New insights into the regulation of inflammation by adenosine

Abstract

Adenosine, long known as a regulator of cardiovascular function, has recently been identified as a significant paracrine inhibitor of inflammation that acts primarily by activation of A_2A adenosine receptors (A_2AARs) on lymphoid or myeloid cells. In this issue of the JCI, Yang et al. describe a proinflammatory phenotype resulting from deletion of the gene encoding the A_2B adenosine receptor (A_2BAR) in the mouse, suggesting that activation of the A_2BAR can also have antiinflammatory effects (see the related article beginning on page 1913). Nevertheless, the role of the A_2BAR remains enigmatic since its activation can either stimulate or inhibit the release of proinflammatory cytokines in different cells and tissues.

There is growing interest in elucidating the mechanisms by which adenosine inhibits the immune system, since these inhibitory adenosine receptors and their downstream signaling pathways are promising targets for new antiinflammatory therapies. By signaling through the A_2A adenosine receptor (A_2AAR), adenosine suppresses the immune system, primarily by inhibiting lymphoid or myeloid cells including neutrophils (1), macrophages (2), lymphocytes (3, 4), and platelets (5). These responses are amplified by rapid induction of A_2AAR mRNA in macrophages and T lymphocytes in response to inflammatory or ischemic stimuli (2, 3, 6, 7). The A_2B adenosine receptor (A_2BAR) also appears to mediate antiinflammatory effects in macrophages by inhibiting the production of TNF-α and IL-1β, stimulating IL-10 and inhibiting macrophage proliferation (8–11) (Figure 1A). Macrophage A_2BAR signaling increases during inflammation, as the macrophage-activating cytokine IFN-γ causes induction of macrophage A_2BAR mRNA (12). However, the A_2BAR is somewhat unusual in that it is dually coupled to the generally antiinflammatory G protein G_s and the generally proinflammatory G protein G_q (13); in addition, numerous proinflammatory effects mediated by activation of the A_2BAR have also been described (Table 1).

Figure 1. Anti- and proinflammatory effects mediated by activation of the A_2B AR.

(A) In macrophages, the release of TNF-α and other proinflammatory cytokines is inhibited as a result of activating either the A_2AAR or the A_2BAR. These effects are amplified if TNF-α production is stimulated by LPS, which signals through TLR4 and other Toll-like receptors. Activation of the A_2BAR also stimulates production of the antiinflammatory cytokine IL-10. (B) In endothelial cells and other cells noted in Table 1, activation of the A_2BAR stimulates IL-6 and other proinflammatory cytokines. Hypoxia increases the intracellular production of adenosine, which is transported outside the cell by nucleoside transport proteins. Hypoxia and possibly activation of the A_2BAR stimulate the release of ATP and the production of the A_2BAR and ectoenzymes (CD39 and CD73) that convert ATP to adenosine. Vasodilation in response to A_2BAR activation may increase shear stress to stimulate ATP release.

Table 1 .

Proinflammatory effects of activating A_2BR

Inflamed phenotype of the A_2B AR knockout mouse

Given the opposing cellular effects mediated by A _2BAR activation, the phenotype that would result from deletion of the gene encoding A _2BAR was uncertain. In this issue of the JCI, Yang and coworkers use a receptor knockout/reporter gene knock-in approach to confirm the expected expression pattern of the A _2BAR transcript, noting high levels in monocytes/macrophages and the vasculature (14). Interestingly, they report a moderately inflamed phenotype in the A _2BAR knockout mouse at baseline, including elevated adhesion molecule expression on vascular endothelial cells and elevated cytokine production. The inflammatory phenotype is exaggerated in mice given endotoxin. The baseline response to A _2BAR knockout is surprising since the A _2BAR is recognized to have a low affinity for adenosine, yet the data imply that baseline adenosine levels are high enough to produce some activation of the A _2BAR. A well-recognized limitation of the global knockout approach is that the resulting phenotype may reflect a developmental adaptation to the gene knockout and not an acute direct consequence of deleting the gene product. In order to address this possibility, Yang et al. examined the effect of deleting the gene that codes for A _2BAR on mRNA levels of the other adenosine receptor subtypes and found no changes. However, other possible adaptations have yet to be examined. It has recently been demonstrated that hypoxia triggers coordinate upregulation on endothelial cells of mRNA for the A _2BAR and ectoenzymes involved in the conversion of extracellular adenine nucleotides to adenosine (15) (Figure 1B). In addition, the release of adenine nucleotides from various cells, including neutrophils and endothelial cells, is a regulated process (16). It will be of interest in future studies to determine whether deletion of the gene encoding A _2BAR influences either nucleotide release or extracellular nucleotide metabolism. Such effects could indirectly influence inflammation by altering the availability of adenosine to other adenosine receptor subtypes, particularly the A _2AAR. Deletion of either the ecto-apyrase CD39 (also known as NTPDase1; ref. 17) or the ecto–5′-nucleotidase (also known as CD73; refs. 18, 19) results in a proinflammatory phenotype.

Future questions

It will also be of interest in future studies to examine the response of the A _2BAR knockout mouse in the setting of hypoxia or ischemia, which elicit the accumulation of large levels of adenosine. A selective A _2BAR antagonist blocks myocardial preconditioning when applied to the isolated rabbit heart after ischemia (20). This could occur either because A _2BARs on the heart mediate cardioprotection or, as discussed above, because A _2BAR activation facilitates the release and/or metabolism of adenine nucleotides to indirectly enhance adenosine production. The latter scheme is consistent with the observation that myocardial preconditioning has a remote adenosine-mediated effect on platelet function (21).

In conclusion, the study by Yang et al. (14) bolsters the conclusion that activation of A_2BARs on certain cells, particularly macrophages, inhibits inflammation. A proinflammatory phenotype noted at rest is somewhat unexpected, and the mechanism underlying this inflammation is not yet known. In view of a number of previous reports indicating that A_2BAR activation can be proinflammatory (Table 1), it will be of interest to use the newly available A_2BAR knockout mouse in order to determine whether A_2BAR activation on different cells can elicit both pro- and antiinflammatory responses.