Adult DETCs produce IL-13 but not IFN-γ upon short-term PMA/ionomycin stimulation (22). To quantify cytokine levels secreted by adult DETCs upon TCR stimulation, we purified DETCs from adult ear epidermal cells without TCR ligation by positive magnetic selection using an anti-integrin β7 mAb. The purified DETCs were >95% pure Vγ3⁺ T cells and contained <1% non-Vγ3⁺ T cells and <1% IA⁺ Langerhans cells ( Figure 1A ). Although DETCs were reported to constitutively produce IL-13 at steady state (22), purified DETCs cultured without TCR stimulation did not secrete IL-13 or other cytokines ( Figure 1B ). The purified DETCs predominantly secreted IL-13 during the first 24 hours of TCR stimulation, and IFN-γ secretion increased after 24 hours. Consistent with a previous study (39), DETCs secreted a small amount of IL-17A upon TCR stimulation, but IL-4 secretion was not detected ( Figure 1B ).

RT–PCR analysis revealed that purified naive DETCs constitutively expressed Gata3, the master transcription factor of type 2 cytokine-producing cells, but Tbx21 was expressed only after activation ( Figure 1C ). Eomes, which encodes eomesodermin that also regulates IFN-γ production in type 1 cytokine-producing T cells, was not expressed in DETCs even after activation ( Figure 1C ). Consistent with the IL-17A-producing capacity of a subpopulation of DETCs, the expression of Rorc, which encodes RORγt, the master transcription factor of IL-17A-producing cells, was detected in naive DETCs ( Figure 1C ). The lack of IL-4 production by DETCs might be explained by the constitutive expression of Zbtb32 ( Figure 1C ), which encodes ZBTB32 (Repressor of GATA, ROG) that represses Il4 but not Il13 gene activation in type 2 CD8⁺ cytotoxic T lymphocytes (43).

These results confirmed that under steady-state conditions, DETCs in the adult epidermis predominantly produce IL-13 during the early phase of activation. However, DETCs can produce IFN-γ during the later phase of activation. Therefore, unlike type 2 CD4⁺ helper T cells (44), the IFN-γ-producing capacity of DETCs would not be repressed by stable epigenetic modifications, as is the case for IL-17A-producing CD27^- γδ T cells, which can produce IFN-γ in a certain inflammatory microenvironment (45).

Similar to embryonic day 17 (E17) mature Vγ3⁺ fetal thymocytes, circulating Vγ3⁺ T cells in the blood at E17 expressed CD27 ( Figure 2A ). As E17 fetal dermal Vγ3⁺ T cells did not express CD27 ( Figure 2A ), Vγ3⁺ T cells lost CD27 expression immediately after entering the dermis. A candidate that induces CD27 downregulation on Vγ3⁺ T cells is extracellular ATP, which would be abundant in the perinatal skin, because CD27 on T cells is rapidly shed and downregulated upon treatment with ATP (46). ATP treatment of E17 Vγ3⁺ fetal thymocytes resulted in rapid downregulation of CD27 ( Figure 2B ). As Vγ3⁺ T cells in day 1 (D1) neonatal epidermis that lacked CD27 expression ( Figure 2A ) could produce IFN-γ ( Figure 3 ), CD27 expression and IFN-γ-producing capacity were not perfectly correlated in Vγ3⁺ T cells.

To clarify when and where Vγ3⁺ T cells functionally switch from preprogrammed IFN-γ producers to rapid IL-13 producers, we analyzed cytokines produced by Vγ3⁺ T cells upon short-term stimulation during ontogeny. Vγ3⁺ T cells in the E18 fetal epidermis still predominantly produced IFN-γ when stimulated with PMA/ionomycin for 4 hours ( Figure 3 ). While IFN-γ-producing cells gradually decreased after birth, IL-13-producing cells gradually increased ( Figure 3 ). As cells producing both IFN-γ and IL-13 transiently appeared in the neonatal epidermis ( Figure 3 ), Vγ3⁺ T cells differentiated from preprogrammed IFN-γ producers into rapid IL-13 producers via intermediate IFN-γ/IL-13 producers during this period in the epidermis.

Thymic maturation of Vγ3⁺ T cells to IFN-γ-producing cells requires TCR signaling. As the proliferation of Vγ3⁺ T cells in the perinatal epidermis also relies on TCR signaling (47–49), we hypothesized that TCR signaling also induces the IL-13-producing capacity of Vγ3⁺ T cells in the perinatal epidermis.

To determine the role of TCR signaling in the induction of the IL-13-producing capacity of epidermal T cells, we analyzed cytokines produced by resident epidermal T cells of adult TCR δ-chain-deficient Tcrd ^-/- mice and TCR Vδ1-chain-deficient Tcrd-V1 ^-/- mice. In Tcrd ^-/- mice lacking all γδ T cells, the epidermal niches of DETCs are replaced by αβ T cells, but the αβ TCRs expressed on these epidermal T cells cannot recognize the self-ligand on epidermal keratinocytes (27, 30). In contrast, epidermal T cells of Tcrd-V1 ^-/- mice express diverse γδ TCRs that can recognize the self-ligand on epidermal keratinocytes (50).

Epidermal αβ T cells of adult Tcrd ^-/- mice produced IFN-γ but not IL-13 or IL-17A upon PMA/ionomycin stimulation for 4 hours ( Figure 4 ). In contrast, both epidermal Vγ3⁺ and Vγ2⁺ T cells of adult Tcrd-V1 ^-/- mice, the latter of which are biased to produce IL-17A in the dermis of wild-type mice (38, 51, 52), predominantly produced IL-13 but not IFN-γ or IL-17A ( Figure 4 ). These results suggest that ligand-dependent TCR signaling in the epidermis is needed for the induction and/or maintenance of the rapid IL-13-producing capacity of epidermal T cells.

To determine the role of TCR signaling in the functional differentiation of DETCs more directly, we used an in vitro differentiation model of neonatal DETCs. Neonatal epidermal cells were cultured under TCR stimulation in the presence of various inhibitors of signaling pathways downstream of TCR, and cytokine production by Vγ3⁺ T cells was analyzed after restimulation with PMA/ionomycin for 4 hours. Among the various inhibitors, only the p38 MAPK inhibitor SB203580 blocked the differentiation of IL-13-producing cells from IFN-γ-producing cells ( Figure 5A ). Therefore, it was found that TCR signaling through the p38 MAPK pathway promotes the functional switch of DETCs.

Interestingly, the addition of MEK1/2-ERK1/2, PI3K, mTORC1, and mTORC1/2 inhibitors increased IL-13-producing cells with minimal impact on IFN-γ-producing cells ( Figure 5A ). As both ERK and PI3K activate mTORC1 (53), mTORC1 activation might suppress differentiated IL-13-producing cells.

In the adult epidermis, TCRs on DETCs are triggered at steady state (30). Therefore, TCR-p38 MAPK signaling may also play a role in the maintenance of the IL-13-producing capacity of DETCs in the adult epidermis. Intradermal administration of the p38 MAPK inhibitor SB203580 24 hours before analysis decreased DETCs with the IL-13-producing capacity in the adult epidermis ( Figures 5B, C ). IL-13 mRNA expression in epidermal cells upon short-term stimulation with PMA/ionomycin in vitro ( Figure 5D ) and in response to tape-stripping in vivo ( Figure 5E ) was suppressed by SB203580 pretreatment. As DETCs are the only cells in the epidermis that produce IL-13 (22), these results suggest that continuous TCR signaling through the p38 MAPK pathway is also needed for maintaining the rapid IL-13-producing capacity of adult DETCs under steady-state conditions.

A recent study showed that IFN-γ-producing γδ T cells use glycolysis for proliferation and to maintain effector functions, but IL-17A-producing γδ T cells are dependent on mitochondrial oxidative phosphorylation and fatty acid oxidation (54). As mTORC1, which regulates cellular metabolism, suppressed IL-13-producing cells in our in vitro differentiation model ( Figure 5A ), the functional switch of DETCs from IFN-γ-producing cells to IL-13-producing cells may be associated with the metabolic switch from mTORC1-dependent metabolic pathways to those less dependent on mTORC1.

To analyze the metabolic profiles of DETCs at different ages, we used the recently developed flow cytometry-based method SCENITH™ (55). D2 neonatal and D14 DETCs displayed higher glucose dependence and higher glycolytic capacity than D21 and D28 adult DETCs ( Figure 6A ). By D21, DETCs showed low glucose dependence and low glycolytic capacity, with a subsequent high dependence on mitochondrial oxidative phosphorylation and fatty acid/amino acid oxidation ( Figure 6A ). This metabolic reprogramming took place between 2 and 4 weeks after birth ( Figure 6A ) and thus occurred later than the functional switch observed during the perinatal period ( Figure 3 ). As DETCs have been shown to cease massive proliferation in the epidermis by 2 weeks after birth in C57BL/6 mice (37), mTORC1 activity may be attenuated at this time and maintained at low levels in adult DETCs under steady-state conditions.

To determine whether mTORC1 activity is diminished in adult DETCs under steady-state conditions, we compared the levels of phosphorylated 4E-BP1 (p4E-BP1), an mTORC1 downstream target, between neonatal and adult DETCs. Compared with D2 neonatal DETCs, p4E-BP1 levels were lower in adult DETCs ( Figures 6B, C ). Therefore, mTORC1 activity was found to be attenuated in adult DETCs under steady-state conditions.

Finally, we determined whether cellular metabolism regulates cytokine production by adult DETCs. In contrast to adult DETCs under steady-state conditions ( Figure 6A ), adult DETCs stimulated with anti-TCR mAb for 3 days in vitro showed high glycolytic capacity ( Figure 7A ). Although adult DETCs predominantly produce IL-13 during the early phase of activation, adult DETCs stimulated for 3 days in vitro produced both IL-13 and IFN-γ ( Figures 7B, C ). Inhibition of glycolysis with 2-deoxy-D-glucose (2-DG) decreased IFN-γ-producing cells but not IL-13-producing cells in adult DETCs stimulated with anti-TCR mAb for 3 days in vitro ( Figures 7B, C ). As TCR-stimulated DETCs cultured in the presence or absence of 2-DG had equivalent levels of p4E-BP1 ( Figures 7D, E ), inhibition of glycolysis did not alter mTORC1 activity in adult DETCs. Therefore, it was found that glycolysis acts downstream of mTORC1 in the regulation of DETC cytokine production.

C57BL/6J mice were purchased from Japan SLC (Hamamatsu, Japan). Tcrd ^-/- mice (63) were purchased from the Jackson Laboratory (Bar Harbor, ME). Tcrd-V1 ^-/- mice (50) were a gift from Yasunobu Yoshikai (Division of Host Defense, Medical Institute of Bioregulation, Kyusyu University, Fukuoka, Japan). TSLP-R ^-/- mice (64) were a gift from Steven F. Ziegler (Benaroya Research Institute, Seattle, WA). All mice were bred and maintained on a C57BL/6 background in the animal facility of Kagoshima University under specific pathogen-free conditions. Female mice at 4-12 weeks of age were used as adult mice. Fetuses and postnatal mice younger than 4 weeks were used irrespective of sex. Fetuses were obtained from timed pregnant mice. The plug date was defined as embryonic day 0 (E0).

To isolate epidermal cells, the skin was floated dermal-side down on 1% trypsin (Gibco, Waltham, MA) in PBS for 30 minutes at 37°C. The epidermis was separated and collected in Iscove’s modified Dulbecco’s medium (IMDM, Gibco) supplemented with 10% FCS (Sigma–Aldrich, St. Louis, MO) and 0.025% DNase I (Sigma–Aldrich). Single-cell suspensions were obtained by mechanical agitation and sequential filtration through 70- and 30-μm nylon meshes.

DETCs were purified from epidermal cells by positive magnetic selection using an anti-integrin β7 mAb because DETCs are the only cells in the normal epidermis that express the integrin β7 chain (65). After preincubation with anti-CD16/CD32 mAb (clone 2.4G2; BD Biosciences, Franklin Lakes, NJ), cells were stained with PE-conjugated anti-integrin β7 mAb (clone M293, BD Biosciences), followed by incubation with magnetic particles conjugated with anti-PE mAb (BD IMag™ Anti-R-PE Magnetic Particles-DM, BD Biosciences). The labeled cells were isolated using the BD IMag™ Cell Separation Magnet (BD Biosciences) according to the manufacturer’s instructions.

To isolate dermal cells, the skin was floated dermal-side down on 1.2 U/mL dispase II (Roche Diagnostics, Basel, Switzerland) in IMDM for 30 minutes at 37°C. After removing the epidermis, small pieces of the dermis were digested in IMDM containing 0.01% DNase I and 250 U/mL collagenase IV (Sigma–Aldrich) for 30 minutes in a shaking water bath at 37°C. The digested dermis was filtered through 70- and 30-μm nylon meshes.

Fetal thymocytes were obtained by teasing the thymic lobes with fine forceps and filtering through a 70-μm nylon mesh.

Blood was collected in 20 mM EDTA in PBS. Lymphocytes were isolated by density gradient centrifugation on Lympholyte™-M Cell Separation Media (Cedarlane Laboratories, Burlington, Canada) for 15 minutes at 1000 × g.

Cells were resuspended in PBS supplemented with 2% FCS and 0.1% NaN₃. After preincubation with anti-CD16/CD32 mAb, cells were stained with the following mAbs: FITC-, PE-, BD Horizon™ BB700-, or biotin-conjugated anti-TCR Vγ3 (clone 536, BD Biosciences or BioLegend, San Diego, CA), FITC-conjugated anti-CD3 (clone 145-2C11, eBioscience, Waltham, MA), FITC-conjugated anti-IA^b (clone AF6-120.1, BD Biosciences), biotin-conjugated anti-TCR Cβ (clone H57-597, BD Biosciences), biotin-conjugated anti-CD27 (clone LG.7F9, eBioscience), BB700-conjugated anti-TCR Vγ2 (clone UC3-10A6, BD Biosciences), and FITC-, PE-, BB700-, or biotin-conjugated isotype control mAbs (BD Biosciences or eBioscience). Biotin-conjugated mAbs were visualized with FITC- or PE-Cy5™-conjugated streptavidin (SouthernBiotech, Birmingham, AL or BD Biosciences).

For intracellular cytokine staining, cells were stimulated with 25 ng/mL PMA (Sigma–Aldrich) and 1 μg/mL ionomycin (Sigma–Aldrich) in the presence of brefeldin A (GolgiPlug™, BD Biosciences) for 4 hours at 37°C. TCR-stimulated cells were incubated with brefeldin A for the last 4 hours. After surface staining, the cells were fixed and permeabilized using Cytofix/Cytoperm™ (BD Biosciences) for 20 minutes at 4°C. Cells were washed and stained in Perm/Wash™ buffer (BD Biosciences) with Alexa Fluor™ 488-conjugated anti-IFN-γ (clone XMG1.2, BD Biosciences), PE-conjugated anti-IL-13 (clone eBio13A, eBioscience), PE-conjugated anti-IL-17A (clone TC11-18H10, BD Biosciences), and Alexa Fluor™ 488- or PE-conjugated isotype control mAbs (BD Biosciences or eBioscience).

For intracellular p4E-BP1 staining, after surface staining, cells were fixed and permeabilized using Cytofix/Cytoperm™ and washed and stained in Perm/Wash™ buffer with Alexa Fluor™ 488-conjugated anti-p4E-BP1 (Thr37/46) mAb (clone 236B4, Cell Signaling Technology, Danvers, MA) or Alexa Fluor™ 488-conjugated isotype control mAb (Cell Signaling Technology).

After gating on forward and side scatters and viable cells as previously described (29, 65), cells were analyzed on a CytoFLEX flow cytometer with CytExpert software (Beckman Coulter, Brea, CA), and the data were analyzed using FlowJo™ software (Tree Star, Ashland, OR).

Purified DETCs were stimulated on 96-well plates (5 × 10⁴ cells/well) coated with 10 μg/mL anti-TCR Cδ mAb (UC7-13D5, BD Biosciences) or isotype control mAb (BD Biosciences) in IMDM supplemented with 10% FCS and 50 μM 2-mercaptoethanol (Nacalai Tesque, Kyoto, Japan) for 24 hours at 37°C. Culture supernatants were harvested and replaced with fresh culture medium, and DETCs were stimulated for an additional 24 hours. Cytokine levels in the supernatants were determined using Quantikine™ ELISA kits (R&D Systems, Minneapolis, MN).

Epidermal cells were stimulated on 24-well plates (1 × 10⁶ cells/well) coated with 10 μg/mL anti-TCR Cδ mAb in IMDM supplemented with 10% FCS, 50 μM 2-mercaptoethanol, and 10 ng/mL recombinant mouse IL-2 (R&D Systems) in the presence of DMSO (Sigma–Aldrich) or 3 mM 2-DG (MedChemExpress, Monmouth Junction, NJ) for 3 days at 37°C. At the time of use, DETCs were harvested by incubation with 1 mM EDTA in PBS for 3 minutes.

Total RNA was extracted from the cells using the RNeasy™ Plus Mini kit (Qiagen, Venlo, Netherlands) and reverse transcribed using the SuperScript™ III First-Strand Synthesis System for RT–PCR (Invitrogen, Waltham, MA) with random hexamers. The cDNA was subjected to quantitative real-time PCR in triplicate using Thermal Cycler Dice™ Real Time System (Takara, Kusatsu, Japan) with FastStart Universal SYBR™ Green Master (Roche Diagnostics). All primers were purchased from Takara. Primer sequences are available upon request. The cycling conditions were 95°C for 10 minutes, followed by 40 cycles of 95°C for 15 seconds and 60°C for 1 minute. Threshold cycle (Ct) values were determined, and mRNA expression relative to that of the Actβ mRNA was calculated as 2^-ΔCt.

E17 fetal thymocytes were cultured on 24-well plates (2 × 10⁶ cells/well) in IMDM supplemented with 10% FCS and 50 μM 2-mercaptoethanol in the presence of PBS or 3 mM ATP (Sigma–Aldrich) for 30 minutes at 37°C.

Epidermal cells isolated from day 1 neonatal mice were stimulated on 24-well plates (1 × 10⁶ cells/well) coated with 10 μg/mL anti-TCR Cδ mAb in IMDM supplemented with 10% FCS, 50 μM 2-mercaptoethanol, and 10 ng/mL recombinant mouse IL-2 for 5 days at 37°C. Various inhibitors of signaling pathways downstream of TCR were added during this period. Cells were harvested and rested on uncoated plates in the same culture medium without inhibitors for 2 days at 37°C to allow the recovery of TCR expression before restimulation with PMA/ionomycin followed by intracellular cytokine staining.

Optimal concentrations of the inhibitors were predetermined as the maximum concentrations that did not affect viable Vγ3⁺ T-cell yields after the cultures. The following inhibitors were used at the indicated concentrations: calcineurin inhibitor cyclosporin A (0.01 μM; Cell Signaling Technology), MEK1/2-ERK1/2 inhibitor U0126 (5 μM, Cell Signaling Technology), p38 MAPK inhibitor SB203580 (10 μM, Cell Signaling Technology), JNK inhibitor SP600125 (5 μM, Cell Signaling Technology), PKCθ inhibitor sotrastaurin (0.1 μM; Abcam, Cambridge, UK), PI3K inhibitor LY294002 (5 μM, Cell Signaling Technology), mTORC1 inhibitor rapamycin (5 ng/mL, Sigma–Aldrich), and mTORC1/2 inhibitor Torin 1 (0.05 μM, Cell Signaling Technology).

PBS or 10 μg SB203580 in 25 μL PBS was administered by intradermal injections into the dorsal and ventral sides of the ear pinna using a 29-gauge needle under inhalation anesthesia. Epidermal cells were isolated 24 hours after the injection and stimulated for 4 hours with PMA/ionomycin. To activate DETCs in situ by mild tissue abrasion induced by tape-stripping, the stratum corneum was removed from both sides of the earlobe by application and removal of cellophane tape (Scotch™, 3M, St. Paul, MN) seven times (22, 24) 24 hours after the injection. Epidermal cells were isolated 4 hours after tape stripping.

SCENITH™ was performed as previously described (55) using the SCENITH™ kit containing all reagents and protocols (obtained from www.scenith.com/try-it). Briefly, cells were treated on 24-well plates (1 × 10⁶ cells/well) with DMSO (control), 100 mM 2-DG, 1 μM oligomycin, or a combination of 2-DG and oligomycin for 40 minutes (for resting cells) or 30 minutes (for activated cells) at 37°C. Puromycin (10 μg/mL) was added for 40 minutes (for resting cells) or for the last 15 minutes (for activated cells) at 37°C. After surface staining, the cells were fixed and permeabilized using Cytofix/Cytoperm™ and washed and stained in Perm/Wash™ buffer with Alexa Fluor 488-conjugated anti-puromycin mAb (clone R4743L-E8) for 30 minutes at 4°C. The impact of the various metabolic inhibitors was quantified as previously described (55).

Differences between the two groups were evaluated by t-test. Dunnett’s test was used for multiple comparisons to a control. All reported P values are two-tailed, with a P value < 0.05 considered significant. Statistical calculations were performed using JMP™ software (SAS Institute, Cary, NC).