The Sphingosine 1-Phosphate Receptor Agonist FTY720 Potently Inhibits Regulatory T Cell Proliferation In Vitro and In Vivo¹²

Naturally occurring regulatory T cells (Treg)⁴ have emerged as a unique CD4⁺ T cell subset mainly characterized by coexpression of high CD25 levels together with the Forkhead transcription factor FoxP3 (1, 2, 3). Treg play a decisive role in the maintenance of immunological tolerance (2). Various target cells such as CD4⁺ and CD8⁺ T cells (4, 5, 6) and NK cells (7), as well as NKT cells (8) and dendritic cells (9), and different immunosuppressive mechanisms including TGF-β (10) and CTLA-4 (11) have been described to date. The proliferative potential of Treg is well documented and is assumed to be of critical importance for counter-regulation of an ongoing effector T cell activation, as a rapid expansion of the Treg pool tips the numerical balance in favor of a more immunosuppressive milieu (12). Among various others, IL-2 has been proven to play a critical role for Treg survival and proliferation (13, 14). Accordingly, the expansion of Treg rather than that of effector T cells in cancer trials testing IL-2 as an immunostimulant support this concept (15). Finally, data from murine models further corroborate these clinical observations, as IL-2 is indispensable for peripheral maintenance of naturally occurring Treg (16).

In addition, the immunosuppressive potential of Treg also depends on direct cell-to-cell contact with target cells. Thus, their migration to sites of immune responses has to be tightly regulated. Secondary lymphoid organs (SLO) are considered to be an important environment for Treg-mediated immune regulation (17), as lymph nodes (LN) as well as the spleen are critical for orchestrating T cell-mediated immune responses (18). Obviously, in case of an exaggerated immune activation within target organs, Treg also have to enter end organs to induce local immune regulation. In line with this idea, it has recently been demonstrated in an experimental model of acute graft-vs-host disease (GvHD) that LN-homing, CD62L-expressing Treg are superior to CD62L-negative Treg in terms of disease control (19, 20). In contrast, in a model of chronic GvHD that rather reflects a situation of ongoing T cell activation within damaged end organs, it appeared to be mainly the CD103⁺ Treg population that was inhibiting GvHD (21). These observations fit perfectly the recent description of various Treg subpopulations characterized by a distinct chemokine and cell surface receptor expression pattern, i.e., CCR7- (22), CCR5- (23), CXCR4- (23), CD62L- (20, 24), and CD103-expressing (21) Treg. In addition to the data derived from the GvHD models, we could recently demonstrate the potent nephroprotective effects of adoptively transferred Treg in an accelerated model of anti-glomerular basement membrane (GBM) glomerulonephritis (GN) (25). In this particular model, Treg do not infiltrate the inflamed end organ but primarily migrate to the SLO, which most likely induces inhibition of tissue-destructive effector T cells. Thus, especially in the early phase of inflammatory processes, efficient LN occupancy as well as a high proliferative potential of Treg is required for efficient immunosuppression.

FTY720 (2-amino-2-(2-[4-octylphenyl]ethyl)-1,3-propanediol hydrochloride) is a sphingosine analog and an immunomodulatory agent derived from the Chinese fungus Iscaria sinclarii. FTY720 has been proven to prolong allograft survival in both rodents and men (26, 27) by acting as a sphingosine-1 phosphate (S1P) agonist inducing a prolonged down-regulation of the cell surface expression of the S1P receptor. The latter enables a blockade of S1P-mediated T cell egress from SLO, which is normally induced by a S1P gradient from blood and lymph fluid to tissue (28). Among its various target cell effects, the trapping of lymphocytes in SLO and the subsequent induction of lymphopenia is thought to be the predominant mechanism that prevents effector T cell exit into the blood stream and subsequent entry into target tissues (29). Interestingly, immunosuppressive function after systemic long-term application of FTY720 has at least in part been attributed to the induction of FoxP3 in naive T cells, thereby inducing T cells with a regulatory phenotype (30). However, this is just one among the plethora of mechanisms induced by FTY720 in vivo, as it has also been shown to modify lymphatic endothelial (28), dendritic (31), and B cell function (32).

We provide in this report first evidence that the S1P receptor agonist FTY720 potently inhibits Treg proliferation in vitro and in vivo by inhibiting IL-2 mediated mitogenic signals. Despite the fact that FTY720 induced the trapping of Treg in inflamed SLO, the immunosuppressive effects of FTY720-exposed and subsequently adoptively transferred Treg were abrogated in two murine models of inflammation.

C57BL/6 and BALB/c mice (6–12 wk of age) were purchased from Harlan Laboratories. For bone marrow transplant experiments, luciferase-expressing transgenic C57BL/6 L2G85 mice (33) were used. All experiments used age and sex-matched control mice and were performed in accordance with protocols approved by the Institutional Animal Care and Use Committee of the Innsbruck Medical University, Innsbruck, Austria (35/9185.81/G-07/19, Austrian Federal Ministry for Education, Science and Culture (BMBWK)-66.011/0087-BrGT/2005).

CFSE and TAMRA were purchased from Molecular Probes. FTY720 was a gift from Dr. V. Brinkmann (Novartis Pharma) and SEW2871 was purchased from Biomol. Anti-CD3 was purchased from BD Biosciences. The annexin V detection kit was from R&D Systems and IL-2, IL-7, and IL-15 were from Peprotech.

Spleens and LN were removed and gently dissociated to single cell suspensions. CD4⁺CD25⁺ T cells (Treg) were magnetically separated using a Treg separation kit (Milteny Biotec) achieving purity of >95%. The selected cells were incubated for 1 h at 37°C with 0.5 μg/ml FTY720-treated FTY-Treg) or solvent-treated Treg (solv-Treg) followed by two washing steps before further use. In some experiments, Treg were incubated with increasing FTY720 or SEW2871 concentrations as indicated.

Isolated Treg were exposed to increasing concentrations of FTY720 ranging from 0.5 μg/ml to 4 μg/ml, washed, and taken into culture using anti-CD3 together with 100 U/ml IL-2. Induction of cell death was determined by detection of annexin V and propidium iodide using FACS analysis at the indicated time points.

To determine the regulatory properties of Treg, cocultures of CD4⁺CD25⁻ T cells together with equal numbers of either FTY720- or solv-Treg were performed in anti-CD3-precoated (5 μg/ml) wells. As control, the respective T cell population was cultured alone. Proliferation was measured on day 5 by [³H]thymidine incorporation in a beta scintillator.

CFSE-labeled (0.5 μM) Treg at 1 × 10⁶/ml were cultured with 100 ng/ml plate-bound anti-CD3 together with 100 U/ml IL-2 for 2–14 days. Stimulation with anti-CD3 plus IL-2 results in cell division with distinct CFSE fluorescence peaks that allows for discrimination between cycled (at least one division) and noncycled cells by FACS analysis.

All FACS Abs (except the anti-FoxP3 mAb) were purchased from BD Biosciences. Phosphospecific flow cytometry was performed strictly according to the manufacturer’s manual (protocol 3; BD Biosciences) after stimulation using plate bound anti-CD3 (5 μg/ml; BD Biosciences) together with one of the following common γ-chain cytokines: IL-2 (100 U/ml), IL-7 (10 ng/ml), or IL-15 (25 ng/ml). S1P specificity of FTY720-mediated effects on STAT-5 phosphorylation was tested by using SEW2871 at concentrations ranging from 100 to 1000 nM instead of FTY720. Flow cytometric analyses were performed on a FACSCalibur (BD Biosciences) and data were analyzed with Cell Quest software (BD Biosciences). FoxP3 was stained by APC-labeled anti-mouse FoxP3 mAb (eBioscience).

Treg (5 × 10⁶) were lysed on ice and lysates were analyzed by electrophoresis on 10% SDS-polyacrylamide gels after being blotted on hydrophobic polyvinylidene difluoride membrane (GE Healthcare). Mouse anti-phosphorylated STAT-5 Ab (clone 8-5-2, dilution 1/1000; Millipore) was added and binding was subsequently detected by HRP-conjugated goat anti-mouse IgG (dilution 1:2000; DakoCytomation). Chemiluminescent reaction was induced using a Lumigen TMA-6 kit (GE Healthcare) and the blots were analyzed by using a ChemiDoc XRS⁺ molecular imaging system (Bio-Rad). To evaluate expression of total STAT-5 protein, the blots were incubated with 1× mild Ab stripping solution (Chemicon) and with mouse anti-STAT-5 Ab (clone 89/STAT-5, dilution 1/400; BD Biosciences).

To detect IL-2 and IFN-γ in supernatants, FTY720 or solvent-exposed T cells were restimulated with PMA/ionomycin. Cytokines were detected by ELISA (BD Biosciences).

FTY- and solv-Treg suspensions were differentially labeled with CFSE (green) or TAMRA (red) dyes for adoptive transfer strictly according to the manufacturer’s recommendations and subsequently injected i.v. (5 × 10⁶ of each Treg population) to follow their in vivo migration behavior under steady state or inflammatory conditions. The latter were induced by s.c. injection of 0.02 g of nonviable desiccated Mycobacterium tuberculosis H37a (Difco Laboratories) dissolved in IFA (Sigma-Aldrich) into the footpad. Seven days after the induction of inflammation, differentially labeled FTY- and solv-Tregs were adoptively transferred. Peripheral blood, spleen, draining (popliteal), and nondraining (popliteal, contralateral to footpad injection) LN were harvested 4–48 h after Treg transfer and subsequently prepared either as single cell suspension (for FACS) or snap frozen (for preparation of tissue sections). Images were taken with an upright Leica MZ 16 FA stereomicroscope equipped with a spot camera and operated via MetaMorph software (system implemented by Visitron Systems). Original magnification was ×10.

Acute GvHD (aGvHD) was induced as described previously (34). In brief, recipients were exposed to lethal irradiation with 850 cGy on day 0 followed by i.v. injection of 5 × 10⁶ T cell-depleted bone marrow cells after 1 day. To induce aGvHD, 1 × 10⁶ CD4⁺/CD8⁺ T cells in the C57BL/6→BALB/c model, respectively, were injected i.v. on day 2. Treg were derived from donors of the same genetic background as those of conventional T cells (Tconv) and injected i.v. on day 0 at the same dosage as the Tconv. Slides of small bowel and large bowel samples collected on day 7 were stained with H&E and scored by a blinded and experienced pathologist (U.V.G.) according to a previously published histopathology scoring system.

BLI was performed as previously described (33). Briefly, mice were injected i.p. with luciferin (10 μg/g body weight). Ten minutes later, mice were imaged using an IVIS200 charge-coupled device imaging system (Xenogen) for 5 min. Imaging data were analyzed and quantified with Living Image software (Xenogen) and IgorProCarbon (WaveMetrics).

An accelerated model of nephrotoxic nephritis was induced as previously described (35). In brief, mice were preimmunized s.c. with 2 mg/ml rabbit IgG (Jackson ImmunoResearch Laboratories) dissolved in IFA and nonviable desiccated Mycobacterium tuberculosis H37a (Difco Laboratories). Two days later, animals received either 3 × 10⁶ solv-Treg or equal numbers of FTY-Treg i.v. As control, mice received 3 × 10⁶ CD4⁺CD25⁻ T cells. After 3 days, 5 mg of heat-inactivated rabbit anti-mouse GBM antiserum per 20 g of body weight was injected. Twenty-four-hour urine samples were collected in metabolic cages on days −1, 1, 7, and 14 after the induction of GN. Urinary albumin was determined by a double-sandwich ELISA (Abcam), and urinary creatinine was quantified spectrophotometrically using a commercially available kit (Sigma-Aldrich).

Formalin-fixed tissue was embedded in paraffin, cut in 4-μm sections, and stained with periodic acid Schiff (PAS) for histologic analysis. Sections cut from frozen tissue (4–25 μm) were used for LN migration studies as well as for immunoperoxidase staining of macrophages and T cells in the kidney (25) using a rat anti-mouse macrophage Ab (Clone F4/80; Serotec), a rat anti-mouse CD4 mAb (clone YTS191.1; Serotec), and an IgG2a isotype Ab as negative control. Biotin-conjugated goat anti-rat IgG Ab (Jackson ImmunoResearch Laboratories) was used as a secondary Ab, followed by incubation with an avidin-biotin complex and subsequent development and HE counterstaining. T cells were quantified by counting the number of cells in six adjacent high-power fields of the renal cortex and medulla. A semiquantitative scoring system was performed for the quantification of macrophages: 0, 0 to 4 cells stained positive; 1+, 5 to 10 cells; 2+, 10 to 50 cells; 3+, 50–200 cells; and 4+, >200 cells stained positive per low-power field.

FTY720 was administered i.p. in a dose of 1 mg/kg body weight on two consecutive days, whereas control mice received solvent. Mice were sacrificed 36 h after the first i.p. injection for isolation of Treg and CD4⁺CD25⁻ control cells from spleen and LN suspensions. The isolated cells were cultured on plates precoated with 100 ng/ml anti-CD3 together with 100 U/ml IL-2 for 5 days and proliferation was determined by [³H]thymidine incorporation.

Results represent data from at least three independent biological experiments. After ANOVA, Student’s t test was used. Values of p < 0.05 were considered significant. Figures show means ± SEM. Statistical analysis was performed using GraphPad Prism.

Only little information is available regarding the direct effects of FTY720 on isolated Treg. Thus, we analyzed the influence of the drug on characteristic Treg properties. First, we tested the effects of FTY720 on Treg viability. Using increasing doses (0.5–4 μg/ml) of FTY720, we observed an increased rate of cell death only after 72 h at the highest dose level of 4 μg/ml, which is known to be toxic for various cell types (Fig. 1). We therefore used the concentration of 0.5 μg/ml for our subsequent in vitro and in vivo experiments. We next tested whether the characteristic properties of Treg function are modulated by short-term FTY720 exposure. The suppressive potential of FTY-Treg, as measured by standard in vitro suppression assays, remained unaltered as compared with solv-Treg, even when FTY-Treg were mixed with CD4⁺CD25⁻ responder T cells in different ratios ranging from 1:1 to 1:10 (Fig. 2,A; ∗, p < 0.05 for FTY- and solv-Treg vs responder cells alone). Moreover, FTY-Treg suppressed the production of IFN-γ and IL-2 by CD4⁺CD25⁻ T cells as effectively as their solvent-treated counterparts (Fig. 2,B; ∗, p < 0.05). Finally, FTY-Treg showed a comparable cell surface marker expression (CD62L, CD103, CCR7) as well as intracellular FoxP3 levels after short-term FTY720 exposure remained unaltered (Fig. 2 C). Thus, typical functional and phenotypical hallmarks of bona fide Treg are preserved upon short-term exposure to FTY720 in vitro.

The critical role of IL-2 for Treg proliferation is well documented (36). Thus, we next focused on in vitro Treg expandability upon stimulation via the TCR together with IL-2. In contrast to solv-Treg, which can be readily expanded, FTY-Treg exhibit a markedly impaired proliferation in vitro (Fig. 3,A; ∗, p < 0.05). These data were corroborated in vivo by measuring CFSE dilution upon adoptive transfer of either CFSE-labeled FTY- or solv-Treg into mice suffering from footpad inflammation. In this model, FTY-Treg also showed a severely reduced expansion in the draining LN when compared with adoptively transferred solv-Treg. As depicted in Fig. 3 B, only 24% of transferred FTY720-treated, CFSE-positive Tregs divided 48 h after i.v. injection in vivo (∗, p < 0.05), whereas 70% of solv-Treg divided at least once during this time period. This effect was solely detectable in draining LN but not in the spleen, where Treg did not proliferate within 48 h.

On the molecular level, we found that FTY720 pre-exposure of Treg dose-dependently impaired IL-2-induced phosphorylation of STAT-5, which is a critical component of the IL-2 signal transduction pathway (Fig. 4, A and B; ∗, p < 0.05). In contrast, solvent pre-exposure did not alter IL-2-induced phosphorylation of STAT-5. Reduced sensitivity to IL-2-induced signals is not mediated by down-regulation of the IL-2 receptor α-chain upon FTY720 exposure (Fig. 4,C). However, in line with the reduced activation of Treg by IL-2 after FTY720 exposure (which is mirrored by inhibition of proliferation), further up-regulation of the mean fluorescence intensity of CD25 throughout the culture period (at 72 and 96h after culture initiation) is prevented by FTY720 (the percentage of CD25-expressing cells remains unchanged in both fractions; data not shown). It is well documented that the IL-2/STAT-5 pathway plays a decisive role for FoxP3 maintenance in Treg (37). Thus, it is not surprising that the maintenance of FoxP3 expression during in vitro culture of initially FTY720-exposed Treg is affected (Fig. 4 D).

S1P receptor 1 specificity of the FTY720-mediated effects was proven by using increasing doses of the S1P receptor 1 specific agonist SEW2871 instead of FTY720. In line with our previous data, SEW2871 also dose dependently inhibited IL-2 induced STAT-5 phosphorylation (Fig. 5,A). Of note, phosphorylation of STAT-5 induced by other cytokines of the common γ-chain family (i.e., IL-7 and IL-15) was not prevented by FTY720 (Fig. 5 B).

We subsequently tested the consequence of impaired Treg proliferation induced by the S1P receptor agonist FTY720 in adoptive Treg transfer models. First, we used an established model for aGvHD as an alloimmune model in which Treg are capable of reducing lethality (34). In vivo expansion of aGvHD-inducing T cells was monitored by detecting photons via bioluminescence imaging, as the donor effector T cells carry the luciferase transgene. The extent of the emitted light is known to correlate with T cell expansion and aGvHD severity. In line with previous data, cotransfer of solv-Treg almost completely prevented the expansion of luciferase transgenic, allogeneic Tconv (Fig. 6, A and B). In contrast, short-term exposure of Treg to FTY720 before injection markedly reduced their suppressive cues, which is reflected by a massive expansion of luciferase-positive effector T cells comparable to that seen in Tconv-transferred animals (Fig. 6, A and B). This reduced suppressor function also resulted in significantly reduced overall survival of the group that received Tconv along with FTY-Treg compared with the group receiving Tconv together with solv-Treg (Fig. 6,C; ∗, p = 0.001). Histopathology scoring revealed a significantly higher GvHD score in the group that received Tconv together with FTY-Treg as compared with the group receiving Tconv plus solv-Treg (Fig. 6 D; ∗, p < 0.05). Overall, these results demonstrate that pre-exposure of Treg to FTY720 impairs Treg function in vivo.

Second, anti-GBM nephritis was selected as an autoimmune model to test FTY-Treg in vivo. According to our previous data (25), transfer of solv-Treg into mice suffering from anti-GBM GN potently inhibited the development of renal inflammation as reflected by significantly reduced proteinuria (Fig. 7,A; ∗, p < 0.05). Histological evaluation revealed only occasional PAS-positive deposits, and the majority of the glomeruli displayed no pathological changes (Fig. 7,C). According to the data from the aGvHD model, ex vivo FTY720 treatment of Treg also abrogated their suppressive effect on nephritic inflammation as demonstrated by significantly elevated albumin/creatinine urine levels when compared with animals receiving solv-Treg (Fig. 7,A). Kidneys displayed classical histological signs of GN such as hypercellularity and focal deposition of PAS-positive material (Fig. 7,D), as well as enhanced infiltration of CD4⁺ T cells and macrophages (Fig. 7,B; ∗, p < 0.05), which was comparable to mice receiving CD4⁺CD25⁻ control T cells (Fig. 7 E).

To exclude the possibility that additional mechanisms contribute to the observed loss of Treg-mediated immunosuppression in vivo, we next focused on FTY720 effects on Treg migration upon adoptive transfer. Using steady state conditions as well as a well defined model of rapid local inflammation (i.e., footpad injection of mycobacteria together with Freund’s adjuvant), we i.v. coinjected TAMRA-labeled FTY-Treg together with equal numbers of CFSE-labeled solv-Treg (Fig. 8,A). Under steady state conditions, 6–12 h after transfer increased numbers of FTY-Treg were found in the spleen, but not in peripheral LN (Fig. 8,B). Notably, despite their loss of immunosuppression in vivo in the inflammatory models, inflamed LN occupancy of Treg is even increased by ex vivo exposure of Treg to FTY720 (Fig. 8 C; ∗, p < 0.05). No significant increase of FTY-Treg was detectable in the contralateral nondraining LN under inflammatory conditions.

Next, we analyzed the positioning of FTY720-exposed Treg in T cell areas of SLO upon adoptive transfer, because an altered localization might also interfere with proper Treg function in vivo. Therefore, we again coinjected differentially labeled FTY720- and solvent-exposed Treg i.v. into mice under steady state as well as inflammatory conditions. As shown in Fig. 8 D, TAMRA labeled FTY-Treg (red fluorescence) perfectly colocalized with CFSE-labeled solv-Treg (green fluorescence) in T cell areas of the spleen and LN after i.v. transfer into mice with local footpad inflammation. This finding was corroborated in healthy mice showing similar positioning patterns of FTY- and solv-Treg in SLO under steady state conditions (data not shown).

Finally, we were interested in determining whether short-term (i.e., 36 h) systemic FTY720 administration also affects Treg proliferative potential. In accordance with the defective proliferation of isolated Treg exposed to FTY720 in vitro, restimulation of Treg isolated from either solvent- or FTY720-exposed animals using anti-CD3 together with IL-2 resulted in a ∼40% reduced proliferation of Treg isolated from FTY720-treated animals (Fig. 9).

The sphingosine analog FTY720 is a promising immunomodulatory agent with proven therapeutic efficacy in various experimental models of autoimmunity (26, 27). The drug is currently tested as an immunosuppressive agent in various clinical trials and will soon be approved for treatment of relapsing multiple sclerosis (38). Immunosuppressive action of FTY720 is suggested to be primarily mediated via the trapping of lymphocytes in SLO, thereby preventing effector T cell exit into the blood stream and subsequent target tissue entry (29). Interestingly, immunosuppressive functions after continuous systemic application of FTY720 or exposure of the whole PBMC fraction to the drug have at least in part been attributed to the induction of Treg (30, 39, 40). However, the fact that FTY720 induces a plethora of mechanisms in various target cells (i.e., lymphatic endothelial, dendritic, T, and B cells) (31, 32) limits the interpretation of direct FTY720-induced effects on Treg. The latter have been shown to play a decisive role in the maintenance of tolerance and have been considered as cellular immunosuppressants for the treatment of exaggerated autoimmune and alloimmune responses (41, 42, 43).

We provide in this report for the first time evidence that the S1P receptor agonist FTY720 is a potent inhibitor of Treg proliferation in vitro and in vivo. Treg expand very efficiently when stimulated appropriately in vitro via the TCR together with IL-2 (14). Accordingly, clinical studies using IL-2 as immunostimulant described a preferential expansion of Treg (15), which can be explained by a low IL-2 receptor signaling threshold that supports Treg development and maintenance (44). Pre-exposure of isolated Treg to FTY720 in vitro almost completely prevented Treg proliferation induced by TCR stimulation and IL-2. Previous reports, which have already demonstrated that S1P inhibits the proliferative response of CD4⁺ T cells (45, 46, 47), support the concept of a T cell inhibitory effect of S1P receptor agonists in addition to their well characterized effects on lymphocyte migration (29). In line with our in vitro observation, we could further show that FTY720 exposure also profoundly impairs Treg expansion in vivo after adoptive transfer under inflammatory conditions (i.e., a model of local LN inflammation). Recent reports have already defined the critical role of IL-2 for Treg maintenance and expansion (44, 48, 49). On the molecular level, FTY720, as well as the S1P receptor 1-specific agonist SEW2871, severely impairs critical components of the IL-2 signaling pathway in Treg (50), as phosphorylation of STAT-5 and the subsequent IL-2-induced maintenance of FoxP3 expression are inhibited. Interestingly, this inhibitory effect appears to be specific for IL-2, as we were not able to detect any inhibitory effect of the compound on STAT-5 phosphorylation induced by other family members of the common γ-chain cytokines, such as IL-7 and IL-15. Inhibition of FoxP3 maintenance during the culture period is most likely due to the fact that STAT-5 binding sites are located in the promoter region of FoxP3 (37, 50), linking IL-2 signaling to the maintenance of FoxP3 expression. It is of particular importance that FTY720 does not simply reduce IL-2R expression on Treg, suggesting that the modulation of IL-2 signals is indeed mediated by the interference of FTY720 with the intracellular IL-2 signaling pathway. Moreover, exposure of Treg to the FTY720 doses we used for our in vitro and in vivo experiments did not affect their in vitro suppressive function (including anergy, suppression of target T cell proliferation, and cytokine production) and did not affect Treg survival in vitro. However, using two different experimental models of inflammation (i.e., aGvHD as an alloimmune model and anti-GBM GN as an autoimmune model), both of which can be inhibited by adoptive Treg transfer (25, 33, 51), we could clearly demonstrate that FTY-Treg loose their immunosuppressive potential in both models in vivo. This observation was particularly surprising to us, because systemic FTY720 was shown to be additive with the application of in vitro expanded Treg for GVHD inhibition (52), and we also provide evidence that FTY720 induces the efficient trapping of adoptively transferred FTY-Treg in SLO. The latter observation actually suggests that the increased LN occupancy of Treg (especially under inflammatory conditions) would tip the balance in favor of a more immunosuppressive milieu. Using a model of localized footpad inflammation, we show that the homing of FTY-Treg occurred preferentially in inflammatory LN and spleen. These data regarding the effects of FTY720 on Treg migration complement a recent report demonstrating that systemic application of FTY720 primarily affects non-Treg T cell populations (39). The authors described an increased number of Treg in blood and spleens but not in LN after systemic FTY720 application (39). In contrast, we did choose an experimental setup that allowed the evaluation of direct FTY720-induced effects on Treg, which revealed that FTY720 obviously modifies Treg migration in vivo. Due to the fact that a blockade of S1P receptors by FTY720 has been shown to destabilize LN architecture with the displacement of marginal zone B cells (32, 53), we next checked for the localization of FTY-Treg in SLO. However, cotransfer studies of differentially labeled FTY720- and solvent-treated Treg proved the proper occupancy of T cell areas in SLO. These findings are of particular interest, because despite the retention of Treg in inflammatory SLO, their in vivo immunosuppressive activity was lost. We know from the GN model that adoptively transferred Treg primarily enter SLO but not the end organ (25). In addition, it is the LN-homing, CD62L-expressing Treg fraction that is most potent in terms of aGvHD inhibition (19, 20), supporting the concept that Treg LN occupancy is critical for the immunosuppressive function of Treg (17). Thus, it is most likely the FTY720-induced inhibition of Treg proliferation that is responsible for the loss of Treg effects in vivo. Finally, we could further corroborate our previous findings by showing that FTY720 also exerts Treg inhibitory effects when applied in a systemic fashion. In line with our in vitro data, Treg isolated from animals exposed to FTY720 for a short term are hyporesponsive to IL-2 plus TCR-induced proliferation. Our data are in line with a very recent report showing that short-term systemic FTY720 treatment enhanced virus-specific CD4⁺ and CD8⁺ T cell responses in a chronic lymphocytic choriomeningitis virus infection model (54). This observation was primarily explained by improved T cell priming within SLO induced by transiently increased LN occupancy of T cells by FTY720, but the effect on Treg function in vivo was not addressed in this report. Our data suggest that concomitant inhibition of immunosuppressive Treg might also account for the observed phenomenon. This observation is supported by a very recent report presented by Liu and colleagues (55) demonstrating that S1P delivers via S1P receptor 1 an intrinsic negative signal activating the Akt/mammalian target of rapamycin pathway leading to inhibition of the thymic generation, peripheral maintenance, and suppressive activity of Treg.

In summary, we show for the first time that the S1P receptor agonist FTY720 induces Treg trapping in inflammatory SLO but concomitantly abrogates their in vivo immunosuppressive potential by inhibition of Treg expansion, which is due to the inhibition of IL-2 induced STAT-5 activation. Our data highlight the need of proper Treg proliferation after adoptive Treg transfer to induce an efficient immunosuppressive impact on effector T cells (i.e., achievement of an optimal Treg to autoreactive or alloreactive T cell ratio) and suggest that the S1P receptor agonism induced by pharmacological agents might represent a rational approach for Treg inhibition. Finally, our data suggest that the inhibition of Treg by S1P receptor agonists might represent an innovative adjuvant for cancer immunotherapy (e.g., in combination with dendritic cell vaccines).

FTY720 was generously provided by Dr. Volker Brinkmann of Novartis. We are indebted to Prof. Gottfried Baier for critically reading the manuscript and for helpful comments.

D.W. receives speaker honoraria from Novartis and is a Novartis Advisory Board member.