-17+ and IL-17 cells were recultured for 6 days in vitro, either without adding or blocking cytokines, or adding IL-23, or under Th1 or Th2 polarizing conditions (Fig. 5B). In the absence of added antibodies or cytokines 72% of the IL-17+ cells re-expressed IL-17. In the presence of added IL-23, 83% re-expressed IL-17. Ex vivo-isolated IL-17+ Th cells were refractory to Th1 and Th2 polarizing signals. Under Th1 conditions the frequency of IFN-γ-expressing cells was 14%. Under Th2 polarizing conditions, 4% of IL-4-expressing cells were observed. About 75 and 68% of the IL-17+ cells re-expressed IL-17 under Th1 and Th2 conditions, respectively. The expression of RORγt and RORα was downregulated 5-fold under Th2 conditions. GATA-3 expression was not induced in IL-17+ cells under any condition. The expression of T-bet was upregulated 2-fold under Th1-inducing conditions in the IL-17+ cells. In IL-17 cells, RORγt and RORα were not upregulated. T-bet and GATA-3 were induced under neutral and Th1 or Th2 conditions. IL-23 receptor, IL-22 and IL-17F were highly expressed in IL-17+ cells and their expression was downregulated upon in vitro culture. IL-21 expression was upregulated at least 2-fold upon in vitro culture, except when cultured under Th2 conditions, which led to a 2-fold reduction in IL-21 expression (Fig. 5C).

Figure 5.

In vivo-generated Th17 cells have a stable memory for IL-17. Memory CD4+CD62Llow cells from 6-months-old DO11.10 mice were isolated and either restimulated for (A) direct cytokine analysis or (B) IL-17 secretion assay. The IL-17+ and IL-17 cells were cultured without (stim) or with IL-23 and under Th1- or Th2-polarizing conditions for 6 days. The cytokine expression was analyzed by intracellular staining after restimulation with PMA/Ionomycin for 5 h. Data are representative of three experiments. (C) mRNA of ex vivo-isolated IL-17+ and IL-17 cells directly after isolation and cells cultured for 6 days was isolated after 2 h restimulation with PMA/ionomycin, reverse transcribed and quantified by quantitative real-time PCR. Data are representative of two experiments.

Discussion

The expression of IL-17 by Th lymphocytes has been originally described many years ago 24, 29. Recently IL-17-expressing Th lymphocytes have been recognized as a separate lineage of Th cell differentiation, distinct from the Th1 and Th2 lineages 23. Stability and plasticity of the cytokine memory of Th17 memory effector cells has been a matter of debate, in particular in light of reports of Th cells expressing both IL-17 and IFN-γ 4, 24–26.

Here we describe a cytometric cytokine secretion assay for murine IL-17 and its use to analyze the memory of IL-17-expressing Th cells for expression of IL-17. Isolated IL-17+ or IL-17 Th cells showed the same proliferation and survival upon reculture. IL-17-expressing cells generated in vitro by stimulation of activated naïve Th cells with TGF-β, IL-6 and IL-23, and blocking of IFN-γ and IL-4 with anti-IFNγ and anti-IL-4, failed to re-express IL-17 upon later reactivation, when the original inducing signals were lacking, or IFN-γ and IL-4 were not neutralized. IL-17+ Th cells, even after 3 wk of repeated instruction for IL-17 expression could still be converted into IFN-γ-expressing Th1 cells with IL-12 or into IL-4-expressing Th2 cells with IL-4. In contrast to in vitro-generated Th17 cells, IL-17-expressing Th cells isolated ex vivo maintained a memory for IL-17 expression in vitro, even in the presence of IL-12 or IL-4.

For the cytokine memory of Th1 and Th2 cells, molecular mechanisms have been described, which prevent the differentiation of Th1 into Th2 cells and vice versa. Such mechanisms include the mutual inhibition of the master transcription factors T-bet and GATA-3 30, the downregulation of receptors for costimulatory signals 31 and the epigenetic silencing of cytokine genes (reviewed in 32). Here we show that Th1 and Th2 effector memory cells also cannot be converted into Th17 cells, at least not by TGF-β, IL-6 and IL-23 signals, blocking IL-4 and IFN-γ. Interestingly, in Th1 cells, under Th17 polarizing conditions RORα and RORγt were upregulated 2- and 6-fold, respectively. Apparently this upregulation of Th17 lineage master transcription factors is not sufficient for the induction of IL-17 expression in such Th1 cells. This may be due to even further upregulation of T-bet in Th1 cells under Th17 polarizing conditions. T-bet has been described as a negative regulator of Th17 differentiation 33, 34. In Th2 cells, GATA-3 is downregulated 2–3-fold and RORγt upregulated 4-fold upon restimulation in a Th17 inducing cytokine milieu. The expression of IL-23 receptor and RORα was not upregulated. In apparent contradiction to our results, it has been shown that ectopic expression of RORγt in combination with RORα in Th1 and Th2 cells can lead to expression of IL-17 14. However, RORγt and RORα in those experiments were expressed in Th cells perhaps not fully committed to the Th1 or Th2 lineage. Future analysis of the Il17 gene with respect to epigenetic silencing in Th1 and Th2 cells will clarify whether exclusion of IL-17 expression in Th1 and Th2 cells is analogous to the reciprocal silencing of the Il4 gene in Th1 or the Ifnγ gene in Th2 cells 35, 36. The methylation of single CpG sites in cytokine gene promoters has been shown to silence gene expression by preventing binding for TcR responsive transcription factors 36, 37.

The expression of IL-17F in the in vitro-generated as well as in the ex vivo-isolated Th17 cells did not correlate with the re-expression of IL-17 after reculture. IL-17F is highly homologous to IL-17 and the Il17f gene is adjacent to the Il17 gene 38, suggesting a coordinated expression. While being highly expressed initially in cells sorted for IL-17 expression, IL-17F expression was only maintained in the presence of TGF-β and IL-6 (Fig. 3B). Interestingly, IL-17F expression was not stable in ex vivo-isolated IL-17-expressing cells, either, indicating that the memory for IL-17F re-expression is conditional and depends on different or additional signals than that for IL-17 re-expression, and that maintained RORγt and RORα expression may be required 14 but not be sufficient for IL-17F expression.

Our results suggest that the currently available in vitro protocols for the induction of IL-17 expression in naïve Th lymphocytes lack signals for the induction of a stable cytokine memory for IL-17 re-expression. Although IL-17 expression was induced very efficiently in vitro by TGF-β, IL-6, IL-23 and anti-IFNγ and anti-IL-4, failure to block IFN-γ or IL-4 during subsequent restimulation and culture led to the loss of IL-17 re-expression in cells which once had expressed IL-17. Both IFN-γ and IL-4 had been described as negative regulators of IL-17 expression 23, 39. In in vitro-generated Th17 cells, the expression of RORγt and RORα was downregulated in conditions under which the cells did not re-express IL-17 (Fig. 3B) and in particular under Th1 or Th2 polarizing conditions (Fig. 4B). Unlike the memory for IL-10 expression in Th2 cells, which requires multiple rounds of stimulation by IL-4 for stability 18, 19, repeated in vitro stimulation in the presence of TGF-β, IL-6 and IL-23 did not lead to a stable commitment for IL-17 expression (Fig. 4B). Such cells were still plastic and responded to the presence of IL-12 or IL-4 with differentiation into IFN-γ-expressing Th1 or into IL-4-expressing Th2 cells, respectively. It remains to be shown how the functional imprinting of in vivo-generated Th17 cells is encoded on the molecular level, to what extent the transcription factors RORγt and RORα, STAT-3 and IRF-4 12–14, 40 are involved in the imprinting of the Il17 locus and whether their genes themselves are imprinted.

Our results are in apparent contrast to previously published data of Harrington et al. 23 and Park et al. 39 with regard to the apparent stability of in vitro-generated Th17 cells when stimulated with IFN-γ or IL-4. In both reports, the authors generated Th17 cells from bulk CD4+ (Harrington et al. 23) or CD4+CD62L “memory” Th cells (Park et al. 39) from spleen and lymph nodes, not from purified naïve Th cells. Th17 cells were then “generated” by IL-23. As we know now, through the work of Veldhoen et al. 6, IL-23 does not induce naïve Th cells to become Th17 cells, but rather expands pre-existing Th17 cells. In view of the work of Veldhoen et al.6 and in view of our present data, the data of Harrington et al. 23 and Park et al. 39 have to be reinterpreted as showing the stability of in vivo-generated Th17 memory cells, and their selective expansion by IL-23 in vitro.

We here have demonstrated that IL-17-expressing Th cells generated in vivo are a stable lineage of effector memory cells, distinct from Th1 and Th2 cells, and functionally imprinted for re-expression of IL-17 upon TCR stimulation, even in the presence of Th1- or Th2-inducing conditions. IL-17-expressing cells generated in vitro are not functionally imprinted for IL-17 re-expression. Their re-expression of IL-17 depends on the continued presence of the canonical Th17-inducing signals.

Materials and methods

Mice

BALB/c and OVA-TCRtg/tg DO11.10 mice (kind gift of Dennis Y. Loh and Kenneth Murphy, Washington University School of Medicine, St. Louis, MO) were bred under specific pathogen-free conditions in our animal facility. IFN-γR−/− mice were kindly provided by Thomas Schüler (Charité, Campus Benjamin Franklin, Berlin, Germany). The mice were sacrificed by cervical dislocation. All animal experiments were performed in accordance with institutional, state and federal guidelines.

Antibodies

All antibodies used in these experiments were either conjugated in-house or purchased as indicated. Anti-IL-4 (11B11), anti-IL-12 (C17.18), anti-IFN-γ(AN17.18.24) antibodies were purified from hybridoma supernatants at the German Rheumatism Research Center and used at 10 μg/mL final concentration. FITC-conjugated anti-CD4 (GK1.5), PE-conjugated anti-IL-17 (TC11-18H10; BD Pharmingen, San Diego, CA) and Cy5-conjugated anti-IL-4 (11B11) were used for all intracellular cytokine stainings.

Cell culture conditions and Th differentiation in vitro

CD4+CD62L+ cells from 6–8-wks-old OVA-TCRtg/tg DO11.10 mice were isolated as described previously 15. All cultures and assays were carried out in RPMI supplemented with 10% FBS (Sigma Chemicals, St. Louis, MO), 100 U/mL of penicillin, 0.1 mg/mL of streptomycin, 0.3 mg/mL of glutamine (Invitrogen) and 10 μM β-mercaptoethanol at 37°C in 5% CO2. Cell cultures were set up at 3×106 cells/mL and stimulated in the presence of 0.5 mM cognate peptide OVA323–339 with irradiated (30 Gy) splenocytes isolated from BALB/c mice as APC. For Th1 differentiation, cells were stimulated in the presence of recombinant IL-12 (5 ng/mL; R&D Systems, Minneapolis, MN) and anti-IL-4 (11B11) antibody for 6 days. For Th2 differentiation, cells were stimulated in the presence of IL-4 (100 ng/mL, culture supernatant of HEK293 T cells transfected with murine IL-4 cDNA), anti-IL-12 (C17.8) and anti-IFN-γ(AN18.17.24) antibodies. For Th17 differentiation, cells were stimulated in the presence of TGF-β1 (1 ng/mL), IL-6, IL-23 (20 ng/mL) (all from R&D Systems), anti-IL-4 and anti-IFN-γ.

Isolation of IL-17-secreting cells in vitro and ex vivo

Cells were cultured under IL-17-inducing conditions for 6 days, or CD4+CD62Llow cells were isolated from spleen of 6-months-old DO11.10 or BALB/c mice. Cells were harvested and restimulated with 10 ng/mL of PMA (Sigma Chemicals) and 1 μg/mL of ionomycin (Sigma Chemicals) for 1.5 h. The cells were washed twice in ice-cold PBS with 0.5% w/v BSA (PBS/BSA). Cells were labeled for 5 min at 4°C with an IL-17-specific high-affinity capture matrix, i.e., bi-specific Ab–Ab conjugates of an anti-CD45 antibody with an anti-IL-17 antibody (Miltenyi Biotec GmbH, Bergisch Gladbach, Germany). Cell samples were taken for low control (kept on ice) and high control (incubated with recombinant IL-17 (0.5 μg/mL; Peprotech, Hamburg, Germany), washed after 10 min and kept on ice. The rest of the cells were transferred into 37°C warm RPMI medium at a low density (105 cells/mL) and placed at 37°C. Every 5 min the cells were mixed gently. After 30 min, the cells were transferred into ice-cold PBS/BSA and kept on ice for 10 min. The captured IL-17 was detected with an anti-IL17 biotin conjugated antibody followed by staining with an APC-conjugated anti-biotin antibody (Miltenyi Biotec). The IL-17 producing cells and the IL-17 non-producing cells were separated by FACSAria™ cell sorter (BD Biosciences). After sorting, the purity of the sort was confirmed with a FACSCalibur (BD Biosciences). Specificity of the IL-17 secretion assay was confirmed by intracellular staining.

Flow cytometry

A total of 4×106 cells/ml were stimulated in RPMI medium with 10 ng/mL of PMA and 1 μg/mL of ionomycin. An aliquot of 5 μg/mL Brefeldin A (Sigma Chemicals) was added after 1 h. After 5 h of stimulation, the cells were washed with PBS and fixed in 2% formaldehyde in PBS for 15 min at room temperature. The cells were stained for intracellular cytokines, as described previously 41. FACS analysis was performed with a FACSCalibur, using CellQuest (BD Biosciences) and FlowJo (TreeStar, Ashland, OR) software.

RNA quantification

RNA preparation and cDNA synthesis was performed as previously described 18. The expression of each gene was normalized to the expression of HPRT. Primer sets for the real-time PCR were as follows: HPRT up: 5′- GCT GGT GAA AAG GAC CTC T-3′, HPRT rev: 5′- CAC AGG ACT AGA CCT GC -3′; RORγ t up: 5′-TGC AAG ACT CAT CGA CAA GG -3′, RORγ trev: 5′- AGG GGA TTC AAC ATC AGT GC -3′; IL-17 for: 5′- TCC AGA AGG CCC TCA GAC TA -3′, IL-17 rev: 5′-AGC ATC TTC TCG ACC CTG AA -3′; IL-17F for: 5′- CAA AAC CAG GGC ATT TCT GT -3′, IL-17F rev: 5′- ATG GTG CTG TCT TCC TGA CC -3′; IL-22 for: 5′- GTC AAC CGC ACC TTT ATG CT -3′, IL-22 rev: 5′- CAT GTA GGG CTG GAA CCT GT -3′; IL-21for: 5′- ATC CTG AAC TTC TAT CAG CTC CAC -3′, IL-21 rev: 5′- GCA TTT AGC TAT GTG CTT CTG TTT C -3′; IL-21R for: 5′- TGT CAA TGT GAC GGA CCA GT -3′, IL-21R rev: 5′- CAC GTA GTT GGA GGG TTC GT -3′; RORα for: 5′- CCC CTA CTG TTC CTT CAC CA -3′, RORα rev: 5′- AGC TGC CAC ATC ACC TCT CT -3′; IL-23R for: 5′- AAC ATG ACA TGC ACC TGG AA -3′, IL-23R rev: 5′-TCC ATG CCT AGG GAA TTG AC-3′; Gata3 for: 5′-CCT ACC GGG TTC GGA TGT AAG T-3′, Gata3 rev: 5′-AGT TCG CGC AGG ATG TCC-3′; Tbet for: 5′-TCC TGC AGT CTC TCC ACA AGT -3′ and Tbet rev: 5′CAG CTG AGT GAT CTC TGC GT -3′.

Acknowledgements

This work has been supported by the German Science Foundation (DFG SFB421).

Conflict of interest: The authors declare no financial or commercial conflict of interest.

Ancillary