CD8+ T cells that express retinoic acid–related orphan receptor (ROR)γt (TC17 cells) have been shown to promote procarcinogenic inflammation and contribute to a tolerogenic microenvironment in tumors. We investigated their phenotype and functional properties in relationship to the pathogenesis of human distal bile duct cancer (DBDC). DBDC patients had an elevated level of type 17 immune responses and the frequency of CD8+RORγt+ T cells (TC17 cells) was increased in peripheral blood. The CD8+RORγt+ T cells represented a highly activated subset and produced IL-17A in equal amount as CD4+RORγt+ T cells (TH17 cells). Most CD8+RORγt+ T cells coexpressed T-bet, a lineage transcription factor for TH1 and TC1 development, suggesting that CD8+RORγt+ T cells undergo plasticity toward a TC17/1-like phenotype with coproduction of IL-17A and INF-γ. In comparison with CD8+RORγt T cells, the CD8+RORγt+ T cells had a higher level of TCR signaling and were terminally differentiated and exhausted. These cells also had impaired ability to re-express perforin after degranulation and reduced cytotoxic immune function. A subset of CD8+RORγt+ T cells expressing a low level of programmed cell death protein 1 and a high level of OX40 were associated with reduced patient survival. In conclusion, CD8+RORγt+ T cells are proinflammatory and functionally impaired and may contribute to the pathogenesis of DBDC.

I nfiltrating the tumor microenvironment in cancer patients, CD8+ T cells play pivotal, but also paradoxical, roles in the antitumor immunity that ultimately affects the clinical outcome (1). Among tumor-infiltrating CD3+ T cells, CD8+ T cells (CTLs or TC cells) are crucial in the antitumor immune responses, and the presence of CD8+ T cells is mostly associated with improved patient survival (1).

Distal bile duct cancer (DBDC) is a devastating disease with a poor prognosis that originates from the extrahepatic biliary tract epithelium (2). In DBDC, CD8+ T cells, CD4+ T cells, CD4+FOXP3+ regulatory T cells (Tregs) and macrophages infiltrate the tumor tissue. However, most of the T cells are CD8+ T cells, and despite the high level of tumor-infiltrating CD8+ T cells in DBDC patients, the tumor cells often persist and metastasize (3). At present, the lineage-specific phenotype and functional relevance of CD8+ T cells in DBDC has not yet been fully revealed.

The expression of the type 17 master regulator retinoic acid–related orphan receptor (ROR)γt regulates the generation of TC17 (CD8+) cells and TH17 (CD4+) cells and is characterized by the production of the proinflammatory cytokine IL-17A (4). Although most of the circulating IL-17A–producing T cells belong to TH17 cells that are involved in the pathogenesis of various human cancers, recent studies also suggest that TC17 cells are involved in the pathogenesis of several human malignancies and particularly in upper gastrointestinal cancers (57). The differentiation of TC17 cells is facilitated by tumor-activated monocytes that secrete IL-1β, IL-6, TGF-β, and IL-23 (6, 7). Consumption of IL-2 by FOXP3+ Tregs also promotes the differentiation of TC17 cells (8). The TC17 cells are less cytotoxic (9, 10), and the secretion of IL-17 appears to promote the recruitment of myeloid-derived suppressor cells (7). This suggests that TC17 cells not only contribute to inflammation, but they also impede antitumor immunity.

The aim of this study was to provide a comprehensive characterization of CD8+ T cells in DBDC patients with well-defined anatomical and histological cancer subtypes. We found that DBDC patients have increased type 17 responses and a high frequency of CD8+RORγt+ T cells with an activated phenotype. The CD8+RORγt+ T cells coexpressed RORγt and T-bet and have a distinct cytokine profile. In vitro functional analysis revealed that the CD8+RORγt+ T cells represented a subset with a hyperactivated TCR signaling signature that resulted in enhanced proliferation and a strong proinflammatory phenotype, but with an impaired ability to sustain cytotoxic immune responses. We conclude that these cells likely contribute to DBDC progression by enhancing the proinflammatory immune response combined with inefficient antitumor immunity.

The study was approved by the Regional Ethics Committee of the South-Eastern Norway Regional Health Authority. Peripheral blood samples were collected from 53 untreated patients preoperatively under patient consent, and 17 of the samples showed pathologically confirmed distal bile duct cancer with pancreaticobiliary histology. Clinical characteristics of all DBDC patients are described in Table I. Patients with other tumor origin, histology, and advanced tumors were excluded from this analysis. Blood from age-matched healthy donors (HDs; six male and six female donors; median age, 66 y; range, 65–68 y) were obtained from Oslo University Hospital Blood Center (Oslo, Norway). PBMCs were isolated by Ficoll density gradient centrifugation (Axis-Shield). PBMCs were cryopreserved in FCS with 10% DMSO.

Table I.
The clinicopathological characteristics and immunological parameters of 17 patients with DBDC
ParametersNo. of Patients
Number (n17 
Age (y) at diagnosis, median (range) 69 (53–81) 
Sex (male/female) 8/9 
Tumor grade (II/III) 14/3 
Tumor size (cm, <25/≥25) 14/3 
Histological type (non-/pancreaticobiliary) 0/17 
Tumor (T) invasion (T1 + T2/T3 + T4) 0/17 
Lymphoid nodal (N) status (N0 + N1/N2 + N3) 17/0 
Distance metastasis (M) status (M0/M1) 16/1 
TNM stage (I + II/III + IV) 16/1 
Lymphatic/vascular invasion (absent/present) 5/12 
Neoadjuvant/adjuvant therapy 0/0 
CA19-9 (U/ml) 177 (5–1963) 
TC17 cell percentagea (median, range) 38.4 (25.3–55.4) 
ParametersNo. of Patients
Number (n17 
Age (y) at diagnosis, median (range) 69 (53–81) 
Sex (male/female) 8/9 
Tumor grade (II/III) 14/3 
Tumor size (cm, <25/≥25) 14/3 
Histological type (non-/pancreaticobiliary) 0/17 
Tumor (T) invasion (T1 + T2/T3 + T4) 0/17 
Lymphoid nodal (N) status (N0 + N1/N2 + N3) 17/0 
Distance metastasis (M) status (M0/M1) 16/1 
TNM stage (I + II/III + IV) 16/1 
Lymphatic/vascular invasion (absent/present) 5/12 
Neoadjuvant/adjuvant therapy 0/0 
CA19-9 (U/ml) 177 (5–1963) 
TC17 cell percentagea (median, range) 38.4 (25.3–55.4) 
a

TC17 cell percentage was obtained by gating on CD8+RORγt+ T cells in CD8+ T cells.

Prior to immunophenotyping and cell purification, PBMCs were thawed and rested overnight at 37°C with 5% CO2 in RPMI 1640 with GlutaMAX supplemented with 10% FCS, 1% penicillin-streptomycin (Life Technologies), 1% sodium pyruvate, and 1% minimum nonessential amino acids. The CD8+ T cells were purified using human CD8 MicroBeads and a MACS column (Miltenyi Biotec) according to the manufacturer’s instructions. For cytokine-specific CD8+ T cell expansion, bead-sorted CD8+ T cells were stimulated with PMA (50 ng/ml) and ionomycin (1 μg/ml) for 4 h, and the IL-17A–secreting cells were labeled with a IL-17A cell enrichment and detection kit (PE) (Miltenyi Biotec). The IL-17A+/−CD8+ T cells were sorted using a BD FACSAria IIu and were expanded in complete RPMI 1640 containing CD3 (1 μg/ml), CD28 (1 μg/ml), IL-2 (10 ng/ml), and allogeneic irradiated PBMCs for 14 d as described earlier (11).

All surface markers were stained first and then the intracellular markers were stained after fixation and permeabilization using a human FOXP3 buffer set (BD Biosciences). The bead-sorted CD8+ T cells (5 × 105 cells/well in 96-well plates) were incubated with anti-CD2/CD3/CD28–coated beads (T cell activation/expansion kit; Miltenyi Biotec) in a 1:5 bead/cell ratio for 24 or 96 h, the culture supernatants were collected for multiplex cytokine assays, and the cells were analyzed by multicolor flow cytometry. For intracellular cytokine staining, the PBMCs were stimulated for 4 h with PMA (50 ng/ml) and ionomycin (1 μg/ml) and brefeldin A (10 μg/ml). Dead cells were excluded with fixable viability stain 700 or 7-aminoactinomycin D (7-AAD; BD Biosciences). The apoptotic cells were analyzed by staining with annexin binding buffer that contained annexin V, 7-AAD (BD Biosciences), and RNAse (10 μg/ml) for 15 min at room temperature. All multicolor flow cytometry data were acquired with a BD LSRFortessa and analyzed using FlowJo version 10 (Tree Star).

PBMCs (1 × 106 cells/well in 96-well plates) were stimulated with biotinylated anti-CD2 (5 μg/ml), anti-CD28 (5 μg/ml), anti-CD3 (1 μg/ml), and avidin (25 μg/ml). Thereafter, cells were fixed with prewarmed BD Phosflow fix buffer I (BD Biosciences) and barcoded three-dimensionally with Pacific Blue, Pacific Orange, and Alexa Fluor 488 (Life Technologies) as described earlier (12). The barcoded cells were permeabilized with BD Phosflow perm buffer III (BD Biosciences) and stained for indicated phosphorylation site-specific Abs. The flow cytometry data were acquired with a BD LSRFortessa and analyzed using Cytobank. Data were presented as the arcsinh ratio of medians as previously described (13).

The anti-CD3 (clone OKT3) and humanized anti-HER2 (trastuzumab [Herceptin]) was heteroconjugated and the bispecific Ab (BsAb) dimer was purified as described earlier by gel filtration chromatography (14). HER2+ SKBR-3 cells were used as targets, and the cytotoxicity was measured using a Cytotoxicity Detection KitPLUS (lactate dehydrogenase) (Promega). Bead-sorted total CD8+ T cells were activated for 4 d with plate-bound CD3 (10 μg/ml), soluble CD28 (1 μg/ml), and IL-2 (10 ng/ml), and the expanded IL-17A+ and IL-17A expanded CD8+ T cells were used as effector cells. In a 3-d incubation, effector cells were cocultured (5 × 104 cells/well in 96-well plates) with target cells at a 5:1 ratio in 200 μl of complete RPMI 1640 containing CD3/HER2 BsAb (100 ng/ml) at 37°C with 5% CO2, following which 50 μl of supernatant was used for a lactate dehydrogenase–release assay according to the manufacturer’s instructions. The cell lysis was calculated as specific cytotoxicity (%) = [(experimental release − effector spontaneous release − target spontaneous release)/(target maximum release − target spontaneous release)] × 100. Each assay was performed in triplicates.

Plasma and culture supernatants were stored at −80°C until analyzed. The indicated cytokines from plasma were quantified using a high-sensitivity magnetic Luminex high-performance assay multiplex kit (R&D Systems) and Bio-Plex manager software version 6.1 (Bio-Rad Laboratories) according to the manufacturers’ instructions.

Abs used for phenotyping anti-CD3 PerCP-Cy5.5 (UCHT1), anti-CD8 PerCP-Cy5.5 (RPA-T8), anti-CD8 Brilliant Violet 786 (RPA-T8), anti-CD8 PE-Cy7 (RPA-T8), anti-CD4 allophycocyanin-H7 (RPA-T4), anti-CD45RA Brilliant Violet 510 (HI100), anti-CD127 Brilliant Violet 786 (HIL-7R-M21), anti-ICOS PE (DX29), anti–HLA-DR PE (TU36), anti-CXCR3 PE-Cy7 (1C6/CXCR3), anti-CCR6 allophycocyanin (11A9), anti-CD38 allophycocyanin (HIT2), anti–IL-17A Brilliant Violet 421 (N49-653), anti–INF-γ PE-Cy7 (B27), anti–Ki-67 Brilliant Violet 421 (B56), anti–Ki-67 PE-Cy7 (B56), anti–granzyme B PE (GB11), anti–granzyme B Brilliant Violet 421 (GB11), anti-CD28 PerCP-Cy5.5 (CD28.2), anti–GATA-binding protein 3 (GATA3) Alexa Fluor 647 (L50-823), anti-CD107a FITC (H4A3), anti-CD69 PE (FN50), annexin V V450, 7-AAD, and the human FOXP3 buffer set were from BD Biosciences. Anti-perforin PE-Cy7 (delta G9), anti-CD161 PerCP-Cy5.5 (HP-3G10), and anti-RORγt PE (AFKJS‐9) were from eBioscience. Anti-RORγt PerCP (600380) and anti–lymphocyte activation gene-3 (LAG-3) allophycocyanin (FAB2319A) were from R&D Systems. Anti–latency-associated peptide (LAP) PE (TW4-6H10), anti-TNFRII PE (3G7A02), anti–programmed cell death protein 1 (PD-1) PE-Cy7 (EH12.2H7), anti-CD103 PE-Cy7 (Ber-ACT8), anti-OX40 PE-Cy7 (Ber-ACT35), anti-CCR7 PE-Cy7 (G043H7), anti–CTLA-4 allophycocyanin (L3D10), anti-CCR4 allophycocyanin (L291H4), anti–T cell Ig and mucin domain-3 (Tim-3) allophycocyanin (F38-2E2), anti–T-bet Brilliant Violet 421 (4B10), and anti–IL-4 allophycocyanin (8D4-8) were from BioLegend. Phosphorylation site-specific Abs anti-CD3ζ (pY142) (K25-407.69) Alexa Fluor 647, anti-ZAP70/Syk (pY319/Y352) (17A/ZAP-70 pY319/Syk pY352) Alexa Fluor 647, anti–linker for activation of T cells (pY171) (I58-1169) Alexa Fluor 647, anti-SLP76 (pY128) (J141-668.36.58) Alexa Fluor 647, anti–NF-κB (pS529) (K10-895.12.50) Alexa Fluor 647, anti-MEK1 (pS298) (J114-64) Alexa Fluor 647, anti–retinoblastoma protein (pS807/811) (J112-906) Alexa Fluor 647, and IgG1κ isotype control (MOPC-21) Alexa Fluor 647 were purchased from BD Biosciences. Anti-Akt (pS473) (D9E) Alexa Fluor 647, anti–NF-κB (pS536) (93H1) Alexa Fluor 647, anti-Erk1/2 (pT202/Y204) (E10) Alexa Fluor 647, anti–p38 MAPK (pT180/Y182) (28B10) Alexa Fluor 647, and anti-S6 ribosomal protein (pS235/236) (D57.2.2E) Alexa Fluor 647 were purchased from Cell Signaling Technology. Unconjugated anti-CaMKII (pT286) (polyclonal) was from Cell Signaling Technology, and anti-NFATc2 (pS213/217/221) (polyclonal) was from Santa Cruz Biotechnology. Secondary anti-rabbit IgG (polyclonal) Alexa Fluor 647 and avidin were from Life Technologies. Anti-CD28 (biotin) (CD28.2) and anti-CD2 (biotin) (RPA-2.10) was from eBioscience. Anti-CD3 (biotin) (OKT3) and anti-CD3 (OKT3) was from Diatec. Anti-HER2 (Herceptin) was from Roche.

The statistical significance of differences (p values) between two groups was calculated by the parametric (Student t test) or nonparametric (Mann–Whitney or Wilcoxon signed rank) tests for the respective paired and unpaired groups. Error bars are expressed as mean ± SEM. The horizontal line in scatter plots was expressed as median, and each dot represented one donor. Correlations between parameters were calculated using the Spearman correlation analysis and linear regression analysis as appropriate. Cumulative survival time was calculated using the Kaplan–Meier method, and survival was calculated in months; the log-rank test was used to compare between two groups (SPSS statistical software, version 13.0). All data were calculated as two-tailed tests, and p ≤ 0.05 was considered statistically significant. All statistical analyses were performed using GraphPad Prism version 6 unless specified otherwise.

First, we analyzed the frequency of CD8+ T cells in PBMCs of DBDC patients, which revealed no difference compared with that of HDs (Fig. 1A). Recent studies suggest that the presence of IL-17A–producing CD8+ T cells (TC17 cells) fuel the progression of various gastrointestinal cancers that are anatomically close to DBDC (57). Therefore, we analyzed the expression of the TC17 lineage–specific transcription factor RORγt in the CD8+ T cells and found that the frequency of CD8+ T cells that expressed RORγt was substantially higher in PBMCs from DBDC patients and that the non-RORγt CD8+ T cell population was decreased compared with that of HDs (Fig. 1A, Supplemental Fig. 1). Next, we analyzed the expression of markers for cellular activation and found that HLA-DR, CD38, and Ki-67 were all upregulated in CD8+RORγt+ T cells compared with the levels in CD8+RORγt T cells from PBMCs of DBDC patients (Fig. 1B). Next, we found that the circulating levels of the cytokines IL-1β and IL-6 that promote TC17 differentiation were elevated in DBDC patients (Fig. 1C). Additionally, the circulating level of type 17 effector cytokines IL-17A and TNF-α were also elevated in DBDC patients, where IL-6 and TNF-α are also associated with type 2 responses. However, no changes were observed in the levels of type 1 response effector cytokines INF-γ and IL-2 (Fig. 1C) (4). Type 17 and type 2 responses are procarcinogenic in many solid tumors (1, 15), and we observed that an increased percentage of circulating CD8+RORγt+ T cells correlated with the cell types that contribute to type 2 and type 17 responses such as CD4+RORγt+ T cells (TH17), CD4+GATA3+ T cells (TH2), and CD8+GATA3+ T cells (TC2), but not the type 1 response–related cells such as CD4+T-bet+ T cells (TH1) and CD8+T-bet+ T cells (TC1) and the suppressive cell type CD4+FOXP3+CD127 T cells (Tregs) (Supplemental Fig. 2). Recently, it has been demonstrated that CD8+RORγt+ T cells have a high plasticity toward expression of type 1 master regulator T-box transcription factor TBX21 (T-bet) that regulates the generation of TC1 (CD8+) cells and TH1 (CD4+) cells. The CD8+RORγt+T-bet+ T cells (TC17/1 cells) are present in various autoimmune diseases and gastrointestinal cancers (16). We analyzed the expression of T-bet in CD8+RORγt+ T cells in the DBDC patients and found that >85% of CD8+RORγt+ T cells also expressed T-bet, which was significantly higher than in the CD8+RORγt T cells from PBMCs of DBDC patients (Fig. 1D, left panel). Similarly, the expression of type 2 master regulator GATA3, which regulates the generation of TC2 (CD8+) cells and TH2 (CD4+) cells was also significantly increased in CD8+RORγt+ T cells compared with that of the CD8+RORγt T cells in DBDC patients (Fig. 1D, right panel). Furthermore, coexpression analyses revealed that the T-bet–expressing CD8+RORγt+ T cells were mostly negative for GATA3 expression (Fig. 1E), whereas most GATA3-expressing CD8+RORγt+ T cells did not exist as a separate population and were mostly coexpressing T-bet (Fig. 1E). The expression of T-bet appeared to be much higher than the expression of GATA3 in the CD8+RORγt+ T cells, indicating that these cells preferentially exist as TC17/1 cells. Additionally, we observed that the Treg master regulator FOXP3 was not expressed in the CD8+RORγt+ T cells (Fig. 1F). Taken together, these results suggest that DBDC patients have an elevated type 17 response, and, correspondingly, 25.3–55.4% of the CD8+ T cells (Table I), from peripheral blood of DBDC patients were highly activated and had differentiated into CD8+RORγt+ T cells (TC17 cells) with potential plasticity toward a TC17/1-like phenotype.

FIGURE 1.

Increased frequency of CD8+RORγt+ T cells in peripheral blood of DBDC patients. (A) A representative flow cytometry dot plot and compiled frequencies of total CD8+ T cells, CD8+RORγt+ T cells, and CD8+RORγt T cells in PBMCs from HDs and DBDC patients. (B) Compiled frequencies of HLA-DR+CD38+ and Ki-67+ in CD8+RORγt T cells and CD8+RORγt+ T cells in PBMCs from DBDC patients. (C) Cytokine levels in plasma from HD and DBDC patients. (D) A representative flow cytometry dot plot and compiled frequencies of T-bet (left panel) and GATA3 (right panel) expression in CD8+RORγt T cells and CD8+RORγt+ T cells in PBMCs from DBDC patients. (E) A representative flow cytometry dot plot and compiled frequencies show the coexpression of T-bet in CD8+RORγt+GATA3+ T cells and GATA3 in CD8+RORγt+T-bet+ T cells from DBDC patients. (F) A representative flow cytometry dot plot and compiled frequencies show the expression of CD127 and FOXP3 in CD8+RORγt+ T cells and CD4+ T cells from DBDC patients. HDs (n = 12) and DBDC patients (n = 17). Error bars represent mean ± SEM. Horizontal bar represents median, and each dot represents one patient. *p ≤ 0.05, ***p ≤ 0.001, ****p ≤ 0.0001. ns, not significant.

FIGURE 1.

Increased frequency of CD8+RORγt+ T cells in peripheral blood of DBDC patients. (A) A representative flow cytometry dot plot and compiled frequencies of total CD8+ T cells, CD8+RORγt+ T cells, and CD8+RORγt T cells in PBMCs from HDs and DBDC patients. (B) Compiled frequencies of HLA-DR+CD38+ and Ki-67+ in CD8+RORγt T cells and CD8+RORγt+ T cells in PBMCs from DBDC patients. (C) Cytokine levels in plasma from HD and DBDC patients. (D) A representative flow cytometry dot plot and compiled frequencies of T-bet (left panel) and GATA3 (right panel) expression in CD8+RORγt T cells and CD8+RORγt+ T cells in PBMCs from DBDC patients. (E) A representative flow cytometry dot plot and compiled frequencies show the coexpression of T-bet in CD8+RORγt+GATA3+ T cells and GATA3 in CD8+RORγt+T-bet+ T cells from DBDC patients. (F) A representative flow cytometry dot plot and compiled frequencies show the expression of CD127 and FOXP3 in CD8+RORγt+ T cells and CD4+ T cells from DBDC patients. HDs (n = 12) and DBDC patients (n = 17). Error bars represent mean ± SEM. Horizontal bar represents median, and each dot represents one patient. *p ≤ 0.05, ***p ≤ 0.001, ****p ≤ 0.0001. ns, not significant.

Close modal

The expression of RORγt is essential for the production of IL-17A, whereas T-bet is required for INF-γ production in both CD4+ T cells and CD8+ T cells (1719). We measured the production of these proinflammatory cytokines from CD8+RORγt+ T cells by intracellular cytokine staining after 4 h of in vitro stimulation of PBMCs from DBDC patients (Fig. 2A, 2B, Supplemental Fig. 3). CD8+RORγt+ T cells strictly produced IL-17A whereas CD8+RORγt T cells did not (Fig. 2A, 2B). Furthermore, INF-γ was expressed at significantly higher levels by CD8+RORγt+ T cells than by the CD8+RORγt T cells (Fig. 2A, 2B). The latter is likely due to the increased expression of T-bet in CD8+RORγt+ T cells (Fig. 1D, left panel). In contrast, the expression of the anti-inflammatory cytokine IL-4 was significantly lower in the CD8+RORγt+ T cells (Fig. 2A, 2B). Because most of the CD8+RORγt+GATA3+ T cells that could potentially secrete IL-4 also expressed T-bet (Fig. 1E), the T-bet expression likely antagonized the functions of GATA3 and therefore may have contributed to the observed reduced expression of IL-4 (20). Furthermore, we found that without stimulation the basal level of expression of the membrane-tethered LAP (TGF-β), which also promotes IL-17–mediated proinflammatory responses (21), was strongly upregulated in CD8+RORγt+ T cells compared with that of CD8+RORγt T cells (Fig. 2A, 2B). TGF-β is also associated with FOXP3 expression and suppressive function in CD4+FOXP3+ T cells (Tregs) (21); however, as shown earlier, we did not detect any FOXP3 expression in the CD8+RORγt+ T cells (Fig. 1F). In humans, most of the circulating IL-17A–producing T cells belong to the CD4+RORγt+ T cell compartment (TH17 cells) (22). Therefore, we analyzed the cytokine profile of CD8+RORγt+ T cells compared with that of CD4+RORγt+ T cells. The expression of RORγt was similar in CD8+ T cells as in CD4+ T cells (data not shown), and the corresponding IL-17A production was also at the same level in both subsets (Fig. 2C, left panel). However, most of the IL-17A–producing CD8+RORγt+ T cells coproduced INF-γ, whereas the IL-17A–producing CD4+RORγt+ T cells did not (Fig. 2C, right panel), again indicating a functional plasticity of TC17/1-like phenotype in CD8+RORγt+ T cells in DBDC patients. Next, we analyzed tissue-homing receptors associated with the expression of RORγt and T-bet in the CD8+ T cells and CD4+ T cells (23). The RORγt-associated homing receptors CD161 and CD103 were expressed at a similar level in CD8+RORγt+ T cells and in CD4+RORγt+ T cells (Fig. 2D, left panels), whereas the expression of the key homing receptors CCR6 and CCR4 (4) was largely absent in CD8+RORγt+ T cells but present in CD4+RORγt+ T cells (Fig. 2D, middle panels). In contrast, CD8+RORγt+ T cells expressed high levels of T-bet–associated homing receptors CXCR3 and CCR5 (Fig. 2D, right panels) (24). Taken together, these results demonstrate that CD8+RORγt+ T cells from DBDC patients are substantially different from CD4+RORγt+ T cells in terms of production of proinflammatory cytokines and in the expression of homing receptors.

FIGURE 2.

CD8+RORγt+ T cells differ from CD4+RORγt+ T cells in their chemokine receptor expression. (A) A representative flow cytometry dot plot and (B) compiled frequencies of intracellular cytokine level of IL-17A, INF-γ, and IL-4 after 4 h of stimulation and LAP (TGF-β) expression on the surface of unstimulated cells from CD8+RORγt T cells and CD8+RORγt+ T cells in PBMCs from DBDC patients. (C) Compiled frequencies of intracellular cytokine levels of IL-17A and the level of coproduction of IL-17A and INF-γ after 4 h of stimulation in CD4+RORγt+ T cells and CD8+RORγt+ T cells in PBMCs from DBDC patients. (D) Compiled frequencies of surface expression of CD161, CD103, CCR6, CCR4, CXCR3, and CCR5 homing receptors on CD4+RORγt−/+ T cells and CD8+RORγt−/+ T cells in PBMCs from DBDC patients. DBDC patients (n = 17). Horizontal bar represents median, and each dot represents one patient. *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001. ns, not significant.

FIGURE 2.

CD8+RORγt+ T cells differ from CD4+RORγt+ T cells in their chemokine receptor expression. (A) A representative flow cytometry dot plot and (B) compiled frequencies of intracellular cytokine level of IL-17A, INF-γ, and IL-4 after 4 h of stimulation and LAP (TGF-β) expression on the surface of unstimulated cells from CD8+RORγt T cells and CD8+RORγt+ T cells in PBMCs from DBDC patients. (C) Compiled frequencies of intracellular cytokine levels of IL-17A and the level of coproduction of IL-17A and INF-γ after 4 h of stimulation in CD4+RORγt+ T cells and CD8+RORγt+ T cells in PBMCs from DBDC patients. (D) Compiled frequencies of surface expression of CD161, CD103, CCR6, CCR4, CXCR3, and CCR5 homing receptors on CD4+RORγt−/+ T cells and CD8+RORγt−/+ T cells in PBMCs from DBDC patients. DBDC patients (n = 17). Horizontal bar represents median, and each dot represents one patient. *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001. ns, not significant.

Close modal

TCR/CD28 costimulation and subsequent activation of downstream signaling pathways elicit CD8+ T cell effector functions (2527). Because we found that the CD8+RORγt+ T cells from DBDC patients are highly activated based on the phenotypic markers (Fig. 1B), we investigated the level of TCR signaling by phosphoflow cytometry in CD8+RORγt+ T cells and compared the signaling activity to that of CD8+RORγt T cells in PBMCs from DBDC patients (as described in 2Materials and Methods). The basal phosphorylation level of TCR signaling proteins in CD8+RORγt+ T cells was higher than in CD8+RORγt T cells (Fig. 3A, left panel). Next, we activated the cells by TCR/CD2/CD28 stimulation for 1, 3, and 5 min and analyzed the phosphorylation level of TCR signaling proteins in CD8+RORγt+ and CD8+RORγt T cells. The phosphorylation level of the proximal TCR signaling adaptor proteins (CD3ζ, ZAP70/Syk, linker for activation of T cells, and SLP76), NF-κB, MAPKs (MEK1, ERK1/2, and p38), and proliferation pathways (Akt, retinoblastoma protein, ribosomal protein S6) were significantly increased in CD8+RORγt+ T cells compared with CD8+RORγt T cells (Fig. 3A, 3B). However, the activation of the calcium pathway (calcium-dependent protein kinase II and NFATc2) was found to be similar in both CD8+RORγt+ T cells and CD8+RORγt T cells (Fig. 3B). Next, we investigated whether the enhanced TCR signaling in CD8+RORγt+ T cells was associated with cell proliferation. The total CD8+ T cells were purified from DBDC patients and stimulated in vitro by TCR/CD2/CD28 stimulation. We found that CD8+RORγt+ T cells strongly upregulated the expression of CD69 after 24 h of stimulation compared with that of CD8+RORγt T cells (Fig. 3C), and they subsequently proliferated vigorously as measured by the expression of Ki-67 after 96 h of stimulation (Fig. 3D). A second signal through the CD28 receptor is essential for T cell proliferation; however, the expression of CD28 on CD8+ T cells is downregulated in various chronic inflammatory conditions (28). Therefore, we investigated whether the relative hypoproliferation of CD8+RORγt T cells compared with the hyperproliferation of CD8+RORγt+ T cells was due to the altered expression of CD28 receptor, but we did not observe any difference in CD28 expression between the two subsets (data not shown). Taken together, our results demonstrated that the CD8+RORγt+ T cells appear to hyperproliferate due to enhanced TCR signaling.

FIGURE 3.

CD8+RORγt+ T cells feature enhanced TCR signaling and proliferation in DBDC patients. (A) A representative heat map (left panel) shows phosphorylation status of indicated TCR signaling proteins in CD8+RORγt+ T cells at the basal level when normalized to CD8+RORγt T cells in PBMCs from DBDC patients. The heat map in the right panel shows the TCR signaling in CD8+RORγt T cells and CD8+RORγt+ T cells after stimulation with anti-CD2 (5 μg/ml), anti-CD28 (5 μg/ml), and anti-CD3 (1 μg/ml) for 1–5 min and their respective unstimulated controls used to normalize the data (as described in 2Materials and Methods). (B) Quantified arcsinh ratio of medians shown as bar graph. DBDC patients (n = 10). (C) Representative overlaid flow histograms (left panel) and their compiled frequencies (right panel) show the expression levels of CD69 in CD8+RORγt T cells and CD8+RORγt+ T cells from purified CD8+ T cells that were stimulated for 24 h with anti-CD2/CD3/CD28–coated beads. (D) Representative overlaid flow histograms (left panel) and compiled frequencies (right panel) show the expression levels of Ki-67 in CD8+RORγt T cells and CD8+RORγt+ T cells from purified CD8+ T cells that were stimulated for 96 h with anti-CD2/CD3/CD28–coated beads. DBDC patients (n = 8). Error bars represent mean ± SEM. Horizontal bar represents median, and each dot represents one patient. *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001. CaMKII, calcium-dependent protein kinase II; LAT, linker for activation of T cells; ns, not significant; Rb, ribosomal protein; SLP76, SH2 domain containing leukocyte protein of 76 kDa.

FIGURE 3.

CD8+RORγt+ T cells feature enhanced TCR signaling and proliferation in DBDC patients. (A) A representative heat map (left panel) shows phosphorylation status of indicated TCR signaling proteins in CD8+RORγt+ T cells at the basal level when normalized to CD8+RORγt T cells in PBMCs from DBDC patients. The heat map in the right panel shows the TCR signaling in CD8+RORγt T cells and CD8+RORγt+ T cells after stimulation with anti-CD2 (5 μg/ml), anti-CD28 (5 μg/ml), and anti-CD3 (1 μg/ml) for 1–5 min and their respective unstimulated controls used to normalize the data (as described in 2Materials and Methods). (B) Quantified arcsinh ratio of medians shown as bar graph. DBDC patients (n = 10). (C) Representative overlaid flow histograms (left panel) and their compiled frequencies (right panel) show the expression levels of CD69 in CD8+RORγt T cells and CD8+RORγt+ T cells from purified CD8+ T cells that were stimulated for 24 h with anti-CD2/CD3/CD28–coated beads. (D) Representative overlaid flow histograms (left panel) and compiled frequencies (right panel) show the expression levels of Ki-67 in CD8+RORγt T cells and CD8+RORγt+ T cells from purified CD8+ T cells that were stimulated for 96 h with anti-CD2/CD3/CD28–coated beads. DBDC patients (n = 8). Error bars represent mean ± SEM. Horizontal bar represents median, and each dot represents one patient. *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001. CaMKII, calcium-dependent protein kinase II; LAT, linker for activation of T cells; ns, not significant; Rb, ribosomal protein; SLP76, SH2 domain containing leukocyte protein of 76 kDa.

Close modal

To assess the effect of chronic Ag stimulation through the TCR, we analyzed the coexpression of the lymphoid chemokine receptor CCR7 and the memory marker CD45RA (29, 30). These markers identify four subsets of CD8+ T cells: naive cells (CCR7+CD45RA+), central memory (CM) cells (CCR7+CD45RA), effector memory (EM) cells (CCR7CD45RA), and effector memory cells that re-express CD45RA (EMRAs; CCR7CD45RA+). The EMRAs constitute terminally differentiated cells (31), and we found that >80% of the CD8+RORγt+ T cells had differentiated to an EMRA phenotype (Fig. 4A). The EMRA subset was significantly higher in CD8+RORγt+ T cells compared with CD8+RORγt T cells from PBMCs of DBDC patients (Fig. 4A). In the absence of TCR activation, the differentiation of EMRAs in humans is also induced by homeostatic proliferation that requires the common γ-chain cytokines IL-7 and IL-15 (31). However, we found that the circulating levels of IL-7 and IL-15 in DBDC patients were similar to those in HDs (Fig. 4B), which indicates that the acquisition of a terminally differentiated phenotype in CD8+RORγt+ T cells is due to TCR activation rather than homeostatic proliferation in the DBDC patients. In humans, expression of CTLA-4, PD-1, LAG-3, and Tim-3 in CD8+ T cells marks their activation and differentiation status (32), and it is also associated with exhaustion (33). The expression of CTLA-4 and PD-1 was similar in the CD8+RORγt+ T cell and in the CD8+RORγt T cell compartment, whereas the expression of LAG-3 and Tim-3 was upregulated in CD8+RORγt+ T cells compared with CD8+RORγt T cells (Fig. 4C). Furthermore, our results confirmed that the expression of CTLA-4, PD-1, LAG-3, and Tim-3 was higher in the total CD8+ T cell population in PBMCs of DBDC compared with that of the HDs (data not shown). These results indicated that the CD8+RORγt+ T cells are terminally differentiated and more exhausted than are CD8+RORγt T cells from DBDC patients. Additionally, we found that CD8+RORγt+ T cells strongly downregulated the prosurvival markers CD127 (IL-7R α-chain) (33, 34), OX40 (costimulation receptor) (35), and upregulated TNFRII, which promotes activation-induced cell death in CD8+ T cells (Fig. 4D) (35, 36). We found that the purified CD8+ T cells from DBDC patients to a high degree underwent apoptosis in vitro in the absence of TCR and costimulation (Fig. 4E) and that this predominantly occurred in the CD8+RORγt+ T cell compartment (10.95% median level) (Fig. 4F) compared with direct ex vivo analysis of PBMCs from DBDC patients (38.4% median level) (Fig. 1A). However, in vitro stimulation of purified CD8+ T cells through TCR/CD2/CD28 increased the percentage of CD8+RORγt+ T cells (36.9% median level) compared with CD8+RORγt T cells (Fig. 4F). This is likely due to the enhanced proliferative activity of CD8+RORγt+ T cells associated with higher TCR signaling activity (Fig. 3D). Taken together, these results suggest that CD8+RORγt+ T cells from DBDC patients are highly activated but terminally differentiated, exhausted, and prone to apoptosis without stimulation.

FIGURE 4.

CD8+RORγt+ T cells are terminally differentiated and exhausted in DBDC patients. (A) Representative flow cytometry dot plots (left panel) and compiled frequencies (right panel) show the naive cells, CM cells, EM cells, and EMRAs, defined according to the expression of CCR7 and CD45RA in CD8+RORγt T cells and CD8+RORγt+ T cells in PBMCs from DBDC patients (n = 17). (B) Circulating level of IL-7 and IL-15 in plasma from HDs and DBDC patients. HDs (n = 12) and DBDC patients (n = 17). (C) Compiled frequencies of expression of activation- and exhaustion-related markers (CTLA-4, PD-1, LAG-3, Tim-3) on CD8+RORγt T cells and CD8+RORγt+ T cells from DBDC patients. (D) Compiled frequencies of expression of IL-7R α-chain (CD127) and TNFR family receptors (OX40 and TNFRII) on CD8+RORγt T cells and CD8+RORγt+ T cells from DBDC patients (n = 17). (E) A representative flow cytometry dot plot and compiled bar graph show the apoptotic cells (annexin V+ 7-AAD+) from total CD8+ T cells either unstimulated or stimulated for 96 h with anti-CD2/CD3/CD28–coated beads. DBDC patients (n = 3). (F) Compiled frequencies of CD8+RORγt T cells and CD8+RORγt+ T cells from purified CD8+ T cells that were stimulated for 96 h with anti-CD2/CD3/CD28–coated beads. DBDC patients (n = 8). Error bars represent mean ± SEM. Horizontal bar represents median, each dot represents one patient. ***p ≤ 0.001, ****p ≤ 0.0001. ns, not significant.

FIGURE 4.

CD8+RORγt+ T cells are terminally differentiated and exhausted in DBDC patients. (A) Representative flow cytometry dot plots (left panel) and compiled frequencies (right panel) show the naive cells, CM cells, EM cells, and EMRAs, defined according to the expression of CCR7 and CD45RA in CD8+RORγt T cells and CD8+RORγt+ T cells in PBMCs from DBDC patients (n = 17). (B) Circulating level of IL-7 and IL-15 in plasma from HDs and DBDC patients. HDs (n = 12) and DBDC patients (n = 17). (C) Compiled frequencies of expression of activation- and exhaustion-related markers (CTLA-4, PD-1, LAG-3, Tim-3) on CD8+RORγt T cells and CD8+RORγt+ T cells from DBDC patients. (D) Compiled frequencies of expression of IL-7R α-chain (CD127) and TNFR family receptors (OX40 and TNFRII) on CD8+RORγt T cells and CD8+RORγt+ T cells from DBDC patients (n = 17). (E) A representative flow cytometry dot plot and compiled bar graph show the apoptotic cells (annexin V+ 7-AAD+) from total CD8+ T cells either unstimulated or stimulated for 96 h with anti-CD2/CD3/CD28–coated beads. DBDC patients (n = 3). (F) Compiled frequencies of CD8+RORγt T cells and CD8+RORγt+ T cells from purified CD8+ T cells that were stimulated for 96 h with anti-CD2/CD3/CD28–coated beads. DBDC patients (n = 8). Error bars represent mean ± SEM. Horizontal bar represents median, each dot represents one patient. ***p ≤ 0.001, ****p ≤ 0.0001. ns, not significant.

Close modal

It has previously been reported that the expression of granzyme B and perforin is impaired in CD8+RORγt+ T cells (TC17 cells) and in CD8+RORγt+ T cells that coexpress RORγt and T-bet (TC17/1) (69, 37). However, we found that CD8+RORγt+ T cells had a significantly higher level of preformed granzyme B and perforin compared with CD8+RORγt T cells from DBDC patients (Fig. 5A), which may be related to the T-bet expression and terminally differentiated status (30, 38). We also measured the level of degranulation by CD8+RORγt+ T cells and CD8+RORγt T cells in purified CD8+ T cells from PBMCs of DBDC patients by measuring the surface translocation of the degranulation marker CD107a. After TCR/CD2/CD28 stimulation for 96 h in vitro, both CD8+RORγt+ T cells and CD8+RORγt T cells upregulated the expression of CD107a compared with unstimulated cells (Fig. 5B). However, the degranulation level was significantly higher in CD8+RORγt+ T cells compared with that of CD8+RORγt T cells (Fig. 5B). These results indicate that CD8+RORγt+ T cells are not impaired in release of granzyme B and perforin. To sustain a cytotoxic immune response, the CD8+ T cells have to restore the cytotoxic components after the initial release (39, 40). Our results demonstrated that the stimulated CD8+RORγt T cells were able to restore both granzyme B and perforin to a significantly higher level than the preformed level (unstimulated cells) (Fig. 5C, left panel). In contrast, CD8+RORγt+ T cells were only able to restore granzyme B, but not perforin, as the level of perforin fell below the preformed level (unstimulated cells) (Fig. 5C, right panel). Next, we evaluated the cytotoxic potential of the CD8+RORγt+ T cells by purifying CD8+IL-17A+ T cells and compared them to CD8+IL-17A T cells from DBDC patients (Supplemental Fig. 4A). Using CD3-HER2 BsAb (Supplemental Fig. 4B), the cytokine-specific or total CD8+ T cells irrespective of their Ag specificity were redirected to detect and kill the HER2-expressing SKBR-3 cell line in 72-h coculture experiments. Our results show that the cytotoxic function of total CD8+ T cells from DBDC patients was significantly reduced compared with total CD8+ T cells from HDs (Fig. 5D, left panel). Similarly, the cytotoxic function of CD8+IL-17A+ T cells (CD8+RORγt+ T cells) was significantly impaired compared with CD8+IL-17A T cells (CD8+RORγt T cells) from DBDC patients (Fig. 5D, right panel). We confirmed that the CD8+IL-17A+ T cells maintained the IL-17A expression (Supplemental Fig. 4C), and that the reduced cytotoxicity of total CD8+ T cells and CD8+IL-17A+ T cells from DBDC patients was not due to impaired degranulation as measured by surface expression of CD107a (Supplemental Fig. 4D). Taken together, these results suggest that the CD8+RORγt+ T cells from DBDC patients have impaired targeted cytotoxic activity.

FIGURE 5.

Impaired cytotoxic function of CD8+RORγt+ T cells in DBDC patients. (A) Representative flow cytometry dot plots (left panel) and compiled frequencies of granzyme B (GrzB) and perforin basal store in CD8+RORγt T cells and CD8+RORγt+ T cells in PBMCs from DBDC patients (n = 17). (B) The compiled frequencies show the expression levels of CD107a in CD8+RORγt T cells and CD8+RORγt+ T cells from purified CD8+ T cells that were stimulated for 96 h with anti-CD2/CD3/CD28–coated beads. (C) The compiled frequencies show the expression levels of GrzB (left panel) and perforin (right panel) in unstimulated (basal level) and stimulated (re-expression) in CD8+RORγt T cells and CD8+RORγt+ T cells from purified CD8+ T cells that were stimulated for 96 h with anti-CD2/CD3/CD28–coated beads. DBDC patients (n = 8). (D) The compiled bar graphs show the targeted cytolytic capacity of total CD8+ T cells from HDs and DBDC patients (left panel), as well as purified and expanded IL-17A+ and IL17-A CD8+ T cells from DBDC patients (right panel) that were cocultured with an HER2-expressing SKBR-3 adenocarcinoma cell line in a ratio of 1:5 and stimulated with CD3/HER2 BsAb (100 ng/ml) for 72 h. HDs (n = 3) and DBDC patients (n = 3). The cell lysis was calculated as specific cytotoxicity (%) = [(experimental release − effector spontaneous release − target spontaneous release)/(target maximum release − target spontaneous release)] × 100. Each assay was performed in triplicates. Error bars represent mean ± SEM. Horizontal bar represents median, and each dot represents one patient. *p ≤ 0.05, **p ≤ 0.01, ****p ≤ 0.0001. ns, not significant.

FIGURE 5.

Impaired cytotoxic function of CD8+RORγt+ T cells in DBDC patients. (A) Representative flow cytometry dot plots (left panel) and compiled frequencies of granzyme B (GrzB) and perforin basal store in CD8+RORγt T cells and CD8+RORγt+ T cells in PBMCs from DBDC patients (n = 17). (B) The compiled frequencies show the expression levels of CD107a in CD8+RORγt T cells and CD8+RORγt+ T cells from purified CD8+ T cells that were stimulated for 96 h with anti-CD2/CD3/CD28–coated beads. (C) The compiled frequencies show the expression levels of GrzB (left panel) and perforin (right panel) in unstimulated (basal level) and stimulated (re-expression) in CD8+RORγt T cells and CD8+RORγt+ T cells from purified CD8+ T cells that were stimulated for 96 h with anti-CD2/CD3/CD28–coated beads. DBDC patients (n = 8). (D) The compiled bar graphs show the targeted cytolytic capacity of total CD8+ T cells from HDs and DBDC patients (left panel), as well as purified and expanded IL-17A+ and IL17-A CD8+ T cells from DBDC patients (right panel) that were cocultured with an HER2-expressing SKBR-3 adenocarcinoma cell line in a ratio of 1:5 and stimulated with CD3/HER2 BsAb (100 ng/ml) for 72 h. HDs (n = 3) and DBDC patients (n = 3). The cell lysis was calculated as specific cytotoxicity (%) = [(experimental release − effector spontaneous release − target spontaneous release)/(target maximum release − target spontaneous release)] × 100. Each assay was performed in triplicates. Error bars represent mean ± SEM. Horizontal bar represents median, and each dot represents one patient. *p ≤ 0.05, **p ≤ 0.01, ****p ≤ 0.0001. ns, not significant.

Close modal

We evaluated the prognostic value of the total circulating CD8+RORγt+ T cell population and subsets that expressed PD-1, CTLA-4, and OX40 on the survival of DBDC patients (Fig. 6). Comparing patients with high (more than or equal to the median level) and low percentages (less than the median level) of CD8+RORγt+ T cells and CTLA-4–expressing CD8+RORγt+ T cells, we did not find that the level of either of these populations correlated with clinical outcome and it did not reveal any prognostic value on 41-mo cancer-specific survival. However, a low percentage of PD-1 and a high percentage of OX40-expressing CD8+RORγt+ T cells were both associated with reduced cancer-specific survival of the DBDC patients (Fig. 6).

FIGURE 6.

Subsets of CD8+RORγt+ T cells correlate with DBDC patient survival. Kaplan–Meier curves illustrate the cancer specific survival (in months). Patients were stratified by median percentages of CD8+RORγt+ T cells, CD8+RORγt+CTLA-4+ T cells, CD8+RORγt+PD-1+ T cells, and CD8+RORγt+OX40+ T cells. DBDC patients (n = 17).

FIGURE 6.

Subsets of CD8+RORγt+ T cells correlate with DBDC patient survival. Kaplan–Meier curves illustrate the cancer specific survival (in months). Patients were stratified by median percentages of CD8+RORγt+ T cells, CD8+RORγt+CTLA-4+ T cells, CD8+RORγt+PD-1+ T cells, and CD8+RORγt+OX40+ T cells. DBDC patients (n = 17).

Close modal

A diverse array of innate and adaptive immune cells interacts in shaping the antitumor immunity in cancer patients. The composition of the immune cells and corresponding cytokines, inflammatory mediators, growth factors, and effector molecules determines whether the immune system is able to mount an efficient antitumor immune response. However, the immune system may also stimulate tumor growth and angiogenesis and increase the metastatic potential (41). Recent studies demonstrate that IL-17A mediates a procarcinogenic inflammatory immune response that promotes the progression of preinvasive neoplasia to invasive adenocarcinomas (42, 43), and T cell subsets that express the master transcription factor RORγt and produce IL-17A promote a variety of human cancers. These T cell subsets include CD4+RORγt+ T cells (TH17 cells) (44, 45), CD4+FOXP3+RORγt+ Tregs (Th17-like Tregs) (46), and CD8+RORγt+ T cells (TC17 cells) (16).

The tumor tissue of DBDCs is strongly infiltrated by macrophages, CD8+ T cells, and CD4+ T cells (3). The CD8+ T cells are essential for tumor elimination by their cytotoxic immune activity, and their presence in the tumor microenvironment is associated with a favorable outcome in many invasive solid tumors (1, 47, 48). However, DBDC patients with localized or metastatic cancer have a poor prognosis (49), and the functional and prognostic relevance of CD8+ T cells in these patients is largely unknown. The relevance of CD8+ T cells in DBDC may be further complicated by a functional shift from antitumor immunity toward procarcinogenic properties. Recent studies from gastric and hepatocellular carcinoma show that CD8+ T cells are converted by tumor-associated macrophages into CD8+RORγt+ T cells that produce IL-17A and INF-γ that support cancer progression (6, 7).

In this study, we demonstrate that CD8+ T cells in peripheral blood from DBDC patients significantly upregulate the expression of RORγt. The differentiation of naive CD8+ T cells into CD8+RORγt+ T cells appears to be similar to that of CD4+RORγt+ T cells and requires type 17 polarizing cytokines, such as IL-1β, IL-6, and TGF-β (9, 16). We found that IL-1β and IL-6, as well as the effector cytokines IL-17A and TNF-α, were elevated in peripheral blood of DBDC patients, whereas the type 1 effector cytokines INF-γ and IL-2 were not elevated. This indicates that there is a skewing from an antitumor type 1 response toward a procarcinogenic type 17 response in the DBDC patients. Additionally, we found that the CD8+RORγt+ T cells exhibited a high degree of plasticity by coexpressing RORγt and T-bet (TC17/1 cells). Previous studies have suggested that T-bet is a negative regulator of RORγt expression and that the expression of INF-γ and IL-17A was consequently mutually exclusive. This reciprocal regulation is thought to ensure that the cells have either a type 17 or type 1 phenotype (5052). However, our analyses of intracellular cytokine expression revealed that the CD8+RORγt+ T cells were able to coproduce IL-17A and INF-γ. This indicates that the expression of T-bet did not downregulate RORγt expression in the CD8+RORγt+ T cells, which may point to a phenotypic plasticity between the subsets. This is in line with recent reports that demonstrate that type 1 responses with secretion of IL-12 induce permissive chromatin modifications in the promoters of IL-17A, INF-γ, RORγt, and T-bet, and thereby induce phenotypic plasticity in the CD8+RORγt+ T cells (53, 54).

The CD8+RORγt+ T cells produced both type 17–specific (IL-17A) and type 1–specific (INF-γ) cytokines. However, the type 2–specific cytokine IL-4 was produced in low amounts and was mostly uncoupled from IL-17A expression. Although the expression of GATA-3 was increased in CD8+RORγt+ T cells, their expression was lower than that of T-bet, and analyses of coexpression revealed that most GATA3-expressing CD8+RORγt+ T cells also expressed T-bet. We think that increased expression of T-bet in CD8+RORγt+ T cells may antagonize GATA3 function (20) and induce plasticity toward the TC17/1 phenotype. We also found that the CD8+RORγt+ T cells expressed TGF-β, which is essential for the generation and maintenance of type 17 responses from T cells (21, 55). TGF-β also induces FOXP3 expression and subsequent suppressive function in Tregs (21). However, the CD8+RORγt+ T cells did not express FOXP3. These results further support the notion that CD8+RORγt+ T cells are a T cell subset with inflammatory properties. In humans, it is predominantly circulating CD4+RORγt+ T cells that produce IL-17A (TH17 cells) (21, 22). Our findings in the present study demonstrate that CD8+RORγt+ T cells are capable of expressing equal amounts of IL-17A as the CD4+RORγt+ T cells in the DBDC patients. It has previously been shown that CD8+RORγt+ T cells and CD4+RORγt+ T cells share similar tissue-homing receptors (4). In contrast, our findings in DBDC patients clearly demonstrate that the subsets differ in the expression of tissue-homing receptors. The RORγt-expressing human T cells have been reported to strictly originate from CD161+ T cells (56), and we found that CD8+RORγt+ T cells and CD4+RORγt+ T cells expressed CD161 to a similar level, but the RORγt-associated homing receptors CCR6 and CCR4 were significantly reduced in CD8+RORγt+ T cells (4). Instead, CD8+RORγt+ T cells strongly expressed the T-bet–associated homing receptors CXCR3 and CCR5 (4). Most CD8+RORγt+ T cells and CD4+RORγt+ T cells are destined for epithelial infiltration in the gut, which is consistent with our findings that the expression of the intraepithelial homing receptor CD103 (4) was similar in the CD8+RORγt+ T cells and in the CD4+RORγt+ T cells. This suggests that both the CD8+RORγt+ T cell subset and the CD4+RORγt+ T cell subset in DBDC are capable of migrating to inflamed tissues, albeit with different receptors (3).

Our results also showed that CD8+RORγt+ T cells in the DBDC patients constitute a highly activated subset that expresses HLA-DR and CD38 as opposed to CD8+RORγt T cells. Increased expression of Ki-67 in CD8+RORγt+ T cells further indicated that they are actively proliferating in the DBDC patients. TCR/CD28 costimulation is indispensable for T cell activation and effector functions (2527). We studied the TCR/CD28/CD2 receptor-mediated signaling in CD8+RORγt+ T cells and CD8+RORγt T cells simultaneously using phosphoflow cytometry. We found that the CD8+RORγt+ T cells had a higher basal signaling level, and TCR/CD28/CD2 receptor activation induced higher phosphorylation of TCR proximal, distal, and proliferation-associated proteins compared with the CD8+RORγt T cells. We further substantiated this finding with proliferation assays that indicated that the CD8+RORγt+ T cells strongly upregulated CD69 within 24 h and hyperproliferated in the next 96 h, which is likely due to their enhanced TCR signaling. The expression of CD28 is essential for proliferative responses (26), and impaired expression leads to senescence and a regulatory phenotype in CD8+ T cells (28, 57). Importantly, the reduced signaling in CD8+RORγ T cells was not due to impaired expression of CD28 or to a regulatory phenotype characterized by expression of FOXP3.

Apart from lineage-specific differentiation of human CD8+ T cells driven by either type 1, 2, or 17 polarizing cytokines, they also undergo memory differentiation, which includes naive cells, CM cells, EM cells, and EMRAs (30, 58). Under chronic Ag stimulation, as in cancer, the CD8+ T cells differentiate to terminally differentiated EMRAs that lose expression of CCR7 and are destined for nonlymphoidal organs (30, 33). Our results showed that the hyperactivated CD8+RORγt+ T cells are terminally differentiated. Furthermore, we found that this cell population significantly expressed the activation-induced inhibitory receptors CTLA-4, PD-1, LAG-3, and Tim-3 (30). Expression of these inhibitory receptors indicates that the CD8+RORγt+ T cells are exhausted (33). Exhausted T cells are activated by Ags rather than by homeostatic cytokines, including IL-7, and this is largely due to loss of CD127 (IL-7 receptor) (31, 33, 34). Our results showed that the CD8+RORγt+ T cells lost CD127 expression, indicating that the CD8+RORγt+ T cells might be restricted to TCR activation rather than to homeostatic proliferation in the DBDC patients. Furthermore, we confirmed this finding by measuring the homeostatic cytokines IL-7 and IL-15, which were not increased in the DBDC patients. Interestingly, we also found that the expression of the TNFR family receptor OX40 was downregulated, whereas TNFRII was upregulated in the CD8+RORγt+ T cells, which play a role in constraining clonal expansion (35). In line with this, we found that the ex vivo–isolated CD8+RORγt+ T cells were highly prone to apoptosis in the absence of TCR and costimulation compared with CD8+RORγt T cells, but this could be reversed when the cells were activated. These results indicate that the activated CD8+RORγt+ T cells might have a rapid turnover and are probably not maintained as long-lived memory cells in the DBDC patients.

We further assessed the degranulation of granzyme B and perforin by CD8+RORγt+ T cells compared with CD8+RORγt T cells from DBDC patients. It has previously been reported that CD8+RORγt+ T cells that have been generated in vitro or isolated from tumor tissue have diminished cytotoxic activity (6, 7, 9, 10). Similarly, type 17–associated polarizing conditions have also been shown to disarm the cytotoxic activity of CD8+RORγt+ T cells (8). In contrast to previous reports, we found that CD8+RORγt+ T cells had significantly higher levels of preformed granzyme B and perforin. Moreover, stimulation through TCR/CD28/CD2 induced translocation of the degranulation marker CD107a on both CD8+RORγt+ T cells and CD8+RORγt T cells. This indicated that both subsets were able to release preformed granzyme B and perforin. Presence of preformed granzyme B and perforin indicate an intact immediate cytotoxic potential, whereas de novo synthesis represents the sustainability of a cytotoxic response (39, 40). Our results clearly indicated that although both CD8+RORγt+ T cells and CD8+RORγt T cells degranulate, only the CD8+RORγt T cells were able to upregulate and regain both granzyme B and perforin, whereas the CD8+RORγt+ T cells were only able to restore granzyme B, but not perforin. This is of particular importance because perforin is crucial for the function of many cytotoxic components secreted by CD8+ T cells, and particularly the targeted function of granzyme B is entirely perforin-dependent (40). Furthermore, we observed a significant impairment in cytotoxic activity of the CD8+RORγt+ T cells in a redirected killing assay using a bispecific Ab against CD3 on CD8+IL-17A+ T cells and HER2 on the SKBR-3 adenocarcinoma cell line. These results indicated that the CD8+RORγt T cells are properly functioning cytotoxic CD8+ T cells, whereas CD8+RORγt+ T cells have impaired capacity to sustain a cytotoxic immune response.

To investigate the clinical importance of our findings, we correlated the clinical features with the level of circulating CD8+RORγt+ T cells in DBDC patients. The increased level of CD8+RORγt+ T cells was not associated with decreased survival. However, the amount of the subset of CD8+RORγt+ T cells that expressed low levels of the inhibitory coreceptor PD-1 and high levels of the costimulatory receptor OX40 was associated with significantly reduced survival rate in the DBDC patients. The combination of low PD-1 and high OX40 expression can increase the effector potential of CD8+RORγt+ T cells (59). This is interesting because CD8+RORγt+ T cells correlate with type 17 and type 2 response-related cells such as TH17 cells, TH2 cells, and TC2 cells that are with poor prognosis and reduced survival in human cancers (1), and in particular pancreatic adenocarcinoma (60), which is anatomically close and histologically similar to DBDC (61). Thus, low PD-1 and high OX40 expression may increase the effector activity of CD8+RORγt+ T cells with detrimental effect on prognosis in the DBDC patients. These findings also indicate that the immunotherapeutic approaches using PD-1 inhibitory Abs and OX40 agonists in DBDC patients should be made with caution.

Taken together, the CD8+RORγt+ T cells have a type 17 proinflammatory phenotype with impaired cytotoxic potential and they are associated with protumorigenic responses in DBDC patients. These findings may explain the diminished antitumor immunity in DBDC patients that appears irrespective of the level of CD8+ T cells. Therefore, therapeutic strategies that aim to shift the balance from a pathogenic RORγt+ lineage to a protective RORγt lineage of CD8+ T cells might improve the antitumor immunity and ultimately the prognosis of DBDC patients.

We thank all the patients who participated in this study and the members of K.T. laboratory (Center for Molecular Medicine Norway, University of Oslo) for helpful discussion.

This work was supported by The Research Council of Norway Grants 221938 (to E.M.A.) and 204784 and 187615 (to K.T.), Norwegian Cancer Society Grants 741746 (to E.M.A.) and 419544 (to K.T.), South-Eastern Norway Regional Health Authority Grant 2010038 (to E.M.A.), and by K.G. Jebsen Foundation Grants 2012/21 and 2012/23 (to K.T. and E.M.A.).

The online version of this article contains supplemental material.

Abbreviations used in this article:

7-AAD

7-aminoactinomycin D

BsAb

bispecific Ab

CM

central memory

DBDC

distal bile duct cancer

EM

effector memory

EMRA

effector memory cell that re-expresses CD45RA

GATA3

GATA-binding protein 3

HD

healthy donor

LAG-3

lymphocyte activation gene-3

LAP

latency-associated peptide

PD-1

programmed cell death protein 1

ROR

retinoic acid–related orphan receptor

Tim-3

T cell Ig and mucin domain-3

Treg

regulatory T cell.

1
Fridman
W. H.
,
Pagès
F.
,
Sautès-Fridman
C.
,
Galon
J.
.
2012
.
The immune contexture in human tumours: impact on clinical outcome.
Nat. Rev. Cancer
12
:
298
306
.
2
Blechacz
B.
,
Komuta
M.
,
Roskams
T.
,
Gores
G. J.
.
2011
.
Clinical diagnosis and staging of cholangiocarcinoma.
Nat. Rev. Gastroenterol. Hepatol.
8
:
512
522
.
3
Goeppert
B.
,
Frauenschuh
L.
,
Zucknick
M.
,
Stenzinger
A.
,
Andrulis
M.
,
Klauschen
F.
,
Joehrens
K.
,
Warth
A.
,
Renner
M.
,
Mehrabi
A.
, et al
.
2013
.
Prognostic impact of tumour-infiltrating immune cells on biliary tract cancer.
Br. J. Cancer
109
:
2665
2674
.
4
Annunziato
F.
,
Romagnani
C.
,
Romagnani
S.
.
2015
.
The 3 major types of innate and adaptive cell-mediated effector immunity.
J. Allergy Clin. Immunol.
135
:
626
635
.
5
Saito
H.
,
Yamada
Y.
,
Takaya
S.
,
Osaki
T.
,
Ikeguchi
M.
.
2015
.
Clinical relevance of the number of interleukin-17-producing CD 8+ T cells in patients with gastric cancer.
Surg. Today
45
:
1429
1435
.
6
Kuang
D. M.
,
Peng
C.
,
Zhao
Q.
,
Wu
Y.
,
Zhu
L. Y.
,
Wang
J.
,
Yin
X. Y.
,
Li
L.
,
Zheng
L.
.
2010
.
Tumor-activated monocytes promote expansion of IL-17-producing CD8+ T cells in hepatocellular carcinoma patients.
J. Immunol.
185
:
1544
1549
.
7
Zhuang
Y.
,
Peng
L. S.
,
Zhao
Y. L.
,
Shi
Y.
,
Mao
X. H.
,
Chen
W.
,
Pang
K. C.
,
Liu
X. F.
,
Liu
T.
,
Zhang
J. Y.
, et al
.
2012
.
CD8+ T cells that produce interleukin-17 regulate myeloid-derived suppressor cells and are associated with survival time of patients with gastric cancer.
Gastroenterology
143
:
951
962.e8
.
8
Tsai
J. P.
,
Lee
M. H.
,
Hsu
S. C.
,
Chen
M. Y.
,
Liu
S. J.
,
Chang
J. T.
,
Liao
C. T.
,
Cheng
A. J.
,
Chong
P.
,
Chu
C. L.
, et al
.
2012
.
CD4+ T cells disarm or delete cytotoxic T lymphocytes under IL-17–polarizing conditions.
J. Immunol.
189
:
1671
1679
.
9
Huber
M.
,
Heink
S.
,
Grothe
H.
,
Guralnik
A.
,
Reinhard
K.
,
Elflein
K.
,
Hünig
T.
,
Mittrücker
H. W.
,
Brüstle
A.
,
Kamradt
T.
,
Lohoff
M.
.
2009
.
A Th17-like developmental process leads to CD8+ Tc17 cells with reduced cytotoxic activity.
Eur. J. Immunol.
39
:
1716
1725
.
10
Kondo
T.
,
Takata
H.
,
Matsuki
F.
,
Takiguchi
M.
.
2009
.
Cutting edge: phenotypic characterization and differentiation of human CD8+ T cells producing IL-17.
J. Immunol.
182
:
1794
1798
.
11
Yssel
H.
,
Spits
H.
.
2002
.
Generation and maintenance of cloned human T cell lines.
Curr. Protoc. Immunol.
Chapter 7
:
Unit 7.19
.
12
Kalland
M. E.
,
Oberprieler
N. G.
,
Vang
T.
,
Taskén
K.
,
Torgersen
K. M.
.
2011
.
T cell-signaling network analysis reveals distinct differences between CD28 and CD2 costimulation responses in various subsets and in the MAPK pathway between resting and activated regulatory T cells.
J. Immunol.
187
:
5233
5245
.
13
Irish
J. M.
,
Myklebust
J. H.
,
Alizadeh
A. A.
,
Houot
R.
,
Sharman
J. P.
,
Czerwinski
D. K.
,
Nolan
G. P.
,
Levy
R.
.
2010
.
B-cell signaling networks reveal a negative prognostic human lymphoma cell subset that emerges during tumor progression.
Proc. Natl. Acad. Sci. USA
107
:
12747
12754
.
14
He
Q.
,
Zhang
H.
,
Wang
Y.
,
Ting
H. H.
,
Yu
W.
,
Cao
X.
,
Ge
W.
.
2014
.
Purified anti-CD3 × anti-HER2 bispecific antibody potentiates cytokine-induced killer cells of poor spontaneous cytotoxicity against breast cancer cells.
Cell Biosci.
4
:
70
.
15
Bailey
S. R.
,
Nelson
M. H.
,
Himes
R. A.
,
Li
Z.
,
Mehrotra
S.
,
Paulos
C. M.
.
2014
.
Th17 cells in cancer: the ultimate identity crisis.
Front. Immunol.
5
:
276
.
16
Liang
Y.
,
Pan
H. F.
,
Ye
D. Q.
.
2015
.
Tc17 cells in immunity and systemic autoimmunity.
Int. Rev. Immunol.
34
:
318
331
.
17
Lazarevic
V.
,
Glimcher
L. H.
,
Lord
G. M.
.
2013
.
T-bet: a bridge between innate and adaptive immunity.
Nat. Rev. Immunol.
13
:
777
789
.
18
Weaver
C. T.
,
Hatton
R. D.
,
Mangan
P. R.
,
Harrington
L. E.
.
2007
.
IL-17 family cytokines and the expanding diversity of effector T cell lineages.
Annu. Rev. Immunol.
25
:
821
852
.
19
Raphael
I.
,
Nalawade
S.
,
Eagar
T. N.
,
Forsthuber
T. G.
.
2015
.
T cell subsets and their signature cytokines in autoimmune and inflammatory diseases.
Cytokine
74
:
5
17
.
20
Zhu
J.
,
Yamane
H.
,
Paul
W. E.
.
2010
.
Differentiation of effector CD4 T cell populations.
Annu. Rev. Immunol.
28
:
445
489
.
21
Korn
T.
,
Bettelli
E.
,
Oukka
M.
,
Kuchroo
V. K.
.
2009
.
IL-17 and Th17 cells.
Annu. Rev. Immunol.
27
:
485
517
.
22
Annunziato
F.
,
Cosmi
L.
,
Liotta
F.
,
Maggi
E.
,
Romagnani
S.
.
2012
.
Defining the human T helper 17 cell phenotype.
Trends Immunol.
33
:
505
512
.
23
Sundrud
M. S.
,
Trivigno
C.
.
2013
.
Identity crisis of Th17 cells: many forms, many functions, many questions.
Semin. Immunol.
25
:
263
272
.
24
Sallusto
F.
,
Lanzavecchia
A.
,
Mackay
C. R.
.
1998
.
Chemokines and chemokine receptors in T-cell priming and Th1/Th2-mediated responses.
Immunol. Today
19
:
568
574
.
25
Huppa
J. B.
,
Gleimer
M.
,
Sumen
C.
,
Davis
M. M.
.
2003
.
Continuous T cell receptor signaling required for synapse maintenance and full effector potential.
Nat. Immunol.
4
:
749
755
.
26
Bonnevier
J. L.
,
Mueller
D. L.
.
2002
.
Cutting edge: B7/CD28 interactions regulate cell cycle progression independent of the strength of TCR signaling.
J. Immunol.
169
:
6659
6663
.
27
Iezzi
G.
,
Karjalainen
K.
,
Lanzavecchia
A.
.
1998
.
The duration of antigenic stimulation determines the fate of naive and effector T cells.
Immunity
8
:
89
95
.
28
Strioga
M.
,
Pasukoniene
V.
,
Characiejus
D.
.
2011
.
CD8+ CD28 and CD8+ CD57+ T cells and their role in health and disease.
Immunology
134
:
17
32
.
29
Bromley
S. K.
,
Thomas
S. Y.
,
Luster
A. D.
.
2005
.
Chemokine receptor CCR7 guides T cell exit from peripheral tissues and entry into afferent lymphatics.
Nat. Immunol.
6
:
895
901
.
30
Fuertes Marraco
S. A.
,
Neubert
N. J.
,
Verdeil
G.
,
Speiser
D. E.
.
2015
.
Inhibitory receptors beyond T cell exhaustion.
Front. Immunol.
6
:
310
.
31
Geginat
J.
,
Lanzavecchia
A.
,
Sallusto
F.
.
2003
.
Proliferation and differentiation potential of human CD8+ memory T-cell subsets in response to antigen or homeostatic cytokines.
Blood
101
:
4260
4266
.
32
Legat
A.
,
Speiser
D. E.
,
Pircher
H.
,
Zehn
D.
,
Fuertes Marraco
S. A.
.
2013
.
Inhibitory receptor expression depends more dominantly on differentiation and activation than “exhaustion” of human CD8 T cells.
Front. Immunol.
4
:
455
.
33
Wherry
E. J.
2011
.
T cell exhaustion.
Nat. Immunol.
12
:
492
499
.
34
Boyman
O.
,
Purton
J. F.
,
Surh
C. D.
,
Sprent
J.
.
2007
.
Cytokines and T-cell homeostasis.
Curr. Opin. Immunol.
19
:
320
326
.
35
Wortzman
M. E.
,
Clouthier
D. L.
,
McPherson
A. J.
,
Lin
G. H. Y.
,
Watts
T. H.
.
2013
.
The contextual role of TNFR family members in CD8+ T-cell control of viral infections.
Immunol. Rev.
255
:
125
148
.
36
Kim
E. Y.
,
Teh
S.-J.
,
Yang
J.
,
Chow
M. T.
,
Teh
H.-S.
.
2009
.
TNFR2-deficient memory CD8 T cells provide superior protection against tumor cell growth.
J. Immunol.
183
:
6051
6057
.
37
Yen
H.-R.
,
Harris
T. J.
,
Wada
S.
,
Grosso
J. F.
,
Getnet
D.
,
Goldberg
M. V.
,
Liang
K.-L.
,
Bruno
T. C.
,
Pyle
K. J.
,
Chan
S.-L.
, et al
.
2009
.
Tc17 CD8 T cells: functional plasticity and subset diversity.
J. Immunol.
183
:
7161
7168
.
38
Knox
J. J.
,
Cosma
G. L.
,
Betts
M. R.
,
McLane
L. M.
.
2014
.
Characterization of T-bet and eomes in peripheral human immune cells. [Published erratum appears in 2016 Front. Immunol. 7: 337.]
Front. Immunol.
5
:
217
.
39
Makedonas
G.
,
Banerjee
P. P.
,
Pandey
R.
,
Hersperger
A. R.
,
Sanborn
K. B.
,
Hardy
G. A. D.
,
Orange
J. S.
,
Betts
M. R.
.
2009
.
Rapid up-regulation and granule-independent transport of perforin to the immunological synapse define a novel mechanism of antigen-specific CD8+ T cell cytotoxic activity.
J. Immunol.
182
:
5560
5569
.
40
Voskoboinik
I.
,
Smyth
M. J.
,
Trapani
J. A.
.
2006
.
Perforin-mediated target-cell death and immune homeostasis.
Nat. Rev. Immunol.
6
:
940
952
.
41
Sasada
T.
,
Suekane
S.
.
2011
.
Variation of tumor-infiltrating lymphocytes in human cancers: controversy on clinical significance.
Immunotherapy
3
:
1235
1251
.
42
McAllister
F.
,
Bailey
J. M.
,
Alsina
J.
,
Nirschl
C. J.
,
Sharma
R.
,
Fan
H.
,
Rattigan
Y.
,
Roeser
J. C.
,
Lankapalli
R. H.
,
Zhang
H.
, et al
.
2014
.
Oncogenic Kras activates a hematopoietic-to-epithelial IL-17 signaling axis in preinvasive pancreatic neoplasia.
Cancer Cell
25
:
621
637
.
43
Wang
K.
,
Kim
M. K.
,
Di Caro
G.
,
Wong
J.
,
Shalapour
S.
,
Wan
J.
,
Zhang
W.
,
Zhong
Z.
,
Sanchez-Lopez
E.
,
Wu
L. W.
, et al
.
2014
.
Interleukin-17 receptor a signaling in transformed enterocytes promotes early colorectal tumorigenesis.
Immunity
41
:
1052
1063
.
44
Su
X.
,
Ye
J.
,
Hsueh
E. C.
,
Zhang
Y.
,
Hoft
D. F.
,
Peng
G.
.
2010
.
Tumor microenvironments direct the recruitment and expansion of human Th17 cells.
J. Immunol.
184
:
1630
1641
.
45
Qi
W.
,
Huang
X.
,
Wang
J.
.
2013
.
Correlation between Th17 cells and tumor microenvironment.
Cell. Immunol.
285
:
18
22
.
46
Du
R.
,
Zhao
H.
,
Yan
F.
,
Li
H.
.
2014
.
IL-17+Foxp3+ T cells: an intermediate differentiation stage between Th17 cells and regulatory T cells.
J. Leukoc. Biol.
96
:
39
48
.
47
Kim
H. J.
,
Cantor
H.
.
2014
.
CD4 T-cell subsets and tumor immunity: the helpful and the not-so-helpful.
Cancer Immunol. Res.
2
:
91
98
.
48
Jochems
C.
,
Schlom
J.
.
2011
.
Tumor-infiltrating immune cells and prognosis: the potential link between conventional cancer therapy and immunity.
Exp. Biol. Med. (Maywood)
236
:
567
579
.
49
Liu
X. F.
,
Tang
K.
,
Sui
L. L.
,
Xu
G.
.
2014
.
Cholangiocarcinoma: present status and molecular aspects of diagnosis.
Oncol. Res.
22
:
177
183
.
50
Lazarevic
V.
,
Chen
X.
,
Shim
J.-H.
,
Hwang
E.-S.
,
Jang
E.
,
Bolm
A. N.
,
Oukka
M.
,
Kuchroo
V. K.
,
Glimcher
L. H.
.
2011
.
T-bet represses TH17 differentiation by preventing Runx1-mediated activation of the gene encoding RORγt.
Nat. Immunol.
12
:
96
104
.
51
Harrington
L. E.
,
Hatton
R. D.
,
Mangan
P. R.
,
Turner
H.
,
Murphy
T. L.
,
Murphy
K. M.
,
Weaver
C. T.
.
2005
.
Interleukin 17-producing CD4+ effector T cells develop via a lineage distinct from the T helper type 1 and 2 lineages.
Nat. Immunol.
6
:
1123
1132
.
52
Gökmen
M. R.
,
Dong
R.
,
Kanhere
A.
,
Powell
N.
,
Perucha
E.
,
Jackson
I.
,
Howard
J. K.
,
Hernandez-Fuentes
M.
,
Jenner
R. G.
,
Lord
G. M.
.
2013
.
Genome-wide regulatory analysis reveals that T-bet controls Th17 lineage differentiation through direct suppression of IRF4.
J. Immunol.
191
:
5925
5932
.
53
Tajima
M.
,
Wakita
D.
,
Satoh
T.
,
Kitamura
H.
,
Nishimura
T.
.
2011
.
IL-17/IFN-γ double producing CD8+ T (Tc17/IFN-γ) cells: a novel cytotoxic T-cell subset converted from Tc17 cells by IL-12.
Int. Immunol.
23
:
751
759
.
54
Satoh
T.
,
Tajima
M.
,
Wakita
D.
,
Kitamura
H.
,
Nishimura
T.
.
2012
.
The development of IL-17/IFN-γ-double producing CTLs from Tc17 cells is driven by epigenetic suppression of Socs3 gene promoter.
Eur. J. Immunol.
42
:
2329
2342
.
55
Filippi
C. M.
,
Juedes
A. E.
,
Oldham
J. E.
,
Ling
E.
,
Togher
L.
,
Peng
Y.
,
Flavell
R. A.
,
von Herrath
M. G.
.
2008
.
Transforming growth factor-β suppresses the activation of CD8+ T-cells when naive but promotes their survival and function once antigen experienced: a two-faced impact on autoimmunity.
Diabetes
57
:
2684
2692
.
56
Maggi
L.
,
Santarlasci
V.
,
Capone
M.
,
Peired
A.
,
Frosali
F.
,
Crome
S. Q.
,
Querci
V.
,
Fambrini
M.
,
Liotta
F.
,
Levings
M. K.
, et al
.
2010
.
CD161 is a marker of all human IL-17-producing T-cell subsets and is induced by RORC.
Eur. J. Immunol.
40
:
2174
2181
.
57
Simone
R.
,
Zicca
A.
,
Saverino
D.
.
2008
.
The frequency of regulatory CD3+CD8+CD28CD25+ T lymphocytes in human peripheral blood increases with age.
J. Leukoc. Biol.
84
:
1454
1461
.
58
Sallusto
F.
,
Lenig
D.
,
Förster
R.
,
Lipp
M.
,
Lanzavecchia
A.
.
1999
.
Two subsets of memory T lymphocytes with distinct homing potentials and effector functions.
Nature
401
:
708
712
.
59
Linch
S. N.
,
McNamara
M. J.
,
Redmond
W. L.
.
2015
.
OX40 agonists and combination immunotherapy: putting the pedal to the metal.
Front. Oncol.
5
:
34
.
60
Wörmann
S. M.
,
Diakopoulos
K. N.
,
Lesina
M.
,
Algül
H.
.
2014
.
The immune network in pancreatic cancer development and progression.
Oncogene
33
:
2956
2967
.
61
Schmuck
R. B.
,
de Carvalho-Fischer
C. V.
,
Neumann
C.
,
Pratschke
J.
,
Bahra
M.
.
2016
.
Distal bile duct carcinomas and pancreatic ductal adenocarcinomas: postulating a common tumor entity.
Cancer Med.
5
:
88
99
.

The authors have no financial conflicts of interest.

Supplementary data