Requirement of NK cells for selective A_{2A r}eceptor blockade to suppress CD73⁺ tumor metastasis

Abstract

CD73 is becoming an emerging therapeutic target for the prevention of tumor growth and metastasis. However, the mechanism by which CD73 promotes tumor metastasis is unclear. Beavis et al. evaluated the efficacy of A_2A and A_2B adenosine receptor antagonists in inhibiting the metastasis of tumors expressing CD73, either endogenously or ectopically. They demonstrate distinct mechanisms whereby A_2A versus A_2B adenosine receptors could contribute to CD73⁺ tumor metastasis. As A_2Areceptor (R)/A_2BR antagonists have been tested in clinical trials in other disease settings, this study highlights the potential therapeutic application of an A_2AR/A_2BR blockade strategy for treatment of CD73⁺ metastatic tumors.

Tumor-induced immunosuppression is one of the main causes of tumor escape and failure of immunotherapy. Thus, understanding immunosuppressive mechanisms is critical for optimization of cancer immunotherapeutic benefit. CD73-mediated adenosinergic effects can now be viewed as among the most important immunosuppressive regulatory pathways in the tumor microenvironment [1,2]. CD73 is a cell surface ectoenzyme (ecto-5′-nucleotidase) that catalyzes the dephosphorylation of extracellular AMP into adenosine [3,4]. Importantly, this ectoenzymatic cascade, in tandem with CD39 (ecto-ATPase), generates adenosine from ATP that in turn signals through four known cell surface G protein-coupled receptors (A₁, A_2A, A_2B, A₃), among which the A_2A receptor (R) and A_2BR particularly exert potent anti-inflammatory activity [5]. They share a common signaling pathway, both leading to the activation of G proteins and the accumulation of intracellular cyclic AMP. Extensive studies demonstrate that A_2AR/A_2BR activation contributes to tumor growth and metastasis by the immunosuppressive effects of adenosine [6,7].

Upregulation of CD73 expression has been observed in a broad spectrum of solid cancers and leukemias [1]. Moreover, a number of studies suggest that CD73 expression is associated with more aggressive, metastatic behaviors, and could serve as a diagnostic/prognostic marker and/or therapeutic target in certain cancer types [8,9]. In support, the feasibility of potent strategies to harness antitumor immune responses by targeting the important CD39/CD73-adenosine receptor axis in tumors has been recently documented [1,2,10,11]. Several preclinical studies demonstrate a pivotal role of tumor and host CD73-mediated adenosinergic effects on tumor growth and metastasis using an anti-CD73-blocking monoclonal antibody and a small-molecular inhibitor of CD73 activity [12–16]. The tumor-inhibiting advantage of CD73 blockade was due to improved antitumor T-cell immune responses, likely as a result of reduced A_2AR-mediated immunosuppression [6]. However, there is also considerable evidence to support the idea that CD73 promotes tumor growth and metastasis through multiple mechanisms [1,2] independently of antitumor T-cell regulation, that need further investigation.

Summary of methods & results

In their recent paper, Beavis and colleagues investigated the mechanism by which CD73 promotes metastasis. They transduced CD73⁻ murine tumor cells with retrovirus-encoding murine CD73 cDNA to generate CD73⁺ tumor cell lines, and then compared the metastatic capacity of the CD73⁻ versus CD73⁺ cells. They found that CD73⁻ and CD73⁺ tumor cells grew at equal rates in vivo. By contrast, ectopic CD73 expression promoted lung metastasis when tumor cells were injected either subcutaneously (spontaneous metastasis), or intravenously (experimental metastasis).

To investigate whether the prometastatic effect of tumor CD73 is dependent on adenosine receptor activation, a pan adenosine receptor agonist (adenosine analog), 5′-(N-ethylcarboxamido) adenosine (NECA), was used. In this setting, wild-type mice were treated with NECA before receiving intravenously injected CD73⁻ tumor cells. NECA pretreatment significantly increased pulmonary metastasis. To further test which specific adenosine receptor was involved in this process, A_2AR or A_2BR antagonists were given before NECA. The prometastatic effect of NECA was partially reversed by either A_2AR or A_2BR blockade, suggesting that both A_2AR and A_2BR contributed to the prometastatic effect of adenosine signaling. This conclusion was further confirmed by the direct use of A_2AR/A_2BR agonists that increased metastasis. The authors investigated whether A_2AR/A_2BR activation also affected metastasis in immunocompromised RAG^−/−cγ^−/− mice (lacking T, B and NK cells). They found that NECA pretreatment in RAG^−/−cγ^−/− mice did increase metastasis, albeit to a significantly lesser extent than in wild-type mice. The increased metastasis could be partially reversed by A_2BR blockade, but not by A_2AR blockade. The results suggested that adenosine receptor activation promoted metastasis in both immune-dependent and immune-independent manners. It appears that the antimetastatic effect of A_2AR blockade was dependent on immune cells that are deficient in RAG^−/−cγ^−/− mice, while blocking A_2B receptors reduced metastasis through a distinct mechanism.

Notably, CD73 expression on tumor cells was required for the metastasis-promoting effects of adenosine receptor activation because blockade of either A_2AR or A_2BR did not inhibit the metastasis of CD73⁻ tumors. In addition, A_2AR blockade failed to affect metastasis of CD73⁺ tumors in A_2AR^−/− mice, indicating that A_2AR activation on host immune cells, rather than tumor cells, contributed to CD73-mediated metastasis.

To investigate which specific host immune cells were affected by A_2AR activation, the authors focused on the NK cell subset, which is deficient in RAG^−/−cγ^−/− mice. As expected, adding NECA significantly inhibited cytokine production, activation and cytolytic activity of NK cells in vitro. This effect was reversed by A_2AR antagonists, indicating a role for A_2AR activation in suppressing NK cell functions. These results could be also recapitulated in vivo because the activity of NK cells from A_2AR^−/− mice was not modulated by NECA. Moreover, A_2AR blockade failed to reduce both spontaneous and experimental metastasis of CD73⁺ tumors from perforindeficient mice, in which the cytolytic function of NK cells is impaired. Interestingly, further new evidence showed that the proportion of tumorinfiltrating NK cells expressing granzyme B was significantly increased by A_2AR blockade, but not A_2BR blockade. Thus, the authors concluded that A_2AR, rather than A_2BR, activation promoted CD73⁺ tumor metastasis through inhibition of perforin-dependent NK cytotoxicity.

While exploring which immune subsets participate in the A_2BR-mediated prometastatic effect, the authors found that the frequencies of myeloidderived suppressor cells, macrophages and dendritic cells in metastatic lungs were unchanged following A_2BR blockade, suggesting that A_2BR stimulation may enhance metastasis through a mechanism independent of these immune cells. Indeed, A_2BR activation on tumor cells was reported to reduce CD73⁺ tumor metastasis and invasiveness [14,17]. Together, these data suggested CD73 promotes metastasis by A^2AR and A^2BR activation, invoking both immune-dependent and immune-independent mechanisms.

Discussion & significance

CD73 expression on tumor cells has been shown to promote tumor growth and metastasis. The mechanism of CD73-mediated tumor growth has been well studied, and an inhibition of antitumor T-cell responses by CD73 has been revealed [12–16]. However, the underlying mechanism by which CD73 influences tumor metasstasis remains largely unknown. The current study demonstrates that tumor CD73 enhances metastasis through both immune-dependent and immune-independent mechanisms. The immune-dependent mechanism relies on A_2AR-mediated suppression of perforin-dependent NK cell cytotoxicity. Conversely, the contribution of the A_2BR to CD73-promoted tumor metatastasis likely occurs through its activation on tumor cells. Although the majority of current CD73-focused cancer therapy studies have been concentrated on CD73-mediated adenosinergic effects, it appears there may also be an adenosine- independent role of CD73. This derives from a more recent study [18], which showed that an antihuman CD73 monoclonal antibody inhibited development of metastasis in a spontaneous animal model of human metastatic breast cancer by causing the clustering and internalization of CD73. This appears to limit the ability of circulating tumor cells to extravasate and colonize, suggesting a nonenzymatic role for CD73 in cell adhesion. Therefore, CD73 promotes tumor growth and metastasis in a multifactorial manner, high-lighting that CD73 is a promising novel target for effective cancer therapy. As there are several adenosine receptor antagonists already in clinical trials for other disease settings, a more rapid transition in anti-CD73 cancer therapy from bench to bedside is warranted.

Future perspective

The most significant result reported here is the demonstration that the metastasis of CD73⁺ tumors can be suppressed by restoring NK cell activity through A_2AR blockade. In contrast to A_2AR blockade, Beavis et al. also showed that A_2BR blockade alone was sufficient to inhibit metastasis in a NK cell-independent manner. Although A_2BR blockade did not change the percentage of myeloid-derived suppressor cells, macrophage and dendritic cells in metastatic tissue, we cannot formally exclude the possibility that A_2BR antagonists reduced CD73⁺ tumor metastasis through these immune cell activities. Indeed, A^2BR blockade has previously been shown to modulate the differentiation and function of these myeloid cells [19,20]. Further investigations with A_2BR-deficient mice are certainly needed. Moreover, it would be interesting to dissect the contribution of tumor A_2BR versus host A_2BR to CD73-mediated tumor metastasis. Finally, exploring the possible correlative analysis of A_2AR/A_2BR expression between the phenotypic characterization of varied immune infiltrates and clinical outcome in a large cohort of different cancer specimens would be very informative to validate the clinical implication for the A_2AR/A_2BR antagonists.

Executive summary.

• CD73 expression on tumor cells is required for A_2Areceptor (R)/A_2BR-mediated tumor metastasis.

• A_2AR blockade inhibits CD73⁺ tumor metastasis by reversing NK cell functions.

• A_2BR blockade inhibits CD73⁺ tumor metastasis by a distinct mechanism which may be independent of immune regulation.

• Targeting the CD73 A_2AR/A_2BR axis represents a new way to effectively treat cancer metastasis.