Osmotic swelling-induced ATP release: a new role for tyrosine and Rho-kinases?

Abstract

Nucleotides such as adenosine and uridine di- and triphosphates are important extracellular signalling molecules that participate in intercellular communication and regulate a broad range of physiological responses including vascular tone, muscle contraction, cell proliferation, mucociliary clearance, platelet aggregation and neurotransmission. Once released, they interact with G-protein coupled purinergic receptors (P2Y1, 2, 4, 6, 11) and ligand-gated ion channel receptors (P2X1-7), all of which can trigger rapid changes in cytosolic [Ca2+]. Exocytotic release of ATP occurs at synapses and from activated blood platelets, chromaffin cells and mast cells. Mechanical perturbations such as shear stress, membrane stretch, media change or hypo-osmotic cell swelling induce significant ATP release from endothelial and epithelial cells, from smooth muscle, fibroblasts, erythrocytes and hepatocytes through mechanisms that remain poorly understood. Except for certain pathological conditions that lead to irreversible membrane damage, increasing evidence suggests that mechanically induced release of ATP is a regulated process that does not involve cellular lysis. Stretch-induced ATP release by urinary bladder epithelium was proposed as a mechanism for sensing when the bladder is full (Ferguson et al. 1997). ATP release from airway epithelial cells is exquisitely sensitive to mechanical perturbations and can be evoked by gentle mixing of the bath solution (Grygorczyk & Hanrahan, 1997). Such mechanically induced release could have a host-defence role in vivo, since luminal ATP could trigger fluid secretion that helps clear the epithelial surface of particulates. While mechanically induced release of ATP is an intriguing phenomenon, the intracellular pool of ATP used, efflux mechanisms and regulatory pathways remain obscure. The report by Koyama et al. in this issue of The Journal of Physiology provides the first evidence that tyrosine kinase and Rho-kinase signalling pathways are involved in hypotonic stress-induced ATP release (Fig. 1). Osmotic swelling of bovine aortic endothelial cells stimulated release of ATP, which, by autocrine/ paracrine action on purinergic receptors, induced oscillations of intracellular Ca2+. Both effects were prevented by the tyrosine kinase inhibitors herbimycin A and tyrphostin 46, although inhibition of ATP release was only partial. This may indicate that additional tyrosine kinase-independent mechanisms are also involved. Indeed, osmotic cell swelling and mechanical load are known to activate multiple signalling cascades including tyrosine and MAP kinase, phosphatidylinositol 3-kinase (PI 3-kinase) and Rho, a monomeric GTPase involved in organising the actin cytoskeleton, endo/exocytosis and in forming focal adhesions. Koyama et al. (2001) found that inhibiting Rho directly with Clostridium botulinum C3 exoenzyme, or downstream at the level of Rho-kinase with Y-27632, diminished osmotic stress-induced ATP release and Ca2+ oscillations. Interestingly, the PI 3-kinase inhibitor wortmannin did not suppress hypotonic stress-induced, ATP-mediated oscillations of intracellular Ca2+. Although data showing a direct effect of wortmannin on ATP release are not presented, this result contrasts with earlier studies on liver cells that indicated volume-sensitive ATP release requires activation of PI 3-kinase (Ferenchak et al. 1998). This difference suggests the existence of distinct, cell-specific pathways for regulating ATP release. Another important phenomenon noted by Koyama et al. (2001) is the basal release of ATP by resting, unperturbed endothelial cells, which confirms similar observations with other cell types (Lazarowski et al. 2000). Resting levels of extracellular nucleotide tonically activate purinergic receptors and establish the set point for signal transduction pathways (Ostrom et al. 2000). Figure 1 Schematic model of signalling pathways implicated in swelling-induced ATP release The present findings provide no clues as to the pathway of ATP release, and several may operate in the same cell. For example, basal ATP release could result from exocytosis during constitutive membrane recycling. Stimulated release could, in addition, involve ATP transporter(s) analogous to those of the inner mitochondrial membrane, or perhaps ATP-permeable channels analogous to mitochondrial porin VDAC. Indeed, ATP channels have been implicated in cellular ATP release, but remain to be identified (Braunstein et al. 2001). The tyrosine kinase and Rho-kinase pathways studied by Koyama et al. (2001) are also known to regulate volume-sensitive Cl− channels (Nilius et al. 1999). Cl− channels are unlikely to provide the pathway for swelling-induced ATP release, however, since Hazama et al. (1999) clearly showed that, at least in intestinal cells, the ATP release pathway was distinct from volume-sensitive Cl− channels. This work by Koyama et al. (2001) greatly strengthens the notion that mechanically induced ATP efflux is an important, cell-regulated process. Identifying the ATP efflux pathway at the molecular level should provide mechanistic insight into how these signalling pathways control ATP release.