DGKα^−/−, DGKζ^−/−, and ERCre mice were generated as previously described (38, 39, 56). DGKζ^f/f mice were generated by introducing two LoxP sites that flank exons 10–14 of the Dgkz locus (57). TCR transgenic OT2 mice were purchased from the Jackson Laboratory and were cross-bred with DGKα^−/−ζ^f/f ERCre mice to generate DGKα^−/−ζ^f/f OT2 ERCre mice in specific pathogen-free facilities at Duke University Medical Center. The experiments in this study were performed according to a protocol approved by the Institutional Animal Care and Usage Committee of Duke University. DGKα^−/−ζ^f/f or DGKα^−/−ζ^f/f OT2 ERCre mice were intraperitoneally injected with tamoxifen (100 mg/kg body weight) on the first, second, and fifth day to delete DGKζ, and mice were then euthanized for experiments on the eighth day.

Iscove's modified Dulbecco's medium (IMDM) was supplemented with 10% (vol/vol) FBS, penicillin/streptomycin, and 50 μM 2-mercaptoethanol (IMDM-10). Fluorescence-conjugated anti-mouse antibodies CD4 (GK1.5), TCRVα2 (B20.1), CD44 (IM7), CD62L (MEL-14), Thy1.1 (OX-7), Thy1.2 (58-2.1), T-bet (4B10), IFN-γ (XMG1.2), IL-4 (11B11), IL-17A (TC11-18H10.1), and IL-17F (9D3.1C8) were purchased from BioLegend; anti-mouse antibodies for RORγt (AFKJS-9) and Foxp3 (FJK-16s) were purchased from eBioscience. Cell death was determined by Live/Dead Fixable Violet Dead Cell Stain (Invitrogen).

Standard protocols were used to prepare single cell suspensions from the spleen and lymph nodes of mice (in IMDM containing 10% FBS and antibiotics). Red blood cells were lysed using an ACK buffer. Samples were subsequently stained with antibodies in PBS containing 2% FBS and collected on a BD FACSCanto II cytometer. Intracellular staining for T-bet and RORγt was performed using the eBioscience Foxp3 Staining Buffer Set. Intracellular staining for IFNγ, IL-4, IL-17A, and IL-17F was performed using the BD Biosciences Cytofix/Cytoperm and Perm/Wash solutions.

CD4⁺ T cells were purified from the spleen and LN with anti-CD4 microbeads (Miltenyi Biotec) and then were further sorted as naïve CD4⁺CD62L^hiCD44^loCD25⁻. Sorted cells were activated with plate-bound anti-CD3 (5 μg/ml, 1452C11, Bio Xcell) and soluble anti-CD28 (1 μg/ml, PV1, BioXcell) for 4–5 days with various combinations of cytokines and antibodies. For the non-polarizing (T_H0) condition, naïve cells were cultured in the presence of hIL-2 (100 U/ml, Peprotech). For the T_H1 condition, naïve cells were cultured with hIL-2 (100 U/ml), mIL-12 (20 ng/ml, Peprotech), and anti-mIL4 (10 μg/ml, 11B11, Bio Xcell) for 4 days. For the T_H2 condition, naïve cells were polarized in the presence of hIL-2 (100 U/ml), mIL-4 (20 ng/ml, Peprotech), and anti-IFNγ (10 μg/ml, XMG1.2, BioXcell) for 5 days. For the T_H17 condition, naïve cells were cultured with hTGF-β1 (5 ng/ml, Peprotech), mIL-6 (25 ng/ml, Peprotech), anti-mIL4 (10 μg/ml), and anti-IFNγ (10 μg/ml) for 4 days. For iTreg induction, 100 U/ml of hIL-2 and 1 ng/ml TGFβ (Peprotech) were included in the culture for 4 days, followed by intracellular Foxp3 staining. To assess proliferation, sorted naïve CD4⁺ T cells were labeled with CellTrace™ Violet (CTV, ThermoFisher) before cultured in different polarization conditions. For the inhibition assay, 10 μM S6K inhibitor (PF-4708671, Sigma) and 1 nM rapamycin were added to the T_H1 and T_H17 polarizing conditions at the beginning of culture, and cells were cultured for 4 days. At the end of polarizing, cells were stimulated with PMA (50 ng/ml) and ionomycin (500 ng/ml) in the presence of GolgiPlug (1 ng/ml) for 4–5 h. This was followed by cell surface and intracellular staining for appropriated cytokines.

TCRVα2⁺ cells from splenocytes and LN cells for TCR OTII transgenic mice were enriched using MACS magnetic beads and Miltenyi Biotec LS columns. About 100 million cells in 500 μl of IMDM-10 were incubated with the PE-TCRVα2 antibody (1:100 dilution) and then with anti-PE magnetic beads to isolate TCRVα2⁺ cells according to the manufacturer's protocol. Enriched samples were stained with anti-CD4, −CD44, and −CD62L antibodies and sorted on a MoFlo Astrios sorter to obtain viable CD4⁺TCRVα2⁺CD44⁻CD62L⁺ naïve OT2 T cells. Naïve WT or DGKα^−/−ζ^f/f OT2 cells (Thy1.1⁻Thy1.2⁺, 1.5 × 10⁶ cell/mouse) were intravenously injected into sex-matched recipients (Thy1.1⁺Thy1.2⁺). Recipient mice were immunized by subcutaneous injection in the inguinal region with 100 μg/mouse OVA_323−339 peptide emulsified in the CFA 24 h after adoptive transfer and were euthanized to harvest the spleen and drain inguinal lymph nodes on the seventh day after immunization. Splenocytes and dLN cells were stimulated with PMA and ionomycin in the presence of GolgiPlug for 4–5 h or stimulated with 10 μg/ml OVA_323−339 for 2 days in the presence of 1 ng/ml GolgiPlug in the last 5 h. Cell surface and intracellular staining for appropriated cytokines were subsequently performed.

For airway inflammation, OTII T cell recipient mice were intranasally injected with 25 μl of 2.5 mg/ml OVA_323−339 peptide in PBS daily for 3 consecutive days starting 24 h after adoptive transfer. Mice were euthanized on the eighth day after adoptive transfer for collection of BALF. Lungs were fixed in 10% formalin and thin-sectioned for hematoxylin and eosin (H&E) staining. Spleen and draining mediastinal LNs were harvested for cytokine analysis.

Cultured supernatant or BALF samples were appropriately diluted and IFNγ, IL-4, and IL-17A concentrations were determined using Mouse ELISA max kits (BioLegend) according to the manufacturer's instructions.

Cells were lysed in Trizol for RNA preparation. The first strand cDNA was made using the iScript Select cDNA Synthesis Kit (Biorad). Real-time quantitative PCR was conducted using Eppendorf realplex². Expressed levels of target mRNAs were normalized with β-actin and calculated using the 2^−ΔΔCT method. Primers used in this study are listed as following: DGKα Forward: GATGCAGGCACCCTGTACAAT, Reverse: GGACCCATAAGCATAGGCATCT; DGKζ Forward: CGGCTGCCTGGTGTAGACA, Reverse: GCACCTCCAGAGATCCTTGATG; IFN-γ Forward: GCGTCATTGAATCACACCTG, Reverse: TGAGCTCATTGAATGCTTGG; IL-4 Forward: ACAGGAGAAGGGACGCCA, Reverse: GAAGCCCTACAGACGAGCTCA; IL-17A Forward: GCTCCAGAAGGCCCTCAGA, Reverse: CTTTCCCTCCGCATTGACA; Tbx21 Forward: GGTGTCTGGGAAGCTGAGAG, Reverse: GAAGGACAGGAATGGGAACA; GATA-3 Forward: AACCACGTCCCGTCCTACTA, Reverse: AGAGATCCGTGCAGCAGA; RORc Forward: CGACTGGAGGACCTTCTACG, Reverse: TTGGCAAACTCCACCACATA; RORα Forward: CCATGCAAGATCTGTGGAGA, Reverse: CAGGAGTAGGTGGCATTGCT; β-actin Forward: TGTCCACCTTCCAGCAGATGT, Reverse: AGCTCAGTAACAGTCCGCCTAGA.

In vitro-cultured T_H cells were lysed in lysis buffer (1% Nonidet P-40, 150 mM NaCl, 50 mM Tris, pH 7.4) with freshly added protease and phosphatase inhibitors. Samples were subjected to immunoblotting analysis, and probed with anti-pS6 (S235/236), -pErk1/2, -total S6, -total Erk1/2, and β-actin antibodies (Cell Signaling Technology).

Data are presented as mean ± SEM, and statistical significance was determined by two-tailed Student's t-test. The p-values are defined as follows: ^*p < 0.05, ^**p < 0.01, ^***p < 0.001.

DGKα and ζ are dynamically regulated during T cell development and activation (27, 35, 39, 40). We found that DGKα mRNA was decreased in T_H0, T_H1, T_H2, T_H17, and iTregs compared with naïve CD4⁺ T cells. DGKζ mRNA also was decreased in T_H0, T_H1, and T_H17 cells but not in T_H2 and iTregs compared with naïve CD4⁺ T cells (Figure 1A). Both DGKα and ζ appeared more significantly down-regulated in T_H1 and T_H17 conditions than in T_H0 condition. To examine the role of DGKα and ζ in T_H differentiation, WT, DGKα^−/−, and DGKζ^−/− CD44⁻CD62L⁺ naïve CD4⁺ T cells were cultured in T_H1, T_H2, and T_H17 polarization conditions in vitro for 4–5 days. DGKα^−/− or DGKζ^−/− CD4⁺ T cells displayed impaired differentiation to T_H1 cells, which was indicated by decreases of IFN-γ⁺ cells in both percentages and numbers (Figures 1B,C), IFN-γ concentration in culture supernatants (Figure 1F), and IFN-γ mRNA levels (Figure 1G), accompanying the decreased expression of T-bet (Figure 1H). However, total CD4⁺ T cells numbers were increased in the absence of either DGKα or ζ during T_H1 polarization (Figure 1C), suggesting that impaired T_H1 differentiation of DGKα^−/− or DGKζ^−/− CD4⁺ T cells did not result from decreased expansion. In contrast, T_H2 and T_H17 differentiation was not obviously affected by DGKα or ζ deficiency. This was reflected by similar percentages of IL-4⁺ or IL-17⁺ cells (Figures 1B,D,E) and similar levels of IL-4 or IL-17A proteins in culture supernatants (Figure 1F) and mRNAs (Figure 1G), which correlated with comparable expression of GATA-3 or RORγt (Figure 1H). Both DGKα^−/− CD4⁺ T cells and DGKζ^−/− CD4⁺ T cells displayed slightly improved survival under the T_H1 condition and had similar survival rates under T_H2 and T_H17 conditions (Figure 1I), suggesting that their reduced T_H1 responses were not due increased cell death. Together, these data suggested individual DGKα and DGKζ are required for T_H1 differentiation, but are dispensable for T_H2 and T_H17 development in vitro.

DGKα and ζ promote T cell and iNKT cell maturation synergistically in the thymus (52, 54). To determine if DGKα and ζ exert a synergistic role during T_H differentiation, we generated DGKα^−/−ζ^f/f-ERCre (DKO) mice so that both DGKα and ζ were ablated after tamoxifen-induced deletion of DGKζ. In contrast to DGKα or ζ single-knockout T cells, DKO CD4⁺ naïve T cells showed enhanced capacity to differentiate into both T_H1 and T_H17 cells but similar T_H2 differentiation compared with their WT counterparts (Figures 2A,B), coinciding with increased IFN-γ and IL-17A but not IL-4 concentration in culture supernatants (Figure 2C) and IFN-γ and IL-17A mRNA levels in these cells (Figure 2D). DKO CD4⁺ T cells displayed slightly decreased survival rate under T_H1 but similar survival rate under T_H17 polarization conditions, suggesting that their enhanced T_H1 and T_H17 responses were not due to improved survival (Figure 2E). However, under both T_H1 and T_H17 conditions, DKO CD4⁺ T cells proliferated more vigorously than WT controls, which might contribute to their enhanced T_H1 and T_H17 responses (Figure 2F). In contrast to T_H1 and T_H17 differentiation, iTreg cell induction was not obviously different between WT and DKO naïve CD4⁺ T cells (WT iTreg percentages: 62.49 ± 9.186 n = 7; DKO iTreg percentages: 53.84 ± 8.465 n = 7; P = 0.5022). Together, these results indicated that deficiency of both DGKα and ζ promoted T_H1 and T_H17 differentiation with minimal effects on T_H2 or iTreg cell differentiation in vitro.

To further determine the impact of DGKα and ζ double deficiency on T_H differentiation in vivo, we generated DKO mice carrying the OT2 TCR transgene, which recognizes chicken ovalbumin peptide 323-339 (OVA_323−339) in the context of I-A^b (58) and adoptively transferred WT- or DKO-naïve OT2 T cells (Thy1.1⁻Thy1.2⁺CD4⁺TCRVα2⁺) into congenic Thy1.1⁺Thy1.2⁺ recipients. Recipient mice were immunized with OVA_323−339 peptide emulsified in complete Freund's adjuvant (CFA) 1 day after the transfer. Seven days after immunization, donor-derived DKO OT2 T cells were increased in both percentages and numbers in the spleen and draining lymph nodes (dLNs) compared with WT controls (Figures 3A–C). In addition, higher percentages of DKO OT2 T cells expressed IFN-γ, IL-17A, and IL-17F than WT controls following in vitro PMA and ionomycin stimulation for 4 h (Figures 3D–F). Because of increased DKO OT2 T cell numbers, DKO OT2 T_H1 and T_H17 cell numbers were much greater than WT controls in dLNs and particularly in the spleen (Figures 3G,H). Moreover, DKO OT2 T cells contained more IFN-γ-, IL-17A-, and IL-17F-positive cells, which was detected by intracellular staining (Figures 3I,J), and secreted more cytokines to culture supernatants, which was detected by ELISA (Figures 3K,L), than their WT controls following stimulation with OVA_323−339 peptide for 2 days. Together, these results demonstrated that the deficiency of both DGKα- and ζ-enhanced T_H1 and T_H17 polarization and expansion in vivo via cell intrinsic mechanisms.

T_H17 cells promote airway inflammation and hyper-responsiveness via recruiting neutrophils and induce airway smooth muscle contraction, which contributes to the severe form of asthma (59, 60). To determine if dysregulated T_H responses of DKO CD4⁺ T cells impact airway inflammation, we adoptively transferred naïve WT and DKO OT2 cells (Thy1.2⁺) into WT Thy1.1⁺Thy1.2⁺ congenic mice on day −1 and then intranasally injected OVA_323−339 peptide into the recipient mice on days 0, 1, and 2. On the seventh day, we detected at least four-fold more DKO OT2 cells in both percentages and numbers in the draining mediastinal lymph nodes and spleen in recipient mice than their WT counterparts (Figures 4A–C). DKO donor-derived OT2 cells in both dLNs and spleens produced more IL-17A and IL-17F as well as IFN-γ in response to in vitro stimulation with PMA and ionomycin for 4 h (Figures 4D–H) or with OVA_323−339 peptide for 2 days (Figures 4I–M). Concordantly, both IFN-γ and IL-17A levels in bronchoalveolar lavage fluid (BALF) were elevated in recipients with DKO OT2 T cells compared with those with WT OT2 T cells (Figure 5A). Moreover, DKO OT2 cell recipients contained more neutrophils and lymphocytes than those with WT control in BALF (Figures 5B,C) and in interstitial lung tissues that surround the bronchioles (Figure 5D). Together, these results demonstrated that DGKα and ζ deficiencies in CD4⁺ T cells exacerbated airway inflammation, likely as a result of enhanced T_H17 responses to protein allergens.

T-bet, GATA-3, RORγt, and RORα are transcription factors that play critical roles in T_H1, T_H2, and T_H17 differentiation, respectively. Under the T_H1 polarization condition, DKO CD4⁺ T cells expressed higher levels of T-bet at both mRNA and protein levels than WT controls (Figures 6A,B), which was consistent with their elevated T_H1 responses. In contrast, GATA-3 expression in DKO CD4⁺ T cells was not obviously different from WT controls under the T_H2 polarization condition (Figure 6C), consistent with a minimal effect of DKO on T_H2 responses as shown in Figure 2. Interestingly, Rorc (gene encoding RORγt) mRNA levels were obviously decreased in DKO CD4⁺ T cells under the T_H17 polarization condition (Figure 6D), although RORγt protein was only slightly decreased (Figure 6E). In contrast, RORa mRNA levels were increased in DKO CD4⁺ T cells 24 and 36 h after polarization (Figure 6F). Both RORα and RORγt are important for T_H17 differentiation and RORγt is considered the master regulator of the Th17 lineage (61–63). It is intriguing that DGKα and ζ double deficiency enhanced Th17 differentiation yet downregulated RORγt expression. Increased RORα expression in DKO CD4⁺ T cells might partially compensate for the decrease of RORγt. Additionally, DGKαζ deficiency might alleviate the requirement of RORγt and promote T_H17 differentiation via other mechanisms.

DGKα and ζ negatively control DAG-mediated Ras-Erk1/2 activation in thymocytes and naïve T cells following TCR engagement (36, 38, 54). We further examined how DGKα and ζ double deficiency might affect this pathway during T_H polarization. As shown in Figure 7A, Erk1/2 phosphorylation was obviously enhanced in DKO CD4⁺ T cells under T_H0, T_H1, T_H2, and T_H17 conditions, suggesting that DGKα and ζ negatively controlled Erk1/2 activation during effector CD4⁺ T cell differentiation. Previous studies have found that DAG-mediated RasGRP1-Ras-Erk, PI3K-Akt, and PKCθ-CARMA1 pathways participate in TCR-induced mTORC1 activation and DGKα and ζ double deficiency but not DGKα or ζ single deficiency leads to enhanced mTOR signaling in developing thymocytes (36, 64, 65) and that mTOR plays important roles in Th differentiation (65–69). Although, S6 phosphorylation, an mTORC1/S6K1-dependent event, in T_H1 cells appeared unaffected by DGKα and ζ double deficiency, it was obviously increased in DKO CD4⁺ T cells under T_H0, T_H2, and T_H17 polarization conditions, suggesting that DGKα and ζ negatively controlled mTORC1 signaling in T_H0, T_H2, and T_H17 cells. Treatment of WT and DKO CD4⁺ T cells with either rapamycin or the S6K1 inhibitor PF-4708671 caused about 50% reduction of IFNγ⁺ cells in both cell types but DKO CD4⁺ T cells still contained higher percentages of IFNγ⁺ cells than WT controls. Thus, DKO CD4⁺ T cells were partially sensitive to mTORC1/S6K1 inhibition (Figures 7B,C), suggesting that additional mechanisms might contribute to enhanced T_H1 differentiation in these cells. In contrast, T_H17 differentiation of both DKO and WT CD4⁺ T cells was potently inhibited by either rapamycin or PF-4708671 (Figures 7D,E). Although, we could not rule out potential off-target effects of PF-4708671 and rapamycin, our data suggested that enhanced mTORC1/S6K1 signaling might contribute to the elevated T_H17 responses of DKO CD4⁺ T cells.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.