Extracellular adenosine and ATP exert important functions in physiology and pathophysiology. They for instance play key role in heart and vascular function, during pregnancy (1) and in immune responses (2). Outside the cell, extracellular ATP (eATP) acts as danger-associated molecular patterns (DAMPs) and can bind to purinergic receptors to trigger signaling cascades to induce an inflammatory response (3). However, adenosine is a potent immune-suppressor of cells that express A2 and A3 receptors, such as lymphocytes (4). To avoid ATP-induced pathological effects, ATP can be hydrolyzed into adenosine and phosphate by a cascade of enzymes, of which CD39 is the most important.

CD39, the NTPDase (ecto-nucleoside triphosphate diphosphohydrolase), regulates immune responses balance by hydrolyzing ATP and ADP. It is now again becoming a newly recognized “immune checkpoint mediator” that interferes with antitumor or anti-inflammatory immune response (5, 6). Moreover, some recent research has revealed a number of neoteric functions of CD39, which display its close relation with Tregs (7–9), Th17 cells (10, 11), γδ T cells (12), and Bregs (13). It follows from this reasoning that CD39 acts as a key molecule in inflammation (14–16) and tumor immunity (17–20).

In the present review, we will discuss the current knowledge on the role of CD39 expressed on different types of cells and explore its potential in inflammation and tumor immunity.

Cellular ATP serves as the main energy currency, driving virtually all cell functions. Self-evidently, intracellular ATP plays important pathophysiological roles. In terms of eATP, it is ubiquitously used for cell–cell communication in physical setting. Low concentration of eATP-detected surrounding resting cells indicates the presence of neighboring living cells, especially in nervous (21) and vascular systems (22). And also, this biochemical substance can be poured into extracellular space upon, for instance, tissue stress such as necrosis, apoptosis, hypoxia, or inflammation (23). Release from activated or apoptotic cells is done via two mechanisms as follows: transport via membrane-bound channels or transporters and exocytosis of intracellular vesicles (24). Additional work showed that kindred purinergic signaling pathways regulate critical aspects of many other physiological processes, including immune response (25).

Recent years, the role of eATP in immune system has broaden our horizon. ATP release in response to inflammatory mediators is a basic mechanism required for neutrophil activation and immune defense (26). The steady-state cytosolic concentration of ATP is 3–10 mM, whereas eATP is only~10 nM (27), which is maintained as a result of the activities of extracellular ecto-apyrases and CD39. The enzymatic activities of CD39 and CD73 play paramount roles in calibrating the particularity, duration, and magnitude of purinergic signals via the conversion of ATP/ADP to AMP and AMP to adenosine, respectively. The ATP-CD39–CD73–adenosine cascade is strictly controlled by enzymatic activity, in which CD39 serves as the rate-limiting enzyme (28).

Balance between eATP and adenosine (see Figure 1) is crucial in immune homeostasis since eATP is a danger signal released by injured or apoptosis cells that acts to prime immune responses through the ligation of P2 receptors (2). There are two subsets of P2 receptors: P2X or P2Y receptors (23). Seven P2X receptors plus eight P2Y receptors have been identified in humans (26, 29). Responses to low eATP are mediated by P2 receptors with high or intermediate affinity for eATP (EC50 <20 µM), while responses to high eATP are mediated by P2X7 (EC50 >100 μM) (27). As for a further explanation, eATP functions as DAMPs and then binds to P2 receptors, resulting in heightened inflammation and regulatory cell inhibition in most cases (30).

ATP has varieties of pro-inflammatory effects. Since sundry immune cells express most of the ATP receptors, ATP can affect most immune cells. For example, eATP released by damaged cells can activate the immune system via the stimulation of P2X7 receptors on DCs and then promote the secretion of IL-1β and IL-18 (31). Next, IL-1β will facilitate macrophages maturation and their cytokine production increase (32). Similarly, IL-18 would boost NK cells proliferation and strengthen IFN-γ production plus cytotoxicity (33). While in T cells, ATP activates T cells by inducing IL-2 production and cytotoxicity. Moreover, it induces differentiation toward pro-inflammatory Th17 cells, while it inhibits the differentiation toward Tregs. From a micro perspective, eATP stimulates Ca²⁺ entry through P2X purinergic receptor channels and Ca²⁺ mobilization to facilitate Ca²⁺-calmodulin-dependent activation and nuclear translocation of nuclear factor of activated T cells (NFAT), which stimulates the production of IL-2, pannexin 1 channels, and other NFAT targets. Autocrine ATP release helps to sustain P2 purinergic receptor signaling and NFAT activation (34).

Adenosine—the counterpart of ATP, which is produced by breakdown of ATP—is nothing of a novelty. As early as 1980s, adenosine has been used to slow down the heart rate of patients suffering from supraventricular tachycardia (22). As for immune system, rather than activating T cells responses, adenosine conversely inhibits T cells responses including Tregs and Th cells. But it is not contradictory that adenosine inhibits the differentiation toward Th17 cells while it promotes differentiation toward Tregs. It is actually regarded as a key endogenous molecule that regulates tissue function by activating four G-protein-coupled P1 receptors, denoted A1, A2A, A2B, and A3 (35). A1 and A2A are high-affinity receptors, while A2B and A3 are low-affinity receptors (36). Meanwhile, A2A and A2B stimulate adenylyl cyclase, while A1 and A3 inhibit adenylyl cyclase (36). P1 receptors are expressed on kinds of immune cells such as macrophages, dendritic cells, and lymphocytes. There are now varieties of promising emerging therapeutic approaches centered on the modulation of adenosine in the immune system (37). Triggering different receptors can have different consequence. Importantly, A2A receptors are closely related to cyclic adenosine monophosphate (cAMP) response element-binding protein, which eventually leads to the transcription of the CEBPβ gene (38). While CEBPβ protein binds to the IL-10 gene promoter, which triggers IL-10 transcription, and subsequently leads to the release of IL-10 (38). IL-10, human cytokine synthesis inhibitory factor, was reported to suppress cytokine secretion, antigen presentation, and CD4⁺ T cell activation (39). One interesting research to mention is that resilient individuals seem to have a better anti-inflammatory response compared to posttraumatic stress disorder patients since they present higher IL-10 levels (40).

The ultimate effect of ATP and adenosine during immune responses depends on the balance between the two molecules. Of note, though CD39 and CD73 expressed on immune cells decrease local ATP levels while increase local adenosine levels, the substrate and catalytic product of ATP/ADP–CD39–AMP–adenosine pathway, are not actually simple yin and yang in immune responses (41). In fact, both ATP and adenosine may have dual effects on immune responses, depending on concentration, the duration of the exposure, and the conditions of the in vivo environment. Under pronged exposure or at low concentrations, as mentioned before, responses to eATP are mediated by P2R with EC50 <20 μM (41). Then, it will reduce secretion of inflammatory cytokine including IL-1β, IL-6, IL-12, and TNF-α, etc., of macrophages or mature DCs (42).

The purinergic system is a high-level delicate system adjusted to fine-tune immune cell functions. Varieties of cell types can release ATP or ADP from intracellular compartments into the extracellular spaces, besides apoptotic cells. eATP functions as a dangerous signal triggering activation of purinergic P2 receptors and hence subsequently a series of pro-inflammation responses (23, 96). This pro-inflammation cascade is terminated by conversion to suppressive molecule adenosine via CD39 and CD73. Similar to ATP, adenosine activates purinergic P1 receptors. During the process of inflammation, there are also several pathogenic microorganisms capable of inducing an adenosine-rich milieu, which favors them to escape host immune surveillance. By contrast, there is also evidence that overexpression of CD39 in mouse airways promotes bacteria-induced inflammation (97). The author speculated that CD39 might limit the desensitization of P2 receptors, which helps to promote airway inflammation in response to bacterial challenge. Consistent with the above results, CD39 expression and activity are elevated in chronic obstructive pulmonary disease (COPD) patients by quantifying CD39 expression and soluble ATPase activity in sputum and bronchoalveolar lavage fluid cells (98). Additionally, Tan et al. showed T cell-related CD39 expression is higher in acute exacerbations of COPD (AECOPD) patients than stable COPD and healthy controls (5). They prospected blocking CD39 would be a novel approach to the control of AECOPD, reducing the dependency on antibiotics.

In addition, some microorganisms themselves are equipped with high nucleotide metabolic versatility, which assist them with dissemination and invasion in the host (99). For example, Fan et al. reported that an ecto-5′-nucleotidase similar to human CD39 on cell surface of Streptococcus sanguinis contributes to its virulence (100). This analog takes effects by means of slowing the platelet aggregation response in vitro and reducing the accumulation of platelets on infected heart valves in vivo (100).

Hence, CD39, a pivotal enzyme between ATP and adenosine, is critical in preventing excessive P2R-mediated inflammation, but its function maybe turns detrimental for the appropriate clearance of apoptotic debris or by generating an immunosuppressive environment, which might promote the development or progression of cancer (101).

Interactions between tumor cells and their immunological microenvironment are essential for the pathophysiology of tumor (102). CD39, the rate-limiting enzyme in the generation of immune-suppressive adenosine, undoubtedly plays a pivotal role in tumor progression. For example, CD39⁺ Tregs, mentioned above, inhibited NK cell antitumor immunity both in vitro and in vivo (20). Besides this, CD39 is expressed at significantly higher rates in tumor-infiltrating tissue such as ovarian, pancreatic, and testicular tumors, etc., in contrast to paired peritumoral tissue (55). Overexpression of CD39 was reported as a predictor of poor outcome for gastric cancer patient following radical resection (103). Additionally, CD39 expressed on Tregs was shown to play a permissive role in a mouse model of hepatic metastasis (104). Therefore, CD39 penetrates deep into to the modulation of tumor cell growth, differentiation, invasion, and migration (105, 106).

Importantly, after co-incubating these tumor cells with POM-1, the CD39 inhibitor, tumor-induced inhibition of CD4⁺ and CD8⁺ T cell proliferation was alleviated and CTL- or NK cell-mediated cytotoxicity increased. Consequently, treatment with a CD39 inhibitor or blocking antibody may become a promising strategy for ameliorating tumor cell-mediated immunosuppression. These data indicate CD39 to be a probable checkpoint in tumor immunotherapy and inhibiting CD39 may restore antitumor responses or boost the efficacies of other antitumor strategies. For example, treatment with ARL6715, another ectoATPase inhibitor, resulted in improving T cell responsiveness (107). The link between CD39 and tumor will become clearer since tumor-related research focused on CD39 is emerging in endless stream.

The enzymatic activity of CD39, combined with CD73, plays a non-ignorable part in the shift from an ATP-mediated pro-inflammatory milieu to an immunosuppressive setting driven by adenosine. Some recent discovery about its new features such as hydrolyzing pAgs and catalyzing sphingomyelin suggest that CD39 have more unrecognized physical functions. Either way, in-depth study of CD39 will provide unique insights into the workings of immune network.

First, in basic research, CD39 can be considered as a novel bridge among immune cells. It is expressed on varieties of immune cells and there exists close relationship between CD39 and their functions. In the context of a subgroup, classification of immune cells based on CD39 may reflect their functions better. It is remarkable that CD39 is increasingly appreciated as a regulatory marker other than an activation marker. We can identify not only CD4⁺ Tregs but also CD8⁺ Tregs, Bregs, and γδ-Tregs based on CD39. As for its newly known feature about hydrolyzing pAgs, this may represent another mechanism by which CD4⁺ Tregs suppress γδ T cells. Besides “classical” functions such as cytokine production and cytotoxicity, recent studies suggest that γδ T cells are equipped with additional efficiencies such as regulatory activity (20) and—quite excitingly—“professional” antigen-presenting capacity (108). We should breathe calmly that it is absolutely worth a great number of costs on this fewer proportion.

Another emerging discovery about CD39 is closely related to CD161—they altogether modulate human Th17 cells responsiveness through alterations in ASM (10). CD39 and CD161 serve as potent surface markers of Th17 cells (10, 11), and furthermore, the latter has been identified as the top favorable pan-cancer prognostic molecule (109). There is still a long way to go to expound on the connection between CD39 and CD161. And what’s more, γδ T cells are also found to be a very important source of IL-17 in some disease models, particularly at early stage (110). Whether CD39/CD161-ASM-IL-17 chain also exists in γδ T cells has not been evaluated. Additionally, expression of CD39 is particularly associated with exhaustion of CD8⁺ T cells and identification of CD8⁺ Tregs, Bregs, both of which are being in the ascendant among immunological studies.

Second, the diagnostic potential of CD39 is only beginning to unfold. There have been already some studies proposing CD39 as a prognostic marker such as pancreatic cancer (105). In the setting of chronic lymphocytic leukemia, both CD4⁺ T cells and CD8⁺ T cells express high-level CD39 and furthermore, CD39 is associated with advanced disease stage (111). Besides, CD39 is also closely related to both Th17 cells and Tregs. In the context of inflammation or autoimmune diseases such as rheumatoid arthritis, Th17 cells stand for a pro-inflammatory subpopulation while Tregs have the antagonist effect. Hence, the Th17/Tregs balance can affect the outcome of immune responses. CD39, surprisingly, participates in the identification of both (see above). Fan et al. proposed identifying CD4⁺ T cell-derived CD161⁺CD39⁺ and CD39⁺CD73⁺ microparticles as new biomarkers for rheumatoid arthritis evaluation (11).

Finally, immunotherapy will be the direction of our long-term efforts and moreover, the possibly ultimate way out to inflammation and cancer. Immunomodulatory effects of CD39 brought a new dawn for immunotherapy. Since CD39 not only hydrolyzes eATP but also increases the concentration of anti-inflammatory adenosine, the administration of exogenous CD39, in nanoparticles or other forms may provide a new approach to limiting inflammation, which is likely to be more efficient than existing strategies aimed at blocking P2X. Alternatively, given the increased CD39 expression in apoptosis-prone T cells (51, 88), blocking CD39 with specific reagents might provide a novel checkpoint to induce a reaction cascade of protective responses in infection or tumor immunotherapy.

In summary, human immune system is a tight-knit social network while CD39 has correlation to various immune cells. It is now becoming increasingly appreciated that CD39 is a promising therapeutic target. Increasing or inhibiting CD39 can interfere with the abnormal pathophysiological process of disorders, especially inflammation and tumor. We are hopeful that the extensive impact of CD39 on the operation of immune response will be profoundly illustrated and its enormous therapeutic potential for a broad spectrum of diseases will be better exploited.

Each author has participated sufficiently in the work to take public responsibility for appropriate portions of the content.

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