in addition to their roles in intracellular energy metabolism and nucleic acid synthesis, ATP and other nucleotides play important functions as extracellular signaling molecules (6, 15). Burnstock (3) first proposed that extracellular nucleotides can be used for signal transduction at nerve endings in diverse tissues. Consistent with this concept of “purinergic” neurotransmission were observations that neurons and neuroendocrine cells release ATP via classic mechanisms involving exocytotic release of nucleotides copackaged with biogenic amines or other neurotransmitters within specialized secretory vesicles or granules (19). Research during the past decade has resulted in the cloning and characterization of at least 14 distinct ATP/nucleotide receptors (8, 15), 4 adenosine receptors (15), and at least 9 different ectonucleotidases (21, 22). Virtually every mammalian cell type appears to express one or more subtype of nucleotide or nucleoside receptor along with one or more of the ectonucleotidases used for scavenging extracellular nucleotides. However, the vast majority of these cells lack direct physical proximity to neurons or neuroendocrine cells. Thus the identification of alternative sources of extracellular nucleotides and the elucidation of mechanisms underlying this nonsynaptic nucleotide release have become significant areas of current investigation. In the studies described in the current article in focus (Ref. 17; see p. C289 in this issue), Schwiebert et al. provide significant new insights regarding the role of vascular endothelial cells (EC) as both sources of extracellular ATP and autocrine targets of this released ATP.
Previous studies determined that multiple types of vascular EC, including human umbilical vein EC (2, 12), EC from rat aorta and caudal artery (18), porcine aortic EC (14), and guinea pig cardiac EC (20), can release nucleotides via noncytolytic but otherwise undefined mechanisms. Two types of stimuli were shown to trigger ATP release from these various EC types: mechanical stress during changes in flow and activation of Ca2+-mobilizing receptors for different vasoactive hormones/neurotransmitters. These nucleotide release agonists included bradykinin, acetylcholine, thrombin, norepinephrine, and ATP itself. Schwiebert et al. (17) have now extended these earlier findings in several ways.
First, these authors utilized on-line, luciferase-based luminometry to directly record ATP release from human EC maintained as polarized and confluent monolayers on permeable filter supports. This particular methodological approach was useful in two ways. Previous studies of ATP release from EC utilized “off-line” methods that involved removal of bathing medium at various times following stimulation of the EC with agonists or mechanical stress, followed by enzymatic or HPLC-based analyses of nucleotide content within these extracellular samples (2, 12, 14, 18, 20). Although sensitive, such off-line assays suffer from a limited capacity for temporal resolution of transient ATP release events from EC. This can be a critical limitation because ATP released from EC will be rapidly catabolized to AMP and adenosine by the combined actions of CD39-family ectoapyrases and the CD73 ecto-5′-nucleotidase. By including luciferase (an enzyme with exquisite sensitivity to, and selectivity for, ATP) and luciferin in the extracellular medium bathing their EC monolayers, Schwiebert et al. (17) were able to record minute-to-minute variations in ATP content at the EC extracellular surface. Moreover, by selectively including luciferase/luciferin in the extracellular medium bathing either the apical or basolateral EC surfaces, they were able to document a marked polarization of basal and mechanically stimulated ATP release to the apical surface. This would suggest a similar polarized localization of the ATP release mechanism(s), be it exocytosis of ATP-laden vesicles or facilitated efflux of cytosolic ATP.
Another significant finding was that human EC released ATP at a significant rate even in the absence of obvious mechanical stimulation or addition of Ca2+-mobilizing agents. By directly measuring basal ATP release at the surface of unstimulated EC with high sensitivity and temporal resolution, Schwiebert et al. (17) provided a third experimental approach for complementing similar findings from other investigators working with different cell models. Previous data from two other experimental protocols have provided convincing support for this notion of constitutive ATP release. Several groups observed that treatment of nominally unstimulated cells with extracellular nucleotide scavenger enzymes, such as potato apyrase or hexokinase, can reduce the cytosolic levels of second messengers, such as cAMP and inositol trisphosphate (10, 13, 16). Ostrom et al. (13) provided particularly salient evidence that this reduction in second messenger levels by nucleotide scavengers indicated that endogenous ATP (and/or UTP) is released from cells in amounts sufficient to produce low-level activation of various G protein-coupled P2Y receptors. Schwiebert et al. similarly found that treatment of EC with extracellular nucleotide scavengers induced a rapid and reversible decrease in the basal concentration of cytosolic [Ca2+]. A second type of experimental support for constitutive ATP release was provided by Lazarowski et al. (9), who employed conventional isotopic tracer methods to measure the steady-state rates of extracellular nucleotide metabolism by four different cell lines. Those authors observed that the four cell types steadily maintained extracellular ATP at the 1–10 nM range for many hours under standard, serum-free tissue culture conditions. The cells were then pulsed with extracellular [γ-32P]ATP at tracer levels that did not significantly change the extracellular ATP concentration. Significantly, the [γ-32P]ATP tracer was rapidly and completely metabolized. This finding demonstrated that the steady-state level of extracellular ATP in the cell cultures reflected a constitutive rate of ATP release that was balanced by ATP hydrolysis measured at 20–200 fmol · min−1 · cell−1. Significantly, the basal rate of ATP release from EC estimated by the direct luciferase-based assay of Schwiebert et al. was similar in magnitude.
Schwiebert et al. (17) performed additional ATP release studies at various temperatures and in the presence of various pharmacological reagents. This approach provided suggestive evidence for involvement of an exocytotic mechanism in the ATP release triggered by Ca2+-mobilizing agents and mechanical stress but not in the constitutive ATP release process. These findings indicated that distinctive cellular mechanisms may underlie constitutive ATP release vs. stress-stimulated release. Although EC contain exocytotic granules in the form of the Weibel-Palade bodies, no studies have evaluated whether these granules compartmentalize and release nucleotides (4). However, even vesicles involved in the constitutive, Ca2+-independent release of secreted proteins are likely to contain low levels of nucleotides due to important roles for intravesicular, nucleotide-dependent enzymes that catalyze the covalent modification and maturation of secreted proteins (7). The likely presence of nucleotides within vesicles that comprise the constitutive exocytotic machinery may have important implications for understanding mechanisms for basal ATP release and autocrine activation of P2 receptors in different cell types. Indeed, the recent studies of Maroto and Hamill (11) on ATP externalization from singleXenopus oocytes suggested that brefeldin A-sensitive exocytotic pathways were involved in both basal and mechanically stimulated nucleotide release in that cell type. It will be interesting to adapt the methods of Schwiebert et al. to determine whether similar brefeldin A-sensitive pathways might play similar roles in either the basal or stimulated ATP release observed in EC monolayers.
Thus the use of on-line luciferase-based assays can clearly illuminate (pun intended) a variety of possible pathways for ATP release from intact cells. It should be stressed that use of this assay system for measuring ATP release can be compromised by the spurious effects of pharmacological reagents on the intrinsic ATP sensitivity or catalytic rate of the luciferase present within the extracellular medium. As noted by both Schwiebert et al. (17) and other investigators (1, 11), a variety of P2 receptor antagonists, ectonucleotidase inhibitors, and modulators of various transporters/channels can affect luciferase activity. Thus changes in luminescent intensity upon addition of such reagents can reflect decreased luciferase activity at a constant level of extracellular ATP. Care must be taken to discriminate such changes in luciferase activity from bona fide changes in extracellular ATP concentration. This is a significant issue because there is disagreement among different laboratories regarding the effects of certain ATP release blockers, such as Gd3+, on luciferase activity. Whereas Schwiebert et al. have observed no significant inhibitory effects of Gd3+on luciferase activity, others (11) have reported that Gd3+ can significantly reduce the ATP sensitivity of luciferase. Whether this reflects differences in particular assay conditions is an important methodological concern that requires clarification. The same caveat applies to the known effects of ionic milieu on intrinsic luciferase activity (5). Schwiebert et al. stressed that the high Cl− concentration in normal extracellular saline results in reduced catalytic activity of luciferase. However, the authors did not discuss how the acute reduction in extracellular Cl− concentration, which necessarily accompanies their hypotonic stress stimulus, should result in enhanced luciferase activity and light output even in the presence of unchanged ATP concentration. Thus the rapid increase in luminescence triggered by hypotonic stimulation of the EC may reflect on the combined effects of mechanically induced ATP release and increased sensitivity of the extracellular luciferase to ambient ATP. Establishing protocols for deconvoluting such luciferase-based signals would be a further improvement to such on-line assays of ATP release and extracellular metabolism. As shown by these informative studies of Schwiebert et al. on primary endothelial cells, the excellent temporal resolution and high sensitivity of this method should provide a powerful approach for testing multiple models of nucleotide efflux or secretion from diverse eukaryotic cells.
Address for reprint requests and other correspondence: G. R. Dubyak, Dept. of Physiology and Biophysics, Case Western Reserve Univ., Cleveland, OH 44106 (E-mail:).
- Copyright © 2002 the American Physiological Society