Adenosine and related compounds counteract tumor necrosis factor-alpha inhibition of neutrophil migration: implication of a novel cyclic AMP-independent action on the cell surface.

Human rTNF-alpha (greater than or equal to U/ml) decreased PMN nondirected and directed migration to FMLP to approximately 50% of control. Adenosine (100 microM) almost completely restored hrTNF-inhibited migration (nondirected from 54 to 92% and directed migration to from 54 to 93% of control). The lowest concentration of adenosine that restored hrTNF-inhibited migration was 3 microM, and the adenosine analogue, 5'-(N-cyclopropyl)-carboxamido-adenosine (CPCA) was more potent than adenosine. Although CPCA binds to A2-receptors and stimulates adenylate cyclase, the reversal of hrTNF-inhibited chemotaxis was found to be independent of both PMN cAMP content and binding to A2-receptors, because neither 8-Br-cAMP nor pertussis adenylate cyclase restored hrTNF-inhibited PMN chemotaxis and the A2-receptor antagonist, 1,3-dipropyl-7-methylxanthine decreased CPCA stimulated cAMP but enhanced CPCA-restoration of hrTNF-inhibited chemotaxis. The effect of adenosine could be augmented by inhibition of adenosine uptake and decreased by adenosine deamination. Pentoxifylline, (3,7 dimethyl-1-[5 oxo-hexyl] xanthine), like adenosine also restored PMN chemotaxis inhibited by hrTNF. The adenosine receptor antagonist, 1,3-dipropyl-8(phenyl-p-acrylate)-xanthine (BW A1433U), decreased restoration of hrTNF-inhibited chemotaxis by CPCA or pentoxifylline. Thus, the inhibitory effect of hrTNF on PMN migration can be counteracted by adenosine, CPCA, pentoxifylline, and compounds that increase adenosine availability to the surface of the PMN. Inasmuch as an A1-selective agonist N6-cyclopentyladenosine was less active, and the action of the A2-selective agonist CPCA was enhanced by an A2-receptor antagonist, we hypothesize that neither A1 or A2 receptors are involved in adenosine restoration of hrTNF-inhibited chemotaxis. Further, increased cAMP, an A2-regulated event, does not cause the effect, and adenosine restoration of hrTNF-inhibited migration does not appear to be mediated by changes in PMN [F-actin], FMLP receptor expression, or cytosolic calcium. Hence, the restoration of hrTNF-inhibited chemotaxis is controlled by a novel cyclic AMP-independent action on the PMN surface.