IL-17–secreting T cells represent a distinct CD4+ effector T cell lineage (Th17) that appears to be essential in the pathogenesis of numerous inflammatory and autoimmune diseases. Although extensively studied in the murine system, human Th17 cells have not been well characterized. In this study, we identify CD4+CD45RO+CCR7−CCR6+ effector memory T cells as the principal IL-17-secreting T cells. Human Th17 cells have a unique cytokine profile because the majority coexpress TNF-α but not IL-6 and a minor subset express IL-17 with IL-22 or IL-17 and IFN-γ. We demonstrate that the cytokines that promote the differentiation of human naive T cells into IL-17-secreting cells regulate IL-17 production by memory T cells. IL-1β alone or in association with IL-23 and IL-6 markedly increase IL-17+ CCR6+ memory T cells and induce IL-17 production in CCR6− memory T cells. We also show that T cell activation induces Foxp3 expression in T cells and that the balance between the percentage of Foxp3+ and IL-17+ T cells is inversely influenced by the cytokine environment. These studies suggest that the cytokine environment may play a critical role in the expansion of memory T cells in chronic autoimmune diseases.
Interleukin-17 or IL-17A is the founding member of a family of cytokines, comprising six members IL-17A-IL-17F. IL-17 is a proinflammatory cytokine that induces the production of various inflammatory mediators including IL-6, IL-8, GM-CSF, and PGE2 from fibroblasts, epithelial, and endothelial cells. IL-17 has been implicated in mediating protection against extracellular microbes and in playing a critical role in the pathogenesis of several inflammatory and autoimmune disorders. In both experimental autoimmune encephalomyelitis and type II collagen-induced arthritis, the animal models for multiple sclerosis and rheumatoid arthritis, respectively, it was shown that IL-17-secreting CD4+ effector T cells are highly pathogenic and essential for the establishment of organ-specific autoimmunity (1, 2).
The IL-17-secreting Th cells (Th17 cells) have been described in mice as a distinct subset of effector cells whose differentiation from naive T cells are promoted by IL-6 and TGF-β and require the transcription factor RORγt (3, 4, 5, 6). IFN-γ and IL-4 inhibit the differentiation of naive T cells into Th17 cells, whereas IL-23 is necessary for the expansion and survival of IL-17-secreting memory T cells (7, 8). In experimental autoimmune encephalomyelitis, IL-23 was found to be essential in the expansion of pathogenic, autoreactive CD4+ T cells, which secreted IL-17A, IL-17F, IL-6, and TNF-α (1).
Since the identification of murine Th17 cells, extensive efforts have been made to characterize the differentiation of naive CD4+ IL-17-secreting T cells in humans. Recently, one group reported that for human naive CD4+ T cells, expression of the transcription factor RORγt and Th17 polarization were induced by IL-1β and IL-6 and suppressed by TGF-β and IL-12, whereas another group showed that IL-1β and IL-23 were critical in the development of human Th17 cells (9, 10). Very little work has been done on the characterization of the IL-17-secreting memory T cells and the factors that regulate IL-17 production. Initially IL-17 mRNA expression was reported to be restricted to CD4+CD45RO+ T cells (11), although no protein data were shown. IL-17 expression in CD8+ memory T cells was shown to require costimulatory signals from accessory cells because polyclonal activation of CD8+CD45RO+ T cells alone did not result in IL-17 production (12). Recently, IL-17-secreting T cells were detected in both CD4+CD45RO+CCR7+ central memory and CD4+CD45RO+CCR7− effector memory T cell subsets, with central memory T cells having a 3-fold higher number than effector memory T cells (13). Further analysis showed that the IL-17+ T cells expressed both CCR6 and CCR4 chemokine receptors. Moreover, a number of studies have not been able to detect IL-17-producing T cells in the peripheral blood of healthy individuals only in patients with inflammatory conditions (14, 15). The cellular phenotype and regulation of human memory T cells secreting IL-17 has not fully been elucidated.
In this study, we demonstrate that effector memory T cells that express the chemokine receptor CCR6 are the principal IL-17-secreting cells and that IL-1β markedly enhances IL-17 production by increasing the percentage of CCR6+ IL-17+ T cells and by inducing IL-17 production in CCR6− memory T cells. Human Th17 cells have a unique cytokine profile as the majority coexpress TNF-α but not IL-6 and a minor subset expresses IL-17 with IL-22 or IL-17 and IFN-γ. Moreover, we show that T cell activation induces Foxp3 expression in T cells and that the balance between the percentage of Foxp3+ and IL-17+ T cells is inversely influenced by the cytokine environment.
Materials and Methods
Isolation of T cell subsets
Research protocols were approved by the Institutional Review Board of New Jersey Medical School in accordance with regulations mandated by the Department of Health and Human Services. Informed consent was obtained from each subject. PBMC were isolated by Ficoll-Hypaque gradient centrifugation of heparinized venous blood obtained from healthy individuals. All isolations of T cell subsets were performed using magnetic beads and reagents from Miltenyi Biotec. CD4+ T cells were isolated from PBMC by immunomagnetic depletion of non-Th cells (CD4+ T Cell Isolation kit) according to the manufacturer’s instructions. Memory CD45RO+ T cells were purified from the CD4+ T cells using magnetic sorting with CD45RO beads, and the purity was >95% in all experiments. Naive CD4+CD45RA+ T cells were obtained by immunomagnetic depletion of CD45RO+ T cells. For isolation of central and effector memory T cell subsets, CD4+ T cells were magnetically separated using CD45RA microbeads and the CD45RA− fraction (CD45RO+) was labeled with FITC-conjugated CCR7 Ab (R&D Systems) followed by anti-FITC microbeads and passed through the magnetic sorting column. This yielded a population of purified central memory T cells (CD4+CD45RO+CCR7+) and effector memory T cells (CD4+CD45RO+CCR7−). To obtain CCR6+/− effector and central memory T cells, we purified CD4+ T cells by magnetic sorting, stained the cells with a PerCP-conjugated CD4, PE-conjugated CD45RO, allophycocyanin-conjugated anti-CCR6, and FITC-conjugated anti-CCR7 and than sorted on a FACSVantage SE System (BD Biosciences). The purity of the sorted CCR6+/− effector and central memory T cells was over 99%. CD8+ T cells were isolated by positive selection using CD8 microbeads to >98% purity. The purity of each separated population was assessed by immunofluorescence flow cytometry.
T cell cultures
Purified CD4+ T cells (2 × 105/well), resuspended in RPMI 1640 supplemented with 10% heat-inactivated FCS (HyClone Laboratories), penicillin, streptomycin, l-glutamine, and CD28 Ab (2 μg/ml; BD Pharmingen) were cultured in 96-well U-bottom plates coated with anti-CD3 Ab (5 μg/ml; BD Pharmingen) for 3 and 6 days. In some experiments, rIL-23 (50 ng/ml; R&D Systems), rIL-6 (40 ng/ml; BioSource International), rIL-1β (10 ng/ml; gift from Hoffmann-LaRoche), and rTGF-β1 (5 ng/ml; R&D Systems) were added to the cultures either alone or in various combinations. Where indicated cultures were supplemented with 5 μg/ml anti-IFN-γ mAb (BD Pharmingen) and 0.5 μg/ml of a polyclonal anti-IL-4 (R&D Systems). Supernatants were collected and stored at −80°C until levels of IL-17, IFN-γ, and IL-4 were determined by ELISA.
ELISA measurement of cytokine levels
IL-17 was measured by ELISA using a plate-bound capture mAb (clone no. 41809, 2 μg/ml; R&D Systems) and a biotinylated polyclonal IL-17 detection Ab (75 ng/ml; R&D Systems), with quantification by reference to a rIL-17 standard (R&D Systems). IFN-γ and IL-4 were quantitated using the OptEIA human IFN-γ kit and IL-4 kit (BD Pharmingen) as per the manufacturer’s instructions. The sensitivity of the ELISA was 7.8 pg/ml for IL-17 and IL-4 and 4.7 pg/ml for IFN-γ.
Intracellular cytokine staining
Cultured cells were harvested on day 6, adjusted to 1 × 106/ml and pulsed with 20 ng/ml PMA (Sigma-Aldrich) and 100 ng/ml ionomycin (Sigma-Aldrich) for 5 h. During the last 2 h, brefeldin A (Sigma-Aldrich) was added to cultures to prevent protein secretion. Cells were stained for surface markers with fluorochrome-conjugated CD4, CD8, CD45RO, CCR6 (R&D Systems), and CCR7 (R&D Systems) Abs, fixed and permeabilized with 0.5% saponin (for Foxp3 staining methanol was used for permeabilization). Cells were incubated with a biotinylated IL-17 Ab (R&D Systems) plus streptavidin PE-Cy7 or with FITC-conjugated anti-TNF-α or PE-conjugated anti-IFN-γ, anti-IL-6, anti-IL-22, or anti-Foxp3 (Biolegend). All Abs with the exception of IL-17, CCR6, CCR7, and Foxp3 were obtained from BD Pharmingen. Data were analyzed on a FACSCalibur cytofluorometer using CellQuest software (BD Biosciences).
Real-time quantitative PCR
Total RNA was extracted from cells using the Absolutely RNA RT-PCR Miniprep kit (Stratagene) according to the manufacturer’s instructions. The first strand of cDNA was obtained using Superscript First-Strand Synthesis System (Invitrogen) and transcripts were quantified by real-time quantitative PCR on an ABI Prism 7000 Sequence Detection System (Applied Biosystems). The IL-17 primers and probe were purchased from Applied Biosystems and the sequences are IL-17, sense 5′-ACAACCGATCCACCTCACCTT-3′, antisense 5′-CTTTGCCTCCCAGATCACAGA-3′, and probe 5′ 6FAM-CTCCACCGCAATGAGGACCCTGAG-TAMRA-3′. For RORC (Hs01076112_ml), GAPDH (Hs00266705_g1), and HPRT-1 (Hs99999909_ml) expression, predesigned TaqMan Gene Expression Assays (Applied Biosystems) and reagents were used according to the manufacturer’s instructions. Specific gene expression was normalized to the housekeeping genes GAPDH and HPRT-1. Expression of IL-17 mRNA levels was calculated by first determining the average threshold cycle (ΔCt) for each culture, which corresponded to the following: (average IL-17 threshold cycle − average HPRT-1 (or GAPDH) threshold cycle). Triplicate samples were used to calculate the average threshold cycle. The replicate threshold cycle (ΔΔCt) was then calculated with the following formula: (ΔCtwith cytokine − ΔCtwithout cytokine). IL-17 and RORC mRNA were expressed as expression fold value (2−ΔΔCt).
Data were analyzed using the Statistical Package for Social Sciences software program (SPSS). Data involving two groups only were analyzed using the paired Student’s t test, whereas data involving more than two groups were analyzed using one-way ANOVA and Tukey’s multiple comparison test. A value of p < 0.01 was considered statistically significant.
Ligation of the TCR and CD28 on CD4+ T cells induces the secretion of IL-17
Since several studies were not able to detect IL-17-secreting T cells in the peripheral blood of healthy individuals, we initially assessed the kinetics of IL-17 secretion by CD4+ T cells activated with immobilized anti-CD3 and soluble CD28 Ab. IL-17 levels were measured in the cell culture supernatants by ELISA. Cross-linking the TCR and CD28 receptor on CD4+ T cells resulted in secretion of IL-17 with low levels of the cytokine detected in day 3 supernatants (mean ± SEM, 726 ± 254 pg/ml), which substantially increased by day 6 (3357 ± 1530 pg/ml) (Fig. 1,a). Activation of CD4+ T cells with PMA and ionomycin induced optimal IL-17 production on day 3. IFN-γ was also produced with similar kinetics by the CD4+ T cells activated with anti-CD3 and anti-CD28 as well as with PMA/ionomycin. No IL-17 was detected in the culture supernatants of the CD8+ T cells activated with anti-CD3 and anti-CD28, whereas PMA/ionomycin induced IL-17 secretion in the CD8+ T cells, although much less than in the CD4+ T cells (Fig. 1,b). CD8+ T cells secreted IFN-γ regardless of which activation method was used. Intracellular cytokine staining of the day 6 cultured cells confirmed the ELISA data and showed that the CD4+ IL-17-secreting T cells were a distinct population from the CD4+ IFN-γ-secreting T cells, although a small percentage of T cells secreted both IL-17 and IFN-γ (Fig. 1 c). No CD8+ IL-17-secreting T cells were detected, although high levels of CD8+ IFN-γ-secreting T cells were seen.
CD4+CD45RO+CCR7−CCR6+ effector memory T cells are the principal IL-17-secreting population
To show that CD4+ memory T cells secrete IL-17, we first purified CD4+ T cells into naive CD45RA+ and memory CD45RO+ T cells and measured IL-17 production after TCR and CD28 cross-linking. IL-17 and IFN-γ were both secreted by CD4+CD45RO+ T cells, whereas IFN-γ but not IL-17 was detected in the supernatants of the activated CD4+CD45RA+ T cells (Fig. 2,a). PMA/ionomycin also stimulated production of IL-17 by CD4+CD45RO+ T cells but not by CD4+CD45RA+ T cells (data not shown). Intracytoplasmic immunofluorescence data confirmed that the IL-17+ T cells were of the CD4CD45RO phenotype (Fig. 2 b). Similar to the results with the unfractionated CD4+ T cells, the IL-17-secreting memory T cells were distinct from the IFN-γ-secreting T cells, and a small subpopulation producing both IL-17 and IFN-γ was detected.
To further characterize which memory T cell subset secreted IL-17, we purified the memory T cells into central memory (CD4+CD45RO+CCR7+) and effector memory (CD4+CD45RO+CCR7−) T cells and measured IL-17 production following ligation of the TCR and CD28 receptor. IL-17 was secreted predominantly by the CD4+CD45RO+CCR7− effector memory T cells; although very little amounts of IL-17 were produced by the CD4+CD45RO+CCR7+ central memory T cells (Fig. 2,c). IFN-γ was also secreted predominantly by the effector memory T cells with low levels of IFN-γ secreted by the central memory T cells. Intracellular cytokine staining was performed to corroborate the protein data and the results from one of five representative donors are shown in Fig. 2,d. In the effector memory subpopulation, there was more than a 4-fold increase (4.3 was median fold increase) in IL-17+ T cells as compared with the central memory T cell subset (Fig. 2,d). The majority of the effector memory T cells secreted only IL-17; however, a proportion (13%) of the IL-17+ T cells secreted both IL-17 and IFN-γ. Because CCR7 may be expressed at low levels on effector memory T cells, we examined CCR7 expression on IL-17+ T cells. We activated purified CD4+CD45RO+ T cells and analyzed IL-17 in the CCR7+/− memory T cells. As shown in Fig. 2,e, the CCR7− effector memory T cells had twice as many IL-17-secreting cells compared with the CCR7+ central memory T cells. We than examined CCR7 expression on all memory CD45RO+ IL-17+ T cells. Analysis of the IL-17+ T cells showed that 77.3% of the IL-17-secreting cells were CD45RO+CCR7−, whereas ∼22.7% of the IL-17 producers were CD45RO+ and expressed low levels of CCR7 (Fig. 2,e). To ascertain whether expression of the chemokine receptor CCR6 was associated with IL-17 secretion, we activated purified CD4+CD45RO+ T cells and analyzed IL-17 secretion in CCR6+/− effector memory T cells. As illustrated in Fig. 2 f, the IL-17-producing T cells (13.3%) were predominantly present in the CD4+CD45RO+CCR7−CCR6+ subset with the low percentage of IL-17+ T cells (1.2%) in the CCR6− fraction probably due to contaminating CCR7lowCCR6+ cells. The majority of the CCR6+ subset secreted only IL-17 but no IFN-γ, but a subpopulation (3.9%) secreted both IL-17 and IFN-γ. In contrast, IFN-γ-producing T cells were detected in both the CCR6+ (21.9%) and CCR6− (44.4%) subset with the majority being found in the CCR6− fraction.
Neutralizing IFN-γ and IL-4 does not effect IL-17 secretion by memory T cells
To determine whether IFN-γ or IL-4 had an effect on the production of IL-17 by memory T cells, we activated CD4+CD45RO+ T cells in the presence and absence of IFN-γ and IL-4 Abs. Levels of IL-17 (2123 ± 711 pg/ml) in cultures with anti-IFN-γ and anti-IL-4 were not significantly different from levels (2080 ± 849 pg/ml) in cultures without the Abs (Fig. 3,a). IL-4 was not detected in any of the cultures. Likewise, there was no difference in the number of IL-17-secreting T cells as measured by intracellular cytokine analysis in the cultures activated in the presence or absence of anti-IFN-γ and anti-IL-4 (Fig. 3 b).
IL-17 is coexpressed with TNF-α and to a lesser extent with IL-22 in memory T cells
In mice, Th17 cells have been reported to secrete IL-17, TNF-α, IL-22, and IL-6 (1). To determine whether the same is true in humans, we examined the expression of IL-17, IL-22, IL-6, TNF-α, and IFN-γ in memory T cells. We observed that only 33% of the total IL-17-secreting cells coexpressed IL-22 and that the majority of IL-17-secreting T cells (67%) did not produce IL-22 (Fig. 4,a). Coexpression of IL-22 with IL-17 was very similar to that of IFN-γ and IL-17. In contrast, TNF-α was associated with the majority of IL-17+ T cells (70%). A population of IL-22-secreting T cells that did not produce IL-17 was also detected. IL-6 production was not detected following TCR and CD28 receptor ligation in any of the individuals tested. Similar results were obtained when we examined IL-22 expression in the CD4+CD45RO+CCR7−CCR6+ T cells (Fig. 4 b). A small subset (3.6%) of CCR6+ effector memory T cells produced both IL-17 and IL-22, although the majority of CCR6+ IL-17+ effector memory T cells (15%) did not secrete IL-22. Likewise, 4% of the CCR6+ effector memory T cells secreted IL-17 and IFN-γ, whereas 13% secreted only IL-17.
IL-1β and to a lesser extent IL-6 and IL-23 up-regulate IL-17 expression in effector memory T cells
To determine whether certain cytokines regulate IL-17 expression of memory T cells, we activated purified CD4+CD45RO+ T cells with anti-CD3 and anti-CD28 in the presence of IL-23, IL-1β, IL-6, or TGF-β and measured IL-17 levels in the culture supernatants. Ligation of the TCR and CD28 on CD4+CD45RO+ T cells in the absence of exogenous cytokines resulted in the induction of IL-17 (median value, 2231 pg/ml), which significantly increased when either IL-23 (4550 pg/ml, p < 0.01) or IL-1β (5485 pg/ml, p < 0.02) alone were added to the cell cultures (Fig. 5 a). Addition of IL-6 alone caused a modest, although not significant increase in IL-17 (2989 pg/ml). However, when IL-6 was added together with IL-23, a significant increase in IL-17 levels (4422 pg/ml, p < 0.004) was observed. The greatest effect on IL-17 secretion was observed when IL-1β was added in combination with either IL-23 (11,726 pg/ml, p < 0.004) or IL-6 (8358 pg/ml, p < 0.004). In contrast, addition of TGF-β alone inhibited, albeit not significantly, IL-17 levels (1185 pg/ml). The inhibitory effect of TGF-β on IL-17 production was significantly counteracted in the presence of IL-1β alone (3607 pg/ml, p < 0.03), IL-23 together with IL-6 (3461 pg/ml, p = 0.05), or the combination of IL-1β, IL-6, and IL-23 (4098 pg/ml, p < 0.04).
We then examined whether these cytokines had an effect on IL-17 mRNA expression by real-time PCR. Similar to the effect observed on protein levels, IL-17 mRNA expression was increased by IL-23, IL-6, and IL-1β alone, whereas IL-23 and IL-1β together markedly augmented IL-17 transcription (Fig. 5,b). Addition to the cell cultures of IL-6 and IL-23, IL-6 and IL-1β, IL-6, IL-23, and TGF-β, or IL-6, IL-1β, IL-23, and TGF-β also enhanced IL-17 mRNA expression. No effect on IL-17 transcription was detected in the presence of TGF-β alone. We than assessed the expression of RORC, the human ortholog of mouse RORγt, a transcription factor that is involved in Th17 differentiation, in these cultures. RORC mRNA expression was detected in the activated CD4 memory T cells (Fig. 5 c). Addition of IL-1β with either IL-23 or IL-6 to the cultures markedly augmented IL-17 mRNA expression but had no affect on RORC mRNA expression. Unexpectedly, expression of RORC mRNA was increased in the presence of TGF-β.
To further examine the role of these individual cytokines on IL-17 secretion, we performed intracellular cytokine analysis of CCR6+/− effector memory T cells activated in the presence and absence of IL-1β, IL-6, and IL-23. We ascertained the effect of these cytokines on three populations of memory T cells: those secreting only IL-17, those secreting IL-17 plus IFN-γ, and those secreting IL-17 plus IL-22. As expected, in the absence of exogenously added cytokines, we detected CCR6+ IL-17+ (11.7%), CCR6+ IL-17+ IFN-γ+ (3.6%), and CCR6+ IL-17+ IL-22+ (3.4%) T cells (Fig. 5 d). Activation with IL-1β led to a 3-fold increase in CCR6+ IL-17+ T cells (33.7%) and a 4-fold augmentation of CCR6+ IL-17+ IFN-γ+ (13.9%) and CCR6+ IL-17+ IL-22+ (13.5%) T cells. When IL-23 or IL-6 was added alone during activation, there was a doubling in the percentage of all three IL-17-secreting subsets. The addition of exogenous IL-23 or IL-6 either alone or together, with IL-1β did not further enhance the number of IL-17-secreting cells over the number detected with IL-1β alone. IL-1β and IL-6 slightly increased the percentage of total CCR6+ T cells, whereas IL-23 slightly decreased the percentage (data not shown). A slight decrease in the percentages of CCR6+ IFN-γ-secreting T cells was seen in the presence of all of these cytokines, whereas no effect was seen in the percentages of CCR6+ IL-22+ T cells.
Surprisingly, in the presence of these three cytokines, IL-17-producing T cells were seen in the CCR6− population. In the absence of exogenously added cytokines, very low numbers of CCR6− IL-17+ (1.4%), CCR6− IL-17+ IFN-γ+ (1.1%), and CCR6− IL-17+ IL-22+ (0.5%) T cells were observed (Fig. 5 d). Activation in the presence of IL-1β induced a 10-fold increase in CCR6− IL-17+ (13.8%), CCR6− IL-17+ IFN-γ (9.7%) and CCR6− IL-17+ IL-22+ (5.2%) memory T cell subsets. Addition of IL-23 resulted in almost a 6-fold enhancement in the number of all three subsets, whereas IL-6 led to a 3-fold increase. The addition of exogenous IL-6 or IL-23 with IL-1β resulted in a slight increase in the percentage of CCR6− IL-17+-secreting T cells compared with the addition of IL-1β alone. IL-1β and IL-6 slightly decreased the percentage of total CCR6− T cells, whereas IL-23 slightly increased the percentage (data not shown). No effect was observed on the frequency of CCR6− IFN-γ+ T cells, although the frequency of CCR6− IL-22+ T cells was slightly increased in the presence of these cytokines.
We than sorted CD4+CD45RO+ T cells based on the expression of CCR6 and CCR7, activated them with anti-CD3 and anti-CD28 with or without IL-1β and IL-23, and measured IL-17 levels. In the presence of IL-1β and IL-23, there was a marked increase in IL-17 secretion in both the CCR6+ and CCR6− effector T cells (Fig. 5 e). The CCR6+ effector memory T cells secreted 2222 pg/ml IL-17, whereas the CCR6− effector memory T cells produced 327 pg/ml, which was a 9- and 17-fold increase, respectively, over levels in the absence of exogenous cytokines. No IL-17 was secreted by the CCR6+/− central memory T cells in the absence or presence of IL-1β and IL-23.
Ligation of the TCR and CD28 leads to Foxp3 expression
Several studies have reported that TCR ligation of human CD4+CD25− T cells induces expression of Foxp3 (16, 17, 18). We investigated whether Foxp3-expressing T cells were present in our cultures along with IL-17+ T cells and whether IL-1β, IL-6, IL-23, and TGF-β affected Foxp3 expression. We activated CD4+CD45RO+CD25− T cells with anti-CD3 and anti-CD28 in the presence or absence of these cytokines and assessed the percentage IL-17+ T cells and Foxp3+ T cells using intracellular cytokine staining and flow cytometry. Cell cultures were gated on surface expression of CD45RO and examined for intracellular expression of IL-17 and Foxp3. In the absence of exogenous cytokines, we detected a low percentage of CD4+CD45RO+ IL-17+ T cells (1.7%) and a high number of CD4+CD45RO+ Foxp3+ T cells (11.9%) following cross-linking of the TCR and CD28 receptor (Fig. 6). Addition of either IL-23 or IL-1β alone led to a doubling in the percentage of IL-17+ T cells but a decrease in the percentage of Foxp3+ T cells, whereas IL-6 had no effect on IL-17 expression but similarly decreased Foxp3 expression. In contrast, activation with TGF-β resulted in a slight decrease in IL-17+ T cells (1.3%) and an increase in Foxp3+ T cells (16.0%). Activation with IL-1β and IL-23 together resulted in the greatest effect on the expression of both IL-17 and Foxp3 with a marked increase in IL-17-secreting cells (11.4%) and a concomitant decrease in Foxp3 expression from (5.3%). Similar but to a lesser extent were the effects of activating with either IL-23 and IL-6 or IL-1β and IL-6. The addition of exogenous TGF-β with IL-1β, IL-6, and IL-23 either alone or in combination led to similar results, which were an increase in IL-17 and decrease in Foxp3 expression.
To further investigate the relationship between IL-17 and Foxp3 expression, we compared the percentage of IL-17+ T cells and the percentage of Foxp3+ cells upon activation under the different culture conditions. We observed that TGF-β promotes the greatest increase in percentage of Foxp3+ cells and decrease in percentage of IL-17+ T cells (Fig. 7). Addition of IL-1β, IL-6, or IL-23 alone or of IL-6 with IL-23 to TGF-β steadily decreases the percentage of Foxp3+ cells with a slight increase in IL-17+ T cells. The percentage of Foxp3+ cells continues to steadily decline in the presence of IL-1β, IL-6, IL-23 alone or IL-6 with either IL-1β and IL-23, and under these conditions the percentage of IL-17+ T cells and of Foxp3+ cells are very similar. The highest percentage of IL-17+ T cells is observed upon addition of IL-1β and IL-23 to the cultures, which induces the greatest drop in the percentage of Foxp3+ cells. The cytokines that promote the greatest increase in IL-17 cause the greatest decrease in Foxp3 expression and conversely the cytokine that induces the greatest increase in Foxp3 promotes the greatest decrease in the percentage of IL-17+ T cells. No association between the percentage of IFN-γ+ T cells and Foxp3+ cells under any of these culture conditions was observed. The percentage of IFN-γ-expressing T cells always greatly exceeded the percentage of Foxp3+ T cells. Thus, the balance between the percentage of Foxp3+ cells and the percentage of IL-17+ T cells is inversely influenced by the cytokine environment.
In the present study, we have characterized the phenotype, cytokine profile, and the factors regulating human memory T cells secreting IL-17. We have identified the CD4+CD45RO+CCR7−CCR6+ effector memory T cell as the principal IL-17-secreting T cell. In contrast, IFN-γ-secreting T cells were found in both CCR6+/− effector memory subsets with the majority being in the CCR6− fraction. Although IL-17 mRNA expression had been initially reported in human memory T cells, no protein data had been shown. During the preparation of this study, a report was published showing the human CD4+ memory T cells secrete IL-17 (19). Recently, it has been reported that CCR6 expression identifies human IL-17-secreting memory T cells and that both central and effector memory T cells produce IL-17 with a higher percentage of IL-17+ T cells detected within the central memory subset (13). Our data are partially in agreement with this study because we found that only CCR6+ effector memory T cells secreted IL-17 and the low percentage of IL-17-secreting T cells detected in the CCR6+ central memory population represent cells with low levels of CCR7 expression. Effector memory T cells have been described as not expressing CCR7 or expressing very low levels of this chemokine receptor.
In mice, Th17 cells have been reported to secrete IL-17, TNF-α, IL-22, and IL-6 (1, 20). Cytometric single-cell analysis of the cytokines produced by human memory IL-17+ T cells revealed coexpression of TNF-α in the majority of IL-17-secreting T cells, whereas IL-22 was only coexpressed in a small subset of the IL-17+ T cells. Similar to the results with IL-22, IFN-γ was coexpressed with IL-17 in a small subset of memory T cells. We consistently observed three subpopulations of memory T cells: IL-17+ memory T cells, IL-17+ and IL-22+ memory T cells, and IL-17+ and IFN-γ+ memory T cells. In fact, it is possible that the IL- 17+ IL-22+ cells are IFN-γ+. Recently, it has been shown that most of the human, naive T cells activated in the presence of IL-1β and IL-6 secrete IL-17 and TNF-α and approximately one-half secrete IL-17 and IL-22 (9). Our observations of memory T cells secreting the same cytokines suggest that these subpopulations represent stable phenotypes of long-lived memory T cells. Human Th17 and IFN-γ-producing Th17 cells have been detected in the gut of subjects with Crohn’s disease and in the blood of healthy individuals (15, 19). IL-6 was not induced following TCR and CD28 cross-linking in our experiments. Whether murine Th17 actually secrete IL-6 is unclear because the initial studies showing that IL-23 expands a population of IL-17, TNF, and IL-6 cells were performed by gene expression analysis and may have contained contaminating IL-6-secreting dendritic cells (1).
In a number of studies, IL-17-producing T cells were not detected in the peripheral blood of healthy individuals only in patients with inflammatory conditions (11, 14). We have consistently detected IL-17 after TCR and CD28 ligation of CD4+ T cells from healthy subjects. We observed that peak secretion of IL-17 occurs on day 6 following TCR and CD28 receptor ligation of CD4+ T cells and on day 3 for PMA/ionomycin activation. In previous studies, IL-17 levels were only measured early, between 24 and 72 h, after CD3/CD28 cross-linking and not at later time points (11, 14, 21, 22). One explanation for the failure to detect IL-17+ T cells in peripheral blood of healthy individuals could be that IL-17 levels were measured at early and not at later times following activation. Whether activated CD8+ T cells secrete IL-17 was unclear because IL-17 mRNA was found to be highly expressed in CD8+ T cells activated with PMA/ionomycin but no protein data were available (12). Another study showed that CD8+ T cells secreted very low levels of IL-17 (22). In our experiments, IL-17 was not detected by cross-linking of CD3 and CD28 on CD8+ T cells, although IFN-γ was produced by these cells. PMA and ionomycin induced IL-17 production by CD8+ T cells; however, the levels were ∼100-fold less than that secreted by CD4+ T cells.
Our study is the first to characterize the cytokines and the mechanisms used to regulate IL-17-secreting memory T cells. We demonstrate that the cytokines that promote the differentiation of human naive T cells into IL-17-secreting cells regulate IL-17 production by memory T cells. IL-1β, IL-23, and to a lesser extent IL-6 up-regulated IL-17 secretion with an additive effect seen when IL-1β is added together with IL-23 or IL-6. IL-23 with IL-6 enhanced IL-17 production but the effect was similar to that seen with IL-23 alone. In contrast, TGF-β inhibited IL-17 when added alone or together with either IL-1β, IL-23, or IL-6. The regulatory effects of the cytokines were seen both at the protein level and at mRNA transcription. In contrast to IL-17, RORC expression was not affected by these cytokines, suggesting that IL-17 expression in memory T cells is likely regulated by other transcription factors. Cytometric single-cell analysis of CCR6+/− memory T cells revealed that the one of the mechanisms by which these cytokines augment IL-17 secretion is by increasing the number of CCR6+ IL-17+ memory T cells. IL-1β had the greatest effect, whereas both IL-23 and IL-6 had more modest effects. Moreover, these cytokines induced IL-17 secretion in CCR6− memory T cells with IL-1β displaying the strongest effect followed by IL-23 and IL-6. Whether IL-1β, IL-6, and IL-23 induce IL-17 production in memory T cells that are truly CCR6− or whether they act on memory T cells that have down-regulated CCR6 is not certain. Our results suggest that in the presence of IL-1β and IL-6, CCR6− acquire the capacity to produce IL-17 and that IL-23 may function in down-regulating CCR6 expression. Furthermore, IL-17 and IL-22 as well as IL-17 and IFN-γ were coordinately regulated by IL-1β, IL-23, and IL-6. Thus, IL-1β, IL-23, and IL-6 regulate IL-17 secretion by increasing the frequency of CCR6+ IL-17+ memory T cells and by inducing IL-17 production in CCR6− memory T cells.
Another potential mechanism by which these cytokines regulate IL-17 secretion is by modulating Foxp3 expression. We have observed that cross-linking of the TCR and CD28 coreceptor on CD4+CD45RO+CD25− T cells results in the expression of Foxp3+ as well as IL-17+ T cells and that IL-1β, IL-6, IL-23, or TGF-β concomitantly modulated Foxp3 and IL-17 expression. IL-1β, IL-6, and IL-23 either alone or in various combinations up-regulated IL-17 and down-regulated Foxp3 expression, whereas TGF-β modestly decreased IL-17 and up-regulated Foxp3 expression. Moreover, we found that the balance between the percentage of Foxp3+ and IL-17+ T cells is inversely influenced by these cytokines. It has been shown that TCR stimulation of human CD4+CD25− T cells results in expression of Foxp3 and that these Foxp3+ cells have regulatory function (16). Other studies reported that the induced Foxp3 expression did not correlate with suppressive function (18, 23). The Foxp3 mAb used in our experiments is clone 259D, which has been shown to specifically bind to the Foxp3 transcription factor (23). Further studies will determine whether the Foxp3-expressing T cells detected in our culture system are “truly” T regulatory cells.
Regulation of established Th17 effector memory cells might be of relevance in autoimmune diseases where the amounts of IL-17 may determine the degree of tissue damage. Our study suggests that there are a number of mechanisms by which IL-17 secretion by memory T cells may be enhanced and perpetuated. IL-17 induces the production of a number of proinflammatory cytokines such as TNF, IL-1β, and IL-6 that may act in an autocrine loop and enhance IL-17 secretion of memory cells. IL-1β with the assistance of IL-23 or IL-6 enhances IL-17 secretion by increasing the frequency of CCR6+ memory T cells as well as inducing IL-17 secretion from CCR6− memory T cells. Moreover, these cytokines down-regulate expression of Foxp3, which results in increased IL-17 levels. The coexistence of IL-17+ and IL-22+ and IL-17+ and IFN-γ+ memory T cells suggests that these cells may be involved in tissue inflammation and damage. We propose that there is one memory T cell subset that is IL-17+ IL-22+ IFN-γ+ and that this population plays a critical role in autoimmunity. During acute experimental autoimmune encephalomyelitis, it has been shown that ∼50% of the T cells infiltrating the CNS coexpress IL-17 and IFN-γ (24). Recently, one study reported that numerous CD45RO+ cells immunopositive for IL-17 or IL-22 were found in multiple sclerosis lesions (25). Further studies characterizing the role of the IL-17 subsets and the Foxp3-expressing cells in human autoimmune conditions are needed.
We thank Tammy Galenkamp for performing the cell sorting.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported by Grant NS34245 from the National Institutes of Health (to C.R.-K.).