The tumor microenvironment induces immunosuppression via recruiting and expanding suppressive immune cells such as regulatory T cells (Treg) to promote cancer progression. In this study, we documented that tumor-infiltrating CD73+ γδTregs were the predominant Tregs in human breast cancer and exerted more potent immunosuppressive activity than CD4+ or CD8+ Tregs. We further demonstrated that cancer-associated fibroblast (CAF)–derived IL6, rather than TGFβ1, induced CD73+ γδTreg differentiation from paired normal breast tissues via the IL6/STAT3 pathway to produce more adenosine and become potent immunosuppressive T cells. CD73+ γδTregs could in turn promote IL6 secretion by CAFs through adenosine/A2BR/p38MAPK signaling, thereby forming an IL6–adenosine positive feedback loop. CD73+ γδTreg infiltration also impaired the tumoricidal functions of CD8+ T cells and significantly correlated with worse prognosis of patients. The data indicate that the IL6–adenosine loop between CD73+ γδTregs and CAFs is important to promote immunosuppression and tumor progression in human breast cancer, which may be critical for tumor immunotherapy.

The tumor microenvironment (TME) links closely with the initiation, promotion, and progression of breast cancer via diverse mechanisms such as inducing immunosuppression and angiogenesis. The TME is composed of extracellular matrix (ECM) and stromal cells such as fibroblasts, pericytes, immune cells, and endothelial cells (1). Immune cells are the core components of the TME. However, different profiles of infiltrating immune cells are correlated with differential clinical outcomes in breast cancer (2). Within the TME, stromal cells can recruit and differentiate immune cells into an immunosuppressive status, including regulatory T cells (Treg), through secretion of various cytokines and metabolites, and thereby inhibits antitumor effector T cells, which subsequently promotes tumor progression (3).

Studies have documented that increased frequencies of tumor-infiltrating Tregs are significantly associated with worse outcome in types of cancer, including ovarian and pancreatic cancer (4, 5). However, the role of CD4+ Tregs in human breast cancer is still not conclusive (6, 7), implicating that CD4+ Tregs are not the predominant Tregs in the breast cancer microenvironment. Tregs are a heterogeneous population (8). However, the whole suppressive T-cell landscape in the TME of breast cancer has not been extensively investigated. Our previous data have shown that tumor-derived CD39+ γδTregs are the predominant Tregs that promote tumor progression in human colorectal cancer (9). Therefore, it prompted us to speculate that there may be a subpopulation in γδT cells that plays a pivotal role in facilitating breast cancer progression and immune evasion.

In this study, we found that γδT cells had high CD73 expression (approximately 60%) in tumor tissues. Tumor-infiltrating CD73+ γδT cells were the predominant Tregs in human breast cancer and exhibited potent, direct immunosuppressive function on CD4+ and CD8+ T cells (CD73+ γδTregs). Those CD73+ γδTregs could be differentiated by cancer-associated fibroblast (CAF)–derived IL6 rather than TGFβ1 via the IL6/STAT3 pathway to generate more exogenous adenosine and exert inhibitory function. Activated CD73+ γδTregs could in turn promote IL6 secretion by CAFs through the adenosine/A2BR/p38MAPK signaling pathway, thereby forming an IL6–adenosine positive feedback loop. The infiltration of CD73+ γδTregs impaired the tumoricidal functions of CD8+ T cells and significantly correlated with worse clinical outcome, including overall survival (OS) and disease-free survival (DFS), of patients with breast cancer. Thus, we unraveled a predominant regulatory γδT-cell subset and the positive feedback loop between CD73+ γδTregs and CAFs in human breast cancer that promote immunosuppression and tumor progression.

Clinical specimens

Fresh tumor (homogeneous cellularity, without foci of necrosis) and paired normal tissues were obtained from 62 patients with breast cancer who underwent surgical resection at the Second Affiliated Hospital, Zhejiang University School of Medicine (Zhejiang, China). Normal autologous tissue was obtained from a macroscopically normal part of the excised breast tissue, at least 2 cm away from the tumor. Paraffin-embedded breast cancer specimens were obtained from 516 patients at this hospital. None of the patients had received chemotherapy or endocrine therapy before operation. The clinicopathologic characteristics of 516 patients are presented in Supplementary Table S1. The median follow-up duration of the patients was 89.6 months (range, 24.0–120.0 months). Peripheral blood samples were obtained from healthy donors from the Zhejiang Blood Center, all of whom were negative for antibodies against hepatitis C virus, hepatitis B virus, HIV, and syphilis. All samples were anonymously coded in accordance with local ethical guidelines (as stipulated by the Declaration of Helsinki), and written informed consent was obtained and the protocol was approved by the Review Board of the Second Affiliated Hospital of Zhejiang University School of Medicine. OS was defined as the time from the date of the first curative operation to the date of the last follow-up, or death from any cause; DFS was the time from the date of the first curative surgery to the date of the first locoregional or systemic relapse, or death without any type of relapse.

T-cell isolation and culture

Freshly excised tumor and paired normal breast tissues were cut into small pieces and then digested with the human Tumor Dissociation Kit (Miltenyi Biotec) containing Enzyme H, R, and A, using a gentleMACS Octo-Dissociator with Heaters (Miltenyi Biotec) under the program of “37C_h_TDK_3” according to the manufacturer's instructions for 3 to 4 hours at 37°C. Collagenase type I (1 mg/mL, Gibco) was added when paired normal breast tissues were digested. After the dissociation, single-cell suspensions were filtered through the cell strainer (40 μm, SORFA). Total γδT cells, CD73+ γδT cells, CD4+ Tregs (CD4+CD25+CD127low), CD73+CD4+, and CD73+CD8+ T cells in single-cell suspensions were sorted with fluorochrome-coupled antibodies (Supplementary Table S2) by Aria II cell sorter (BD Biosciences). For CD4+ and CD8+ T-cell isolation, peripheral blood mononuclear cells (PBMC) were derived from peripheral blood of healthy donors with Lymphocyte Separation Medium (Human, Ficoll–Paque, Tbdscience) and were labeled with the human CD4+ T (Miltenyi Biotec, #130-096-533) or CD8+ T-Cell Isolation Kit (Miltenyi Biotec, #130-096-495) and separated by magnetic-activated cell sorting following the manufacturer's instructions. The purity of all sorted cells was greater than 90%.

Fibroblast (CAF and normal tissue–associated fibroblast) isolation

Single-cell suspensions (about 0.5 × 109 cells) from tumor or paired normal breast tissues were prepared as indicated above and subsequently seeded into large flasks containing HAM/F-12 medium (Gibco) supplemented with 10% FBS (Gibco) and 1% penicillin/streptomycin (Tbdscience). After 12 hours, the nonadherent cells were washed away, and the adherent cells were continuously cultured in HAM/F-12 medium (Gibco) supplemented with 10% FBS (Gibco) and 1% penicillin/streptomycin (Tbdscience) for approximately 2 weeks (tumor tissue) or 3 to 4 weeks (paired normal breast tissues). The medium was changed every other day to remove nonadherent cells until large groups of fibroblasts (morphologically spindle-shaped cells, 1 × 106 cells) became apparent.

Flow cytometry

For extracellular staining of immune markers, we prepared single-cell suspensions by mechanic dispersion and enzymatic digestion of normal and tumor tissues as indicated above. We preincubated dissociated tissue cells (1 × 106 cells/mL) in a mixture of PBS, 2% FCS, and 0.1% (w/v) sodium azide (Sigma) with an FcgIII/IIR-specific antibody (BD Pharmingen) to block nonspecific binding for 15 minutes and stained with different combinations of fluorochrome-coupled antibodies (Supplementary Table S2) at room temperature. For intracellular staining of IL2, IL6, IL8, IL10, IL17A, S100A9, IFNγ, TGFβ1, GM-CSF, perforin, granzyme B, and FoxP3, we followed the manufacturer's protocol after a 6-hour incubation in the presence of Leukocyte Activation Cocktail (BD Pharmingen, 4 μL in 1 × 106 cells/mL). Fluorescence data were collected on a FACSCanto II system (BD Biosciences) and analyzed with FlowJo software (Tree Star; Supplementary Fig. S1).

Immunofluoresence

Breast cancer tissue sections were subjected to antigen retrieval in EDTA Buffer (1 mmol/L EDTA, 0.05% Tween 20, pH 8.0) after deparaffinization, sections were incubated with primary antibody against CD73 (ab133582, Abcam, 1:5,000). After washing in PBS, sections were incubated with a horseradish peroxidase (HRP)–labeled secondary antibody (GB23301, Servicebio Technology, 1:500), and then incubated with CY3-TSA (G1223, Servicebio Technology). After washing in EDTA Buffer, the sections were incubated with another primary antibody against EpCAM (ab223582, Abcam, 1:3000), HRP-labeled secondary antibody (GB23301, Servicebio Technology, 1:500), and CY5-TSA (G1224, Servicebio Technology) in proper sequence. After washing in EDTA buffer, sections were incubated with third primary antibody TCRγδ (AB171110, Abcam, 1:500). Finally, sections were counterstained with 4′,6-diamidino-2-phenylindole (DAPI; G1012, Servicebio Technology).

For double staining, sections were incubated with primary antibody against CD8 (66868-1-Ig, Proteintech, 1:500). After washing in PBS, sections were incubated with secondary antibodies (cy3 goat anti-rabbit, GB21303, Servicebio Technology, 1:300). After washing in EDTA buffer, the sections were incubated with another primary antibody against EpCAM (ab223582, Abcam, 1:300) and secondary antibodies (488-goat anti-mouse, GB25301, Servicebio Technology, 1:400). Finally, sections were counterstained with 4′,6-diamidino-2-phenylindole (DAPI; G1012, Servicebio Technology).

Sections were imaged with Pannoramic Digital Slide Scanners (Pannoramic MIDI, 3DHISTECH) to obtain high-power field images. For each section, the number of double-positive (CD73+TCRγδ+) cells or CD8+ cells were analyzed in at least 10 randomly high-power fields in tumor stroma (EpCAM-negative area). We acquired the median number of CD73+ γδT cells and CD8+ T cells for all sections of 516 patients with statistical analysis, respectively. Above the median number was regarded as high density.

siRNA silencing

siRNA-mediated gene silencing was carried out with validated control or IL6-specific siRNAs according to the manufacturer's instructions (Ribobio). The siRNA specific for human IL6 (1# 5′-CCCAGGAGAAGATTCCAAAGATGTA-3′, 2# 5′-GGACAUGACAACUCAUCUCTT-3′) and control siRNA were synthesized by Ribobio. Transfection of siRNA was conducted in 6-well plates with Lipofectamine 3000 (Invitrogen) as described in the manufacturer's instructions. After 48 hours of transfection, the cells were collected to perform real-time PCR analysis to confirm the knockdown efficiency, as indicated below.

RNA extraction and gene expression using quantitative real-time PCR

Tumor and paired normal tissue, CAF, vehicle-pretreated CAF, siRNA-transfected CAF, normal tissue–associated fibroblast (NAF), and tumor cell RNAs were extracted using the RNeasy Mini Kit (Qiagen). Total RNA was quantified by NanoDrop 2000 (Thermo Fisher Scientific), and 500 ng was reverse transcribed into cDNA using the PrimeScript RT Reagent Kit with gDNA Eraser (Takara). The relative mRNA of IL6 in samples was determined by real-time RT-PCR and normalized to GAPDH. 20 μL of each reaction contained 10 μL of iTaq Universal SYBR Green Mix (BIO-RAD), 10 μL of forward and reverse primers (Supplementary Table S3), 2 μL of diluted cDNA, and 6 μL of DEPC water.

For populations with a small number of cells, including CD73 γδT cells from paired normal tissue and CAFs induced by CD73+ γδT cells from tumor tissue, RNAs were reverse transcribed into cDNA by TaqMan PreAmp Cells-to-CT Kit (Life Technologies) according to the manufacturer's instructions. TaqMan Gene Expression Assays (NT5e/CD73, Assay ID: Hs00159686_m1; IL6, Assay ID: Hs00985641_m1) were used to preamplify (10 cycles) objective genes, as well as the housekeeping gene GAPDH (Assay ID: Hs99999905_m1). The preamplified product was diluted 1:5 up to 250 μL in TE buffer, and 5 μL of this dilution was used as template for real-time PCR analysis. Real-time RT-PCR was conducted with the same TaqMan Gene Expression Assays used in the preamplification step and was performed with an ABI 7500 Fast Real-time PCR System (Applied Biosystems). Thermocycling parameters were set as recommended by the instruction manual. Data were calculated by the comparative ΔΔCT method with the GAPDH gene as the endogenous control. Three biological replicates were conducted for each sample, with three technical replicates per repeat.

CD4+ T-cell proliferation assay

For assessing CD73+ γδT cell–mediated CD4+ T-cell suppression, sorted tumor-infiltrating CD73+ γδT (1.0 × 104), CD73 γδT (1.0 × 104), CD4+ Treg (1.0 × 104), CD73+ CD4+ T (1.0 × 104), CD73+ CD8+ T cells (1.0 × 104), and normal tissue–derived CD73+ γδT cells (1.0 × 104) were cocultured with CFSE (Invitrogen)-labeled allogeneic peripheral blood (PB) CD4+ T cells (5.0 × 104) in RPMI1640 Media (Gibco) supplemented with 10% FBS (Gibco) and 1% penicillin/streptomycin (Tbdscience) in the presence of anti-CD3 (Clone HIT3a, 10 μg/mL) and anti-CD28 (Clone CD28.2, 10 μg/mL). At day 6, cells were harvested, and CFSElow CD4+ T cells were detected by flow cytometry, as described above.

ELISAs

For assessing CD73+ γδT cell–mediated CD8+ T-cell suppression, sorted tumor-infiltrating CD73+ γδT (1.0 × 104), CD73 γδT (1.0 × 104), CD4+ Treg (1.0 × 104), CD73+ CD4+ T (1.0 × 104), CD73+ CD8+ T cells (1.0 × 104), and normal tissue–derived CD73+ γδT cells (1.0 × 104) were cocultured with allogeneic PB CD8+ T cells (5.0 × 104) in RPMI1640 Media (Gibco) supplemented with 10% FBS (Gibco) and 1% penicillin/streptomycin (Tbdscience) in 96-well cell culture cluster (Corning Costar) in the presence of anti-CD3 (Clone HIT3a, 10 μg/mL) and anti-CD28 (Clone CD28.2, 10 μg/mL). At day 6, supernatants (300–500 μL, without diluting) were collected for IFNγ and perforin detection by human IFNγ (ab46025, Abcam) or perforin ELISA Kit (ab46068, Abcam).

For determining CD73+ γδT cell–mediated IL6 expression in CAFs, sorted tumor-infiltrating CD73+ γδT cells were cocultured with CAFs (5.0 × 104) at different effector-to-target (E:T) ratios (0:1, 0.1:1, 0.2:1, 0.5:1, 1:1) in HAM/F-12 medium (Gibco) supplemented with 10% FBS (Gibco) and 1% penicillin/streptomycin (Tbdscience) in a 48-well cell culture cluster (Corning Costar) in the presence of anti-CD3 (Clone HIT3a, 10 μg/mL) and anti-CD28 (Clone CD28.2, 10 μg/mL). At day 6, supernatants (300–500 μL, without diluting) were collected for IL6 detection by human IL6 ELISA Kit (ab46027, Abcam). CAFs (adherent cells) were washed and harvested. IL6 mRNA in CAFs was detected by real-time PCR. IL6 concentration in the supernatants derived from tumor, normal tissues, tumor cells, and NAFs were also measured by human IL6 ELISA Kit (Abcam, ab46027).

Blocking assay

Blocking anti-IL10 (Clone JES3-19F1, 0.1 μg/mL, BioLegend), anti-TGFβ (ab50716, 0.1 μg/mL, Abcam), anti-CD73 (Clone 4G4, 20 μg/mL, Abcam), anti–CTLA-4 (Clone L3D10, 0.5 μg/mL, BioLegend), or anti–LAG-3 (Clone 17B4, 0.5 μg/mL, GeneTex); or A2A (SCH58261, CAS No. 160098-96-4, 0.1 mmol/L, Tocris) and A2B (PSB603, CAS No. 1092351-10-4, 0.05 mmol/L, Tocris) adenosine receptor antagonists were used to block the inhibitory function of CD73+ γδT cells in the CD4+ T-cell proliferation assay and CD8+ T-cell ELISAs (IFNγ and perforin secretion) as indicated above.

Extracellular adenosine detection

Adenosine concentrations were measured using a high-performance liquid chromatography system (LC-20) equipped with an Inertsil ODS-SP C18 chromatogram column (4.6 mm × 250 mm, 5 μm; temperature: 30°C) using a mobile phase consisting of acetonitrile and 0.04 mol/L potassium dihydrogen phosphate at the volume ratio of 5:95, and at the flow rate of 0.5 mL/minute, under the equal concentration elution program. Identification and quantification of adenosine peaks were done by comparison with retention times of known standards and peak integration and normalization.

In vitro CD73+ γδT-cell induction

CD73-negative γδT cells (1.0 × 104) isolated from normal breast tissues were cocultured with recombinant human IL6 (concentration gradient: 0, 40, 200, 1,000, 2,000, 5,000, 10,000, 20,000 pg/mL; PeproTech) or TGFβ1 (concentration gradient: 0, 100, 500, 1,000, 2,000, 5,000, 10,000, 20,000 pg/mL; PeproTech) in RPMI1640 Media (Gibco) supplemented with 10% FBS (Gibco) and 1% penicillin/streptomycin (Tbdscience) in a 96-well cell culture cluster (Corning Costar) in the presence of anti-CD3 (Clone HIT3a, 10 μg/mL) and anti-CD28 (Clone CD28.2, 10 μg/mL). Half of the liquid volume was renewed every 3 days. At day 6, NT5e/CD73 expression in γδT cells was detected by real-time PCR or by flow cytometry as indicated above. pSTAT3+ γδT cells were also detected by flow cytometry.

Next, CD73-negative γδT cells (2.0 × 104) isolated from normal breast tissues were cultured in RPMI1640 Media (Gibco) supplemented with 10% FBS (Gibco) and 1% penicillin/streptomycin (Tbdscience), recombinant human IL6 (2,000 pg/mL, PeproTech), tumor supernatants (TS), TS with neutralizing anti-IL6 (Clone 13A5, 0.1 μg/mL, MabTech), or TS with control antibody in a 96-well cell culture cluster (Corning Costar) in the presence of anti-CD3 (Clone HIT3a, 10 μg/mL) and anti-CD28 (Clone CD28.2, 10 μg/mL). pSTAT3 inhibitor (STX-0119, CAS No. 573126, 0.1 mmol/L, Merck) was added to some cultures. Half of the liquid volume was renewed every 3 days. At day 6, CD73+ γδT cells in each well were detected by flow cytometry.

Finally, CD73+ γδT cells (2.0 × 104) isolated from normal breast tissues were cocultured with CAFs (1.0 × 105) in the presence of neutralizing anti-IL6 (Clone 13A5, 0.1 μg/mL, MabTech), control mAb or pSTAT3 inhibitor (STX-0119, CAS No. 573126, 0.1 mmol/L, Merck), vehicle-pretreated CAFs (1.0 × 105), and control siRNA– or IL6-siRNA–transfected CAFs (1.0 × 105) in HAM/F-12 medium (Gibco) supplemented with 10% FBS (Gibco) and 1% penicillin/streptomycin (Tbdscience) in the presence of anti-CD3 (Clone HIT3a, 10 μg/mL) and anti-CD28 (Clone CD28.2, 10 μg/mL). At day 4, discarding the CAFs (adherent cells), CD73+ γδT cells (1.0 × 104) were washed and harvested, then cocultured with CFSE-labeled allogeneic PB CD4+ T cells (5.0 × 104) in the presence of anti-CD3 (Clone HIT3a, 10 μg/mL) and anti-CD28 (Clone CD28.2, 10 μg/mL) for additional 6 days. Supernatants were collected for the adenosine assay, and the CFSElow CD4+ T cells were detected by flow cytometry as described previously.

In vitro CD73+γδTreg regulation of CAFs

CD73+ γδT cells and fibroblasts isolated from tumor tissue were cocultured in HAM/F-12 medium (Gibco) supplemented with 10% FBS (Gibco) and 1% penicillin/streptomycin (Tbdscience) at different E:T ratios (0:1, 0.1:1, 0.2:1, 0.5:1, 1:1) in the presence of anti-CD3 (Clone HIT3a, 10 μg/mL) and anti-CD28 (Clone CD28.2, 10 μg/mL). A1 (KW3902, CAS No. 136199-02-5, 0.05 mmol/L, Tocris), A3 (MRS1220, CAS No. 183721-15-5, 0.05 mmol/L, Tocris), A2A (SCH58261, CAS No. 160098-96-4, 0.1 mmol/L, Tocris), and A2B (PSB603, CAS No. 1092351-10-4, 0.05 mmol/L, Tocris) adenosine receptor antagonists; p38 MAPK (SB203580, CAS No. 152121-47-6, 0.05 mmol/L, Abmole), ERK1/2 (SCH772984, CAS No. 942183-80-4, 0.01 mmol/L, Abmole); or JNK (SP600125, CAS No. 129-56-6, 0.05 mmol/L, Abmole) inhibitors; and anti-CD73 were added to some cultures. At day 6, supernatants were collected for IL6 detection by ELISA as described previously. IL6 mRNA in CAFs was detected by real-time PCR as indicated above, and p38+, p-ERK1/2+, or p-JNK+ CAFs were detected by flow cytometry.

Statistical analysis

Results were exhibited as means ± SEM. Statistical analysis was performed using GraphPad Prism software version 7. The statistical significance of differences between tumor and paired normal tissue groups was determined by paired Student t test, otherwise unpaired Student t test was applied. Kaplan–Meier survival curves were used to look for correlations with survival and were compared with the use of the log-rank statistic. All data were analyzed using two-tailed tests unless otherwise specified, and a P < 0.05 was considered statistically significant.

CD73+ γδT cells are abundant, with potent immunosuppressive activity

In human breast cancer tissues, we found that γδT cells expressed high CD73 (approximately 60%; Fig. 1A) and low CD39 (about 20%; Supplementary Fig. S2), and CD73+ γδT cells were significantly increased in the tumor tissues compared to the paired normal tissues (Fig. 1B). We also found that CD73+ γδT cells were abundant in the tumor tissues, and their absolute numbers within CD3+ T cells was significantly more than conventional CD4+ Tregs (Supplementary Fig. S3), CD73+CD4+ and CD73+CD8+ T cells (Fig. 1C). CD73+ γδT cells were almost CD4 negative (more than 95%; Supplementary Fig. S4), predominately expressed a Vδ1 TCR (Fig. 1D), and expressed high GARP (Glycoprotein A Repetitions Predominant), LAG-3 (Lymphocyte Activation Gene-3), and cytotoxic T-like antigen-4 (CTLA-4; Fig. 1E).

Figure 1.

Tumor-infiltrating CD73+ γδT cells are abundant and express higher immunosuppression-related molecules in human breast cancer. A, Representative flow cytometric analysis of CD73 expression in γδT cells in tumor tissues. n = 28. Flow plots were gated on CD45+CD3+TCRγδ+ cells. B, Representative flow cytometric analysis of CD73+ γδT cells in the tumor and paired normal tissues. Plots were gated on CD45+CD3+ T cells. Bar diagram summarizes the percentages of CD73+ γδT cells in CD45+CD3+ cells. Data, mean ± SEM. n = 28; ***, P < 0.001; paired Student t test. N, normal tissue; T, tumor tissue. C, Representative flow cytometric analysis of CD73+ γδT, CD4+CD73+, and CD8+CD73+ T cells in tumor tissues. Flow plots were gated on CD45+CD3+TCRγδ+, CD45+CD3+CD4+, or CD45+CD3+CD8+ T cells. Bar diagram summarizes the percentages of CD73+ γδT, CD4+CD73+, and CD8+CD73+ T cells within CD45+CD3+ cells, respectively. Data, mean ± SEM. n = 7; *, P < 0.05; **, P < 0.01; unpaired Student t test. D, Single-cell suspensions from tumors were stained with a panel of antibodies and analyzed by flow cytometry. Flow plots were gated on TCRγδ+CD73+ cells. Bar diagram summarizes the percentages of TCRVδ1+ and TCRVδ2+ cells in tumor-infiltrating CD73+ γδT cells. Data, mean ± SEM. n = 6; **, P < 0.01; unpaired Student t test. Abs, antibodies. E, Single-cell suspensions from tumors were stained with a panel of antibodies and analyzed by flow cytometry. Flow plots were gated on CD45+CD3+TCRγδ+CD73+ T cells. Representative histograms indicate the percentages of GARP+ cells, LAG-3+ cells, and CTLA-4+ cells within CD73+ γδT cells, respectively. n = 5. F, Sorted CD73+ γδT cells (1.0 × 104) from tumor tissue were cocultured with CFSE-labeled allogeneic CD4+ T cells (5.0 × 104) in the presence of anti-CD3 and anti-CD28. CD4+ T-cell proliferation was evaluated on day 6 by flow cytometry. Bar diagram summarizes the percentages of proliferated CD4+ T cells (CFSElow). Data, mean ± SEM. n = 5; ***, P < 0.001; unpaired Student t test.

Figure 1.

Tumor-infiltrating CD73+ γδT cells are abundant and express higher immunosuppression-related molecules in human breast cancer. A, Representative flow cytometric analysis of CD73 expression in γδT cells in tumor tissues. n = 28. Flow plots were gated on CD45+CD3+TCRγδ+ cells. B, Representative flow cytometric analysis of CD73+ γδT cells in the tumor and paired normal tissues. Plots were gated on CD45+CD3+ T cells. Bar diagram summarizes the percentages of CD73+ γδT cells in CD45+CD3+ cells. Data, mean ± SEM. n = 28; ***, P < 0.001; paired Student t test. N, normal tissue; T, tumor tissue. C, Representative flow cytometric analysis of CD73+ γδT, CD4+CD73+, and CD8+CD73+ T cells in tumor tissues. Flow plots were gated on CD45+CD3+TCRγδ+, CD45+CD3+CD4+, or CD45+CD3+CD8+ T cells. Bar diagram summarizes the percentages of CD73+ γδT, CD4+CD73+, and CD8+CD73+ T cells within CD45+CD3+ cells, respectively. Data, mean ± SEM. n = 7; *, P < 0.05; **, P < 0.01; unpaired Student t test. D, Single-cell suspensions from tumors were stained with a panel of antibodies and analyzed by flow cytometry. Flow plots were gated on TCRγδ+CD73+ cells. Bar diagram summarizes the percentages of TCRVδ1+ and TCRVδ2+ cells in tumor-infiltrating CD73+ γδT cells. Data, mean ± SEM. n = 6; **, P < 0.01; unpaired Student t test. Abs, antibodies. E, Single-cell suspensions from tumors were stained with a panel of antibodies and analyzed by flow cytometry. Flow plots were gated on CD45+CD3+TCRγδ+CD73+ T cells. Representative histograms indicate the percentages of GARP+ cells, LAG-3+ cells, and CTLA-4+ cells within CD73+ γδT cells, respectively. n = 5. F, Sorted CD73+ γδT cells (1.0 × 104) from tumor tissue were cocultured with CFSE-labeled allogeneic CD4+ T cells (5.0 × 104) in the presence of anti-CD3 and anti-CD28. CD4+ T-cell proliferation was evaluated on day 6 by flow cytometry. Bar diagram summarizes the percentages of proliferated CD4+ T cells (CFSElow). Data, mean ± SEM. n = 5; ***, P < 0.001; unpaired Student t test.

Close modal

Studies have demonstrated that CD73 expressed on T cells endows them with immunosuppressive properties (10, 11). Therefore, we next examined their immunosuppressive function on effector T cells. We sorted CD73+ γδT cells from tumor tissues and cocultured them with allogeneic peripheral blood CD4+ T cells. The proliferative property of CD4+ T cells significantly decreased, suggesting that CD73+ γδT cells exerted potent immunosuppressive activity (Fig. 1F).

CD73+ γδT cells express higher immunosuppression-related molecules

Because CD73+ γδT cells have not been reported in breast cancer tumor tissues, we then examined their phenotype compared with the paired normal tissues. Tumor-infiltrating CD73+ γδT cells had higher expression of GARP, LAG-3, CTLA-4, CD103, CD122, CD39, as well as CD44, whereas they expressed significantly lower immune effector marker NKp44, NKp30, costimulation marker CD83, and CD26/ADA (Fig. 2A and B). Other markers such as FOXP3, glucocorticoid-induced tumor necrosis factor receptor related gene (GITR), PD-1, CD25, CD38, and CD203a were not significantly different (Supplementary Fig. S5A).

Figure 2.

Phenotype of CD73+ γδT cells in breast cancer versus paired normal tissue. A, Representative flow cytometric analysis of CD73+ γδT-cell phenotype in tumor and paired normal tissues. n = 5. Flow plots were gated on CD45+CD3+TCRγδ+CD73+ cells. B, Bar diagrams summarize the percentages of the indicated markers in CD73+ γδT cells. A and B, Data, mean ± SEM. n = 5; *, P < 0.05; **, P < 0.01; ***, P < 0.001; paired Student t test. C, Representative flow cytometric analysis of cytokine production by CD73+ γδT cells in tumor and paired normal tissues. n = 5. Flow plots were gated on CD45+CD3+TCRγδ+CD73+ cells. D, Bar diagrams summarize the percentages of the indicated cytokines in CD73+ γδT cells. C and D, Data, mean ± SEM. n = 5; *, P < 0.05; **, P < 0.01; paired Student t test. N, normal tissue; T, tumor tissue.

Figure 2.

Phenotype of CD73+ γδT cells in breast cancer versus paired normal tissue. A, Representative flow cytometric analysis of CD73+ γδT-cell phenotype in tumor and paired normal tissues. n = 5. Flow plots were gated on CD45+CD3+TCRγδ+CD73+ cells. B, Bar diagrams summarize the percentages of the indicated markers in CD73+ γδT cells. A and B, Data, mean ± SEM. n = 5; *, P < 0.05; **, P < 0.01; ***, P < 0.001; paired Student t test. C, Representative flow cytometric analysis of cytokine production by CD73+ γδT cells in tumor and paired normal tissues. n = 5. Flow plots were gated on CD45+CD3+TCRγδ+CD73+ cells. D, Bar diagrams summarize the percentages of the indicated cytokines in CD73+ γδT cells. C and D, Data, mean ± SEM. n = 5; *, P < 0.05; **, P < 0.01; paired Student t test. N, normal tissue; T, tumor tissue.

Close modal

We also found that tumor-infiltrating CD73+ γδT cells produced significantly more IL6, GM-CSF, IL8, IL17A, S100A9, and less IFNγ (Fig. 2C and D). However, CD73+ γδT cells from tumor and paired normal tissues secreted lower IL10, TGFβ1, IL2, perforin, and granzyme B (Supplementary Fig. S5B).

CD73+γδT are the predominant Tregs in human breast cancer

Our initial data indicated that tumor-infiltrating CD73+ γδT cells had potent immunosuppressive activity, so we next examined the suppressive T-cell landscape in tumor tissues. We found that tumor-derived CD73+ γδT cells exerted more inhibitory function on CD4+ T-cell proliferation than conventional CD4+ Tregs, CD73+CD4+ and CD73+CD8+ T cells, and CD73 γδT cells, and cells derived from paired normal tissues (Fig. 3A and B). We also found that tumor-infiltrating CD73+ γδT cells inhibited perforin (Fig. 3C) and IFNγ (Supplementary Fig. S6) production by CD8+ T cells compared with conventional CD4+ Tregs, CD73+CD4+ T cells, and CD73+CD8+ T cells. Taken together, these data suggest that CD73+ γδT cells are the predominant Tregs in human breast cancer, not only in quantity but also in immunosuppressive ability.

Figure 3.

CD73+ γδT cells are the predominant Tregs in breast cancer and exhibit direct immunosuppression via the adenosine-mediated pathway. A and B, Sorted CD73+ γδT cells, CD4+ Tregs (CD45+CD3+CD4+CD25+CD127low), CD4+CD73+ and CD8+CD73+ T cells, and CD73 γδT cells (1.0 × 104 cells, respectively) from tumor tissues and CD73+ γδT cells (1.0 × 104) from paired normal tissue were cocultured with CFSE-labeled allogeneic CD4+ T cells (5.0 × 104) in the presence of anti-CD3 and anti-CD28. n = 5. A, CD4+ T-cell proliferation was evaluated on day 6 by flow cytometry. B, Bar diagram summarizes the percentages of proliferated CD4+ T cells (CFSElow). Data, mean ± SEM. n = 5; **, P < 0.01; ***, P < 0.001; paired Student t test for the analysis between CD73+ γδT cells from tumor and paired normal tissue and unpaired Student t test for the analysis among other cells from tumor tissue. C, Sorted CD73+ γδT, CD4+ Tregs, CD4+CD73+ and CD8+CD73+ T cells, and CD73 γδT cells (1.0 × 104 cells, respectively) from tumor tissues and CD73+ γδT cells (1.0 × 104) from paired normal tissue were cocultured with allogeneic CD8+ T cells (5.0 × 104) in the presence of anti-CD3 and anti-CD28. Concentrations of perforin in the supernatants were detected on day 6 by ELISA. Data, mean ± SEM. n = 5; *, P < 0.05; ***, P < 0.001; paired Student t test for the analysis between CD73+ γδT cells from tumor and paired normal tissue and unpaired Student t test for the analysis among other cells from tumor tissue. D and E, Sorted CD73+ γδT cells (1.0 × 104) from tumor tissues were cocultured with CFSE-labeled allogeneic CD4+ T cells (5.0 × 104) in the presence of anti-CD3 and anti-CD28 and pretreated with anti-CD73, an antibody cocktail, control antibody, or A2A (SCH58261) and A2B (PSB603) adenosine receptor antagonists. n = 5. D, CD4+ T-cell proliferation was evaluated on day 6 by flow cytometry. n = 5. E, Bar diagram summarizes the percentages of proliferated CD4+ T cells (CFSElow). Data, mean ± SEM. n = 5; ns, no significance; ***, P < 0.001; unpaired Student t test. mAb cocktail, the mixture of antibodies for LAG-3, CTLA-4, IL10, and TGFβ; N, normal tissue; T, tumor tissue.

Figure 3.

CD73+ γδT cells are the predominant Tregs in breast cancer and exhibit direct immunosuppression via the adenosine-mediated pathway. A and B, Sorted CD73+ γδT cells, CD4+ Tregs (CD45+CD3+CD4+CD25+CD127low), CD4+CD73+ and CD8+CD73+ T cells, and CD73 γδT cells (1.0 × 104 cells, respectively) from tumor tissues and CD73+ γδT cells (1.0 × 104) from paired normal tissue were cocultured with CFSE-labeled allogeneic CD4+ T cells (5.0 × 104) in the presence of anti-CD3 and anti-CD28. n = 5. A, CD4+ T-cell proliferation was evaluated on day 6 by flow cytometry. B, Bar diagram summarizes the percentages of proliferated CD4+ T cells (CFSElow). Data, mean ± SEM. n = 5; **, P < 0.01; ***, P < 0.001; paired Student t test for the analysis between CD73+ γδT cells from tumor and paired normal tissue and unpaired Student t test for the analysis among other cells from tumor tissue. C, Sorted CD73+ γδT, CD4+ Tregs, CD4+CD73+ and CD8+CD73+ T cells, and CD73 γδT cells (1.0 × 104 cells, respectively) from tumor tissues and CD73+ γδT cells (1.0 × 104) from paired normal tissue were cocultured with allogeneic CD8+ T cells (5.0 × 104) in the presence of anti-CD3 and anti-CD28. Concentrations of perforin in the supernatants were detected on day 6 by ELISA. Data, mean ± SEM. n = 5; *, P < 0.05; ***, P < 0.001; paired Student t test for the analysis between CD73+ γδT cells from tumor and paired normal tissue and unpaired Student t test for the analysis among other cells from tumor tissue. D and E, Sorted CD73+ γδT cells (1.0 × 104) from tumor tissues were cocultured with CFSE-labeled allogeneic CD4+ T cells (5.0 × 104) in the presence of anti-CD3 and anti-CD28 and pretreated with anti-CD73, an antibody cocktail, control antibody, or A2A (SCH58261) and A2B (PSB603) adenosine receptor antagonists. n = 5. D, CD4+ T-cell proliferation was evaluated on day 6 by flow cytometry. n = 5. E, Bar diagram summarizes the percentages of proliferated CD4+ T cells (CFSElow). Data, mean ± SEM. n = 5; ns, no significance; ***, P < 0.001; unpaired Student t test. mAb cocktail, the mixture of antibodies for LAG-3, CTLA-4, IL10, and TGFβ; N, normal tissue; T, tumor tissue.

Close modal

Previous studies have shown that CD73, as an ectonucleotidase, can hydrolyze extracellular AMP to adenosine (12, 13), which can alter the activity of immune cells (14). We found that tumor-derived CD73+ γδT cells could produce more adenosine than cells from paired normal tissues (Supplementary Fig. S7A). Expectedly, in vitro blocking assays indicated that CD8+ (Supplementary Fig. S8) and CD4+ T-cell inhibition by CD73+ γδTregs was almost completely blocked by two adenosine receptor antagonists, A2A (SCH58261) and A2B (PSB603), or by anti-CD73 (Fig. 3D and E), suggesting that the CD73/adenosine-mediated signaling pathway played a critical role in immunosuppression of CD73+ γδTregs. Cancer cells derived from tumor tissue had low expression of CD73 compared with γδT cells, and the CD73 MFI of tumor-infiltrating CD73+ γδTregs was higher than that of CD73+CD4+ T cells, CD73+CD8+ T cells, CD73+ cancer cells, and normal tissue–derived CD73+ γδT cells (Supplementary Fig. S7B and S7C). Taken together, these results implied that γδT cell-derived CD73 might possibly make a significant contribution to extracellular adenosine production in the TME of breast cancer.

IL6, rather than TGFβ1, induces CD73 expression in γδT cells

A previous study has demonstrated that IL6 and TGFβ1 play an important role in inducing ectonucleotidase CD73 (encoded by Nt5e) expression on Th17 cells from mice (15). To determine the role of these two cytokines in NT5e expression in CD73 γδT cells, we directly stimulated CD73 γδT cells from the normal tissues with varying concentrations of IL6 and TGFβ1, respectively. We noted that at least 2 ng/mL of IL6 was required to induce NT5e expression; however, higher doses of IL6 could not further enhance NT5e expression (Fig. 4A). Unexpectedly, we noted that TGFβ1 was not required in inducing NT5e expression in CD73 γδT cells (Supplementary Fig. S9).

Figure 4.

IL6 rather than TGFβ1 in the tumor microenvironment induces CD73+ γδTregs via IL6/STAT3 pathway. A, Sorted CD73 γδT cells (1.0 × 104) from paired normal tissue were stimulated with varying concentrations of IL6 in the presence of anti-CD3 and anti-CD28. Cells were harvested on day 6, and the relative mRNA expression of NT5e was determined by RT-PCR and normalized to GAPDH. Data, mean ± SEM. n = 5; ns, no significance; **, P < 0.01; unpaired Student t test. B, The relative mRNA level of IL6 in the tumor and paired normal tissues was determined by RT-PCR and normalized to GAPDH. C, Concentrations of IL6 in the tumor and paired normal tissue–derived supernatants were detected by ELISA. B and C, Data, mean ± SEM. n = 5; **, P < 0.01; paired Student t test. D, Sorted CD73 γδT cells (2.0 × 104) from paired normal tissue were cultured in different conditions [medium or varying concentrations of IL6 (1–2 ng/mL)] in the presence of anti-CD3 and anti-CD28 for 6 days. The expression of p-STAT3 in γδT cells was analyzed by flow cytometry. Data, mean ± SEM. n = 5; ns, no significance; ***, P < 0.001; paired Student t test. E, Sorted CD73 γδT cells (2.0 × 104) from paired normal tissue were cultured in different conditions [medium, TS, TS with anti-IL6, TS with control antibody, IL6 (2 ng/mL), or IL6 with STX-0119] in the presence of anti-CD3 and anti-CD28. The expression of CD73 in γδT cells was analyzed by flow cytometry. Data, mean ± SEM. n = 5; ns, no significance; **, P < 0.01; ***, P < 0.001; unpaired Student t test. N, normal tissue; NS, normal (breast tissue) supernatant; STX-0119, p-STAT3 inhibitor; T, tumor tissue; TS, tumor supernatant.

Figure 4.

IL6 rather than TGFβ1 in the tumor microenvironment induces CD73+ γδTregs via IL6/STAT3 pathway. A, Sorted CD73 γδT cells (1.0 × 104) from paired normal tissue were stimulated with varying concentrations of IL6 in the presence of anti-CD3 and anti-CD28. Cells were harvested on day 6, and the relative mRNA expression of NT5e was determined by RT-PCR and normalized to GAPDH. Data, mean ± SEM. n = 5; ns, no significance; **, P < 0.01; unpaired Student t test. B, The relative mRNA level of IL6 in the tumor and paired normal tissues was determined by RT-PCR and normalized to GAPDH. C, Concentrations of IL6 in the tumor and paired normal tissue–derived supernatants were detected by ELISA. B and C, Data, mean ± SEM. n = 5; **, P < 0.01; paired Student t test. D, Sorted CD73 γδT cells (2.0 × 104) from paired normal tissue were cultured in different conditions [medium or varying concentrations of IL6 (1–2 ng/mL)] in the presence of anti-CD3 and anti-CD28 for 6 days. The expression of p-STAT3 in γδT cells was analyzed by flow cytometry. Data, mean ± SEM. n = 5; ns, no significance; ***, P < 0.001; paired Student t test. E, Sorted CD73 γδT cells (2.0 × 104) from paired normal tissue were cultured in different conditions [medium, TS, TS with anti-IL6, TS with control antibody, IL6 (2 ng/mL), or IL6 with STX-0119] in the presence of anti-CD3 and anti-CD28. The expression of CD73 in γδT cells was analyzed by flow cytometry. Data, mean ± SEM. n = 5; ns, no significance; **, P < 0.01; ***, P < 0.001; unpaired Student t test. N, normal tissue; NS, normal (breast tissue) supernatant; STX-0119, p-STAT3 inhibitor; T, tumor tissue; TS, tumor supernatant.

Close modal

We also found that the mRNA and the protein levels of IL6 in the breast cancer tissues were significantly higher than those in the paired normal tissues (Fig. 4B and C). Expectedly, tumor supernatants, as well as IL6, could induce CD73 expression on CD73 γδT cells from normal tissues (Fig. 4E). CD39 was significantly upregulated, whereas CD26/ADA was downregulated in CD73+ γδT cells, and there were no significant differences on expression of other ectonucleotidases, including CD38 and CD203a, in CD73+ γδT cells when treated with tumor supernatants (Supplementary Fig. S10). We also found that p-STAT3 expression in CD73 γδT cells was significantly upregulated when stimulated with 2 ng/mL of IL6 (Fig. 4D), and further in vitro blocking assays demonstrated that IL6-mediated CD73 expression in CD73 γδT cells could be almost completely inhibited by a specific STAT3 inhibitor (STX-0119; Fig. 4E). Taken together, these data suggest that the IL6/STAT3 signaling pathway is required for the induction of CD73+ γδT cells.

CAF-derived IL6 induces CD73+γδTreg differentiation

IL6 can be secreted by different cell subsets including tumor cells (16) and fibroblasts (17). Indeed, breast CAFs expressed higher mRNA and protein levels of IL6 than NAFs, even more than tumor cells from breast cancer tissues. Knockdown of IL6 by siRNA significantly reduced both the mRNA and protein expression levels of IL6 in CAFs (Fig. 5A and B). CD73+ γδT cells from normal tissues cocultured with CAFs inhibited CD4+ T-cell proliferation and had significantly more adenosine secretion, whereas an IL6-neutralizing antibody, IL6 siRNA knockdown, or addition of a specific STAT3 inhibitor (STX-0119) significantly decreased the property of CAFs in the induction of CD73+ γδTregs (Fig. 5C–E). Taken together, these data suggest that CAF-derived IL6 induces CD73+ γδTregs via the IL6/STAT3 pathway with increased exogenous adenosine production, thus exhibiting more potent immunosuppressive activity.

Figure 5.

CAF-derived IL6 plays a pivotal role in the differentiation of CD73+ γδTregs. A, The relative mRNA expression of IL6 in NAFs, TCs, CAFs, vehicle-pretreated CAFs, control siRNA–transfected CAFs, and IL6-siRNA–transfected CAFs was determined by RT-PCR and normalized to GAPDH. B, Concentrations of IL6 in supernatants derived from NAFs, TCs, CAFs, vehicle-pretreated CAFs, control siRNA–transfected CAFs, and IL6-siRNA–transfected CAFs were detected by ELISAs. A and B, Data, mean ± SEM. n = 5; ns, no significance; **, P < 0.01, ***, P < 0.001; paired Student t test for expression of IL6 in NAFs and CAFs and unpaired Student t test for analysis among other cells. TC, tumor cell. C–E, Sorted CD73+ γδT cells (2.0 × 104) from paired normal tissue were cocultured with CAFs (1.0 × 105) in the presence of anti-IL6 or control antibody, vehicle-pretreated CAFs, and control siRNA– or IL6-siRNA–transfected CAFs in the presence of anti-CD3 and anti-CD28. CD73+ γδT cells were harvested on day 4, and cocultured with CFSE-labeled allogeneic CD4+ T cells in the presence of anti-CD3 and anti-CD28 for additional 6 days. pSTAT3 inhibitor (STX-0119, 0.1 mmol/L) was added to some cultures. C, CD4+ T-cell proliferation was evaluated by flow cytometry. D, Bar diagram summarizes the percentages of proliferated CD4+ T cells (CFSElow). E, The concentrations of adenosine in the supernatants of cocultures were detected by high-performance liquid chromatography. Data, mean ± SEM. n = 5; ns, no significance; **, P < 0.01; ***, P < 0.001; unpaired Student t test.

Figure 5.

CAF-derived IL6 plays a pivotal role in the differentiation of CD73+ γδTregs. A, The relative mRNA expression of IL6 in NAFs, TCs, CAFs, vehicle-pretreated CAFs, control siRNA–transfected CAFs, and IL6-siRNA–transfected CAFs was determined by RT-PCR and normalized to GAPDH. B, Concentrations of IL6 in supernatants derived from NAFs, TCs, CAFs, vehicle-pretreated CAFs, control siRNA–transfected CAFs, and IL6-siRNA–transfected CAFs were detected by ELISAs. A and B, Data, mean ± SEM. n = 5; ns, no significance; **, P < 0.01, ***, P < 0.001; paired Student t test for expression of IL6 in NAFs and CAFs and unpaired Student t test for analysis among other cells. TC, tumor cell. C–E, Sorted CD73+ γδT cells (2.0 × 104) from paired normal tissue were cocultured with CAFs (1.0 × 105) in the presence of anti-IL6 or control antibody, vehicle-pretreated CAFs, and control siRNA– or IL6-siRNA–transfected CAFs in the presence of anti-CD3 and anti-CD28. CD73+ γδT cells were harvested on day 4, and cocultured with CFSE-labeled allogeneic CD4+ T cells in the presence of anti-CD3 and anti-CD28 for additional 6 days. pSTAT3 inhibitor (STX-0119, 0.1 mmol/L) was added to some cultures. C, CD4+ T-cell proliferation was evaluated by flow cytometry. D, Bar diagram summarizes the percentages of proliferated CD4+ T cells (CFSElow). E, The concentrations of adenosine in the supernatants of cocultures were detected by high-performance liquid chromatography. Data, mean ± SEM. n = 5; ns, no significance; **, P < 0.01; ***, P < 0.001; unpaired Student t test.

Close modal

CD73+γδTregs promote IL6 expression in CAFs

Previous data have shown that CAFs can interact with immune cells, such as tumor-associated macrophages (TAM), thereby promoting tumor progression of neuroblastoma and prostate cancer (18, 19). Our prior work has demonstrated that CAFs could induce CD73+ γδTreg differentiation. We next investigate whether CD73+ γδTregs could interact with CAFs in the TME. We sorted CD73+ γδT cells and fibroblasts from tumor tissues and cocultured them at different ratios. Tumor-infiltrating CD73+ γδTregs significantly promoted IL6 expression in CAFs both at the mRNA and protein levels, at least under the ratio of 0.5:1 (Fig. 6A and B).

Figure 6.

CD73+ γδTregs upregulate IL6 expression in CAFs via the adenosine/A2BR/p38MAPK pathway. A and B, Sorted CD73+ γδT cells from tumor tissue were cocultured with CAFs (5.0 × 104) in different E:T ratios (0:1, 0.1:1, 0.2:1, 0.5:1, or 1:1) in the presence of anti-CD3 and anti-CD28 for 6 days. n = 5; unpaired Student t test. A, The relative mRNA expression of IL6 in CAFs was determined by RT-PCR and normalized to GAPDH. n = 5; unpaired Student t test. B, Concentrations of IL6 in supernatants were detected by ELISAs. Data, mean ± SEM. n = 5; ns, no significance; **, P < 0.01; unpaired Student t test. C–E, Sorted CD73+ γδT cells from tumor tissues were cocultured with CAFs (5.0 × 104) in different E:T ratios (0:1, 0.2:1, or 0.5:1) in the presence of anti-CD3 and anti-CD28 for 6 days. The expression of p38 (C), ERK1/2 (D), and JNK (E) in CAFs was analyzed by flow cytometry. Bar diagrams summarize the percentages of p38, ERK1/2, and JNK in CAFs. Data, mean ± SEM. n = 5; ns, no significance; **, P < 0.01; unpaired Student t test. F and G, Sorted CD73+ γδT cells (2.5 × 104) from tumor tissues were cocultured with CAFs (5.0 × 104) in different conditions [medium, A1 (KW3902), A3 (MRS1220), A2A (SCH58261), and A2B (PSB603) adenosine receptor antagonists, p38 MAPK (SB203580), ERK1/2 (SCH772984), or JNK (SP600125) inhibitor, anti-CD73 or control antibody] in the presence of anti-CD3 and anti-CD28 for 6 days. F, The relative mRNA expression of IL6 in CAFs was determined by RT-PCR and normalized to GAPDH. G, Concentrations of IL6 in supernatants were detected by ELISAs. Data, mean ± SEM. n = 5; ns, no significance; **, P < 0.01.

Figure 6.

CD73+ γδTregs upregulate IL6 expression in CAFs via the adenosine/A2BR/p38MAPK pathway. A and B, Sorted CD73+ γδT cells from tumor tissue were cocultured with CAFs (5.0 × 104) in different E:T ratios (0:1, 0.1:1, 0.2:1, 0.5:1, or 1:1) in the presence of anti-CD3 and anti-CD28 for 6 days. n = 5; unpaired Student t test. A, The relative mRNA expression of IL6 in CAFs was determined by RT-PCR and normalized to GAPDH. n = 5; unpaired Student t test. B, Concentrations of IL6 in supernatants were detected by ELISAs. Data, mean ± SEM. n = 5; ns, no significance; **, P < 0.01; unpaired Student t test. C–E, Sorted CD73+ γδT cells from tumor tissues were cocultured with CAFs (5.0 × 104) in different E:T ratios (0:1, 0.2:1, or 0.5:1) in the presence of anti-CD3 and anti-CD28 for 6 days. The expression of p38 (C), ERK1/2 (D), and JNK (E) in CAFs was analyzed by flow cytometry. Bar diagrams summarize the percentages of p38, ERK1/2, and JNK in CAFs. Data, mean ± SEM. n = 5; ns, no significance; **, P < 0.01; unpaired Student t test. F and G, Sorted CD73+ γδT cells (2.5 × 104) from tumor tissues were cocultured with CAFs (5.0 × 104) in different conditions [medium, A1 (KW3902), A3 (MRS1220), A2A (SCH58261), and A2B (PSB603) adenosine receptor antagonists, p38 MAPK (SB203580), ERK1/2 (SCH772984), or JNK (SP600125) inhibitor, anti-CD73 or control antibody] in the presence of anti-CD3 and anti-CD28 for 6 days. F, The relative mRNA expression of IL6 in CAFs was determined by RT-PCR and normalized to GAPDH. G, Concentrations of IL6 in supernatants were detected by ELISAs. Data, mean ± SEM. n = 5; ns, no significance; **, P < 0.01.

Close modal

It is known that all adenosine receptors (e.g., A1, A2A, A2B, and A3) physiologically couple to MAPK pathways, which include extracellular signal-regulated kinase 1/2 (ERK1/2), Jun N-terminal kinase (JNK), and p38 MAPK (20), and all these adenosine receptors are expressed on fibroblasts (21). We noted that the expression of p38 MAPK in CAFs, which expressed very low CD73 and CD39 (Supplementary Fig. S11), significantly increased when cocultured with tumor-infiltrating CD73+ γδT cells, whereas the ERK1/2 and JNK were not upregulated (Fig. 6CE). These data suggested that p38 MAPK, rather than ERK1/2 and JNK, might be involved in the adenosine-mediated pathway.

We next further confirmed the involvement of p38 MAPK and the adenosine receptor in IL6 expression in CAFs induced by CD73+ γδTregs. An in vitro blocking assay showed that either a p38 MAPK inhibitor (SB203580) or an adenosine A2B receptor antagonist (PSB603) could significantly decrease the mRNA and protein expression of IL6, whereas the other adenosine receptor antagonists A1 (KW3902), A2A (SCH58261), or A3 (MRS1523) and ERK1/2 (SCH772984) or JNK (SP600125) inhibitors could not block IL6 expression in CAFs. We also observed that a CD73 blocking antibody almost completely inhibited the mRNA and protein expression of IL6 in CAFs induced by CD73+ γδTregs (Fig. 6F and G). Taken together, these results showed that CD73+ γδTregs upregulated IL6 expression in CAFs via the adenosine/A2BR/p38MAPK signaling pathway.

Tumor-infiltrating CD73+γδTregs predict worse clinical outcome in breast cancer

Because tumor-infiltrating CD73+ γδTregs were the predominant Tregs in human breast cancer, we next investigated the association between these cells and clinical outcomes, including OS and DFS. We analyzed CD73+ γδTregs (Fig. 7A) and clinical data from 516 patients. Patients with a low density of these cells in the tumor stroma had significantly longer OS and DFS compared with those with a high density of these cells. (Fig. 7C and D) In stratified analyses based on molecular typing of breast cancer, the results indicated that the infiltration of CD73+ γδTregs significantly reduced OS and DFS in triple-negative, Her-2–positive, and Luminal B breast cancer subtypes (Supplementary Fig. S12).

Figure 7.

Tumor-infiltrating CD73+ γδTregs predict worse clinical outcome and impede the prognostic impact of CD8+ T cells in patients with breast cancer. A, Breast cancer tissue sections of 516 samples were stained with antibodies for TCR γδ (green), EpCAM (dark pink), and CD73 (red). Nuclei were stained with DAPI (blue). Representative image indicates that cells positive for both CD73 and TCRγδ in EpCAM-negative areas were considered as stromal CD73+ γδTregs. White arrows indicate target cells (stromal CD73+ γδTregs). Scale bar, 50 μm. B, Breast cancer tissue sections of 516 samples were stained with antibodies for CD8 (green) and EpCAM (pink). Nuclei were stained with DAPI (blue). Representative image indicates that cells positive for CD8 in EpCAM-negative areas were considered as stromal CD8+ T cells. White arrows indicate target cells (stromal CD8+ T cells). Scale bar, 50 μm. C and D, The Kaplan–Meier survival curves/log-rank tests were used to compare OS (C) and DFS (D) in groups with high and low numbers of CD73+ γδTregs; median: 16.8 cells. E and F, Kaplan–Meier survival curves for the CD73+ γδTreg–high cohort of patients with breast cancer. Log-rank tests were used to compare OS (E) and DFS (F) in groups with high and low numbers of CD8+ T cells; median: 22.3 cells. G and H, Kaplan–Meier survival curves for the CD73+ γδTreg–low cohort of patients with breast cancer. Log-rank tests were used to compare OS (G) and DFS (H) in groups with high and low numbers of CD8+ T cells; median: 22.3 cells.

Figure 7.

Tumor-infiltrating CD73+ γδTregs predict worse clinical outcome and impede the prognostic impact of CD8+ T cells in patients with breast cancer. A, Breast cancer tissue sections of 516 samples were stained with antibodies for TCR γδ (green), EpCAM (dark pink), and CD73 (red). Nuclei were stained with DAPI (blue). Representative image indicates that cells positive for both CD73 and TCRγδ in EpCAM-negative areas were considered as stromal CD73+ γδTregs. White arrows indicate target cells (stromal CD73+ γδTregs). Scale bar, 50 μm. B, Breast cancer tissue sections of 516 samples were stained with antibodies for CD8 (green) and EpCAM (pink). Nuclei were stained with DAPI (blue). Representative image indicates that cells positive for CD8 in EpCAM-negative areas were considered as stromal CD8+ T cells. White arrows indicate target cells (stromal CD8+ T cells). Scale bar, 50 μm. C and D, The Kaplan–Meier survival curves/log-rank tests were used to compare OS (C) and DFS (D) in groups with high and low numbers of CD73+ γδTregs; median: 16.8 cells. E and F, Kaplan–Meier survival curves for the CD73+ γδTreg–high cohort of patients with breast cancer. Log-rank tests were used to compare OS (E) and DFS (F) in groups with high and low numbers of CD8+ T cells; median: 22.3 cells. G and H, Kaplan–Meier survival curves for the CD73+ γδTreg–low cohort of patients with breast cancer. Log-rank tests were used to compare OS (G) and DFS (H) in groups with high and low numbers of CD8+ T cells; median: 22.3 cells.

Close modal

Given the potent immunosuppressive function of tumor-infiltrating CD73+ γδTregs on CD8+ T cells, we next coanalyzed the two T-cell subpopulations in the tumor stroma of all patients, and discovered that for tumors with a high density of CD73+ γδTregs, the presence of CD8+ T cells (Fig. 7B) was no longer associated with favorable prognosis (Fig. 7E and F), whereas in patients with a low density of CD73+ γδTregs, CD8+ T cells significantly improved survival (Fig. 7G and H). These data indicated that CD73+ γδTregs hindered the prognostic impact of CD8+ T cells. Hence, the presence of tumor-infiltrating CD73+ γδTregs is a prognostic factor in human breast cancer.

Tregs are considered to be a pivotal mediator of immune suppression in the TME. In this study, we identified CD73+ γδTregs to be significantly increased in human breast cancer. They exhibited more potent immunosuppressive activity than conventional CD4+ Tregs or CD8+ Tregs. CAF-derived IL6, rather than TGFβ1, induced CD73+ γδTreg differentiation with more exogenous adenosine production via the IL6/STAT3 signaling pathway, and CD73+ γδTregs could in turn promote IL6 secretion by CAFs through activating the adenosine/A2BR/p38MAPK pathway thereby forming a positive feedback loop. The infiltration of CD73+ γδTreg significantly correlated with worse clinical outcomes and hindered the prognostic impact of CD8+ T cells in human breast cancer. These findings suggest that tumor-infiltrating CD73+ γδTregs are critical Tregs, and the positive feedback loop between these cells and CAFs is important to promote immune suppression and tumor progression in human breast cancer.

The TME promotes tumor progression and metastasis through multiple mechanisms such as the induction of Tregs and immune evasion (22, 23). Conventional CD4+ Tregs are thought to be the major immunosuppressive T cells in many types of cancer; however, the role of CD4+ Tregs in human breast cancer is not conclusive. Although previous data shows γδ1T cells are the major immunosuppressive T cells in quantity in breast cancer, these cells are just from generated T-cell clones rather than directly isolated from the tumor tissues. More importantly, the authors did not compare the suppressive properties of γδ1T cells with other Tregs (24). In this study, we examined the suppressive T-cell landscape in the TME of breast cancer and showed that CD73+ γδTregs were the predominant Tregs, not only in quantity but also in quality. CD73+ γδTregs were expanded in the TME, and the data support that they were induced and differentiated by CAF-derived IL6 through activation of the transcription factor STAT3. CD73 γδT cells of paired normal breast tissues could be differentiated into CD73+ γδTregs by direct stimulation with IL6 rather than TGFβ1, and this induction could be blocked by neutralizing anti-IL6 or a p-STAT3 inhibitor. We also showed that cocultures of CD73+ γδT cells from normal breast tissues with IL6-producing CAFs, but not IL6-knockdown CAFs or the addition of CAFs, induced CD73+ γδTregs. This is different from previous studies showing that circulating CD73+ Vγ9Vδ2 T cells from healthy donors are induced by IL21 (25), and CD73 expression in γδT cells derived from colon tissues can be induced by TGFβ1 (9), further highlighting the differences between breast and other tissues.

Studies have demonstrated that CAFs can induce Tregs in the TME through secretion of cytokines and chemokines (26–28). However, there is only one study showing that Tregs can affect CAFs, arresting CAFs at the G2–M phase with their metabolites (29). In this study, we found that CD73+ γδTregs from breast cancer tissues produced a large amount of adenosine, and could induce IL6 expression in CAFs. Adenosine signaling pathway plays an important role in immunity and tumor progression (30). However, different adenosine receptors stimulated by adenosine activate differential downstream pathways in immune cells in specific microenvironments (31). More importantly, the effect of adenosine on CAFs and its mechanism has not been reported. In this study, we demonstrated that the CD73+ γδTregs promoted IL6 secretion by CAFs through producing high extracellular adenosine, which activated the A2B receptor and p38 MAPK and were independent of other adenosine receptors or ERK1/2 or JNK. This is different from the previous data showing IL6 secretion by CAFs via mTOR pathway (32), and that fibroblasts are activated by adenosine through activating A2A receptor in collagen production (33) and pathogenesis of diseases such as cardiac fibrosis (34).

Adenosine plays a crucial role in establishing the immunosuppressive environment directly through activation of its receptor (35, 36). In this study, we demonstrated that CD73 expressed on tumor-infiltrating γδT cells functioned as a ectonucleotidase (37), and the inhibitory effect of CD73+ γδTregs was found to function via CD73 and A2A or A2B receptor (38, 39). This is consistent with a previous report that CD73-expressing γδT cells functioned via an adenosine-mediated pathway in experimental autoimmune uveitis (40), but different from the conventional view that Tregs suppress effector T cells in an IL10- and/or TGFβ-dependent manner (41, 42).

Previous work has demonstrated that the infiltration of CD8+ T cells improves survival of patients with breast cancer (43, 44). In this study, we documented that increased number of CD73+ γδTregs significantly decreased survival. The presence of stromal CD8+ T cells did not correlate with favorable prognosis when a high density of CD73+ γδTregs existed in the TME, suggesting that CD73+ γδTregs impaired the tumoricidal functions of CD8+ T cells in human breast cancer.

In summary, our study identified CD73+ γδTregs as the predominant Tregs and a key mediator in human breast cancer, and the IL6-adenosine positive feedback loop between CD73+ γδTregs and CAFs is very important to promote immunosuppression and tumor progression. Our findings suggest that CD73+ γδTregs are a prognostic factor of human breast cancer, and eradication of these cells may have a potential for effective treatment.

No potential conflicts of interest were disclosed.

G. Hu: Conceptualization, data curation, formal analysis, supervision, validation, investigation, methodology, writing–original draft, project administration, writing–review and editing. P. Cheng: Formal analysis, validation, investigation, methodology. J. Pan: Formal analysis, investigation. S. Wang: Formal analysis, writing–review and editing. Q. Ding: Formal analysis, investigation. Z. Jiang: Resources. L. Cheng: Resources. X. Shao: Investigation. L. Huang: Conceptualization. J. Huang: Conceptualization, writing–review and editing.

This work was funded by the National Natural Science Foundation of China (grant no. 81702803, to G. Hu). The work was also partly funded by the National Natural Science Foundation of China (Key Program, grant no. 81930079, to J. Huang; Youth Program, grant no. 81902629, to P. Cheng) and Shaoxing Bureau of Science and Technology (2018C30055, to L. Huang).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Mao
Y
,
Keller
ET
,
Garfield
DH
,
Shen
K
,
Wang
J
. 
Stromal cells in tumor microenvironment and breast cancer
.
Cancer Metastasis Rev
2013
;
32
:
303
15
.
2.
Finak
G
,
Bertos
N
,
Pepin
F
,
Sadekova
S
,
Souleimanova
M
,
Zhao
H
, et al
Stromal gene expression predicts clinical outcome in breast cancer
.
Nat Med
2008
;
14
:
518
27
.
3.
Mantovani
A
,
Allavena
P
,
Sica
A
,
Balkwill
F
. 
Cancer-related inflammation
.
Nature
2008
;
454
:
436
44
.
4.
Knutson
KL
,
Maurer
MJ
,
Preston
CC
,
Moysich
KB
,
Goergen
K
,
Hawthorne
KM
, et al
Regulatory T cells, inherited variation, and clinical outcome in epithelial ovarian cancer
.
Cancer Immunol Immunother
2015
;
64
:
1495
504
.
5.
Liu
L
,
Zhao
G
,
Wu
W
,
Rong
Y
,
Jin
D
,
Wang
D
, et al
Low intratumoral regulatory T cells and high peritumoral CD8(+) T cells relate to long-term survival in patients with pancreatic ductal adenocarcinoma after pancreatectomy
.
Cancer Immunol Immunother
2016
;
65
:
73
82
.
6.
Li
Z
,
Dong
P
,
Ren
M
,
Song
Y
,
Qian
X
,
Yang
Y
, et al
PD-L1 expression is associated with tumor FOXP3(+) regulatory T-cell infiltration of breast cancer and poor prognosis of patient
.
J Cancer
2016
;
7
:
784
93
.
7.
Yeong
J
,
Thike
AA
,
Lim
JC
,
Lee
B
,
Li
H
,
Wong
SC
, et al
Higher densities of Foxp3(+) regulatory T cells are associated with better prognosis in triple-negative breast cancer
.
Breast Cancer Res Treat
2017
;
163
:
21
35
.
8.
Wu
D
,
Wu
P
,
Qiu
F
,
Wei
Q
,
Huang
J
. 
Human gammadeltaT-cell subsets and their involvement in tumor immunity
.
Cell Mol Immunol
2017
;
14
:
245
53
.
9.
Hu
G
,
Wu
P
,
Cheng
P
,
Zhang
Z
,
Wang
Z
,
Yu
X
, et al
Tumor-infiltrating CD39+gammadeltaTregs are novel immunosuppressive T cells in human colorectal cancer
.
Oncoimmunology
2017
;
6
:
e1277305
.
10.
Bono
MR
,
Fernandez
D
,
Flores-Santibanez
F
,
Rosemblatt
M
,
Sauma
D
. 
CD73 and CD39 ectonucleotidases in T cell differentiation: beyond immunosuppression
.
FEBS Lett
2015
;
589
:
3454
60
.
11.
Regateiro
FS
,
Cobbold
SP
,
Waldmann
H
. 
CD73 and adenosine generation in the creation of regulatory microenvironments
.
Clin Exp Immunol
2013
;
171
:
1
7
.
12.
Robson
SC
,
Sevigny
J
,
Zimmermann
H
. 
The E-NTPDase family of ectonucleotidases: structure function relationships and pathophysiological significance
.
Purinergic Signal
2006
;
2
:
409
30
.
13.
Antonioli
L
,
Pacher
P
,
Vizi
ES
,
Hasko
G
. 
CD39 and CD73 in immunity and inflammation
.
Trends Mol Med
2013
;
19
:
355
67
.
14.
Linnemann
C
,
Schildberg
FA
,
Schurich
A
,
Diehl
L
,
Hegenbarth
SI
,
Endl
E
, et al
Adenosine regulates CD8 T-cell priming by inhibition of membrane-proximal T-cell receptor signalling
.
Immunology
2009
;
128
:
e728
37
.
15.
Chalmin
F
,
Mignot
G
,
Bruchard
M
,
Chevriaux
A
,
Vegran
F
,
Hichami
A
, et al
Stat3 and Gfi-1 transcription factors control Th17 cell immunosuppressive activity via the regulation of ectonucleotidase expression
.
Immunity
2012
;
36
:
362
73
.
16.
Taher
MY
,
Davies
DM
,
Maher
J
. 
The role of the interleukin (IL)-6/IL-6 receptor axis in cancer
.
Biochem Soc Trans
2018
;
46
:
1449
62
.
17.
Kalluri
R
. 
The biology and function of fibroblasts in cancer
.
Nat Rev Cancer
2016
;
16
:
582
98
.
18.
Hashimoto
O
,
Yoshida
M
,
Koma
Y
,
Yanai
T
,
Hasegawa
D
,
Kosaka
Y
, et al
Collaboration of cancer-associated fibroblasts and tumour-associated macrophages for neuroblastoma development
.
J Pathol
2016
;
240
:
211
23
.
19.
Comito
G
,
Giannoni
E
,
Segura
CP
,
Barcellos-de-Souza
P
,
Raspollini
MR
,
Baroni
G
, et al
Cancer-associated fibroblasts and M2-polarized macrophages synergize during prostate carcinoma progression
.
Oncogene
2014
;
33
:
2423
31
.
20.
Hoskin
DW
,
Mader
JS
,
Furlong
SJ
,
Conrad
DM
,
Blay
J
. 
Inhibition of T cell and natural killer cell function by adenosine and its contribution to immune evasion by tumor cells (review)
.
Int J Oncol
2008
;
32
:
527
35
.
21.
Lu
D
,
Insel
PA
. 
Cellular mechanisms of tissue fibrosis. 6. Purinergic signaling and response in fibroblasts and tissue fibrosis
.
Am J Physiol Cell Physiol
2014
;
306
:
C779
88
.
22.
Ghiringhelli
F
,
Puig
PE
,
Roux
S
,
Parcellier
A
,
Schmitt
E
,
Solary
E
, et al
Tumor cells convert immature myeloid dendritic cells into TGF-beta-secreting cells inducing CD4+CD25+ regulatory T cell proliferation
.
J Exp Med
2005
;
202
:
919
29
.
23.
Liu
VC
,
Wong
LY
,
Jang
T
,
Shah
AH
,
Park
I
,
Yang
X
, et al
Tumor evasion of the immune system by converting CD4+CD25- T cells into CD4+CD25+ T regulatory cells: role of tumor-derived TGF-beta
.
J Immunol
2007
;
178
:
2883
92
.
24.
Peng
G
,
Wang
HY
,
Peng
W
,
Kiniwa
Y
,
Seo
KH
,
Wang
RF
. 
Tumor-infiltrating gammadelta T cells suppress T and dendritic cell function via mechanisms controlled by a unique Toll-like receptor signaling pathway
.
Immunity
2007
;
27
:
334
48
.
25.
Barjon
C
,
Michaud
HA
,
Fages
A
,
Dejou
C
,
Zampieri
A
,
They
L
, et al
IL-21 promotes the development of a CD73-positive Vgamma9Vdelta2 T cell regulatory population
.
Oncoimmunology
2017
;
7
:
e1379642
.
26.
Liao
Z
,
Tan
ZW
,
Zhu
P
,
Tan
NS
. 
Cancer-associated fibroblasts in tumor microenvironment - Accomplices in tumor malignancy
.
Cell Immunol
2019
;
343
:
103729
.
27.
Costa
A
,
Kieffer
Y
,
Scholer-Dahirel
A
,
Pelon
F
,
Bourachot
B
,
Cardon
M
, et al
Fibroblast heterogeneity and immunosuppressive environment in human breast cancer
.
Cancer Cell
2018
;
33
:
463
79
.
28.
Kato
T
,
Noma
K
,
Ohara
T
,
Kashima
H
,
Katsura
Y
,
Sato
H
, et al
Cancer-associated fibroblasts affect intratumoral CD8(+) and FoxP3(+) T cells via IL6 in the tumor microenvironment
.
Clin Cancer Res
2018
;
24
:
4820
33
.
29.
Fu
Z
,
Zuo
Y
,
Li
D
,
Xu
W
,
Li
D
,
Chen
H
, et al
The crosstalk: tumor-infiltrating lymphocytes rich in regulatory T cells suppressed cancer-associated fibroblasts
.
Acta Oncol
2013
;
52
:
1760
70
.
30.
Antonioli
L
,
Blandizzi
C
,
Pacher
P
,
Hasko
G
. 
Immunity, inflammation and cancer: a leading role for adenosine
.
Nat Rev Cancer
2013
;
13
:
842
57
.
31.
Antonioli
L
,
Fornai
M
,
Blandizzi
C
,
Pacher
P
,
Hasko
G
. 
Adenosine signaling and the immune system: when a lot could be too much
.
Immunol Lett
2019
;
205
:
9
15
.
32.
Duluc
C
,
Moatassim-Billah
S
,
Chalabi-Dchar
M
,
Perraud
A
,
Samain
R
,
Breibach
F
, et al
Pharmacological targeting of the protein synthesis mTOR/4E-BP1 pathway in cancer-associated fibroblasts abrogates pancreatic tumour chemoresistance
.
EMBO Mol Med
2015
;
7
:
735
53
.
33.
Perez-Aso
M
,
Fernandez
P
,
Mediero
A
,
Chan
ES
,
Cronstein
BN
. 
Adenosine 2A receptor promotes collagen production by human fibroblasts via pathways involving cyclic AMP and AKT but independent of Smad2/3
.
FASEB J
2014
;
28
:
802
12
.
34.
Vecchio
EA
,
White
PJ
,
May
LT
. 
Targeting adenosine receptors for the treatment of cardiac fibrosis
.
Front Pharmacol
2017
;
8
:
243
.
35.
Stagg
J
,
Divisekera
U
,
McLaughlin
N
,
Sharkey
J
,
Pommey
S
,
Denoyer
D
, et al
Anti-CD73 antibody therapy inhibits breast tumor growth and metastasis
.
Proc Natl Acad Sci U S A
2010
;
107
:
1547
52
.
36.
Desmet
CJ
,
Gallenne
T
,
Prieur
A
,
Reyal
F
,
Visser
NL
,
Wittner
BS
, et al
Identification of a pharmacologically tractable Fra-1/ADORA2B axis promoting breast cancer metastasis
.
Proc Natl Acad Sci U S A
2013
;
110
:
5139
44
.
37.
Allard
B
,
Longhi
MS
,
Robson
SC
,
Stagg
J
. 
The ectonucleotidases CD39 and CD73: novel checkpoint inhibitor targets
.
Immunol Rev
2017
;
276
:
121
44
.
38.
Cekic
C
,
Linden
J
. 
Adenosine A2A receptors intrinsically regulate CD8+ T cells in the tumor microenvironment
.
Cancer Res
2014
;
74
:
7239
49
.
39.
Zarek
PE
,
Huang
CT
,
Lutz
ER
,
Kowalski
J
,
Horton
MR
,
Linden
J
, et al
A2A receptor signaling promotes peripheral tolerance by inducing T-cell anergy and the generation of adaptive regulatory T cells
.
Blood
2008
;
111
:
251
9
.
40.
Liang
D
,
Zuo
A
,
Zhao
R
,
Shao
H
,
Born
WK
,
O'Brien
RL
, et al
CD73 expressed on gammadelta T cells shapes their regulatory effect in experimental autoimmune uveitis
.
PLoS One
2016
;
11
:
e0150078
.
41.
Strauss
L
,
Bergmann
C
,
Szczepanski
M
,
Gooding
W
,
Johnson
JT
,
Whiteside
TL
. 
A unique subset of CD4+CD25highFoxp3+ T cells secreting interleukin-10 and transforming growth factor-beta1 mediates suppression in the tumor microenvironment
.
Clin Cancer Res
2007
;
13
:
4345
54
.
42.
Ghiringhelli
F
,
Menard
C
,
Terme
M
,
Flament
C
,
Taieb
J
,
Chaput
N
, et al
CD4+CD25+ regulatory T cells inhibit natural killer cell functions in a transforming growth factor-beta-dependent manner
.
J Exp Med
2005
;
202
:
1075
85
.
43.
Peng
GL
,
Li
L
,
Guo
YW
,
Yu
P
,
Yin
XJ
,
Wang
S
, et al
CD8(+) cytotoxic and FoxP3(+) regulatory T lymphocytes serve as prognostic factors in breast cancer
.
Am J Transl Res
2019
;
11
:
5039
53
.
44.
Ali
HR
,
Provenzano
E
,
Dawson
SJ
,
Blows
FM
,
Liu
B
,
Shah
M
, et al
Association between CD8+ T-cell infiltration and breast cancer survival in 12,439 patients
.
Ann Oncol
2014
;
25
:
1536
43
.