IL-7R signaling in regulatory T cells maintains peripheral and allograft tolerance in mice

Significance

In mammals, regulatory T cells establish and maintain immune responses toward self- and non–self-antigens by suppression of specific T cells. Several mechanisms of suppression were previously described including starvation for IL-2, a cytokine required for T-cell expansion. Here we show that regulatory T cells (T_reg) respond to IL-7, a T-cell survival factor produced by lymph node fibroblast reticular cells, by enhancing their IL-2 sensitivity, thereby limiting expansion of reactive T cells during inflammation. These findings highlight a previously unidentified and unexpected mechanism by which IL-7R signaling in T_reg cells maintains peripheral and allograft tolerance. This work may contribute to design new therapeutic approaches using lower IL-2 doses in humans by enhancing the functional competence of regulatory T cells with IL-7.

Abstract

Foxp3⁺CD4⁺ regulatory T cells (T_reg) have a crucial role in controlling CD4⁺ T-cell activation, proliferation, and effector function. However, the molecular mechanisms regulating T_reg function remain poorly understood. Here we assessed the role of IL-7, a key cytokine regulating T-cell homeostasis, in suppressor capacity of T_reg. Using a skin allograft model in which transplant acceptance is controlled by the number of transferred T_reg, we find that T_reg impair the proliferation of allogeneic CD4⁺ T cells, decrease production of IFNγ by effector T cells, and prevent early and increase late IL-7 induction by lymph node stromal cells. Increased IL-7 availability enhanced T_reg survival, stabilized T_reg molecular signature, enhanced surface IL-2Rα expression, and improved IL-2 binding of T_reg, which diminished proliferation of alloreactive CD4⁺ T cells. Sequestration of IL-7 or impairment of IL-7R signaling after allograft transplantation abolished T_reg-mediated tolerance by limiting their suppressive capacity. Aged Il7rα-ΔT_reg mice displayed mild symptoms of autoimmunity correlating with impaired expansion of effector T_reg in response to IL-2. Thus, IL-7R signaling on T_reg supports the functional activity of effector T_reg by increasing their IL-2 sensitivity in the lymph node during peripheral and allograft tolerance.

Regulatory T cells (T_reg) express the transcription factor Foxp3 and maintain peripheral tolerance through the suppression of potentially self-reactive T cells (1, 2). T_reg deficiency and mutations in FoxP3 result in the development of autoimmune diseases in both humans and mice. The molecular mechanisms underlying the function and stability of the T_reg compartment have not been fully elucidated (2–5). T_reg constitutively express IL-2Rα (6), in which IL-2 has been shown to regulate T_reg survival, expansion, and suppressive function under both homeostatic and activation conditions (7, 8). T_reg impaired in IL-2Rα signaling fail to suppress effector T-cell responses (7). Because T_reg themselves are unable to produce their own IL-2, due to FoxP3-dependent repression of the interleukin-2 (IL-2) gene (9), T_reg are susceptible to IL-2 concentrations and need to compete for it in secondary lymphoid organs (SLO) (10). Therefore, competition for IL-2 with conventional T cells is one of the proposed suppressive mechanism of T_reg (6).

Recently, it was shown that central T_reg (cT_reg), defined by CD44^low and CD62L^high expression, completely depend on IL-2 during homeostatic and inflammatory conditions, whereas effector T_reg (eT_reg), defined by CD44^high and CD62L^low expression, require inducible T-cell costimulator (ICOS) signaling (11). Because T_reg not only use IL-2 for expansion and proliferation, but also for survival and pool maintenance, it is unclear if IL-2 competition occurs at low IL-2 concentration (as suggested in ref. 12). Moreover, IL-2 signaling via STAT5 (13) stabilizes FOXP3 expression and increases expression of suppressive T_reg molecules including ICOS, glucocorticoid-induced TNF receptor-related protein (GITR), and cytotoxic T lymphocyte-associated protein 4 (CTLA4) (14–16).

IL-7, a member of the γc-cytokine family, is constitutively produced by stromal cells in SLO and can be further induced in response to various inflammatory mediators (17, 18). Production of IL-7 by lymph node (LN) stromal cells (LNSC) is important for naïve T-cell survival and activation during adaptive immune responses (19, 20). Interaction between IL-7 and the IL-7Rα/γc chain activates STAT5 phosphorylation and induces the expression of antiapoptotic molecules, such as Bcl-2, which play a key role in T-cell development and proliferation (reviewed in ref. 21). Alternatively, continuous IL-7 signaling in naïve T cells can lead to activation and cell death when T-cell receptor (TCR) signaling is insufficient (22). The role of IL-7 in controlling Foxp3⁺ T_reg is less clear. The injection of soluble IL-7 complexes has been shown to increase the number of T_reg (23, 24) and optimizes their reactivity to IL-2 (25). IL-7Rα^−/− mice have a reduced T_reg population (26), and IL-7R–deficient T_reg demonstrate impaired proliferation (27). Although these studies imply an important role for IL-7 in the maintenance and activation of T_reg, the role of IL-7 during inflammatory T-cell responses in the periphery is still poorly understood.

To study the role of IL-7 and IL-7R signaling in T_reg homeostasis and function, we generated a skin allograft tolerance model in which specific graft rejection is prevented by cotransfer of T_reg. Our data indicate that T_reg inhibited strong allogeneic CD4⁺ T-cell responses, thereby ensuring their maintenance within the LN. T_reg responded to IL-7 by enhancing CD25 expression and their sensitivity to IL-2, thereby limiting IL-2 availability for effector T-cell proliferation required for graft rejection. Moreover, IL-7R deficiency had only mild effects on the survival and homeostasis of T_reg under steady-state conditions. A subset of T_reg was dependent on IL-7R signaling in the LN and increased in response to IL-2. IL-7R signaling in T_reg was required to prevent autoimmunity and allograft tolerance in mice. Our findings indicate that IL-7R–sufficient T_reg require stromal cell-derived IL-7 in the LN to maintain their phenotype and function.

Results

T_reg Suppress Rejection of I-A^bm12 Skin Allografts.

To assess whether T_reg can suppress allogeneic CD4⁺ T-cell responses in vivo, we used a newly established murine transplantation model. I-A^bm12 mismatch skin grafts from Rag2^−/− mice are accepted by Rag2^−/− mice, whereas transfer of I-A^bm12–specific transgenic CD4⁺ T cells (ABM) leads to rejection (28). We first confirmed our previous finding that 2 × 10⁴ ABM reject I-A^bm12 skin grafts from transplanted Rag2^−/− mice within 12–14 days (Fig. 1A) (28). Furthermore, coinjection of twice the number of polyclonal GFP-expressing T_reg only delayed graft rejection (2:1 group), whereas 10 times more T_reg induced stable graft tolerance up to 100 days (Fig. 1A).

Fig. 1.

T_reg establish and maintain skin allograft tolerance. I-A^bm12 tail skin transplanted Rag2^−/− mice adoptively transferred with monoclonal transgenic anti–I-A^bm12 CD4⁺ T cells (ABM) alone or at different ratios of polyclonal Foxp3⁺ T_reg:ABM. (A) Percentage of graft survival of transplanted mice after adoptive T-cell transfer [n = 4–6, log-rank (Mantel–Cox) test: P < 0.0001]. (B) Counts of recovered ABM (Left) and Foxp3⁺ T_reg (Right) in the different groups (n = 10, SEM, t test). (C) Ratio of recovered ABM and Foxp3⁺ T_reg (n = 10, SEM, t test) on day 9 after adoptive T-cell transfer in the LN. (D) Frequency of IFNγ⁺ ABM isolated from transplanted mice on day 9 after adoptive T-cell transfer and restimulated with PMA-Ionomycin (n = 4, SEM, t test).

To assess whether T_reg are required to alter initial establishment of tolerance, we used T_reg from Foxp3-DTR mice to allow their depletion following diphtheria toxin (DT) treatment. Indeed, we found that the absence of T_reg resulted in rejection of I-A^bm12 allografts, indicating a T_reg-dependent tolerance (Fig. S1A). We further characterized the migration kinetics of transferred ABM and T_reg in LN and skin grafts. Both ABM and T_reg were detected in the draining lymph nodes (dLN) on day 3, and their numbers progressively increased over time. On day 9, the number of ABM recovered from the dLN was lower with higher numbers of transferred T_reg (Fig. 1B, Left), whereas the number of T_reg was comparable (Fig. 1B, Right). Interestingly, a ratio of T_reg to ABM higher than 1 in the dLN correlated with long-term tolerance of skin allografts (Fig. 1C). In contrast, in the skin grafts, higher numbers of ABM compared with T_reg were found in the 2:1 and the 10:1 group after T-cell transfer on day 9 (Fig. S1B).

Fig. S1.

T_reg cells do not inhibit ABM cell migration to the graft and activation in the LN. I-A^bm12 tail skin transplanted Rag2^−/− mice adoptively transferred with monoclonal transgenic anti I-A^bm12 CD4⁺ T cells (ABM) alone or at different ratios of polyclonal Foxp3⁺ T_reg:ABM. (A) Percentage of graft survival of transplanted mice treated (i.p.) with 0.4 mg/mouse diphtheria toxin or PBS on days 2 and 4 after adoptive T-cell transfer [n = 6, log-rank (Mantel–Cox) test: P < 0.0001]. (B) Counts of recovered ABM and T_reg cells from skin allograft of transplanted mice on day 9 after adoptive transfer. (C) Surface expression of CD25, CD44, CD62L, and CD69 and (D) proliferation of ABM, determined by eFluor670 proliferation dye staining, on day 9 after adoptive T-cell transfer in the LN (n = 4–6, SEM, t test). The filled gray curve represents the fluorescence of undivided cells at the day of adoptive transfer (day 0). (E) Counts of IFNγ⁺ ABM isolated from transplanted mice on day 9 after adoptive T-cell transfer and restimulated with PMA-Ionomycin (n = 4, SEM, t test). Data pooled from two independent experiments, and each symbol represents an individual mouse (B, C, and E; t test).

It is known that I-A^bm12 skin graft rejection depends on activation and IFNγ production of allogeneic CD4⁺ T cells (28, 29). To assess whether T_reg are able to suppress allogeneic T-cell responses, we analyzed activation, proliferation, and IFNγ production of ABM in the presence of T_reg in I-A^bm12–transplanted Rag2^−/− mice on day 9. Surface expression of CD25, CD44, and CD62L on ABM was not influenced by the number of transferred T_reg, whereas CD69 did (Fig. S1C). Proliferation of ABM was impaired depending on the number of cotransferred T_reg (Fig. S1D). Moreover, IFNγ-producing ABM were lower in frequency and number in the 10:1 group than in the ABM group (Fig. 1D and Fig. S1E). Taken together, T_reg establish allograft tolerance after transplantation by suppressing proliferation and IFNγ production of allogenic CD4⁺ T cells.

T_reg Affect the IL-7 Production of LN-FRC.

Activation of ovalbumin (OVA)-specific CD8⁺ T cells has been shown to modulate the transcription of several molecules in LN fibroblastic reticular cells (LN-FRC) (18). We first assessed the effect of allogeneic T-cell activation in the transcription profile of LN-FRC after transplantation of I-A^bm12 allo- and I-A^b syngrafts on C57BL/6 and IFNγ^−/− mice. ICAM-1, VCAM-1, and IL-7 were increased after allotransplantation of naïve C57BL/6 mice, whereas transcription of ICAM-1, VCAM-1, IL-7, thymic stromal lymphopoietin (TSLP), and programmed death-ligand 1 (PD-L1) were increased in LN-FRC from allografted compared with syngrafted C57BL/6 mice (Fig. S2A) but not in IFNγ^−/− mice (Fig. S2B). Moreover, IL-7 protein was increased in LN lysates from allograft compared with syngraft transplanted and naïve C57BL/6 mice (Fig. 2A). IL-7 transcripts in LN-FRC were higher after treatment of C57BL/6 mice with IFNγ compared with saline (Fig. 2B). These results showed that allogeneic CD4⁺ T-cell activation and release of IFNγ induce IL-7 expression in LNSC. Indeed, ABM transfer induced early IL-7 production in LN-FRC on day 3 (Fig. 2C). Subsequently, we examined the ability of T_reg to modulate the kinetics of LN-FRC–derived IL-7 using our transplantation model. T_reg in the tolerant 10:1 group prevented early up-regulation of IL-7 transcripts on day 3 (Fig. 2C). Unexpectedly, the IL-7 transcripts were higher in the tolerant 10:1 group than in the ABM group on day 9 (Fig. 2 C and D). These results suggest that the presence of T_reg affects IL-7 expression in stromal cells after transplantation.

Fig. S2.

LN-FRC respond to IFNγ in allografts. (A) C57BL/6 mice and (B) IFNγ^−/− mice were allograft transplanted (black bars) or syngraft transplanted (gray bars) and naive (white bars) mice, and RT-PCR was performed for relative mRNA expression of ICAM-1, VCAM-1, I-A^b, IL-7, TSLP, and PD-L1 in LN-FRC isolated on day 9 after transplantation. Data pooled from two independent experiments with n = 6 mice (SEM, Mann–Whitney test).

Fig. 2.

T_reg affect IL-7 production in the LN. (A) IL-7 protein levels in dLN lysates of I-A^bm12 (allograft, n = 5) and I-A^b (syngraft, n = 3) skin transplanted and nontransplanted (dotted line, n = 4) C57BL/6 mice at day 7, measured by ELISA (SEM, t test). (B) Relative expression of IL-7 mRNA in LN-FRC isolated from saline-treated mice (n = 9) and IFNγ-treated C57BL/6 mice (10,000 U/mouse, n = 5) 12 h after treatment (SEM, t test). (C) Relative expression of IL-7 mRNA in LN-FRC isolated from naïve mice (dotted line, n = 4) and on days 3, 6, and 9 after adoptive transfer of mice injected with ABM alone (n = 4) or coinjected with 10 times more Foxp3⁺ T_reg (10:1, n = 4) and (D) their relative expression of IL-7 mRNA on day 9 (n = 6 ABM, n = 4 10:1, SEM, t test).

IL-7 Signaling Stabilizes the Suppressive T_reg Phenotype.

IL-7 stabilizes T_reg numbers in mice (30, 31) and increases T-cell survival by reducing apoptosis in a concentration-dependent manner in vitro (Fig. S3A). We next studied the molecular basis of IL-7R signaling on naïve and CD3/28 activated T_reg at varying IL-7 concentrations. IL-7R down-regulation was observed at increasing IL-7 concentrations with or without TCR stimulation (EC₅₀ values) (Fig. S3B). In addition, surface expression of activation markers, including CD25 and GITR, increased dependent of IL-7 concentrations with or without TCR signaling (EC₅₀ values) (Fig. S3 C and D). Our data indicate that increasing IL-7 levels influence the phenotype of T_reg in vitro. To confirm these results in vivo, we increased IL-7 availability in mice by injecting IL-7 complex (IL-7C). Only a high dose of IL-7C significantly increased T_reg number, induced IL-7R down-regulation, and increased surface CD25 expression (Fig. 3A), indicating that high IL-7 levels induced IL-7R signaling and activation of T_reg in vivo.

Fig. S3.

IL-7 signaling stabilizes the suppressive T_reg phenotype. (A) Frequency of apoptotic (●, n = 6, SEM) and counts of surviving Foxp3⁺ T_reg (○, n = 3, SEM) stimulated with varying IL-7 concentrations for 24 h in vitro. Surface expression of (B) IL-7R, (C) CD25, and (D) GITR on Foxp3⁺ T_reg from Foxp3-eGFP mice stimulated with varying IL-7 concentrations (A) for 24 h in vitro with or without CD3/CD28 beads. EC₅₀ of each group is reported (n = 3, SEM). Data are representative of two and three experiments (A–D).

Fig. 3.

IL-7 signaling stabilizes the suppressive T_reg phenotype. (A) Cell counts, IL-7R, and CD25 expression of Foxp3⁺ T_reg isolated from a pool of brachial, axillary, and inguinal LN of Foxp3-eGFP mice 24 h after a single injection of IL-7 complex (IL-7C) (n = 3–4, SEM, t test). Data are representative of three experiments. (B) Percentage of graft survival for I-A^bm12 tail skin transplanted Rag2^−/− mice adoptively transferred with ABM alone or at different Foxp3⁺ T_reg:ABM ratios, nontreated and treated (i.p.) at alternate days with 12.5 mg/mouse anti–IL-7 (M25) and IgG2b antibody starting at day 6 after adoptive transfer [n = 5–6, log-rank (Mantel–Cox) test: P < 0.0017]. (C) Ratio of Foxp3⁺ T_reg:ABM in M25-treated mice on day 9 after adoptive T-cell transfer (n = 6, SEM). Each symbol represents an individual mouse from at least two experiments.

We next examined the effect of IL-7 signaling on the immunosuppressive function of T_reg in allograft transplantation. The administration of the anti–IL-7 blocking antibody M25 in the 10:1 group reduced IL-7 availability, which resulted in allograft rejection in 40% of the transplanted mice (Fig. 3B). The ratio of T_reg to ABM was higher than 1 in five of six mice in the dLN (Fig. 3C), thereby excluding graft rejection as a result of reduced T_reg survival. These results suggested that the suppressive capacity of T_reg depends on IL-7 availability in the dLN.

IL-7 Sensitizes T_reg for IL-2 to Suppress T_eff Proliferation.

Our data thus far suggested that T_reg respond to IL-7 signaling by increasing their suppressive phenotype and the expression of surface markers. T_reg from IL-7C–treated mice were increased in expression of CD25 (Figs. 3A and 4A) and of CD122, GITR, and ICOS but not CD132 and KI67 (Fig. 4A and Fig. S4A). It is well known that T_reg require IL-2 for their suppressive function (32). To assess the IL-2 sensitivity of IL-7–exposed T_reg in vivo, LN and spleens were analyzed from IL-2C–injected mice pretreated with IL-7C or saline. T_reg from IL-7C–pretreated mice up-regulated the expression of CD132, GITR, ICOS, and Ki67, but not CD25 and CD122 and phosphorylation of STAT5 in response to the IL-2C compared with saline-pretreated mice (Fig. 4A and Fig. S4 A and B). To exclude possible contributions of other cells responding to IL-7 and IL-2, we stimulated IL-7– or saline-pretreated T_reg with IL-2 in the presence or absence of TCR stimulation in vitro (Fig. S4C). IL-7–pretreated T_reg responded to IL-2 by higher CD25 expression, which was further up-regulated by increasing concentrations of IL-2 (Fig. S4C). Although CD25 expression was even higher in TCR stimulated T_reg, IL-7 responses were comparable to those without TCR stimulation, as indicated by similar EC₅₀ values (Fig. S4C). Interestingly, IL-7–pretreated T_reg had more pSTAT5 (Fig. S4D) and had increased binding of IL-2 with or without TCR stimulation (Fig. 4B). These data strongly suggest that IL-7R signaling has an impact in IL-2 sensitivity of T_reg in vitro. The TCR-stimulated proliferation of CD3/CD28-activated ABM was reduced by 2 (2:1 group), whereas the addition of 10 times (10:1 group) more T_reg and exogenous IL-7 further increased T_reg-mediated suppression of ABM in the 2:1 group, but not in the presence of a high number of T_reg in the 10:1 group (Fig. 4C). Moreover, IL-7–pretreated T_reg were able to suppress proliferation of I-A^bm12–activated ABM more efficiently than nonpretreated T_reg (Fig. 4D). These data indicate that IL-7 increases IL-2 sensitivity of T_reg, thereby increasing their competence to suppress allogeneic T-cell proliferation.

Fig. 4.

T_reg exposed to IL-7 enhance their capacity to bind IL-2 and suppress CD4⁺ T-cell proliferation. (A) IL-2C– and saline-induced expression of CD25, GITR, ICOS, and KI67 on T_reg isolated from spleens of Foxp3-eGFP mice 24 h after treatment with IL-7C or saline for 24 h in vivo (n = 2–5, SEM, t test). (B) IL-2 binding by medium and IL-7–stimulated Foxp3⁺ T_reg in the absence or presence of CD3/CD28 bead stimulation. (C) ABM proliferation stimulated with anti-CD3/CD28 beads for 5 d determined by eFluor670 proliferation dye staining in presence of Foxp3⁺ T_reg in a 2:1 or 10:1 (Foxp3⁺ T_reg: ABM) ratio with or without 50 ng/mL IL-7 for 5 d (n = 4, SEM, t test). (D) Percent of proliferating ABM in the presence of two times more Foxp3⁺ T_reg cocultured with irradiated I-A^bm12 splenocytes in the presence or absence of 10 ng/mL mrIL-7 for 3 d (n = 9, SEM, t test). Data are representative of two and three experiments (A–D), and each symbol represents an individual mouse from at least two experiments (A).

Fig. S4.

T_reg exposed to IL-7 enhance CD25 expression and signaling. (A) IL-2C– and saline-induced expression of CD122, CD132, and (B) pSTAT5 on T_reg isolated from spleens of Foxp3-eGFP mice 24 h after treatment with IL-7C or saline for 24 h in vivo (n = 2–5, SEM, t test). (C) CD25 expression on Foxp3⁺ T_reg from Foxp3-eGFP mice pretreated with 10 ng/mL IL-7 (squares) or medium (circles) in the absence (filled) or presence (empty) of CD3/CD28 beads after incubation with increasing IL-2 concentrations for 24 h in vitro. EC₅₀ values are shown (n = 3, SEM). (D) STAT5 phosphorylation of medium and IL-7–preincubated Foxp3⁺ T_reg in response to increasing IL-2 concentrations (see C). EC₅₀ values are shown (n = 3, SEM). Data are representative of two and three experiments (A–D), and each symbol represents an individual mouse from at least two experiments.

IL-7R Signaling Is Required for T_reg Function.

To study the function of IL-7R signaling in T_reg, we generated T_reg-conditional IL-7R KO by crossing Il-7a–floxed mice with Foxp3-YFP-Cre mice. T_reg from Il7ra-∆T_reg mice did not express IL-7R and did not phosphorylate STAT5 in response to IL-7 (Fig. S5 A and B). Conventional CD4⁺ and CD8⁺ T cells (T_con cells) were present in LN of Il7ra-∆T_reg mice and did not differ in frequency or number compared with Foxp3-YFP-Cre mice (Fig. S5C). Moreover, T_reg number and frequency were similar in skin dLN of Il7ra-∆T_reg mice compared with their control littermate (Fig. S5C). To further investigate survival in competition, we cotransferred Il7ra-∆T_reg and Foxp3-YFP-Cre T_reg at a ratio of 1:1 and 3:1 in Rag2^−/− mice (Fig. 5A). Four weeks after T_reg transfer, the number of Il7ra-∆T_reg was 2.5-fold (1:1) and 5-fold (3:1) lower compared with Foxp3-YFP-Cre T_reg, and the initial ratio of 1:1 and 3:1 was not maintained in LN (Fig. 5A). Collectively, these data suggest that homeostasis and in vivo function of T_reg partially depend on IL-7R signaling.

Fig. S5.

Characterization of Il7ra-ΔT_reg mice. (A) IL-7R was deleted in Foxp3⁺ T_reg by crossing Foxp3-YFP-Cre mice with Il7ra^flox/flox mice; representative genotyping for Il7ra^fl/fl, Il7ra^Δ/Δ, and Il7ra^+/+ WT band in YFP⁺ and YFP⁻ CD4⁺ T cells. (B) Representative flow cytometry of phosphorylated STAT5 in Foxp3⁺ T_reg sorted from LN of Foxp3-YFP-Cre and Il7ra-ΔT_reg mice after stimulation with 10 ng/mL rmIL-7 for 5 min. (C) Summary of the frequency and counts of T cells in skin draining LN of Foxp3-YFP-Cre (filled) and Il7ra-ΔT_reg mice (open) (n = 8, SEM, t test). Each symbol represents an individual mouse from at least two experiments.

Fig. 5.

IL-7R increases suppressive capacity of T_reg cells. (A) Ratio of recovered Foxp3⁺ T_reg from LN of Il7ra-ΔT_reg and Foxp3-YFP-Cre mice after adoptive transfer in Rag2^−/− mice on day 30 in different ratios (n = 6, SEM, t test). (B) I-A^bm12 tail skin transplanted Rag2^−/− mice adoptively transferred with monoclonal transgenic anti–I-A^bm12 CD4⁺ T cells (ABM, n = 4) alone and different ratios of polyclonal Foxp3⁺ cells:ABM. Percentage of graft survival for transplanted mice after adoptive transfer of ABM and Foxp3⁺ T_reg from Foxp3-Cre and IL7ra-ΔT_reg mice [n = 6, log-rank (Mantel–Cox) test: P < 0.0001]. (C) In vitro suppression of ABM cells stimulated with I-A^bm12 splenocytes in the presence of T_reg from FoxP3-YFP-Cre and Il7ra-ΔT_reg mice (2:1 Treg:ABM) and exogenous IL-7 (n = 7, SEM, t test). Data are representative of three experiments.

To assess the importance of IL-7R signaling in the function of T_reg, we purified T_reg from LN of Foxp3-YFP-Cre and Il7ra-∆T_reg mice and cotransferred them in I-A^bm12–transplanted mice. IL-7R–deficient T_reg in a 10:1 ratio were less efficient to suppress I-A^bm12 skin graft rejection in allografted Rag2^−/− mice (Fig. 5B). In addition, IL-7R–deficient T_reg were less efficient than Foxp3-eGFP T_reg to suppress proliferation of naïve I-A^bm12–activated ABM in vitro (Fig. 5C), indicating that IL-7R signaling is important for the suppressive capacity of T_reg in vitro and to maintain allograft tolerance in LN of transplanted mice.

eT_reg Require IL-7R Expression to Maintains Their Function.

T_reg are a heterogeneous population that shows a high degree of phenotypic and functional specialization defined as cT_reg (CD44^loCD62L^hi) and eT_reg (CD44^hiCD62L^lo). To understand whether IL-7R signaling facilitates differentiation of eT_reg in mice, we analyzed T_reg subsets both before and after adoptive T-cell transfer in I-A^bm12 transplanted mice. The frequency of cT_reg and eT_reg before transfer were comparable (Fig. S6A, Left); frequency of eT_reg was higher in the dLN of the 10:1 group (Fig. S6A) and increasingly expanded as assessed by eFluor670 dilution on day 9 (Fig. S6B). We confirmed that eT_reg have lower CD25 expression than cT_reg (Fig. 6A, Left) (11) and found higher IL-7R expression in eT_reg in Foxp3-eGFP mice (Fig. 6A, Right). Collectively, these data suggest that effector T_reg accumulate during allo-transplantation.

Fig. S6.

IL-7R⁺ T_reg maintain immunological tolerance in mice. (A) Representative distribution of cT_reg and eT_reg among LN Foxp3⁺ T_reg before (Left) and on day 9 after adoptive transfer of Foxp3⁺ T_reg:ABM (10:1) in I-A^bm12–transplanted mice (Right). (B) Proliferation of cT_reg (gray line) and eT_reg (black line) on day 9 after adoptive transfer of Foxp3⁺ T_reg:ABM (10:1) in I-A^bm12–transplanted Rag2^−/− mice. eFlour670 staining of T_reg before adoptive transfer (light gray line, filled). (C) H&E staining of lung tissue from 19- to 21-wk-old Il7ra-ΔT_reg mice. Data are representative of at least two experiments.

Fig. 6.

IL-7R signaling maintains suppressive function of eT_reg. (A) Flow cytometry of cT_reg (CD44^loCD62L^hi) and eT_reg (CD44^hiCD62L^lo) among CD4⁺Foxp3⁺ T_reg from spleens of Foxp3-eGFP mice assessing the expression of CD25 (Left) and IL-7R (Right) (n = 5, SEM, Mann–Whitney). (B) Expression of CD25 on eT_reg isolated from spleens of Foxp3-eGFP mice after 24 h of IL-7 complex (IL-7 C) injection (n = 3–4, SEM, t test). (C) Summary of frequency of cT_reg and eT_reg from Foxp3-Cre and Il7ra-ΔT_reg mice treated with saline and IL-2 complex (IL-2C) for 24 h (n = 6–7, SEM, Mann–Whitney). (D) Frequency of eT_reg from LN of 19–21 wk old Il7ra-ΔT_reg mice (n = 5, SEM, Mann–Whitney). (E) H&E staining of liver tissue from 19- to 21-wk-old Il7ra-ΔT_reg mice. Data are representative of at least two experiments.

To examine whether eT_reg respond to increased IL-7 levels, we injected IL-7C in Foxp3-eGFP mice and analyzed expression of CD25. Indeed, IL-7C treatment increased CD25 expression in eT_reg (Fig. 6B). We next assessed eT_reg expansion in response to IL-2 by injection of IL-2C and BrdU in Foxp3-YFP-Cre and Il7ra-∆T_reg mice. Interestingly the frequency of eT_reg was specifically increased in Foxp3-YFP-Cre but not in Il7ra-∆T_reg mice (Fig. 6C). These data indicate that increased IL-7 levels enhance IL-2 sensitivity and thereby the specific proliferation and maintenance of highly suppressive eT_reg subsets.

Finally, we analyzed whether the IL-7R–specific signaling in T_reg has an impact on the eT_reg subset under homeostatic conditions in aged mice. In aged mice, the frequency of eT_reg was higher in Foxp3-YFP-Cre than Il7ra-∆T_reg mice (Fig. 6D). The lower number of eT_reg was accompanied by mild symptoms of autoimmunity by induction of follicles in the liver and lung of Il7ra-∆T_reg, but not Foxp3-YFP-Cre mice (Fig. 6E and Fig. S6C). These data indicate that IL-7R signaling in T_reg is required to maintain immunological tolerance during aging in mice.

Taken together, we show that IL-7 levels increase the suppressive capacity of T_reg, both during transplantation and aging, via up-regulation of CD25. As a result, T_reg IL-2 sensitivity is increased for successful IL-2 competition with CD25^hi T cells, thereby maintaining T-cell tolerance.

Discussion

The population of Foxp3⁺ T cells consists of functionally different T_reg subsets that regulate different immune responses (2). Although many studies investigated the function of T_reg, the mechanisms controlling homeostasis and activation of different T_reg subsets are poorly defined. Using the newly established allograft tolerance model, we identified that IL-7 signaling, beside its known role in survival, stabilizes the function of T_reg in the LN during inflammatory T-cell responses. T_reg prevented early and increased late IL-7 induction in LN-FRC. Our data suggest that T_reg require IL-7R signaling to increase their CD25 expression and efficiently compete with other T-cell subsets for IL-2. Improved IL-2 sensitivity stabilized the T_reg phenotype and promoted expansion of eT_reg in vivo. The specific ablation of IL-7R specifically on T_reg reduced their suppressive function, causing allograft rejection and mild symptoms of autoimmunity during aging.

In various allograft models including skin allograft transplantation, T_reg were shown to control immune responses to foreign antigens (33). We established a transplantation model in which I-A^bm12 skin allografts are accepted on Rag2^−/− mice, and the function of polyclonal T_reg and their specialized subsets can be investigated after cotransfer with I-A^bm12–specific CD4⁺ T cells. Transfer of high T_reg numbers established allograft tolerance, which could be reverted by specific T_reg depletion. T_reg-suppressed ABM expansion dependent on the initial ratio and IL-2 production of ABM increased their own and T_reg numbers. T_reg displayed an effector phenotype in the LN. Blockade of IL-7 or transfer of T_reg from Il7ra-ΔT_reg mice caused graft rejection in 40% of tolerant mice, which could be due to reduced survival in the LN. However, the ratio of T_reg/ABM in rejecting mice was in favor of T_reg, which usually correlated with allograft tolerance. Therefore, our results suggest that IL-7R signaling does not exclusively support T_reg survival but also increases the suppressive activity of allo-activated T_reg.

Structural LN-FRC and lymphatic endothelial cells (LEC) are the main source of IL-7 in the LN (19, 20, 34) and increase IL-7 production on inflammatory T-cell responses (18). Consistent with this view, allogeneic CD4⁺ T-cell activation and release of IFNγ increased IL-7 levels in LN-FRC at early time points, which massively declined thereafter in dLN of transplanted Rag2^−/− mice. T_reg inhibited both IFNγ production and expansion of ABM, which correlated with an early reduction of IL-7 transcription in LN-FRC of allograft-tolerant mice. However, at later time points, the presence of T_reg resulted in higher IL-7 transcription. The reasons for this are unclear, and further experiments need to be done to address this. However, T_reg can influence the LN microenvironment to facilitate suppression of T cells, which would allow IL-2 consumption and suppression in close proximity of activated T cells (35), as well as help to maintain immunological tolerance (36).

Recent work has demonstrated that homeostasis and expansion of T_reg depend on the presence of γc-chain cytokines including IL-7 and IL-2 (21). The development of autoimmunity in IL-2– and IL-2R–deficient mice (37), the impairment of LN organogenesis in IL-7– and IL-7R–deficient mice (38), and the dependence of various adaptive immune cells on these survival factors complicate interpretation of the results. Mice deficient in IL-7 signaling have been shown to have low numbers of T_reg, and specific ablation of γc chain signaling in T_reg suggests IL-7 is a critical requirement for survival (24, 31) by inhibiting FOXP3-driven apoptosis (30, 39). These observations were confirmed by our study by showing that IL-7–mediated signals increase the absolute numbers of T_reg in vitro and in vivo after IL-7C treatment in mice, possibly by Bcl2- and Mcl1-dependent mechanisms (40). By cotransferring IL-7R-KO with WT-T_reg, we further demonstrate that ablation of IL-7R expression diminishes T_reg number in the LN after cotransfer with WT-T_reg, indicating that IL-7R signaling is required on T_reg to compete for sufficient IL-7 signals with other responding cells in the periphery (39).

T_reg transmitted IL-7 signals via STAT5 phosphorylation, which led to increased expression of CD25, GITR, and partially ICOS in a dose-dependent manner. These surface markers are part of the genes directing the expression of the T_reg molecular signature (4, 7). Our findings suggest that IL-7 signaling might optimize IL-2 acquisition (25) by increasing CD25 levels on T_reg. CD25 expression in untreated and IL-7–pretreated T_reg was not increased by different IL-2 doses excluding a feedback loop, which further enhances IL-2 binding. Similarly, IL-7C pretreatment increased γc chain expression, together with GITR and ICOS, in response to IL-2C. Thus, the greater IL-2 sensitivity of IL-7–exposed T_reg was due to higher CD25 expression and IL-2R signaling specificity via the STAT5 pathway, which might be required to compete for low IL-2 concentrations in vivo. IL-7 and IL-7R signaling stabilized the functional competence of T_reg populations that was further increased by IL-2 (41). Whether IL-7R signaling and increased IL-2 sensitivity in T_reg cells enable immunological tolerance against memory T cells was not addressed here, but IL-7R expression maintains survival of memory T cells and influences their activation (42), which might lead to resistance of allogeneic memory T cells to T_reg-mediated suppression (43). In addition, an effect of IL-7 on human T_reg should be carefully evaluated because CD127^low T_reg have been described to suppress T-cell responses (44). Therefore, further studies are required to use IL-7 for improved low-dose IL-2 therapy in autoimmunity and inflammatory diseases including allograft transplantations (12).

CD44^loCD62L^hi cT_reg require IL-2 signaling to maintain homeostasis, and CD44^hiCD62L^lo eT_reg depend on ICOS signals (11). cT_reg were higher in CD25 than eT_reg, but our results suggest that the increased expression of IL-7R might compensate for reduced IL-2 availability. Indeed, eT_reg up-regulate their CD25 expression in response to increasing IL-7 levels, thereby facilitating IL-2 access required for proliferation and efficient suppression of T-cell responses. This finding suggests that IL-7R⁺ eT_reg become more sensitive to IL-2. eT_reg were found in LN during allograft responses, indicating that the suppressor function of T_reg is mediated by this particular subset during inflammation. eT_reg develop from T_reg after TCR engagement in the presence of inflammatory cytokines (45). It remains to be tested whether IL-7R signaling promotes eT_reg differentiation from cT_reg or whether eT_reg accumulate in the dLN by proliferation or migration from nonlymphoid tissue.

On the basis of the presented results, IL-7R signaling in eT_reg may change the balance of cT_reg and eT_reg during inflammation due to altered IL-2 sensitivity. cT_reg strictly depend on IL-2 signaling in the LN (11), and improved IL-2 sensitivity of eT_reg might favor their expansion. Indeed, a limitation of IL-2 during aging has been shown to increase CD25^lo T_reg in mice (46). Due to impaired expansion of eT_reg in the absence of IL-7R signaling, Il7ra-ΔT_reg mice displayed mild symptoms of autoimmunity by formation of follicles in the liver and lung.

In conclusion, our work shows that, within LN, T_reg profit from the IL-7 expression of stromal cells by increased survival and IL-2 sensitivity, which is required for sufficient suppression of effector CD4⁺ T-cell responses in allograft tolerance and aging. These findings demonstrate a previously unidentified role for IL-7 in control of the T_reg compartment and highlight important links between T_reg and stromal cells within lymphoid tissues.

Materials and Methods

Mice.

All animal work was approved by the Cantonal Veterinary Office of the City Basel (SI Materials and Methods).

Preparation of T Cells and LN Cells.

Naïve transgenic Vα2Vβ8 CD4⁺ T cells were isolated from LN of ABM mice. T_reg were sorted from Foxp3-eGFP mice. LNSC were isolated according to the protocol (47) (SI Materials and Methods).

Skin Transplantation Model.

Recipient mice were transplanted on the back with full-thickness tail skin from allogenic I-A^bm12 or syngenic I-A^b Rag2^−/− mice under general anesthesia (28) (SI Materials and Methods).

In Vitro T_reg Assays.

For stimulation, T_reg were incubated in RPMI medium containing rmIL-7 or rmIL-2 with or without CD3/CD28 activation beads (Invitrogen) or on splenocytes (SI Materials and Methods).

Statistical Analyses.

Statistical analyses of data were performed using Prism (GraphPad Software) using the unpaired two-tailed Student t test, Mann–Whitney t test, and log-rank (Mantel–Cox) test.

For IL-2 binding assays, flow cytometry, and histology details, please see SI Materials and Methods.

SI Materials and Methods

Mice.

All animals were bred in the Animal Facility of the Department of Biomedicine, University Hospital of Basel, according to the regulations of Swiss veterinary law. WT inbred C57BL/6, Rag2^−/−, IFNγ^−/−, Bm12 Rag2^−/−, and TCR tg mice ABM Rag2^−/− reactive to MHCII I-A^bm12, Foxp3-eGFP, Foxp3-YFP-Cre, Foxp3-DTR, IL-7Rα^fl/fl mice (31), and conditional KO mice with targeted deletion of IL-7Rα in Foxp3⁺ cells were generated by mating IL-7Ra^fl/fl mice with Foxp3-YFP-Cre. Unless stated otherwise, all figures are representative of experiments with healthy 6- to 10-wk-old mice.

Generation of IL-7Rα Deficiency to Foxp3⁺ Cells.

Specific deletion of IL-7Rα in Foxp3⁺ cells at the protein level was assessed by flow cytometry. Recombination of Il7ra at the genomic DNA level was performed on a highly pure sorted cell population to specifically detect Il7ra^fl and deleted Il7ra: IL-7R forward, 5-GGGGCTCTTTTACGAGTGAAATGC-3; IL-7R reverse wt/fl, 5-TGTGAGTCTGAGGTAGATGGCCTGC-3; IL-7R reverse ko/ko, 5-TGTGAGTCTGAGGTAGATGGCCTGC-3.

Antibodies and Reagents.

Monoclonal antibodies recognizing TCR Vα2 (B20.1), TCR Vβ8.1, 8.2 (KJ16-133.18), CD4 (GK1.5), CD8 (53-6.7), CD11c (N418), CD16/32 (93), CD25/IL-2Rα (PC61, PC61.5), CD31 (390), CD44 (IM7), CD45.2 (104), CD62L (MEL-14), CD122/IL-2Rβ (5H4), CD127/IL-7Rα, (A7R34), CD132/γc (TUGm2), CD278/ICOS (7E.17G9), GITR CD357 (DTR-1), podoplanin (gp38, 8.1.1), IFNγ (XMG1.2), FOXP3 (150D), MHCII (AF6-120.1, 25–9-17), pSTAT5 (47), Ki67 (B56), Bcl-2 (3F11), and fluorochrome-labeled streptavidins, and the fixation and permeabilization kit with BD GolgiStop, were purchased from Biolegend, eBioscience, Cell Signaling, and BD Pharmingen. Specificity of staining was confirmed with isotype-matched control antibodies or nonstained samples. The proliferation dyes eFluor670 was purchased from eBioscience, and the BrdU Flow Kit was from BD Pharmingen. Recombinant IL-7 and IFNγ were purchased from Prepro Tech. IL-2 was provided by Matthias Kreuzaler [Department of Biomedicine (DBM), Basel, Switzerland]. The IL-7 complex was formed by incubation of recombinant mouse IL-7 and anti–IL-7 Ab (M25, provided by Daniela Finke, DBM, Basel, Switzerland). The IL-2 complex was formed by incubation of IL-2 and anti–IL-2 (JES-6, provided by Matthias Kreuzaler and Antonius Rolink, DBM Basel, Switzerland). Sandwich ELISAs were purchased from BD Biosciences and R&D Systems. Phorbol 12-myristate 13-acetate and ionomycin were purchased from Sigma, collagenase IV was from Worthington, and DNaseI and collagenase D were from Roche.

Skin Transplantation Model.

Recipient mice were transplanted on the back with full-thickness tail skin from allogenic I-A^bm12 or syngenic I-A^b Rag2^−/− mice under general anesthesia (28). After transplantation of Rag2^−/− mice, naïve ABM and T_reg were adoptively transferred (i.v.) into recipient mice at different ratios. At indicated time points, LN and skin grafts were collected, cells were isolated, and LN extracts were prepared (47). To deplete diphtheria-receptor expressing T_reg after adoptive transfer, transplanted mice were i.p. treated three times with 0.4 mg diphtheria toxin (Merk).

Preparation of T Cells and LN Cells.

Naïve transgenic Vα2Vβ8 CD4⁺ T cells were isolated from LN of ABM mice. T_reg were sorted from Foxp3-eGFP mice. LNSC were isolated from naive, transplanted, or IFNγ treated (10,000 U) mice according to the protocol (47).

Sorting.

GFP- or YFP-positive Treg cells were sorted with a 70-μm tip at a pressure of 60 psi. LN-FRC were stained for CD45, CD11c, gp38, and CD31 and sorted with a 100-µm tip at a pressure of 30 psi. The purity of sorted cells routinely exceeded 98% with an Influx cell sorter (BD).

Flow Cytometry.

Cell suspensions were stained in ice-cold HBSS supplemented with 2% (vol/vol) FCS for surface staining. For intracellular cytokine staining, cells were stained for surface markers, fixed, and then resuspended in permeabilization buffer containing anti-IFNγ. Phosphorylation of STAT5 was assessed in stimulated T cells after fixation and permeabilization by intracellular staining according to protocol (BD Biosciences). Data were acquired on a FACS Canto II and LSRFortessa (BD Biosciences) and analyzed with FlowJo software (TreeStar).

In Vitro T_reg Assays.

For stimulation, T_reg (5 × 10³/well to 5 × 10⁴/well) were incubated in RPMI medium containing rmIL-7 or rmIL-2 (Peprotech) with or without CD3/CD28 activation beads (Invitrogen) or on irradiated splenocytes. To determine cytokine sensitivity, T_reg were incubated with IL-7 or medium for 24 h, washed with HBSS, and then stimulated with different IL-2 and IL-7 concentrations.

In suppression assays, eFluor670-stained ABM and different ratios of T_reg were activated with CD3/CD28 activation beads in the presence of cytokines. Proliferation and up- or down-regulation of surface markers on T cells were measured by flow cytometry.

IL-2 Binding Assay.

IL-2 was biotinylated using EZ-Link Sulfo-NHS-Biotin (Thermo Scientific) and incubated with T_reg treated with IL-7 or medium. After 30 min of stimulation, T_reg were incubated with biotinylated IL-2 followed by staining with streptavidin.

RT-PCR from Cells.

mRNA was isolated from sorted LN stromal cell subsets by a µMACS mRNA Isolation Kit. cDNA was synthesized using a One-step cDNA Kit from Miltenyi Biotech according to the manufacturer’s protocol. Amplification was done in a 10-µL volume with GoTaq qPCR Master Mix (Promega) and analyzed with an Applied Biosystems 7900HT Real-Time PCR System. The following program was used: 40 cycles of 50 °C for 2 min, 95 °C for 2 min, 95 °C for 15 s, and 60 °C for 1 min. Primers for RT-PCR were designed using the Universal ProbeLibrary Assay Design Center from Roche Applied Biosystems and synthesized by Microsynth.

Histology.

Organs were fixed with 4% (wt/vol) formalin, embedded in paraffin, and stained with hematoxilin and eosin for pathological signs.