Cutting Edge: Alternative Signaling of Th17 Cell Development by Sphingosine 1-Phosphate¹

The Th17 subset of effector and memory CD4 T cells develops under the influence of TGF-β1, IL-1 or TNF-α, and IL-6 and expands up to the full potential production of IL-17 in response to IL-23 (1, 2). Th17 cells have major roles in host defense against diverse microbes and in full expression of many different types of inflammatory responses, largely through the secretion of IL-17, IL-17F, and IL-6 (3, 4, 5). At low concentrations, IL-17 also may suppress activities of CD4 T cells and dendritic cells and block the generation of chemokines by immune and other types of cells, thereby leading to the inhibition of established allergic diseases including asthma (6).

Several reports suggest that Th17 cells may develop and function fully in the absence of IL-23, which raises possibilities of the involvement of other T cell-active mediators (7, 8, 9). Sphingosine 1-phosphate (S1P)³ from innate immune cells and erythrocytes acts through type 1 S1P receptors (S1P₁s) of T cells to affect their migration in tissues, lymphoid homing, and proliferation (10, 11, 12). Recently, we have shown that OVA Ag challenge of double transgenic CD4 T cells bearing OTII OVA-TCRs and overexpressing S1P₁ T cell-selectively resulted in the development of more Th17 cells generating higher mean levels of IL-17 in vitro and in vivo than after OVA challenge of single transgenic T cells only bearing OTII OVA-TCRs (13). We therefore further analyzed the possibility that the S1P-S1P₁ axis may have an IL-23-like capacity to enhance the development, expansion, and activities of Th17 cells in a modified in vitro model system (14). The present results prove that the S1P-S1P₁ axis has the same ability as IL-23 to augment the number and IL-17 secretory activity of Th17 cells by a mechanism that is similarly susceptible to immune-specific suppression by IFN-γ and IL-4 and by IL-27.

Cytokines, Abs, pharmacological inhibitors, and other reagents

The following were obtained from the indicated suppliers: murine IL-6, IL-1β, IL-4, and IFN-γ and human TGF-β1 (PeproTech); murine IL-23 and IL-27 (eBioscience); guinea pig purified anti-mouse CD3ε, monoclonal Syrian hamster anti-mouse CD28, and rat IgG1 anti-mouse IL-17 (clone TC11–18H 10.1) Abs (BD Pharmingen); S1P, the PI3K inhibitor wortmannin, and the BAY 11-7082 inhibitor of IκB-α phosphorylation and thus NF-κB (BIOMOL); charcoal- and dextran-treated FBS (HyClone); and goat alkaline phosphatase-conjugated anti-rat IgG Ab and 5-bromo-4-chloro-3-indolyl phosphate/NBT alkaline phosphatase substrate (Calbiochem).

CD4 T cells were isolated from spleens of 6- to 10-wk-old C57BL/6 mice by an immunomagnetic bead method as described (15) and replicate suspensions of 2 × 10⁶/ml RPMI 1640 with 10% charcoal- and dextran-treated FBS, 100 U/ml penicillin G, and 50 μg/ml streptomycin were cultured in 12-well plates that had been precoated with 2 μg each of anti-mouse CD3 and anti-mouse CD28 Abs. The medium was supplemented with 10 ng/ml IL-1β, 10 ng/ml IL-6, and 1 ng/ml TGF-β1 and, in some studies, 10 ng/ml IL-23 and/or 100 ng/ml IL-27 and/or 100 ng/ml IFN-γ plus 1 ng/ml IL-4 and was changed every 48 h. S1P, SEW2871, and/or phospho-FTY720 (P-FTY720) were added to cultures every 24 h.

Supernatant fluid concentrations of IL-4, IFN-γ, and IL-17 were quantified in duplicate by ELISA kits following procedures recommended by the manufacturer (Ready Set-Go kits; eBioscience).

Replicate 0.5-ml suspensions of 0.5–1 × 10⁶ CD4 T cells that had been variously treated were incubated for 2 h at 37°C in 24-well plates precoated with 1 μg/well guinea pig purified anti-CD3ε, washed with PBS, fixed in 1% paraformaldehyde for 30 min at 4°C, washed, and permeabilized with 0.5% Triton X-100 in PBS for 30 min at 20°C. After 30 min at 20°C in PBS with 5% BSA and 0.1% Triton X-100, the CD4 T cells were washed twice with PBS, incubated for 16 h at 4°C in 50 ng of rat monoclonal anti-IL-17 Ab/0.25 ml of PBS with 1% BSA and 0.1% Triton X-100, washed twice with PBS, incubated in 1/5,000 goat alkaline phosphatase-conjugated anti-rat IgG in PBS with 1% BSA and 0.1% Triton X-100 for 2 h at 20°C, and washed twice again with PBS. After incubation for 30 min at 20°C in 5-bromo-4-chloro-3-indolyl phosphate/NBT alkaline phosphatase substrate, the CD4 T cells were rinsed once with PBS and once with distilled water, air dried, and a minimum total of 500 stained and unstained cells per well were counted microscopically.

S1P and IL-23 stimulate development of Th17 cells from splenic CD4 T cells in vitro

Incubation of mouse splenic CD4 T cells in S1P-deleted medium on adherent anti-CD3 plus anti-CD28 Abs with IL-1β, IL-6, and TGF-β1 as recently described (14), but without IL-23, resulted in significant increases in the generation of IL-17 after 5 days that were a mean of 10-fold higher than the medium control with anti-CD3 plus anti-CD28 Abs alone and 12-fold higher with these TCR-directed Abs and the mixture of cytokines together (Fig. 1,A). The generation of IL-17 by the anti-TCR Ab (a-TCR)- and cytokine-stimulated CD4 T cells was increased further by S1P with clear concentration-dependence that attained significance at the physiologically relevant levels of 10⁻⁷ M and 10⁻⁶ M S1P on both days 4 and 5 (Fig. 1,B). Slight increases in IL-17 generation evoked by S1P on day 3 and earlier were not significant. In the same series of studies, a concentration of the S1P₁-selective agonist SEW2871 that is equivalent in signaling activity to 1 μM S1P also significantly augmented the generation of IL-17 by the stimulated CD4 T cells on day 5 (Fig. 1 B). In three of the four studies, an optimal concentration of 10 ng/ml IL-23 increased IL-17 generation up to 470 ± 98 pg/ml (mean ± S.D.; p < 0.01) on day 5. In contrast, neither 10⁻⁹ to 10⁻⁶ M S1P or 10 ng/ml IL-23 increased IL-17 generation by CD4 T cells not conditioned by a-TCR and cytokines.

To further evaluate the effectiveness of S1P relative to IL-23, optimal levels of each were introduced into cultures separately, IL-17 was quantified on days 4 and 5, and the respective IL-17 concentrations were expressed as a ratio (Fig. 1,C). Although IL-23 was a mean of 2.2- to 2.4-fold more active than S1P at enhancing the generation of IL-17 by stimulated CD4 T cells, the differences were not significant because of the level of variation in results. The percentages of CD4 T cells containing intracellular IL-17 also were quantified in these studies by an ELISPOT assay to determine the effects of S1P and IL-23 on the expansion of Th17 cells, as well as their level of secretion of IL-17. The percentage of IL-17-containing CD4 T cells was increased significantly and progressively by 10⁻⁸ to 10⁻⁶ M S1P up to a level indistinguishable from that achieved by the optimal level of IL-23 (Fig. 1,D). In the context of this model system for Th17 cell development and maturation, therefore, S1P is a physiologically complete surrogate for IL-23. After the initial differentiating effects of TCR and cytokine signals on naive CD4 T cells, S1P and IL-23 are similarly able to enhance the expansion and functional activation of Th17 cells. Under optimal conditions, S1P and IL-23 promoted the development of the same number of Th17 cells (Fig. 1,D), but IL-23 appeared to be twice as effective as S1P at augmenting extracellular secretion of IL-17 (Fig. 1 C).

Enhancement of the development of Th17 cells by S1P led to the concurrent suppression of the generation of IL-4 and IFN-γ (Fig. 2). At days 4 and 5 of incubation, 10⁻⁸ to 10⁻⁶ M S1P very significantly suppressed generation of IFN-γ and 10⁻⁷ M and 10⁻⁶ M S1P significantly reduced the generation of IL-4. This effect of S1P on IFN-γ generation had been demonstrated previously in TCR-activated CD4 T cells in the absence of a mixture of differentiating cytokines, but decreases in IL-4 had not been detected in any other culture system of native T cells (16, 17) (13). Thus, enhancement of the Th17 subset in this system apparently is accompanied by concomitant suppression of the Th1 and Th2 subsets.

The decreases in the generation of IFN-γ and IL-4 observed concurrently with the increases in IL-17 raised a possibility that the development of the Th1/2 and Th17 subsets of CD4 T cells is immunoregulated by reciprocal mechanisms in this system. The addition on day 3 of concentrations of exogenous IFN-γ and IL-4 attained in cultures of CD4 T cells stimulated by the adherent anti-CD3 plus anti-CD28 Abs without the cytokine mixture or S1P very significantly prevented the appearance of IL-17 generation (Fig. 3,A). The known inhibitor of IL-23-mediated maturation of Th17 cells, IL-27, also very significantly prevented S1P-mediated development of Th17 cells (Fig. 3,A). An initial exploration of the distinctive biochemical prerequisites of S1P-S1P₁ axis enhancement of Th17 cell generation of IL-17 suggested major roles for PI3K and NF-κB. At optimally effective concentrations, the PI3K inhibitor wortmannin and the inhibitor of IκB-α-phosphorylation BAY11-7082 both strikingly suppressed the development of IL-17 generation by differentiating Th17 cells (Fig. 3 B). Both inhibitors suppress biochemical pathways coupling G protein-coupled receptors (GPCRs) for some inflammatory mediators to the generation of cytokines, including IL-6 and IL-8, that are critical for inflammatory and metabolic responses (18, 19, 20). In particular, the extension of NF-κB involvement to S1P-enhancement of both Th17 cell development and IL-17 generation broadens the spectrum of NF-κB contributions to immunological cytokine-mediated chronic inflammation (21). The potential involvement of the S1P-S1P₁ axis in several autoimmune diseases had been considered to be attributable solely to the effects on lymphocyte trafficking in lymphoid organs and lymphocyte migration in nonlymphoid target organs (22). However, the present and other recent findings imply that the S1P-S1P₁ axis has major roles in the control of T cell proliferation (23) and the mediation of target organ-specific chronic inflammation through the Th17 cell-IL-17 system. The differences in the sources of S1P and IL-23 as well as in expression of S1P₁ and IL-23 receptors suggest that these distinct stimuli represent alternative means of recruiting and activating Th17 cells. S1P in blood, lymph, and lymphoid organs is generated predominantly by mast cells and platelets, whereas erythrocytes function principally as intravascular buffers and reservoirs for S1P delivery. S1P₁ is highly expressed by naive and memory T cells and B cells, with much lower levels in activated effector lymphocytes (15, 24). In contrast, IL-23 is a product of monocytes, macrophages, and activated dendritic cells, and its target cells also are monocytes and macrophages as well as precursors of Th17 cells (5, 25). The immunopathological activity of IL-23 is manifested preferentially in intestinal and other mucosal inflammation mediated by innate and T cell-dependent mechanisms, in contrast to that of IL-12, which is expressed mostly in systemic inflammatory reactions and the resultant cachexia (5). Thus Th17 cells are more likely to be recruited by the S1P-S1P₁ axis in acute systemic and vascular inflammation but would be more effectively mobilized by IL-23 in infections and intestinal inflammatory states.

The immunosuppressive agent FTY720, which has both agonist and down-regulatory activity for S1P₁ and some other S1P GPCRs (26, 27), has proven to be an effective therapy for human transplant rejection and multiple sclerosis (28, 29). In many rodent models of human autoimmunity that are clearly T cell-dependent, FTY720 also suppresses immunopathology and diminishes clinical signs of disease (22). It has become increasingly evident that the proven ability of FTY720 to disrupt the lymphoid organ traffic of T cells, induce blood lymphopenia, and minimize T cell infiltration of target organs by effects on the S1P-S1P₁ axis are only one set of mechanisms for its potent immunosuppression. This realization is supported by findings that its clinical immunosuppressive benefit can be attributed partially to altering effects of the S1P-S1P₁ axis on the nonmigration functions of T cells, including proliferation and cytokine secretion. The introduction of FTY720 into cultures of Th17 cells developing under the influence of S1P substantially suppressed the generation of IL-17 (Fig. 4). These present findings add significantly to our understanding of the potential mechanisms underlying FTY720 therapeutic immunosuppression.

We are grateful for the expert graphics by Robert Chan and the dedicated technical assistance of Yvonne Kong.

The authors have no financial conflict of interest.