The production of CD73-derived adenosine (Ado) by Tregs has been proposed as a resistance mechanism to anti-PD-1 therapy in murine tumor models. We reported that human Tregs express the ectonucleotidase CD39, which generates AMP from ATP, but do not express the AMPase CD73. In contrast, CD73 defined a subset of effector CD4+ T cells (Teffs) enriched in polyfunctional Th1.17 cells characterized by expression of CXCR3, CCR6, and MDR1, and production of IL17A/IFNγ/IL22/GM-CSF. CD39+ Tregs selectively targeted CD73+ Teffs through cooperative degradation of ATP into Ado inhibiting and restricting the ability of CD73+ Teffs to secrete IL17A. CD73+ Teffs infiltrating breast and ovarian tumors were functionally blunted by Tregs expressing upregulated levels of CD39 and ATPase activity. Moreover, tumor-infiltrating CD73+ Teffs failed to express inhibitory immune checkpoints, suggesting that CD73 might be selected under pressure from immune checkpoint blockade therapy and thus may represent a nonredundant target for restoring antitumor immunity.

Significance: Polyfunctional CD73+ T-cell effectors lacking other immune checkpoints are selectively targeted by CD39 overexpressing Tregs that dominate the breast tumor environment. Cancer Res; 78(13); 3604–18. ©2018 AACR.

During infection or tumor development, ATP is released into the extracellular space and can be found at high levels in inflamed tissues, including tumors (1). Extracellular ATP represents a proinflammatory alarmin for the immune system. It induces the chemoattraction of dendritic cells (DC) and the activation of inflammasomes and IL1β secretion by monocytes/macrophages (Mφ) through the engagement of P2 purinergic receptors (for review; ref. 2). Interestingly, the degradation of ATP into adenosine (Ado) by murine CD4+ regulatory T cells (Tregs), coexpressing CD39 and CD73, has been associated with immunosuppression (3), particularly within tumor environment (4, 5). Mechanistically, CD39 (ectonucleoside triphosphate diphosphohydrolase-1, ENTPD1) degrades ATP and ADP into AMP, which is then hydrolyzed by CD73 (5′-ectonucleotidase, NT5E) into Ado. Ado can favor tumor progression by inhibiting the function of immune cells (T cells, monocytes, DCs, and Mφ), through the engagement of its receptors (AdoR) A2a or A2b (for review; ref. 6). These AdoR induce an increased concentration of intracellular cyclic AMP, resulting in the inhibition of proliferation, cytotoxic functions, and cytokine secretion (IL2, TNFα, IFNγ, and IL13) of T cells (for review; ref. 6). Ado can act also on innate immune cells by inhibiting IL12 production by DCs (7, 8), and inducing IL10 and VEGF secretion by Mφ, which favor immunosuppression and angiogenesis (9, 10). Moreover, Ado can act on nonimmune cells, such as tumor cells, through A1 or A3 receptors coupled to Gαi proteins, fostering tumor cell proliferation and migration (for review; ref. 6). The importance of Ado signaling in immunosuppression is evidenced in patients with severe combined immunodeficiency (SCID), in whom accumulation of Ado, due to a lack of adenosine deaminase (ADA) enzyme that degrades Ado into inosine, is observed (11). Remarkably, in CD73-deficient mice, in mice with CD73-deficient Tregs, or in A2a-deficient mice, the antitumor immune response is increased, leading to the tumor rejection and to the inhibition of metastases (12–15). Also, in murine models, targeting of CD73 alone or in combination with immunotherapy (anti-CTLA-4 or anti-PD-1) enhances tumor rejection and blocks metastases, demonstrating that endogenous Ado limits the efficacy of these immunotherapies (16, 17).

Studies investigating the expression of CD73 and CD39 on human T cells subsets remain scarce. In fact, despite the poor prognostic value of Tregs in breast and ovarian tumors (18, 19), the coexpression of CD39 and CD73 on human Tregs remains controversial (20–23). A recent report by Doherty and colleagues also suggests the interconnection between CD73 expression on memory CD4+ T cells and a Th17 profile (24).

Th17 cells, implicated in the response to extracellular pathogens, are characterized by the secretion of IL17A, IL17F, and IL22. Recently, the involvement of a subset of Th17 cells, called Th1.17 coproducing IL17A, IFNγ, and GM-CSF, in the development of autoimmune diseases, was unveiled (25), highlighting the polyfunctionality of these cells (26, 27). In tumor biology, a protumoral role of Th17 was reported and mainly associated with the secretion of IL17A; however, coproduction of IFNγ and GM-CSF was not considered, except in a recent report, demonstrating that Th1.17-cell subset contributes to tumor rejection (28).

In this study, we demonstrate that CD73, not detected on human Tregs, defines a subpopulation of CD4+ effector T cells (Teffs) associated with cytokine polyfunctionality similar to that described in Th1.17 cells. CD73+CD4+ Teffs cooperate with CD39+ Tregs to degrade ATP into immunosuppressive Ado inhibiting and restricting function of CD73+CD4+ Teffs to IL17A secretion. This cooperation does not require cell contact and, in a spaced-out microenvironment, induces specific inhibition of CD73+CD4+ Teffs through autocrine ATP-derived Ado action without affecting distant CD73CD4+ Teffs. Interestingly, we reveal that CD73+CD4+ Teffs are present in breast and ovarian tumor environment together with CD39high Tregs and express only low levels of inhibitory immune checkpoints (i-ICP) highlighting CD73 as a potential resistance mechanism to current immunotherapies, and as a nonredundant target for restoring antitumor immunity.

Human samples

HD-blood was purchased anonymously from the French Blood Service (EFS). In addition, blood and primary tumor samples were obtained from patients with non-pretreated breast tumors and chemotherapy-pretreated ovarian tumors. All of these samples were provided by the tissue bank (BRC) of Léon Bérard Cancer Center (CLB), after approval from the Institutional review board and ethics committee (L-06-36 and L-11-26) and patients' written informed consent, in accordance with the Declaration of Helsinki. Human healthy colon endoscopic biopsies obtained during routine endoscopic check-ups of healthy donors selected on the basis that they were neither under chemotherapy nor taking antibiotics were obtained from the BRC of the CLB after approval from the Institutional review board and ethics committee (French agreement number: AC-2013-1871) and donors' written informed consent.

Purification of the different cell subsets

PBMCs were purified from blood of HDs or patients with cancer by Ficoll density gradient. Memory CD4+ T cells were purified using MagniSort Human CD4+ Memory T Cell Enrichment Kit (eBioscience). CD73CD4+ Teffs (CD4+CD45RACD127+CD25 CD39CD73), CD73+CD4+ Teffs (CD4+CD45RACD127+CD25CD39CD73+) and total Tregs (CD4+CD45RACD127CD25+) or CD39+ and CD39 Treg subsets were sorted from purified memory CD4+ T cells by multiparametric FC (flow cytometry; FACSAria III, BD Biosciences) using antibodies against CD25 (2A3, BD Biosciences), CD45RA (2H4LDH11LDB9, Beckman-Coulter), as well as CD127 (eBioRDR5), CD39 (eBioA1), and CD73 (AD2; all from eBioscience), alongside a viability marker (DAPI). Infiltrating CD4+ T cells from breast and ovarian tumors and healthy colonic tissues were isolated from single-cell suspensions, obtained by enzymatic disaggregation (18), with the Dynabeads Human CD4+ Kit (Life Technologies). Tumor-infiltrating CD4+ Teffs (CD127+CD25) and Tregs (CD127CD25+) were sorted from single-cell suspensions by multiparametric FC using antibodies against CD127, CD25, CD45RA, and a viability marker (Fixable Viability Dye, Biolegend).

Flow cytometry analyses

Multiparametric FC analyses were performed on (i) PBMC from HDs or patients with cancer, or on (ii) single cell suspensions derived either from primary breast and ovarian tumors or healthy colonic tissues. The FC panels used to assess T-cell differentiation relied on the use of anti-human antibodies against CD3 (UCHT1), CD4 (RPA-T4), CD8 (SK1), CD95 (DX2), and CD28 (CD28.2; all from BD Biosciences), CD27 (O323, eBioscience), CCR7 (G043H7, Biolegend), CD45RA, CD39, and CD73 (see above), while those used to evaluate T-cell polarization relied on the use of the anti-human antibodies against CD3, CD4, CD45RA, CD127, CD25, CD39, and CD73 (see above), CCR6 (11A9, BD Biosciences), CXCR3 (G025H7) and CRTH2 (BM16) from Biolegend, and a viability marker. The i-ICPs panel included antibodies against CD3, CD4, CD8 (see above), CD45 (HI30), PD-1 (EH12.1), TIM-3 (7D3; all from BD Biosciences), CD45RA (HI100) and CTLA-4 (L3D10; Biolegend), TIGIT (MBSA43), CD39 and CD73 (eBioscience), and a fixable viability marker (Zombie Fixable Viability Dyes, Biolegend). The production of intracellular cytokines on CD73+CD4+ and CD73CD4+ Teffs was analyzed using the test described previously (29) with some modifications, in particular the evaluation of IL22 (22URTI, BD Biosciences) instead of IL21. Cells were analyzed on a LSR-Fortessa (BD Biosciences) and data were processed using the FlowJo Software (Tree Star). For the coexpression of i-ICPs, data were analyzed using the Boolean method on the FlowJo software and then represented using SPICE v5.3 software.

MDR1 staining and activity assay

Purified memory CD4+ T cells were incubated with Rh123 (1μg/mL, Sigma-Aldrich) for 30 minutes on ice. After washes in PBS, cells were incubated at 37°C for 2 hours (efflux phase). Cells were then washed in PBS and stained with surface markers for FC analysis. In some conditions, a MDR1 inhibitor, Elacridar (1 μmol/L, Tocris), or vehicle (DMSO) were added to cells immediately before the efflux phase. MDR1 expression was assessed using an anti-human MDR1 antibody (UIC2, eBioscience) for 20 minutes at 37°C in the presence of Cyclosporin A (25 μmol/L; R&D Systems), as described previously (30).

Transcriptomic analyses

CD73 and CD73+ CD4+ Teffs were sorted from the PBMCs of six HDs blood samples. Cells were activated in the presence of anti-CD3/anti-CD28 beads (expand beads; Life Technologies, ratio 1 bead to 4 cells) in complete RPMI medium [supplemented with antibiotics, l-glutamine (Life Technologies) and 5% of human serum AB+ (EFS)] at 37°C under 5% CO2. Resting and short-term activated cells (mix of 6-hour and 24-hour activation), were lyzed for mRNA extraction (miRNeasy Micro Kit; Qiagen) and transcriptomic analyses (Human Gene-Expression 8 × 60K Microarray Kit AMADID 039494, Agilent Technologies) were performed.

Western blot analysis

Tregs, CD73, or CD73+ CD4+ Teffs (2 × 106 cells) were isolated and lyzed in RIPA buffer in the presence of proteinase inhibitors. Protein lysates were boiled, loaded on Mini-PROTEAN TGX Precast Gels 5%–20% (Bio-Rad) and transferred to Trans-Blot Turbo Mini PVDF membrane (Bio-Rad). CD73 was detected using a mouse anti-hCD73 antibody (1D7, 1/500, Abcam). Pellets of MDA-MB-231 cell line (LGC Standards) whose genetic profile has been verified (Eurofins Forensic Department) and negative for Mycoplasma (MycoAlert Mycoplasma Detection Kit; Lonza) and human purified HD blood monocytes were used as positive and negative control, respectively. ADA and A2b receptors were assessed either on purified CD73 or CD73+ CD4+ Teffs, or after 1 or 4 days activation with expand beads (ratio 1:4), using a goat anti-A2bR antibody (ab40002, 1/1,000; Abcam) and a mouse anti-hADA antibody (ab54969, 1/500; Abcam). The HRP-coupled secondary antibodies used were goat anti-mouse Ab (12-349, 1/5,000; Upstate) and rabbit anti-goat Ab (P 0449 1/2000; Dako). Membranes were revealed with Luminata Crescendo Reagent (Millipore) and analyzed on Chemidocssystem (Bio-Rad).

Proliferation experiments

CD73CD4+ and CD73+CD4+ Teffs were stained, respectively, with the carboxyfluorescein succinimidyl ester (CFSE; 2 μmol/L; Life Technologies) and CellTrace Violet (CTV; 20 μmol/L; Life Technologies) proliferation markers, while Tregs were costained with both CFSE and CTV according to manufacturer's instructions. CD73CD4+ Teffs or CD73+ CD4+ Teffs (3 × 104) alone, in coculture with Tregs (total or sorted CD39+ and CD39 subsets) at a ratio 1:1 or coculture of CD73CD4+ Teffs, CD73+CD4+ Teffs and Tregs at physiologic ratio (70%:20%:10%) were incubated with expand beads (ratio 1:4) in 96-round-bottom well plates (Falcon) in 200 μL of complete RPMI medium for 4 days at 37°C under 5% CO2. Ado, AMP, or ATP (Sigma-Aldrich) were added every day at the indicated concentration, whereas the CD73 inhibitor (APCP, 50 μmol/L; Sigma-Aldrich), CD39 inhibitor (ARL-67156, 250 μmol/L; Tocris) or rhADA (1 μg/mL; R&D Systems) were preincubated for 30 minutes before the beginning of the culture. At the end of the experimental time course, cells were harvested and stained with a viability marker (LIVE/DEAD Fixable Dead Cell Stains) and fixed in 2% formaldehyde. Cell proliferation was analyzed under an inverted Zeiss microscope (objective 4×) using an Axiovision 4 software (Zeiss), and by FC (LSR Fortessa, BD Biosciences) according to the CFSE and CTV dilution.

For experiments in Transwell Permeable Supports (polycarbonate membrane 0.4-μm pore, Corning), in the 6.5-mm inserts, 6 × 104 CD73+CD4+ Teffs alone or in coculture with Tregs (at a ratio of 1:1) were activated using expand beads (ratio of 1:4) in the presence, in the bottom wells, of 1.5 × 105 CD73CD4+ Teffs alone, or in coculture with Tregs (at a ratio of 1:4) and expand beads (ratio 1:4), in a total volume of 800 μL.

Cytokine analysis

CD73CD4+ or CD73+CD4+ Teffs (5 × 104) were activated with either PMA (50 ng/mL)/ionomcycin (1 μg/mL) (Sigma-Aldrich) for 24 hours or expand beads (ratio of 1:4) for 48 hours in complete RPMI at 37°C under 5% CO2, and in the presence or absence of Ado (75 or 100 μmol/L). Supernatants were harvested and frozen, and cytokines were analyzed by ELISA using Luminex Multiplex Kits (eBioscience; Multiplex 1: IL2/TNFα/IFNγ/IL22/IL17A/IL10/IL13/IL21; Multiplex 2: GM-CSF/IL3).

Nucleotide and nucleoside quantification by HPLC coupled to LC-MS/MS

The capacity of 5 × 104 Tregs, CD73CD4+ Teffs or CD73+CD4+ Teffs alone or coculture of Tregs with either CD73CD4+ Teffs or CD73+CD4+ Teffs (ratio 1:1), to degrade ATP or AMP was analyzed after 2-hour incubation at 37°C under 5% CO2 with labeled ATP13C,15N or AMP13C,15N (all from Sigma- Aldrich) in 200 μL of serum-free RPMI medium supplemented with antibiotics and l-glutamine (Life Technologies). In some cases, cells were preincubated with ARL-67156 (250 μmol/L) or APCP (50 μmol/L) for 30 minutes before the experiment. Cell supernatants were harvested, boiled at 65°C for 5 seconds, and frozen at −20°C. ATP13C,15N, AMP13C,15N, Ado13C,15N, and Inosine13C,15N were quantified in 50 μL of supernatant. Nucleotides and nucleosides were extracted using off-line Oasis-WAX cartridges (60 mg; 3 cc). Briefly, cartridges were conditioned using 2 mL of methanol and 2 mL of water. After sample loading, cartridges were washed with 1 mL of water. Elution was performed thrice using 1 mL of the following mixture: NH4OH 0.25% pH 10.0/water/acetonitrile (30/30/40, v/v). Eluates were pooled and gas dried under nitrogen at 37°C. The residue was suspended in 250 μL of 5 mmol/L HA-0.5% DEA in water, and 10 μL was injected into a liquid chromatograph.

Analysis was performed using LC/MS-MS as described previously by Machon and colleagues (31). Quantification was conducted by adding standard solutions of labelled nucleotides (ATP13C, AMP15N, Ado13C, and inosine15N) to samples prior to the extraction step. ATP15N and GMP13C,15N were used as internal standards. Concentrations of nucleotides in the supernatants were calculated using calibration curves of the corresponding labeled nucleotides. We also verified that ARL-67156 (250 μmol/L) and APCP (50 μmol/L) did not interfere with the ATP, AMP, Ado, and inosine quantifications.

Multi-immunofluorescence stainings on frozen tumor sections

Tissue-Tek O.C.T. (Sakura Finetek) embedded frozen human primary breast tumors were used to generate 6-μm frozen tissue sections with Cryotome (Thermo Fisher Scientific). Sections were fixed in paraformaldehyde, permeabilized in Triton X-100, and stained with murine anti-human CD4-AF488 (Biolegend, Clone RPA-T4), rat anti-human FoxP3-APC (eBioscience, Clone PCH101), and uncoupled rabbit anti-human CD73 (Cell Signaling Technology, Clone D7F9A). CD73 staining was revealed with secondary donkey anti-rabbit antibody (Life Technologies). Immunofluorescence stainings were analyzed on Upright Microscope (Nikon Ni-E) using ImageJ free software.

CD39 and CD73 are expressed by distinct memory CD4+ T cells in humans

We initially observed by FC that human CD4+ T cells in the blood of healthy donors (HD-blood) did not coexpress CD39 and CD73, but exclusive CD39+ or CD73+ cells were observed among CD4+ T-cell populations (Fig. 1A). While naïve CD4+ T cells expressed no or very low levels of CD39 and CD73 (Fig. 1A and B), the expression of CD39 and CD73 was detectable on memory CD4+ T cells (Fig. 1A). Among memory CD4+ T cells, CD39, but not CD73, was expressed at variable levels on human HD-blood Tregs defined by FoxP3 expression (Fig. 1A and C). The absence of CD73 expression on Tregs purified from blood was also confirmed by Western blot analysis (Fig. 1D). In contrast, CD73, but also CD39, can be expressed on distinct populations of CD4+ Teffs defined by the absence of FoxP3 expression (Fig. 1A and E). Of note, the intensity of CD73 expression on CD73+CD4+ Teffs was significantly higher compared with naïve CD73+CD4+ T cells (Fig. 1F). Moreover, the percentage and the intensity of CD39 expression were lower on CD4+ Teffs compared with Tregs [Δ CD39 mean fluorescence intensity (MFI) Tregs/CD4+ Teffs: 1.64 ± 0.33].

Figure 1.

CD39 and CD73 are expressed on distinct subsets of memory CD4+ T cells. A, Representative expression of CD39 and CD73 on CD4+ T-cell subsets from HD-blood analyzed by FC. B, C, and E, Cumulative data from 17 donors of the different subsets defined by the expression of CD39 and CD73 among naïve CD4+ T cells (B), CD4+ Tregs (C), and CD4+ Teffs (E; statistical analysis, Friedman test). D, Tregs were analyzed for CD73 expression by Western blot analysis (CD73, 70 kDa). F, Intensity of CD73 staining on naïve CD4+ T cells and CD4+ Teffs (statistical analysis, Wilcoxon test). ***, P < 0.001; ****, P < 0.0001; ns, nonsignificant.

Figure 1.

CD39 and CD73 are expressed on distinct subsets of memory CD4+ T cells. A, Representative expression of CD39 and CD73 on CD4+ T-cell subsets from HD-blood analyzed by FC. B, C, and E, Cumulative data from 17 donors of the different subsets defined by the expression of CD39 and CD73 among naïve CD4+ T cells (B), CD4+ Tregs (C), and CD4+ Teffs (E; statistical analysis, Friedman test). D, Tregs were analyzed for CD73 expression by Western blot analysis (CD73, 70 kDa). F, Intensity of CD73 staining on naïve CD4+ T cells and CD4+ Teffs (statistical analysis, Wilcoxon test). ***, P < 0.001; ****, P < 0.0001; ns, nonsignificant.

Close modal

Our results demonstrate that in contrast to murine Tregs, human Tregs only express CD39. They also highlight the expression of CD73 on a subset of CD4+ Teffs. Thus, we wondered whether the expression of CD73, favoring the generation of Ado, on a subset of CD4+ Teffs may denote a functional specialization of this population.

CD73 identifies a population of CD4+ Teffs enriched in polyfunctional Th1.17 cells

We analyzed by FC the expression pattern of the chemokine receptors CRTH2, CXCR3, and CCR6 on blood CD73+ versus CD73 CD4+ Teffs to determine the relative percentage of Th1 (CRTH2CXCR3+CCR6), Th2 (CRTH2+CXCR3CCR6), Th17 (CRTH2CXCR3CCR6+), and Th1.Th17 (CRTH2CXCR3+CCR6+; Supplementary Fig. S1A; ref. 26). CD73+CD4+ Teffs contained a similar proportion of Th1, Th17, and Th2 subsets as CD73 Teffs, but exhibited an increased proportion of the Th1.17 subset, which accounted for 27.98% ± 7.99% of total cells, at the expense of other Th cells (Fig. 2A; Supplementary Fig. S1B). Consistent with this Th1.17 enrichment, a higher proportion of CD73+CD4+ Teffs migrated toward a CXCL10 or CCL20 gradient in Transwell chemotaxis assays (Supplementary Fig. S1C). Analysis of the transcriptome of purified CD73+ and CD73 CD4+ Teffs after short-term TCR stimulation revealed an overexpression of Th1.17-related genes (CSF2, ABCB1, IL22, IL3, IFNG, GZMB, IL23R, TBX21, LGALS3; refs. 25, 27) in the former subset, while genes related to Th2 cells (IL4, IL10, IL13, CCR4) and Th17, Tfh, or Tr1 (LRMP, IKZF3, CXCR5, IL21, IL10) were downregulated (Fig. 2B). IL17A and IL17F genes were not detected likely due to insufficient sensitivity of transcriptomic chip used in our conditions.

Figure 2.

CD73+CD4+ Teffs are enriched in polyfunctional Th1.17 cells. A, HD-blood CD73+ and CD73neg CD4+ Teffs were analyzed by FC for their composition in Th subsets. Data represent mean of 7 donors. B, Transcriptomic analysis was performed on resting or short-term activated CD73+ or CD73neg CD4+ Teffs (n = 6; P < 0.05 for the genes presented) and genes related to Th profiles were extracted. C, Percentage of IL2 or TNFα, IFNγ, IL22, and IL17A-producing cells and IFNγ/IL17A coproducing cells was assessed by FC on CD73+ (black) and CD73neg (white) CD4+ Teffs from HD-blood after short-term PMA/ionomycin reactivation (statistical analysis, Wilcoxon test). D, Quantification of cytokines produced by purified HD-blood CD73+ (black) and CD73neg (white) CD4+ Teffs activated for 24 hours with PMA/ionomycin (n = 16 for IL2,TNFα, IFNγ,IL17A, IL10, IL13, and IL21, with exception due to out-of-range quantification, and n = 10 for IL22, GM-CSF, and IL3; statistical analysis, Wilcoxon test). E, MDR1 expression and MDR1 functionality were assessed by FC on CD73+ (black) and CD73neg CD4+ (white) Teffs (statistical analysis, Wilcoxon test). F, CD73+CD4+ Teffs frequency in HD-blood and healthy colonic tissues (statistical analysis, Mann–Whitney test). *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, nonsignificant.

Figure 2.

CD73+CD4+ Teffs are enriched in polyfunctional Th1.17 cells. A, HD-blood CD73+ and CD73neg CD4+ Teffs were analyzed by FC for their composition in Th subsets. Data represent mean of 7 donors. B, Transcriptomic analysis was performed on resting or short-term activated CD73+ or CD73neg CD4+ Teffs (n = 6; P < 0.05 for the genes presented) and genes related to Th profiles were extracted. C, Percentage of IL2 or TNFα, IFNγ, IL22, and IL17A-producing cells and IFNγ/IL17A coproducing cells was assessed by FC on CD73+ (black) and CD73neg (white) CD4+ Teffs from HD-blood after short-term PMA/ionomycin reactivation (statistical analysis, Wilcoxon test). D, Quantification of cytokines produced by purified HD-blood CD73+ (black) and CD73neg (white) CD4+ Teffs activated for 24 hours with PMA/ionomycin (n = 16 for IL2,TNFα, IFNγ,IL17A, IL10, IL13, and IL21, with exception due to out-of-range quantification, and n = 10 for IL22, GM-CSF, and IL3; statistical analysis, Wilcoxon test). E, MDR1 expression and MDR1 functionality were assessed by FC on CD73+ (black) and CD73neg CD4+ (white) Teffs (statistical analysis, Wilcoxon test). F, CD73+CD4+ Teffs frequency in HD-blood and healthy colonic tissues (statistical analysis, Mann–Whitney test). *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, nonsignificant.

Close modal

We also performed short-term in vitro reactivation of PBMCs with PMA/ionomycin to analyze cytokines production by intracytoplasmic staining in CD73+ versus CD73 CD4+ Teffs. We found a higher proportion of cells producing IL2, TNFα, IL17A, IFNγ, and IL22, in CD73+ CD4+ Teffs (Fig. 2C). More precisely, CD73+CD4+ Teffs were enriched in cells coproducing IFNγ/IL17A (Fig. 2C). A multiplex immunoassay performed with supernatants of these in vitro activated subsets confirmed the significantly increased production of all cytokines but also of IL3 and GM-CSF by CD73+CD4+ Teffs (Fig. 2D) except IL2 analyzed by FC (Fig. 2C). In addition, this assay revealed the low production of IL10, IL13, and IL21 (Fig. 2D) compared with CD73CD4+ Teffs in accordance with Th1.17 gene signature (Fig. 2B).

One of the most upregulated genes differentiating CD73+ and CD73 CD4+ Teffs (Fig. 2B) after CSF2 gene (coding for GM-CSF) is the ABCB1 gene, coding for the multidrug transporter MDR1 and recently proposed as a Th1.17-specific marker (27). A significantly increased expression of MDR1 on CD73+CD4+ Teffs compared with CD73CD4+ ones was observed at the protein level by FC (Fig. 2E) and at the functional level with the Rh123 exclusion assay (Fig. 2E; Supplementary Fig. S1D).

Taken together, our results demonstrate that CD73 identifies a subset of CD4+ Teffs enriched in polyfunctional Th1.17 cells. Consistent with this conclusion, a high proportion of CD73+CD4+ Teffs was observed in the colon of healthy individuals (Fig. 2F), a mucosal tissue known to be enriched in Th17 and Th1.17 populations (32).

CD73+CD4+ Teffs cooperate with CD39+ Tregs for Ado production

Next, we evaluated the capacity of CD73+CD4+ Teffs to generate Ado from AMP through the measurement of nucleotides and nucleosides by HPLC (31). AMP degradation and generation of Ado were observed after 2-hour incubation with AMP and were abrogated by a specific CD73 inhibitor (Fig. 3A). Whereas purified human CD73+CD4+ or CD73CD4+ Teffs were not able to degrade ATP (Fig. 3B), purified human Tregs were able to degrade 48.24% ± 6.20% of ATP into AMP without generation of Ado (Fig. 3B). We confirmed that degradation of ATP results from CD39 expression by Tregs as only purified CD39+ Tregs but not CD39 Tregs generated AMP and that ATP degradation was blocked by a CD39 inhibitor (Fig. 3C). The addition of purified CD73+CD4+ Teffs with Tregs favored the generation of Ado following ATP degradation (Fig. 3B).

Figure 3.

CD73+ CD4+ Teffs cooperate with Tregs to degrade ATP into Ado. A, The capacity of HD CD73+ or CD73neg CD4+ Teffs to degrade AMP into Ado was analyzed after a 2-hour incubation with the stable AMP13C,15N isotope (37.5 μmol/L). Residual and generated metabolites were quantified by mass spectrometry–coupled HLPC and the role of CD73 was confirmed by preincubation of cells with the CD73 inhibitor (APCP, 50 μmol/L) for 30 minutes before AMP addition. Experiments were performed in duplicate (n = 3 donors; statistical analysis, Friedman test). B, The capacity of HD-blood Tregs alone or cocultured with CD73+ or CD73neg CD4+ Teffs (ratio 1:1) to degrade ATP into Ado was analyzed after 2-hour incubation with ATP13C,15N isotope (37.5 μmol/L); residual and generated metabolites were quantified by mass spectrometry–coupled HLPC. The role of CD39 and CD73 was confirmed by preincubation of cells with the CD39 inhibitor (ARL-67156, 250 μmol/L) and/or the CD73 inhibitor (APCP, 50 μmol/L) for 30 minutes before ATP addition. Experiments were performed in duplicate (n = 3 donors; statistical analysis, Friedman test). C, Cell-sorted CD39neg and CD39+ Treg capacity to degrade ATP13C,15N isotope (37.5 μmol/L) was analyzed by quantification by mass spectrometry coupled-HLPC of ATP and metabolite (AMP) generated. The role of CD39 was confirmed by preincubation of cells with the CD39 inhibitor (ARL-67156, 250 μmol/L). Experiments were performed on three donors. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, nonsignificant.

Figure 3.

CD73+ CD4+ Teffs cooperate with Tregs to degrade ATP into Ado. A, The capacity of HD CD73+ or CD73neg CD4+ Teffs to degrade AMP into Ado was analyzed after a 2-hour incubation with the stable AMP13C,15N isotope (37.5 μmol/L). Residual and generated metabolites were quantified by mass spectrometry–coupled HLPC and the role of CD73 was confirmed by preincubation of cells with the CD73 inhibitor (APCP, 50 μmol/L) for 30 minutes before AMP addition. Experiments were performed in duplicate (n = 3 donors; statistical analysis, Friedman test). B, The capacity of HD-blood Tregs alone or cocultured with CD73+ or CD73neg CD4+ Teffs (ratio 1:1) to degrade ATP into Ado was analyzed after 2-hour incubation with ATP13C,15N isotope (37.5 μmol/L); residual and generated metabolites were quantified by mass spectrometry–coupled HLPC. The role of CD39 and CD73 was confirmed by preincubation of cells with the CD39 inhibitor (ARL-67156, 250 μmol/L) and/or the CD73 inhibitor (APCP, 50 μmol/L) for 30 minutes before ATP addition. Experiments were performed in duplicate (n = 3 donors; statistical analysis, Friedman test). C, Cell-sorted CD39neg and CD39+ Treg capacity to degrade ATP13C,15N isotope (37.5 μmol/L) was analyzed by quantification by mass spectrometry coupled-HLPC of ATP and metabolite (AMP) generated. The role of CD39 was confirmed by preincubation of cells with the CD39 inhibitor (ARL-67156, 250 μmol/L). Experiments were performed on three donors. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, nonsignificant.

Close modal

Ado suppresses proliferation and restricts polyfunctional CD73+CD4+ Teffs to IL17A secretion

The capacity of CD73+CD4+ Teffs to generate Ado suggests that these cells may be particularly sensitive to Ado. This immunosuppressive molecule exerts its immunosuppressive functions by engaging A2a or A2b receptors but can be degraded into inosine by the enzyme ADA. Our transcriptomic data revealed that ADORA2A, but not ADORA2B, was expressed on resting CD73+ and CD73 CD4+ Teffs, and that both receptors were upregulated following short-term TCR triggering (Fig. 4A). In contrast, ADORA1 (A1R) and ADORA3 (A3R) were not detected as described in the literature (6). ADA mRNA expressed at steady state decreased after activation on both subsets (Fig. 4A). The modulation observed for ADORA2B and ADA at mRNA level was confirmed at the protein level, which gradually increased for A2bR and decreased for ADA until 4 days of activation (Fig. 4B). These data demonstrated the acquisition of an Ado-sensitive phenotype by CD73+ and CD73 CD4+ Teffs upon TCR triggering, as observed in the murine model (3). Otherwise, no significant difference in the cell proliferation capacities of both subsets was observed using same proliferation markers evaluated by FC (Supplementary Fig. S1E).

Figure 4.

Ado suppresses proliferation and restricts polyfunctional CD73+CD4+ Teffs to IL17A secretion. A, Gene expression profiles of ADA and Ado receptors (AdoRa2a and AdoRa2b) were extracted from transcriptomic analyses of purified resting CD73+ or CD73 CD4+ Teffs (statistical analysis, two-way ANOVA). B, AdoRa2b and ADA proteins levels in CD73+ and CD73 CD4+ Teffs before and after 1- or 4-day activation with expand beads were quantified by Western blot analysis. C, The impact of Ado (75 μmol/L/day) on the proliferation of purified CD73+ or CD73neg CD4+ Teffs after activation using expand beads (1:4) was assessed by microscopy after 4 days. In some conditions, cells were preincubated for 30 minutes with rhADA (1 μg/mL) before adding Ado. Data are representative of three donors. D, The impact of metronomic doses of Ado (37.5 μmol/L or 75 μmol/L/day) on cytokine secretion (IL22, IL17A, IFNγ, GM-CSF, IL10, IL13, IL21, and IL3) by purified CD73+CD4+ Teffs was quantified by Multiplex Luminex assay in the supernatant of cells after a 48-hour–culture period with expand beads (1:4; statistical analysis, Friedman test). *, P < 0.05; ns, nonsignificant.

Figure 4.

Ado suppresses proliferation and restricts polyfunctional CD73+CD4+ Teffs to IL17A secretion. A, Gene expression profiles of ADA and Ado receptors (AdoRa2a and AdoRa2b) were extracted from transcriptomic analyses of purified resting CD73+ or CD73 CD4+ Teffs (statistical analysis, two-way ANOVA). B, AdoRa2b and ADA proteins levels in CD73+ and CD73 CD4+ Teffs before and after 1- or 4-day activation with expand beads were quantified by Western blot analysis. C, The impact of Ado (75 μmol/L/day) on the proliferation of purified CD73+ or CD73neg CD4+ Teffs after activation using expand beads (1:4) was assessed by microscopy after 4 days. In some conditions, cells were preincubated for 30 minutes with rhADA (1 μg/mL) before adding Ado. Data are representative of three donors. D, The impact of metronomic doses of Ado (37.5 μmol/L or 75 μmol/L/day) on cytokine secretion (IL22, IL17A, IFNγ, GM-CSF, IL10, IL13, IL21, and IL3) by purified CD73+CD4+ Teffs was quantified by Multiplex Luminex assay in the supernatant of cells after a 48-hour–culture period with expand beads (1:4; statistical analysis, Friedman test). *, P < 0.05; ns, nonsignificant.

Close modal

The metronomic addition of exogenous Ado (75 μmol/L/day) inhibited the proliferation of both subsets as shown by the reduction of cell clusters size (round-bottom well observation; Fig. 4C) and dilution of the CTV proliferation marker (Supplementary Fig. S2) compared with the medium condition. The addition of recombinant ADA (rhADA, 1 μg/mL), restored the proliferative capacity of both T-cell subsets (Fig. 4C; Supplementary Fig. S2).

The impact of Ado was also analyzed on the cytokines production capacity of both purified CD4+ Teffs subsets. To avoid any confounding effects of Ado on proliferation, cytokine secretions were analyzed after 2 days of TCR triggering, determined as the optimal time lapse for cell activation without proliferation. Of great interest, Ado significantly inhibited the secretion of most of the cytokines produced by CD73+CD4+ Teffs except IL17A and only a slight effect was observed on IL22 secretion (Fig. 4D).

These results indicate that Ado suppresses the proliferation of both subsets in the same manner but dramatically restricts the functionality of CD73+CD4+ Teffs to IL17A production.

CD39+ Tregs make CD73+CD4+ Teffs sensitive to ATP-derived Ado in spaced-out environment

We previously observed that CD73+CD4+ Teffs were able to generate Ado from AMP produced by the ATP degradation from CD39+ Tregs (Fig. 3). Moreover, we found that functional CD73 on CD4+ Teffs favored their suppression mediated by AMP-derived Ado similar to that observed in presence of exogenous Ado (Supplementary Fig. S3A and S3B). We then analyzed the biological impact of the cooperation between CD39+ Tregs and CD73+CD4+ Teffs, by coculturing Tregs with purified CD73+ or CD73 CD4+ Teffs in presence or absence of exogenous ATP. As reported previously (18, 33), Tregs did not present any sign of proliferation, due to their anergic status. Conversely, the proliferation of purified CD73+ or CD73 CD4+ Teffs was unaltered by the addition of Tregs, due to the strong TCR triggering signal. Although, ATP alone did not modulate the proliferation of CD4+ Teffs subpopulations, combination of Tregs and ATP strongly suppressed the proliferation of CD73+CD4+ Teffs, but not CD73CD4+ Teffs. This inhibition was reversed by the addition of inhibitors of CD73 (APCP) or CD39 (ARL67156) (Fig. 5A; Supplementary Fig. S4).

Figure 5.

Isolated CD73+ CD4+ Teffs are preferential target of CD39+ Tregs through the generation of Ado. A and B, The impact of ATP (18.75 μmol/L/day) on the proliferation of purified CD73+ or CD73neg CD4+ Teffs alone or in coculture with total Tregs (A), or cell-sorted CD39+ and CD39neg Tregs (B) at a ratio of 1:1 was assessed by microscopy (A) or by CTV dilution (B) 4 days after their activation with expand beads (ration 1:4). C, Similar experiments were performed using Transwell plates. CD73+ (black) and CD73neg (white) CD4+ Teffs were cultured in the top and bottom chambers, respectively, in the presence of with expand beads (ratio 1:4), with or without ATP (18.75 μmol/L/day) or Ado (75 μmol/L/day), and total Tregs (R) were added either in the top or bottom chamber. D and E, The impact of ATP was also assessed on the proliferation of CD73+ and CD73neg CD4+ Teffs after the addition of Tregs, all present at a physiologic ratio (20%/70%/10%), by microscopy and by FC, by measuring the dilution of proliferation markers (CTV: CD73+CD4+ Teffs; CFSE: CD73negCD4+ Teffs). The impact of CD73 and CD39 was assessed by preincubating cells with inhibitors of CD73 (APCP, 50 μmol/L) and CD39 (ARL-67156, 250 μmol/L) for 30 minutes. ****, P < 0.0001; ns, nonsignificant.

Figure 5.

Isolated CD73+ CD4+ Teffs are preferential target of CD39+ Tregs through the generation of Ado. A and B, The impact of ATP (18.75 μmol/L/day) on the proliferation of purified CD73+ or CD73neg CD4+ Teffs alone or in coculture with total Tregs (A), or cell-sorted CD39+ and CD39neg Tregs (B) at a ratio of 1:1 was assessed by microscopy (A) or by CTV dilution (B) 4 days after their activation with expand beads (ration 1:4). C, Similar experiments were performed using Transwell plates. CD73+ (black) and CD73neg (white) CD4+ Teffs were cultured in the top and bottom chambers, respectively, in the presence of with expand beads (ratio 1:4), with or without ATP (18.75 μmol/L/day) or Ado (75 μmol/L/day), and total Tregs (R) were added either in the top or bottom chamber. D and E, The impact of ATP was also assessed on the proliferation of CD73+ and CD73neg CD4+ Teffs after the addition of Tregs, all present at a physiologic ratio (20%/70%/10%), by microscopy and by FC, by measuring the dilution of proliferation markers (CTV: CD73+CD4+ Teffs; CFSE: CD73negCD4+ Teffs). The impact of CD73 and CD39 was assessed by preincubating cells with inhibitors of CD73 (APCP, 50 μmol/L) and CD39 (ARL-67156, 250 μmol/L) for 30 minutes. ****, P < 0.0001; ns, nonsignificant.

Close modal

Similar suppression experiments performed with Treg subsets sorted on the basis of CD39 expression confirmed that CD39+ Tregs, but not CD39 ones, suppressed specifically the proliferation of CD73+CD4+Teffs in presence of exogenous ATP (Fig. 5B).

Transwell experiments enabled us to address the importance of the colocalization of CD73+CD4+ Teffs and Tregs in this Ado-mediated suppression. CD73+CD4+ Teffs were cultured in the top chamber and CD73CD4+ Teffs in the bottom one, whereas Tregs were added in either one or the other. The addition of exogenous ATP, whatever the localization of Tregs, induced a strong inhibition of CD73+CD4+ Teff proliferation, indicating that the colocalization of CD73+CD4+ Teffs and Tregs is not necessary to obtain ATP-mediated inhibition of CD73+CD4+ Teffs (Fig. 5C).

Of importance, these experiments revealed that the proliferation of CD73CD4+ Teffs localized in the bottom chamber was not altered by Ado generated by CD73+CD4+ Teffs in the top one demonstrating that CD73+ cells among CD4+ Teffs were preferentially inhibited by Tregs through Ado generation (Fig. 5C). As Ado similarly inhibited purified CD73 or CD73+ CD4+ Teffs in “U” wells (Fig. 4C) and in Transwell (Fig. 5C), we addressed the importance of both CD4+ Teffs populations' colocalization for Ado-mediated inhibition resulting from ATP degradation. In cultures where Tregs, CD73+CD4+ Teffs, and CD73 CD4+ Teffs, mixed at physiologic ratio observed in blood (10%/20%/70%), were colocalized altogether in “U” wells, the CD73+CD4+ Teffs subset was suppressed by autocrine generation of Ado through collaboration with Tregs, but the CD73population was also suppressed by paracrine action of generated Ado (Fig. 5C–E). This demonstrated that, in a culture system favoring close interactions between T-cell subsets, cooperation of Tregs and CD73+CD4+ Teffs degrading ATP into Ado, inhibited proliferation of CD73+CD4+ Teffs but also surrounding CD73negCD4+ Teffs.

Collectively, these data highlight that the cooperation between Tregs and CD73+CD4+ Teffs does not require cell contact. Furthermore, Ado, acting in an autocrine manner on CD73+CD4+ Teffs, does not affect the proliferation of CD73negCD4+ Teffs except if they are colocalized at high cell density, through collateral paracrine effect. Altogether, our data suggest that the CD73+ population among CD4+ Teffs is selectively targeted by CD39+ Tregs through autocrine Ado production.

CD73+CD4+ Teffs are present in the tumor microenvironment enriched in CD39+ Tregs

We analyzed CD73+CD4+ Teffs and CD39+ Tregs in patients with primary breast or ovarian tumors. The frequency of memory cells among CD4+ T cells (Supplementary Fig. S5A), Tregs among memory CD4+ T cells (Supplementary Fig. S5B), and the CD39/CD73 expression pattern on Tregs and CD4+ Teffs in breast tumor- or ovarian tumor–blood were similar to that observed in healthy donors (Fig. 6A and B). However, in contrast to that observed on CD4+ Teffs, the percentage and intensity of CD39 on Tregs tended to increase in breast tumor-blood (62.31% ± 14.94%) and ovarian tumor-blood (62.79% ± 14.93%), compared with HD-blood (50.01% ± 24.88%). Conversely, there was no modulation of CD73+CD4+ Teffs' proportion (Supplementary Fig. S5C), cytokine production pattern (Supplementary Fig. S5D) or Rh123 efflux capacity (Bossennec, manuscript in preparation), in the blood of patients with cancer.

Figure 6.

Breast and ovarian tumors contain highly functional CD39+ Tregs and CD73+ CD4+ Teff with Th1.17 characteristics. A, Representative FC expression of CD39 and CD73 on Tregs from paired breast tumor and blood sample. B, Representative example of CD39 and CD73 expression on CD4+ Teffs from paired breast tumor and ovarian tumor and blood samples by FC. The proportion (C) and MFI (D) of CD39 expression on Tregs from paired blood and breast tumor (n = 11) or ovarian tumor (n = 4) samples were analyzed by FC (statistical analysis, two-way ANOVA). E and F. Comparison of CD39% (E) and MFI (F) of CD4+ Tregs and Teffs within breast (n = 11) and ovarian (n = 4) tumor environment. G, Purified tumor-infiltrating Tregs were assessed for ATP (37.5 μmol/L) degradation by HPLC and compared with HD-blood Tregs. Thirty minutes preincubation with the CD39 inhibitor (ARL-67156, 250 μmol/L) was used to validate the role of CD39. H, The proportion CD73+CD4+ Teffs from paired blood and breast tumor (n = 11) or ovarian tumor (n = 4) samples analyzed by FC (statistical analysis, two-way ANOVA). I, Single-cell suspensions of breast tumor (n = 12) and ovarian tumor (n = 7) were reactivated with PMA/ionomycin, and single cytokine production (TNFα, IL2, IFNγ, and IL17A) and coproduction of IFNγ and IL17A were analyzed by FC after gating on CD73+ (black circles) or CD73neg (white circles) Teffs (statistical analysis, two-way ANOVA). J, Localization of Tregs (FoxP3+) and CD73+CD4+ Teffs was analyzed in frozen human breast tumor section by multi-immunofluorescence stainings. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, nonsignificant.

Figure 6.

Breast and ovarian tumors contain highly functional CD39+ Tregs and CD73+ CD4+ Teff with Th1.17 characteristics. A, Representative FC expression of CD39 and CD73 on Tregs from paired breast tumor and blood sample. B, Representative example of CD39 and CD73 expression on CD4+ Teffs from paired breast tumor and ovarian tumor and blood samples by FC. The proportion (C) and MFI (D) of CD39 expression on Tregs from paired blood and breast tumor (n = 11) or ovarian tumor (n = 4) samples were analyzed by FC (statistical analysis, two-way ANOVA). E and F. Comparison of CD39% (E) and MFI (F) of CD4+ Tregs and Teffs within breast (n = 11) and ovarian (n = 4) tumor environment. G, Purified tumor-infiltrating Tregs were assessed for ATP (37.5 μmol/L) degradation by HPLC and compared with HD-blood Tregs. Thirty minutes preincubation with the CD39 inhibitor (ARL-67156, 250 μmol/L) was used to validate the role of CD39. H, The proportion CD73+CD4+ Teffs from paired blood and breast tumor (n = 11) or ovarian tumor (n = 4) samples analyzed by FC (statistical analysis, two-way ANOVA). I, Single-cell suspensions of breast tumor (n = 12) and ovarian tumor (n = 7) were reactivated with PMA/ionomycin, and single cytokine production (TNFα, IL2, IFNγ, and IL17A) and coproduction of IFNγ and IL17A were analyzed by FC after gating on CD73+ (black circles) or CD73neg (white circles) Teffs (statistical analysis, two-way ANOVA). J, Localization of Tregs (FoxP3+) and CD73+CD4+ Teffs was analyzed in frozen human breast tumor section by multi-immunofluorescence stainings. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, nonsignificant.

Close modal

In breast and ovarian tumor tissues, we observed a strong enrichment of memory cells among CD4+ T cells (Supplementary Fig. S5A) and of Tregs among memory CD4+ T cells, which frequency reached roughly three times that found in patient's blood as reported previously by our team (18) and others (19). In the breast tumor and ovarian tumor microenvironment, as observed in blood, Tregs expressed CD39 but not CD73 (Fig. 6A). Interestingly, the percentage of CD39+ Tregs was also increased, leading to an increase of CD39+ Tregs frequency among total CD4+ T cells of about four folds compared with blood. Finally, the intensity of CD39 expression was also significantly higher in breast tumor and ovarian tumor compared with matched blood (Fig. 6C and D). In addition, CD39 expression on tumor-infiltrating Tregs was higher than on CD4+ Teffs in terms of percentage and intensity of expression (Fig. 6E and F). The upregulated CD39 expression on tumor Tregs compared with blood Tregs was associated to an increased capacity to degrade exogenous ATP into AMP (Fig. 6G). In contrast, the proportion of CD73+CD4+ Teffs was not significantly modulated in the tumor microenvironment (breast tumor: 4.74% ± 2.64%; ovarian tumor: 4.73% ± 2.71%) compared with blood (HD-blood: 7.37 %± 3.21%; Fig. 6B and H). Tumor-infiltrating CD73+CD4+ Teffs, activated with PMA/ionomycin, presented a higher percentage of IL2, TNFα, IFNγ, IL17A producers, and IFNγ/IL17A coproducers compared with the CD73CD4+ Teffs (Fig. 6I), as observed in breast tumor- and ovarian tumor-blood (Supplementary Fig. S5D) and HD blood (Fig. 2D). Multi-immunofluorescence stainings on primary breast tumor frozen sections confirmed the codetection of Tregs (FoxP3+) and CD73+CD4+ Teffs within the stromal immune infiltrate (Fig. 6J).

The presence of CD73+CD4+ Teffs and the enrichment of CD39+ Tregs in breast tumor and ovarian tumor support the idea that in breast and ovarian tumor environments, CD39/CD73-mediated Ado generation can occur and suppress CD4+CD73+ Teffs activation.

Tumor-infiltrating CD73+CD4+ Teffs have lower expression of inhibitory immune checkpoints

To determine whether the intrinsic capacity to transform AMP into Ado may constitute the main functional regulatory pathway of CD73+CD4+ Teffs, we analyzed, by FC, the coexpression of i-ICPs namely TIGIT, CTLA-4, TIM-3, and PD-1 on CD4+ Teffs (Fig. 7A). i-ICPs expression were mostly observed on tumor-infiltrating CD4+ T cells compared with paired blood. The relative percentage of TIGIT+ or PD-1+ cells was significantly lower in CD73+ cells compared with CD73 ones among CD4+ Teffs from tumor tissues (Fig. 7A and B). Indeed, tumor-infiltrating CD73+CD4+ Teffs contained higher proportion of cells devoid of i-ICPs and fewer cells coexpressing 2, 3, or 4 i-ICPs, compared with CD73CD4+ Teffs (Fig. 7C). Moreover, the intensity of PD-1 expression was lower on tumor-infiltrating CD73+CD4+ Teffs compared with CD73CD4+ Teffs (Fig. 7D).

Figure 7.

CD73+ CD4+ Teffs express less i-ICPs. A, Representative expression of i-ICPs (TIGIT, CTLA-4, TIM-3, PD-1) and CD73 on CD4+ Teffs from patients with cancer blood (blood-Teff) and paired tumor-infiltrating CD4+ Teffs (Ti- CD4+ Teff), analyzed by FC. B, Expression of i-ICPs on CD73+ (black) and CD73neg (white) CD4+ Teffs (statistical analysis, two-way ANOVA). C, Pie chart representing the expression and coexpression of i-ICPs (TIGIT, CTLA-4, TIM-3, and PD-1) on CD73+ and CD73neg CD4+ Teffs from breast tumor (n = 10) and ovarian tumor (n = 10) samples. D, MFI of i-ICPs (CTLA-4, PD-1, TIGIT, and TIM-3) expressed on CD73neg versus CD73+ CD4+ Teffs from breast tumor and ovarian tumor samples (statistical analysis, two-way ANOVA). *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, nonsignificant.

Figure 7.

CD73+ CD4+ Teffs express less i-ICPs. A, Representative expression of i-ICPs (TIGIT, CTLA-4, TIM-3, PD-1) and CD73 on CD4+ Teffs from patients with cancer blood (blood-Teff) and paired tumor-infiltrating CD4+ Teffs (Ti- CD4+ Teff), analyzed by FC. B, Expression of i-ICPs on CD73+ (black) and CD73neg (white) CD4+ Teffs (statistical analysis, two-way ANOVA). C, Pie chart representing the expression and coexpression of i-ICPs (TIGIT, CTLA-4, TIM-3, and PD-1) on CD73+ and CD73neg CD4+ Teffs from breast tumor (n = 10) and ovarian tumor (n = 10) samples. D, MFI of i-ICPs (CTLA-4, PD-1, TIGIT, and TIM-3) expressed on CD73neg versus CD73+ CD4+ Teffs from breast tumor and ovarian tumor samples (statistical analysis, two-way ANOVA). *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, nonsignificant.

Close modal

These findings, associated with the previous evidence that CD73+CD4+ Teffs are functionally suppressed by autocrine Ado generated through cooperation with CD39+ Tregs, strongly suggest that the CD39/CD73/Ado pathway might constitute the main pathway controlling the Th1.17 potency of human CD73+CD4+ Teffs.

Consistent with this conclusion, analyses done on PD-1 expressing CD4+ tumor-infiltrating T cells in breast tumors and ovarian tumors unveiled that the proportion of cells devoid of i-ICPs was reduced to 7% among breast tumor–infiltrating PD1+CD4+ Teffs (and 15.5% for ovarian tumors), contrasting with more than 63% among breast tumor-infiltrating CD73+CD4+ Teffs (and 46.8% for ovarian tumor; Supplementary Fig. S5E).

We have herein demonstrated that CD73 expression on CD4+ T cells delineates polyfunctional memory CD4+ Teffs enriched in Th1.17 cells, whereas Tregs are devoid of CD73 expression. CD73+CD4+ Teffs are targeted by CD39+ Tregs through ATP degradation into Ado inhibiting and restricting their functionality to IL17A secretion. This contact-independent cooperation induces specific inhibition of CD73+ cells among CD4+ Teffs in a spaced-out environment. In breast and ovarian tumor environments, the increased proportion of Tregs overexpressing CD39 and the presence of CD73+CD4+ Teffs almost devoid of i-ICPs strongly suggest that CD73-mediated generation of autocrine Ado represents an essential regulatory mechanism of these potent CD4+ Teffs.

Our results highlight that, compared with CD73CD4+ Teffs, the CD73+ subset secretes higher levels of proinflammatory cytokines (IL17A, IFNγ, GM-CSF, IL22, TNFα, IL2, IL3), and lower levels of anti-inflammatory ones (IL10, IL13, and IL21). Of importance, FC analyses reveal the higher propensity of CD73+CD4+ Teffs to coproduce IFNγ and IL17A and coexpress CXCR3/CCR6, in line with Th1.17 compared with CD73CD4+ Teffs. Our results extend the work of Doherty and colleagues (24), who identified the memory CD73+CD4+ T cells as Th17 cells. Indeed, the exploitation of the transcriptome data from Ramesh and colleagues (27), comparing MDR1+ and MDR1 CD4+ memory populations (Gene Expression Omnibus repository, accession ID: GSE49702), reveals that NT5E mRNA, coding for CD73, is one of the most significantly upregulated genes in the MDR1+ population [Log2-fold change (FC) = 0.867; P = 0.004]. Moreover, CD73+CD4+ Teffs present a number of features of Th1.17 cells recently described by Ramesh and colleagues (27). They express high level of functional MDR1, secrete IL3 and GM-CSF and present higher expression of IL23R in contrast to nonpathogenic Th17 that produce only IL17A, IL22, IL10, and IL21 (26, 27, 34). Altogether, our data emphasize the polyfunctionality of CD73+CD4+ Teffs and their similarity with Th1.17.

Th1.17 cells are known to play a pathogenic role in autoimmune diseases through their production of IFNγ and GM-CSF in addition to IL17A (for review; ref. 35). While, Th1.17 cells, also called pathogenic Th17 induced by Gram+ bacteria influence gut antitumor immune response in murine models (36), information on the impact of this population in breast and ovarian tumors remains limited. In fact, in patients with invasive breast carcinoma, IL17A produced by CD4+ T cells positively correlates with a high histologic grade and with triple-negative molecular subtype, and represents an independent prognostic factor for shorter disease-free survival (37, 38). In line with these observations, neutralization of IL17 in murine mammary tumor models inhibits tumor growth (39). In contrast, in ovarian carcinoma, the detection of IL17 within the tumor microenvironment (40), and more precisely the presence of CD4+ T cells coexpressing IL17 and IFNγ (41), is associated with a good prognosis. Owing to their capacity to produce IFNγ, these cells can exert a potent tumor suppressive activity in synergy with IL17 to reprogram recruited neutrophils and myeloid-derived suppressor cells into potent antitumor effectors, and to neutralize the proangiogenic effects of IL17 through the production of the CXCR3 ligand (CXCL10; ref. 41). This is in line with the strong antitumor response associated to Th1.17 cells in murine ovarian tumor model (28, 41).

Of importance, we observe that Ado blocks all cytokines produced by CD73+CD4+ Teffs, except IL17A. In this context, the analysis of cytokine pattern in 102 human breast and ovarian tumor mechanic disaggregation milieu (STM: soluble tumor milieu) with sensitive method highlight that 51% (n = 52) STM were negative for both IL17A and IFNγ, 28.4% (n = 29) contained only IL17A, whereas 18.6% (n = 19) contained both IL17A and IFNγ and 2.9% (n = 3) IFNγ only. As CD73+CD4+ Teffs were detected in most tumors at a frequency similar to blood and preserve their capacity to coproduce IL17A and IFNγ upon PMA + ionomycin activation (Fig. 6), these STM data suggest that the level of endogenous Ado production may vary from tumor to tumor, a high level of Ado production leading to IL17A only, while a low Ado level allowing both IL17A and IFNγ coproduction. Moreover, 87.5% of STM producing both IL17A and IFNγ contained high concentrations of CXCL10 (>8 ng/mL), whereas only 12.5% of those with IL17A only and none of the other contained significant levels of CXCL10. This represents another argument that IL17/IFNγ found in STM may be coproduced by CCR6+CXCR3+ Th1.17 CD73+ Teffs. Further analyses remain necessary to confirm this.

Importantly, we demonstrate that in the presence of exogenous ATP, CD39+ Tregs potently inhibit the proliferation and cytokine production of purified CD73+CD4+ Teffs through CD73-dependent Ado generation without impacting CD73CD4+ Teffs. These results support the concept that the potent CD73+ Th1.17 cells are selectively inhibited by CD39+ Tregs, through the cooperative production of Ado in their local microenvironment. The localized effect of Ado can be explained by its short half-life in tissues due to the presence of ADA and specific membrane transporters (42).

High ATP concentrations released in the tumor microenvironment (1) play an important role in the initiation of the immune response, whereas Ado, also found in human tumor environment (43), fosters tumor progression through its immunosuppressive functions (for review; ref. 2).

In this study, we confirm the increase in Tregs proportion in breast and ovarian tumor environment (18, 19) and demonstrate an overexpression of CD39 on these tumor-infiltrating Tregs (frequency of CD39+ Tregs within total CD4+ T cells, about 4 folds in tumor compared with blood), which efficiently degrades exogenous ATP into AMP. Consequently, CD73+CD4+ Teffs will be inhibited through autocrine Ado production. These results support the concept that the Th1.17 potency of CD73+CD4+ effectors can be selectively inhibited by tumor-infiltrating CD39+ Tregs through cooperative Ado production. Others CD39-expressing cells such as Mφ or B cells may also contribute to the regulation of this CD73+CD4 Teffs (44, 45).

Of further relevance for therapy, CD73+CD4+ Teffs from blood or tumor present a high expression of the functional multidrug transporter MDR1, suggesting their protection from deleterious effects of chemotherapy, thus strengthening the role they could play in the antitumor immunity. This key property could be evaluated through the analysis of the proportion of these CD73+CD4+ Teffs following neoadjuvant chemotherapy. Furthermore, CD73+CD4+ Teffs express very low levels of i-ICPs (PD-1, TIM-3, TIGIT, and CTLA-4), suggesting the nonredundancy of i-ICPs and the adenosinergic pathway in the regulation of these cells. These results corroborate recent reports, in murine tumor models, demonstrating that Ado limits the efficacy of immunotherapeutic interventions targeting CTLA-4 or PD-1/PDL-1, and that CD73 neutralization enhances the efficacy of these therapies (16, 17). Thus, CD73-mediated generation of Ado is an important mechanism of regulation of these polyfunctional CD73+CD4+ Teffs, and may represent a critical resistance mechanism to current immunotherapies (anti-PD-1/L1 and CTLA-4). These findings should promote the development of combined therapies based on immunotherapies targeting i-ICPs and the neutralization of the enzymatic function of CD73 (46, 47).

Furthermore, clinical investigations are required to evaluate the presence of CD73+CD4+ Teffs in primary or chemotherapy-treated tumors and their clinical impact on the survival of the patients in association with the presence of CD39+ cells such as Tregs or Mφ.

No potential conflicts of interest were disclosed.

Conception and design: N. Gourdin, C. Machon, J. Faget, P. Romero, C. Caux, C. Ménétrier-Caux

Development of methodology: N. Gourdin, C. Rodriguez, S. Vigano, J.C. Marie, J. Guitton, C. Ménétrier-Caux

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): N. Gourdin, M. Bossennec, S. Vigano, C. Machon, C. Jandus, D. Bauché, N. Chopin, O. Tredan, J.C. Marie, J. Guitton

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): N. Gourdin, M. Bossennec, S. Vigano, C. Machon, C. Jandus, C. Ménétrier-Caux

Writing, review, and/or revision of the manuscript: N. Gourdin, M. Bossennec, S. Vigano, C. Jandus, O. Tredan, J.C. Marie, B. Dubois, P. Romero, C. Caux, C. Ménétrier-Caux

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): C. Rodriguez, D. Bauché, I. Durand, C. Ménétrier-Caux

Study supervision: N. Gourdin, C. Caux, C. Ménétrier-Caux

C. Caux, C. Ménétrier-Caux, N. Gourdin, M. Bossennec, B. Dubois, and C. Rodriguez were financially supported by the Breast Cancer Research Foundation, and grants from the Puy de Dôme comity of the “Ligue Contre le Cancer.” J.C. Marie, D Bauché, C. Caux, C. Ménétrier-Caux, B. Dubois, N. Gourdin, M. Bossennec and C. Rodriguez, O. Tredan, I. Durand and N. Chopin were financially supported by the SIRIC project (LYRIC, grant no. INCa_4664). C. Caux, C. Ménétrier-Caux, N. Gourdin, M. Bossennec and C. Rodriguez, P. Romero, C. Jandus and S. Vigano were financially supported by the FP7 European TumAdoR project (grant no. 602200). J.C. Marie, D Bauché, and C. Caux were financially supported by the LABEX DEVweCAN (ANR-10-LABX-0061) of the University of Lyon, within the program “Investissements d'Avenir” organized by the French National Research Agency (ANR). We wish to thank the staff of the core facilities at the Cancer Research Center of Lyon (CRCL) for their technical assistance and the BRC (Biological Resources Centre) of the CLB for providing human samples. Drs. P. Guibert, F. Desseigne, M. Sarabi, and L. Mais are acknowledged for their clinical expertise. We will also thank Marie Laure Thibult (Innate Pharma) for her assistance for flow cytometry sorting. We are very grateful to Justine Berthet for her help and expertise in the development of multi-IF analysis on breast tumor sections. We would also like to thank Dr. B. Manship for critical reading of the manuscript.

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.
Pellegatti
P
,
Raffaghello
L
,
Bianchi
G
,
Piccardi
F
,
Pistoia
V
,
Di
VF
. 
Increased level
of extracellular ATP at tumor sites: in vivo imaging with plasma membrane luciferase
.
PLoS One
2008
;
3
:
e2599
.
2.
Junger
WG
. 
Immune cell regulation by autocrine purinergic signalling
.
Nat Rev Immunol
2011
;
11
:
201
12
.
3.
Deaglio
S
,
Dwyer
KM
,
Gao
W
,
Friedman
D
,
Usheva
A
,
Erat
A
, et al
Adenosine generation catalyzed by CD39 and CD73 expressed on regulatory T cells mediates immune suppression
.
J Exp Med
2007
;
204
:
1257
65
.
4.
Stagg
J
,
Divisekera
U
,
Duret
H
,
Sparwasser
T
,
Teng
MW
,
Darcy
PK
, et al
CD73-deficient mice have increased antitumor immunity and are resistant to experimental metastasis
.
Cancer Res
2011
;
71
:
2892
900
.
5.
Wang
L
,
Fan
J
,
Thompson
LF
,
Zhang
Y
,
Shin
T
,
Curiel
TJ
, et al
CD73 has distinct roles in nonhematopoietic and hematopoietic cells to promote tumor growth in mice
.
J Clin Invest
2011
;
121
:
2371
82
.
6.
Ohta
A
. 
A metabolic immune checkpoint: adenosine in tumor microenvironment
.
Front Immunol
2016
;
7
:
109
.
7.
Hasko
G
,
Kuhel
DG
,
Chen
JF
,
Schwarzschild
MA
,
Deitch
EA
,
Mabley
JG
, et al
Adenosine inhibits IL-12 and TNF-alpha production via adenosine A2a receptor-dependent and independent mechanisms
.
FASEB J
2000
;
14
:
2065
74
.
8.
Hasko
G
,
Kuhel
DG
,
Salzman
AL
,
Szabo
C
. 
ATP suppression of interleukin-12 and tumour necrosis factor-alpha release from macrophages
.
Br J Pharmacol
2000
;
129
:
909
14
.
9.
Hasko
G
,
Szabo
C
,
Nemeth
ZH
,
Kvetan
V
,
Pastores
SM
,
Vizi
ES
. 
Adenosine receptor agonists differentially regulate IL-10, TNF-alpha, and nitric oxide production in RAW 264.7 macrophages and in endotoxemic mice
.
J Immunol
1996
;
157
:
4634
40
.
10.
Le
MO
,
Stordeur
P
,
Schandené
L
,
Marchant
A
,
de Groote
D
,
Goldman
M
, et al
Adenosine enhances IL-10 secretion by human monocytes
.
J Immunol
1996
;
156
:
4408
14
.
11.
Hirschhorn
R
,
Vawter
GF
,
Kirkpatrick
JA
 Jr.
,
Rosen
FS
. 
Adenosine deaminase deficiency: frequency and comparative pathology in autosomally recessive severe combined immunodeficiency
.
Clin Immunol Immunopathol
1979
;
14
:
107
20
.
12.
Ohta
A
,
Gorelik
E
,
Prasad
SJ
,
Ronchese
F
,
Lukashev
D
,
Wong
MKK
, et al
A2A adenosine receptor protects tumors from antitumor T cells
.
Proc Natl Acad Sci U S A
2006
;
103
:
13132
7
.
13.
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
.
14.
Terp
MG
,
Olesen
KA
,
Arnspang
EC
,
Lund
RR
,
Lagerholm
BC
,
Ditzel
HJ
, et al
Anti-human CD73 monoclonal antibody inhibits metastasis formation in human breast cancer by inducing clustering and internalization of CD73 expressed on the surface of cancer cells
.
J Immunol
2013
;
191
:
4165
73
.
15.
Zhang
B
. 
CD73 promotes tumor growth and metastasis
.
Oncoimmunology
2012
;
1
:
67
70
.
16.
Allard
B
,
Pommey
S
,
Smyth
MJ
,
Stagg
J
. 
Targeting CD73 enhances the antitumor activity of anti-PD-1 and anti-CTLA-4 mAbs
.
Clin Cancer Res
2013
;
19
:
5626
35
.
17.
Iannone
R
,
Miele
L
,
Maiolino
P
,
Pinto
A
,
Morello
S
. 
Adenosine limits the therapeutic effectiveness of anti-CTLA4 mAb in a mouse melanoma model
.
Am J Cancer Res
2014
;
4
:
172
81
.
18.
Gobert
M
,
Treilleux
I
,
Bendriss-Vermare
N
,
Bachelot
T
,
Goddard-Leon
S
,
Arfi
V
, et al
Regulatory T cells recruited through CCL22/CCR4 are selectively activated in lymphoid infiltrates surrounding primary breast tumors and lead to an adverse clinical outcome
.
Cancer Res
2009
;
69
:
2000
9
.
19.
Curiel
TJ
,
Coukos
G
,
Zou
L
,
Alvarez
X
,
Cheng
P
,
Mottram
P
, et al
Specific recruitment of regulatory T cells in ovarian carcinoma fosters immune privilege and predicts reduced survival
.
Nat Med
2004
;
10
:
942
9
.
20.
Alam
MS
,
Kurtz
CC
,
Rowlett
RM
,
Reuter
BK
,
Wiznerowicz
E
,
Das
S
, et al
CD73 is expressed by human regulatory T helper cells and suppresses proinflammatory cytokine production and Helicobacter felis-induced gastritis in mice
.
J Infect Dis
2009
;
199
:
494
504
.
21.
Mandapathil
M
,
Szczepanski
MJ
,
Szajnik
M
,
Ren
J
,
Jackson
EK
,
Johnson
JT
, et al
Adenosine and prostaglandin E2 cooperate in the suppression of immune responses mediated by adaptive regulatory T cells
.
J Biol Chem
2010
;
285
:
27571
80
.
22.
Dwyer
KM
,
Hanidziar
D
,
Putheti
P
,
Hill
PA
,
Pommey
S
,
McRae
JL
, et al
Expression of CD39 by human peripheral blood CD4+ CD25+ T cells denotes a regulatory memory phenotype
.
Am J Transplant
2010
;
10
:
2410
20
.
23.
Schuler
PJ
,
Schilling
B
,
Harasymczuk
M
,
Hoffmann
T
,
Johnson
J
,
Lang
S
, et al
Phenotypic and functional characteristics of ATP-hydrolysing CD4(+) CD39+ FOXP3+ and CD4+ CD39+ FOXP3neg T-cell subsets in patients with cancer
.
Eur J Immunol
2012
;
42
:
1876
85
.
24.
Doherty
GA
,
Bai
A
,
Hanidziar
D
,
Longhi
MS
,
Lawlor
GO
,
Putheti
P
, et al
CD73 is a phenotypic marker of effector memory Th17 cells in inflammatory bowel disease
.
Eur J Immunol
2012
;
42
:
3062
72
.
25.
Lee
Y
,
Awasthi
A
,
Yosef
N
,
Quintana
FJ
,
Xiao
S
,
Peters
A
, et al
Induction and molecular signature of pathogenic TH17 cells
.
Nat Immunol
2012
;
13
:
991
9
.
26.
Duhen
T
,
Campbell
DJ
. 
IL-1beta promotes the differentiation of polyfunctional human CCR6+CXCR3+ Th1/17 cells that are specific for pathogenic and commensal microbes
.
J Immunol
2014
;
193
:
120
9
.
27.
Ramesh
R
,
Kozhaya
L
,
McKevitt
K
,
Djuretic
IM
,
Carlson
TJ
,
Quintero
MA
, et al
Pro-inflammatory human Th17 cells selectively express P-glycoprotein and are refractory to glucocorticoids
.
J Exp Med
2014
;
211
:
89
104
.
28.
Muranski
P
,
Boni
A
,
Antony
PA
,
Cassard
L
,
Irvine
KR
,
Kaiser
A
, et al
Tumor-specific Th17-polarized cells eradicate large established melanoma
.
Blood
2008
;
112
:
362
73
.
29.
Dubin
PJ
,
Kolls
JK
. 
Th17 cytokines and mucosal immunity
.
Immunol Rev
2008
;
226
:
160
71
.
30.
Machon
C
,
Bordes
C
,
Cros-Perrial
E
,
Clement
Y
,
Jordheim
LP
,
Lanteri
P
, et al
Use of designed experiments for the improvement of pre-analytical workflow for the quantification of intracellular nucleotides in cultured cell lines
.
J Chromatogr A
2015
;
1405
:
116
25
.
31.
Baecher-Allan
C
,
Brown
JA
,
Freeman
GJ
,
Hafler
DA
. 
CD4+CD25high regulatory cells in human peripheral blood
.
J Immunol
2001
;
167
:
1245
53
.
32.
Zielinski
CE
,
Mele
F
,
Aschenbrenner
D
,
Jarrossay
D
,
Ronchi
F
,
Gattorno
M
, et al
Pathogen-induced human TH17 cells produce IFN-gamma or IL-10 and are regulated by IL-1beta
.
Nature
2012
;
484
:
514
8
.
33.
Korn
T
,
Bettelli
E
,
Oukka
M
,
Kuchroo
VK
. 
IL-17 and Th17 Cells
.
Annu Rev Immunol
2009
;
27
:
485
17
.
34.
Viaud
S
,
Daillère
R
,
Boneca
IG
,
Lepage
P
,
Langella
P
,
Chamaillard
M
, et al
Gut microbiome and anticancer immune response: really hot Sh*t!
.
Cell Death Differ
2015
;
22
:
199
214
.
35.
Benevides
L
,
Cardoso
CR
,
Tiezzi
DG
,
Marana
HR
,
Andrade
JM
,
Silva
JS
. 
Enrichment of regulatory T cells in invasive breast tumor correlates with the upregulation of IL-17A expression and invasiveness of the tumor
.
Eur J Immunol
2013
;
43
:
1518
28
.
36.
Chen
WC
,
Lai
YH
,
Chen
HY
,
Guo
HR
,
Su
IJ
,
Chen
HH
. 
Interleukin-17-producing cell infiltration in the breast cancer tumour microenvironment is a poor prognostic factor
.
Histopathology
2013
;
63
:
225
33
.
37.
Benevides
L
,
da Fonseca
DM
,
Donate
PB
,
Tiezzi
DG
,
De Carvalho
DD
,
de Andrade
JM
, et al
IL17 promotes mammary tumor progression by changing the behavior of tumor cells and eliciting tumorigenic neutrophils recruitment
.
Cancer Res
2015
;
75
:
3788
99
.
38.
Lan
C
,
Huang
X
,
Lin
S
,
Huang
H
,
Cai
Q
,
Lu
J
,
Liu
J
. 
High density of IL-17-producing cells is associated with improved prognosis for advanced epithelial ovarian cancer
.
Cell Tissue Res
2013
;
352
:
351
9
.
39.
Kryczek
I
,
Banerjee
M
,
Cheng
P
,
Vatan
L
,
Szeliga
W
,
Wei
S
, et al
Phenotype, distribution, generation, and functional and clinical relevance of Th17 cells in the human tumor environments
.
Blood
2009
;
114
:
1141
9
.
40.
King
AE
,
Ackley
MA
,
Cass
CE
,
Young
JD
,
Baldwin
SA
. 
Nucleoside transporters: from scavengers to novel therapeutic targets
.
Trends Pharmacol Sci
2006
;
27
:
416
25
.
41.
Blay
J
,
White
TD
,
Hoskin
DW
. 
The extracellular fluid of solid carcinomas contains immunosuppressive concentrations of adenosine
.
Cancer Res
1997
;
57
:
2602
5
.
42.
Kansas
GS
,
Wood
GS
,
Tedder
TF
. 
Expression, distribution, and biochemistry of human CD39. Role in activation-associated homotypic adhesion of lymphocytes
.
J Immunol
1991
;
146
:
2235
44
.
43.
MacKenzie
WM
,
Hoskin
DW
,
Blay
J
. 
Adenosine suppresses alpha(4)beta(7) integrin-mediated adhesion of T lymphocytes to colon adenocarcinoma cells
.
Exp Cell Res
2002
;
276
:
90
100
.
44.
Geoghegan
JC
,
Diedrich
G
,
Lu
X
,
Rosenthal
K
,
Sachsenmeier
KF
,
Wu
H
, et al
Inhibition of CD73 AMP hydrolysis by a therapeutic antibody with a dual, non-competitive mechanism of action
.
MAbs
2016
;
8
:
454
67
.
45.
Giraudon-Paoli
M
,
Augier
S
,
Blemont
MR
,
Rodriguez
C
,
Blanc
HR
,
Chanteux
S
, et al
Discovery and
characterization of new original blocking antibodies targeting the CD73 immune checkpoint for cancer immunotherapy[abstract].
In:
Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16–20; New Orleans, LA
.
Philadelphia (PA)
:
AACR
; 
2016
.
Abstract nr 2344.
46.
Verronese
E
,
Delgado
A
,
Valladeau-Guilemond
J
,
Garin
G
,
Guillemaut
S
,
Tredan
O
, et al
Immune cell dysfunctions in breast cancer patients detected through whole blood multi-parametric flow cytometry assay
.
Oncoimmunology
2016
;
5
:
e1100791
.
47.
Dimeloe
S
,
Frick
C
,
Fischer
M
,
Gubser
PM
,
Razik
L
,
Bantug
GR
, et al
Human regulatory T cells lack the cyclophosphamide-extruding transporter ABCB1 and are more susceptible to cyclophosphamide-induced apoptosis
.
Eur J Immunol
2014
;
44
:
3614
20
.