IL-2 Shapes the Survival and Plasticity of IL-17–Producing γδ T Cells

The cytokine IL-17 is crucial for host defense against a range of bacterial and fungal infections, promoting pathogen clearance by induction of granulopoiesis and neutrophil recruitment (1, 2). It is also implicated in a number of autoimmune and inflammatory pathologies where it is considered a potential target for clinical intervention (3). Although CD4⁺ Th17 cells are a major source of IL-17, their differentiation requires Ag exposure and inflammatory cytokines (4–7), largely excluding them from the initial phase of the immune response. Instead, rapid IL-17 production is the domain of cells with innate qualities such as type-3 innate lymphoid cells, mucosal-associated invariant T cells, NK T cells, and γδ T cells (8).

It has become clear that of these cells, γδ T cells contribute significantly to early IL-17 production in many inflammatory settings. They are the principle source of early IL-17 in several murine models of infection, inflammation, and autoimmunity; particularly at sites such as the lung, skin, and peritoneum (9–16). In humans, IL-17–producing γδ T (γδT-17) cells have been found in increased frequencies in patients with tuberculosis, bacterial meningitis, ankylosing spondylitis, and psoriasis (15, 17–19). Additionally, although γδT-17 cells have shown some antitumor properties (20), they have more recently been found to promote tumor cell proliferation and metastasis via downstream effects on small peritoneal macrophages, myeloid-derived suppressor cells, and neutrophils (21–23). These findings provide a potential explanation for the association of IL-17 and γδ T cells with poor prognosis in breast cancer patients (24, 25), and highlight a previously unappreciated function of γδT-17 cells.

Given their contribution to inflammatory disease and their emerging protumor role, it is crucial to understand what promotes and limits the action of γδT-17 cells. Unlike CD4⁺ Th17 cells, natural γδT-17 cells exit the thymus committed to the IL-17–producing lineage (16, 26–28). They express the transcription factor retinoid-related orphan receptor γt (RORγt) and, as such, do not require exposure to IL-6 in the periphery to initiate its expression (29). Signaling via the TCR is important for their thymic development; however, it has been found that innate signals such as IL-1β and IL-23 largely drive responsiveness of γδT-17 cells in the periphery (14, 30, 31). Being driven primarily by innate signals, rather than TCR specificity, implies the existence of alternate mechanisms to curb unwanted activation. The inhibitory receptor B and T lymphocyte attenuator can limit proliferation and cytokine production by γδT-17 cells via an inhibitory feedback loop involving IL-7 (32). More recently, retinoic acid was found to suppress IL-17A and IL-17F production by γδ T cells by downregulating IL-1R and IL-23R expression (33). However whether other mechanisms exist to keep γδT-17 cells in check is not known.

IL-2 is a cytokine with pleotropic functions. Whereas it was originally discovered as a T cell growth factor, it is now known to influence a much broader range of biological processes. It is key to preserving tolerance, predominantly by maintaining regulatory T (Treg) cells and also by promoting apoptosis of Ag-activated cells by upregulation of the death receptor FAS and FAS ligand (34). It also plays a major role in driving T effector function. For CD8⁺ T cells, the strength of IL-2 signaling influences the differentiation of cells into either apoptosis-prone, short-lived effectors or long-lived memory T cells (35, 36). For CD4⁺ T cells, IL-2 guides effector differentiation, promoting Th1 and Th2, while suppressing Th17 and T follicular helper cell pathways by augmenting subset-specific cytokine receptors (37). Within the γδ T cell population, IL-2R α-chain expression was found on IL-17– but not IFN-γ–producing γδ T (γδT–IFN-γ) cells in the peritoneum (28). This, coupled with the observation that γδT-17 cells were decreased in the peritoneum of 3-wk-old, IL-2–deficient animals, suggested that IL-2 plays a specific role in the maintenance of IL-17– and not γδT–IFN-γ cells (28).

Recently it was found that IL-7 is crucial to γδT-17 cells. It is able to selectively enrich for mouse and human γδT-17 cells (38) and is exclusively required for their peripheral homeostasis (39). Considering that IL-7 alone appears to be able to maintain the γδT-17 population, we decided to revisit the question of the role of IL-2. To our surprise, we found γδT-17 cells to be enriched, instead of depleted, in an IL-2–deficient environment. The absence of IL-2 also correlated with greater IL-17 production and an increase in IL-17⁺IFN-γ⁺ double producers. Although IL-2 does paradoxically induce the proliferation of γδT-17 cells, it also affects their survival via downregulation of the IL-7R. These data suggest that IL-2, particularly in the context of an inflammatory environment, can guide the fate of γδT-17 cells into apoptosis-prone effectors with curtailed plasticity.

C57BL/6 mice were obtained from the Animal Resources Centre (Perth, Australia) and the Australian Bioresources Centre (ABR; Moss Vale, Australia). Blimp1^+/GFP (40), Foxp3–diphtheria toxin receptor (DTR) (41), and IL-2^−/− colonies were maintained at the ABR on a C57BL/6 background (Moss Vale, Australia). Foxp3-DTR mice and Blimp1^+/GFP mice were kindly provided by A. Rudensky (Memorial Sloan Kettering Cancer Center) and S. Nutt (The Walter and Eliza Hall Institute of Medical Research), respectively. C57BL/6 mice were used at 6–12 wk of age. IL-2^−/− mice and their wild-type (WT) littermates were used at 5–6 wk of age, prior to overt signs of disease unless otherwise stated. Animals were housed under conventional barrier protection and handled in accordance with the Garvan Institute of Medical Research and St. Vincent’s Hospital Animal Experimentation and Ethics Committee, which comply with the Australian code of practice for the care and use of animals for scientific purposes.

Single-cell suspensions were prepared from spleens and peripheral lymph nodes (pLN) (inguinal, axillary, brachial, and cervical) by mechanical disruption. Peritoneal exudate cells (PEC) were collected after washing the peritoneal cavity with PBS. Cell suspensions were surface stained for FACS analysis using the following Abs (eBioscience unless stated) specific for: CD25 (PC61), CD122 (TM-β1; BD Biosciences), CD3 (17A2), γδTCR (GL3), Vγ4 (Vγ2, UC3-10A6; BioLegend), Vγ1 (2.11; BioLegend), CD4 (RM4-5; BD Biosciences), CD127 (A7R34), CD95 (Jo2; BD Biosciences), CD27 (LG.7F9), and TCRβ (H57-597; BioLegend). Anti-TCRVγ5Vδ1 (17D1) was kindly provided by A. Hayday (King’s College London, London, United Kingdom). For Vγ6 detection, staining with anti-γδTCR (clone GL3) and 17D1 Abs was performed as previously described (42). For intracellular staining, cells were washed, fixed, and permeabilized with the Foxp3/Transcription Factor Staining Buffer Set (eBioscience) or Cytofix/Cytoperm (BD Biosciences) according to manufacturer’s protocol, and stained with Abs (eBioscience) specific for: RORγ(t) (B2D), Foxp3 (FJK-16s), Ki67 (SolA15), IL-17A (eBio17B7), IFN-γ (XMG1.2), and IL-22 (1H8PWSR). Viability was determined using 7-Aminoactinomycin D (7AAD) (BD Biosciences) or Fixable Viability Dye eFluor 780 (eBioscience). For detection of p-STAT5, cells were surface stained, stimulated with IL-2 (10 ng/ml) for 10 min at 37°C, and then immediately fixed, permeabilized, and stained with pSTAT-5 plus RORγt according to the manufacturer’s instructions. For detection of active caspase-8, the CaspGLOW Fluorescein Active Caspase-8 Staining Kit was used according to the manufacturer’s instructions (eBioscience). Data were acquired on either the FACS CantoII, LSR Fortessa, or LSR II SORP (BD Biosciences). Data were analyzed using FloJo software (Tree Star).

Single cell suspensions of spleen and pLN were first negatively depleted by staining with anti–B220-biotin, anti–CD8-biotin, and anti–CD4-biotin followed by incubation with Dynabeads Biotin Binder (Invitrogen) and magnetic separation. The depleted fraction was stained with anti-γδTCR, anti-TCRβ, anti-CD3, and anti-CD27, and in some cases also anti-Vγ4 and anti-Vγ1. Cell sorting was performed on a FACSAria (BD Biosciences) to obtain a pure population of γδ T cells or γδ T cell subsets for in vitro culture.

Sorted γδ T cells were washed once in 0.1% BSA in PBS, and then labeled with 10 μM CellTrace Violet (CTV) (Molecular Probes) at a density of 1 × 10⁷ cells/ml with 0.1% BSA in PBS for 20 min at 22°C. The reaction was stopped with RPMI 1640 and 10% FCS and washed twice.

Single-cell suspensions from pLN or sorted γδ T cell subsets were cultured at 37°C in complete RPMI-1640 medium with different combinations of recombinant murine (rm) IL-1β (PeproTech), rmIL-23 (R&D Systems), and rmIL-2 (PeproTech) at the concentrations indicated for either 40 or 68 h, and afterward assessed for cytokine production, apoptosis, or expression of cytokine receptors. In some experiments, CTV-labeled, sorted γδ T cell subsets were cultured on plate-bound anti-CD3ε (1 μg/ml; clone 145.2C11) with or without anti-CD28 (10 μg/ml; clone 37.51) (both from eBioscience). Supernatants were taken at 24 h for analysis of IL-2 production by ELISA (BioLegend) and proliferation was analyzed at 68 h by CTV dilution by flow cytometry.

For intracellular cytokine staining, spleen, pLN, or PEC single-cell suspensions or FACS-sorted cells were incubated at 37°C in complete RPMI-1640 medium with PMA (10 ng/ml) and ionomycin (1 μg/ml) for 4 h with monensin. Cells were surface stained, then treated with Cytofix/Cytoperm according to manufacturer’s protocol (BD Biosciences), and stained with IL-17, IFN-γ, or IL-22.

Binding of IL-2 to γδ T cells was assessed using an IL-2–Fc fusion protein consisting of human IL-2 and a mutated mouse Fc (IgG2c) with abolished FcγR interaction. IL-2–Fc was produced recombinantly in HEK293 cells (Expi293 Expression System; Thermo Fisher) and biotinylated using EZ-Link NHS-PEG4-Biotin (Thermo Fisher). Serially diluted IL-2–Fc–biotin was added to single cell suspensions of pLN prestained with fluorescently conjugated Abs for γδ T cell identification, and binding was allowed for 30 min on ice. Cells were washed and stained with allophycocyanin-conjugated streptavidin (eBioscience) prior to flow cytometric analysis.

rmIL-2 (PeproTech) was mixed with anti–IL-2 (JES6-1; eBioscience) and incubated at 37°C for 30 min, as described (43). Mice were injected i.p. with 1 μg cytokine and 5 μg mAb on three consecutive days. IL-2/JES6-1 was used as it preferentially stimulates CD25-expressing cells, while still allowing stimulation of low-affinity, IL-2R–expressing cells. This is in contrast to IL-2/S4B6, where the epitope for CD25 binding is blocked.

To eliminate Tregs in vivo, Foxp3-DTR mice were injected i.p. with 50 μg/kg diphtheria toxin on days 0 and 2. On day 6, pLN were collected for analysis of CD4⁺ and CD8⁺ αβ T cells and γδ T cell subsets.

Results are presented as mean ± SEM. Statistical significance was assessed using the Mann–Whitney U test, one-way ANOVA, or two-way ANOVA with the Tukey multiple comparison test (Prism; Graphpad) as detailed in figure legends.

To determine what role IL-2 plays in γδT-17 cell biology, we first sort to confirm the expression of IL-2R subunits on these cells. In agreement with previous reports (28, 44), we found the IL-2Rα subunit (CD25) to be highly expressed on RORγt-expressing γδ T (γδT-17) cells (Fig. 1A), although still at significantly lower levels than that expressed by Foxp3⁺ Tregs (Fig. 1B). Because CD25 increases the affinity of the IL-2R for IL-2 rather than participating in signaling (45), expression of IL-2Rβ (CD122) is jointly needed if there is to be signal transduction. It is surprising, then, that CD122 is barely detectable on steady-state γδT-17 cells compared with IFN-γ–expressing γδ T cells (RORγt⁻CD45RB⁺) (Fig. 1C) (28) for which CD122 is crucial to their IL-15–responsiveness. However, when cells are exposed to IL-2, or to cytokines that promote γδT-17 function (IL-1β and IL-23) CD122 expression rises to detectable levels (Fig. 1D). Interestingly, exposure to these cytokines also results in a sharp increase in surface CD25 expression (Fig. 1D). In fact, IL-1β and IL-23 are significantly more effective at inducing CD25 than IL-2 alone, with the combination of all three inflammatory cytokines resulting in the highest surface receptor density.

To confirm that the high density of CD25 translated to greater IL-2 binding for γδT-17 cells, we used an ex vivo binding assay. Fluorescently labeled IL-2–fusion protein was indeed found to preferentially bind CD27⁻ γδ T (γδT-17) cells over the remaining CD27⁺ γδ T cells (Fig. 2A, 2B). This also translated to more robust STAT5 signaling for γδT-17 cells after IL-2 exposure (Fig. 2C), implying that the high-density CD25 leads to a functional difference.

In an earlier report, γδT-17 cells were found to be decreased in IL-2–deficient animals compared with WT controls, suggesting a role for IL-2 in their maintenance (28). This result was observed in the PEC of 3-wk-old animals, and so we decided to extend this analysis to peripheral lymphoid organs. In contrast to this report, we failed to find a significant difference in the proportions or absolute number of γδT-17 cells in 3-wk-old IL-2–deficient mice relative to WT mice, either in PEC or spleen (Supplemental Fig. 1). However, there was a marked enrichment of γδT-17 cells in 5- to 6-wk-old mice (Fig. 3A–C). This enrichment was observed in the spleen, pLN, and PEC, with a corresponding decrease in the proportion of γδT–IFN-γ cells (Fig. 3B). However, only absolute numbers of γδT-17 cells, not γδT–IFN-γ cells, were altered significantly (Fig. 3C); implying that the removal of IL-2 has specific consequences only for the γδT-17 subset. In addition to increased proportions, we observed that γδT-17 cells produced more IL-17 per cell, as evidenced by an increase in IL-17 mean fluorescence intensity (MFI) (Fig. 3D). Interestingly, we also observed an emergence of double-producing IL-17⁺IFN-γ⁺ γδ T cells in IL-2–deficient animals (Fig. 3A, 3E). These cells, which have been found to originate from the CD27⁻ subset, can be generated in vitro but are usually rare in WT mice and only observed in greater numbers in particular disease models (46, 47). We confirmed that IL-2 could have a direct cell-intrinsic effect on γδT-17 cells by culturing FACS-purified CD27⁻ cells (Vγ1⁻Vγ4⁻ subset [Vγ6⁺]) with IL-2, IL-1β and IL-23, and then assessing their cytokine production with PMA and ionomycin, an approach used by others to assess plasticity (48). We found that with a 40-h culture, addition of IL-2 to cells cultured with IL-1β and IL-23 lowered the IL-17 MFI and reduced the proportion of double-producing IL-17⁺IFN-γ⁺ and IL-17⁺IL-22⁺ cells (Fig. 4). Similar findings applied with the subset of Vγ4⁺ CD27⁻ cells (Supplemental Fig. 2). (Heilig and Tonegawa mouse Vγ nomenclature used throughout [49]).

Most γδT-17 cells in the mouse are Vγ4⁺ or Vγ6⁺ (26). Vγ6⁺ γδT-17 cells mostly express the invariant TCR Vγ6Vδ1, whereas Vγ4⁺ γδT-17 cells are more polyclonal. Both subsets have demonstrated roles in tissue inflammation, although the Vγ6⁺ subset appears to acquire tissue-homing capabilities in the thymus, whereas the Vγ4⁺ subset may require extrathymic imprinting (50). In IL-2–deficient animals, we found that the γδT-17 population was skewed away from cells expressing Vγ4 to those negative for both Vγ1 and Vγ4 (Fig. 5A, 5B). In fact, in the spleen, the increase in γδT-17 cells was entirely because of an enlargement of the Vγ1⁻Vγ4⁻ population (Fig. 5C). We confirmed that these cells were positive for Vγ6 by staining with a combination of the 17D1 Ab and anti-γδTCR (GL3) (Fig. 5D). (17D1, an Ab that detects the epidermis-associated Vγ5/Vδ1⁺ TCR, will also detect Vγ6/Vδ1⁺ cells if their TCR is first complexed to the anti-γδTCR [clone GL3] Ab [42]). In line with this finding, although both γδT-17 subsets express high levels of CD25, we found that the Vγ6⁺ (Vγ1⁻Vγ4⁻) subset had a higher density of this receptor (Fig. 5E) and showed a correspondingly higher binding of IL-2 in an ex vivo binding assay (Fig. 5F, 5G).

To determine whether perturbed thymic development was responsible for the enrichment of γδT-17 cells, we examined thymic γδ T cells. “Natural” γδT-17 cells (those that exit the thymus precommitted to the lineage) primarily develop in the embryonic and neonatal period (26). Correspondingly, in adult thymus, γδT-17 cells are found in very low proportions (Supplemental Fig. 3A). We did not observe an increase in γδT-17 cells in IL-2^−/− adult thymus (Supplemental Fig. 3B), indicating no extension of the thymic developmental window because of the loss of IL-2. We did, however, observe a skewing of the γδT-17 population toward Vγ6⁺ (Vγ1⁻Vγ4⁻) cells (Supplemental Fig. 3C). However, because the majority of γδT-17 cells were of a mature phenotype (CD24^loCCR6⁺) (Supplemental Fig. 3D), it is likely that this reflects a recirculating population from the periphery. As such, we examined neonatal thymus for a more accurate representation of γδT-17 thymic development. In this study, we found both WT and IL-2^−/− neonatal thymus to contain high numbers of γδT-17 cells, with no differences observed in the ratio of Vγ6⁺ to Vγ4⁺ γδT-17 cells (Supplemental Fig. 3E–G). Further to these data, unlike their peripheral counterparts, thymic γδT-17 cells do not express CD25, further supporting a role for IL-2 in the periphery (Supplemental Fig. 3H) (28).

Considering that an IL-2–deficient environment yielded more and not less γδT-17 cells, we next analyzed the effect of exogenous IL-2. Mice were treated with high-dose IL-2 in the form of IL-2/IL-2 mAb complexes (IL-2c). In this study, we found that IL-2 provoked proliferation, with a more rapid increase in the proportion of Ki67⁺ cells in the RORγt⁺ (γδT-17) subset than other γδ T cells (Fig. 6A). An increase in proliferation was evident in both Vγ4⁺ and Vγ6⁺ (Vγ1⁻Vγ4⁻) γδT-17 cell subsets (Fig. 6B). Intriguingly, although they demonstrated more robust proliferation than other γδ T cells, the proportion of γδT-17 cells decreased in the days following IL-2 injection (Fig. 6C, 6D). Although an increase in γδT-17 cell number was observed, their fold change was dramatically less than other CD25-expressing cells (Tregs), and even less than γδT–IFN-γ cells which only express the CD122-containing, low-affinity IL-2R (Fig. 6E, 6F). Because γδT-17 cells proliferated, but did not accumulate in correspondingly high numbers, we questioned whether IL-2 was triggering increased rates of cell death. Consistent with this, we did indeed observe more apoptotic (as measured by caspase-8) and dead γδT-17 cells after proliferation was induced in the presence of IL-2 compared with cells cultured only with IL-1β and IL-23 (Fig. 6G, 6H).

Considering the impact of IL-2 on γδT-17 cells, it was important to determine if γδT-17 cells produce IL-2 themselves. CD28 costimulatory signals in synergy with TCR activation have been found to induce autocrine IL-2 production by γδ T cells (51). As has been reported by others (52), we found that CD27⁻ γδ T cells (γδT-17) display poor proliferative responses to TCR stimulation in vitro compared with CD27⁺ γδ T cells (Fig. 7A). Additionally, we found that costimulation with CD28 enhanced IL-2 production, but only for the CD27⁺ subset (Fig. 7B). We could not detect IL-2 production by CD27⁻ γδ T cells after either stimulation with anti-CD3 and anti-CD28 or cytokines (Fig. 7B).

Given that expression of B lymphocyte–induced maturation protein-1 (Blimp-1) is known to repress IL-2 transcription (53) and is a hallmark of mature, pathogenic CD4⁺ Th17 cells (54), we decided to investigate its expression in γδ T cells. Using naive Blimp1^+/GFP mice, we detected Blimp-1 expression in CD27⁻, but not CD27⁺ γδ T cells (Fig. 7C, 7D), potentially explaining the dichotomy in IL-2 production between the γδ T cell subsets. Further to this, Blimp-1 expression dramatically increased upon exposure to IL-1β and IL-23 (Fig. 7E), suggesting that an inflammatory environment would promote the activity of Blimp-1 in γδT-17 cells.

The enrichment of γδT-17 cells in IL-2–deficient animals and their lack of accumulation after IL-2c treatment could also be related to the absence or expansion of Tregs in these respective environments. To test whether γδT-17 cells show greater sensitivity to Treg suppression, we transiently depleted Tregs in Foxp3-DTR mice. Using this approach, we observed a three- to sixfold increase in αβ and γδ T cell populations in pLN (Supplemental Fig. 4). RORγt⁺ γδ T cells did not undergo a greater fold expansion than other γδ T cells or αβ T cells, suggesting that they are not uniquely susceptible to the effects of Tregs.

Because IL-7 is crucial for γδT-17 cell survival and homeostasis (39), we analyzed the impact of IL-2 on expression of the IL-7R. Indeed we found that treatment of mice with IL-2 resulted in a significant, yet transient, decrease in the density of IL-7Rα on the surface of γδT-17 cells (Fig. 8A). Interestingly, this decrease was greater and more long lasting for Vγ6⁺ (Vγ4⁻) compared with Vγ4⁺ γδT-17 cells (Fig. 8B), providing further insight into the increased susceptibility of this subset to IL-2 deficiency.

A successful immune response is contingent not only on appropriate immune cell recruitment and activation, but equally on the timely contraction of inflammatory responses to limit host damage. For adaptive immune cells, the availability of Ag is central to both their expansion and subsequent contraction. For innate-like T cells, such as γδT-17 cells which are driven primarily by innate signals rather than Ag specificity, the rules that govern their population dynamics are not clear. In this study we found that IL-2 plays a specific role in determining the functional capacity of γδT-17 cells.

We found that although IL-2 induces γδT-17 cell proliferation, it also negatively impacts on their survival via downregulation of the IL-7R. It also seems to preserve IL-17 single producers rather than promote differentiation into multifunctional cytokine-producing cells. In short, it appears to favor short-lived effectors with limited plasticity. Because IL-2 is produced in large part by CD4⁺ T cells, and is strongly induced after Ag activation, we suggest that IL-2 is part of a system to start curtailing the γδT-17 cell response upon the arrival of the adaptive immune cells. The marked upregulation of CD25 after exposure to combined IL-2, IL-1β and IL-23 supports the idea that IL-2 will have a greater effect on these cells in an inflammatory environment. In an IL-2–deficient environment, we observed an enrichment of γδT-17 cells in 5- to 6-wk-old mice, and not 3-wk-old mice. Although the absence of Tregs (as occurs in IL-2^−/− mice) would undoubtably result in an expansion of γδT-17 cells, our data from Foxp3-DTR mice suggest that this would be no greater than that of other lymphocyte populations. Instead we propose that in these mice, a build up of γδT-17 cells could occur because of excessive homeostatic (IL-7) and inflammatory (IL-1β and IL-23) cytokine-induced proliferation, in the absence of the apoptosis-inducing effect of IL-2.

Further to this, although our experiments suggest that proliferation driven by IL-2 does not lead to enhanced γδT-17 survival, experiments in CD28-deficient mice imply that the combination of CD28 costimulation and IL-2 stimulation could potentially drive a sustainable population of γδT-17 cells (51). As such, the rules that govern γδT-17 population dynamics could differ significantly depending on the contribution to activation from TCR and non-TCR sources.

For γδ T cells, IL-17 and IFN-γ cytokine production was originally found to segregate into mutually exclusive populations along with CD27 expression (52). More recently it was found that the CD27⁻, and not the CD27⁺, subset can exhibit plasticity under inflammatory conditions (46, 47). This can be observed in vitro after stimulation with IL-1β and IL-23, where IL-17A–producing cells start simultaneously producing IFN-γ, IL-22, GM-CSF, and IL-17F (48). It is seemingly not common in vivo, but has been observed in two disease models: a peritoneal cancer model, and oral Listeria monocytogenes infection (46, 47). In our study, we observed that in the absence of IL-2 in vivo, IL-17⁺IFN-γ⁺ γδ T cells start to emerge, and that the addition of IL-2 to culture conditions which favor polyfunctional γδT-17 cells can partially block this differentiation. It has been shown that the capacity for simultaneous IL-17 and IFN-γ production corresponds with an upregulation of T-bet expression (47, 48). It will be of interest to determine if signaling via IL-2 can impact on this and other downstream processes that enable the acquisition of plasticity in γδT-17 cells.

In this study, we found that compared with the Vγ4⁺ subset, Vγ6⁺ γδT-17 cells are more sensitive to the effects of IL-2. These cells express more surface CD25, bind IL-2 with higher affinity, and in response to IL-2 they sustained a greater and more long-lasting downregulation of the IL-7R than Vγ4⁺ cells. Why Vγ6⁺ γδT-17 cells would benefit from an increased sensitivity to IL-2 is not clear. In contrast to Vγ4⁺ γδT-17 cells, the Vγ6⁺ subset is known to preferentially migrate to tissues such as the uterus, vagina, lung, peritoneal cavity, and dermis after thymic development (50, 55). Vγ4⁺ γδT-17 cells are more common in circulation and peripheral lymphoid tissues, and although they are also found in the dermis (50, 56, 57) and the maternal–fetal interface during pregnancy (58), their tissue-homing capabilities are thought to be gained postthymically (50). Because Vγ6⁺ γδT-17 cells are the subset that more commonly reside in tissues, increased sensitivity to IL-2 may allow them to respond to incoming IL-2–producing cells, which may be less plentiful than those in lymphoid tissues. Here, IL-2 reactivity may be a mechanism that helps to tune the response of Vγ6⁺ γδT-17 cells, as in tissues such as the gut, joint, bone, and retina they have the potential to play significant protective or pathogenic roles (47, 59–61).

Interestingly, we found that γδT-17 cells constitutively express Blimp-1, and that its expression is upregulated upon exposure to IL-1β and IL-23. It is most likely the presence of Blimp-1 that inhibits the production of IL-2 by γδT-17 cells, leaving them receptive instead to paracrine IL-2. Because Blimp-1 plays a pivotal role in the generation of pathogenic Th17 cells which coexpress IFN-γ and GM-CSF (54), it is tempting to speculate that its expression in γδT-17 cells not only suppresses IL-2 production, but can affect differentiation into polyfunctional cytokine-producing cells. It will be of interest to determine the function of Blimp-1 in γδT-17 cells and whether its expression contributes to the γδT-17 plasticity observed in in vivo disease models (46, 47).

In conclusion, we have identified IL-2 as an important player in γδT-17 cell fate. Due to high expression of CD25, γδT-17 cells are able to preferentially bind IL-2 over other γδ T cells. Ligation of IL-2R by IL-2 induces proliferation but affects their survival via downregulation of the IL-7R. It also moderates IL-17 production and limits differentiation into multifunctional cytokine-producing cells. These data suggest that, particularly in the presence of IL-1β and IL-23, IL-2 can augment the γδT-17 immune response. This may form part of a process to curtail the innate-like response of γδT-17 cells upon arrival of adaptive immune cells at the site of inflammation.

We thank the staff of the Biological Testing Facility (Garvan Institute), particularly Erin Wray, for help with animal breeding and care. We thank Dr. Robert Salomon and the staff of the Garvan Flow Cytometry Facility for cell sorting (Garvan Institute). We thank Adrian Hayday (King’s College London, London) for providing the anti-TCRVγ5Vδ1 (17D1) Ab, Stephen Nutt (The Walter and Eliza Hall Institute of Medical Research, Melbourne) for providing the Blimp1^+/GFP mice, and Alexander Rudensky (Memorial Sloan Kettering Cancer Center, New York) for providing the Foxp3-DTR mice.

The authors have no financial conflicts of interest.