Faculty Bibliography

We have recently demonstrated that human neutrophils (PMN) possess two different classes of adenosine receptors (A1 and A2) that, when occupied, promote chemotaxis and inhibit the generation of reactive oxygen species (e.g., O2- and H2O2), respectively. We have previously demonstrated that adenosine protects endothelial cells (EC) from injury by stimulated neutrophils (PMN) both by diminishing generation of H2O2 and inhibiting adherence of PMN to EC. We therefore determined whether occupancy of A1 or A2 adenosine receptors regulated adherence of PMN to EC. At concentrations similar to those required to inhibit release of O2- by ligation of A2 receptors, both adenosine (IC50 = 56 nM) and 5'N-ethylcarboxamidoadenosine (NECA, IC50 = 8 nM), the most potent A2 agonist, inhibited adherence to EC by stimulated PMN (FMLP, 0.1 microM). In direct contrast, the specific A1 agonists N6-phenylisopropyladenosine and N6-cyclopentyladenosine (CPA) promoted PMN adherence to EC at concentrations of 1-100 nM. To further investigate the mechanisms by which adenosine receptor agonists affected the adherence of stimulated PMN we examined the effect of NECA (A2) and CPA (A1) on the adherence of PMN to fibrinogen (a ligand for the beta 2 integrin CD11b/CD18) and to gelatin. In a dose-dependent manner (IC50 = 2 nM), NECA inhibited the adherence of FMLP-treated PMN to fibrinogen- but not gelatin-coated plates. In contrast, CPA (A1) promoted adherence of stimulated PMN to gelatin-(EC50 = 13 pM) but not fibrinogen-coated plates. Theophylline (10 microM), an adenosine receptor antagonist, reversed the inhibition by NECA (0.3 microM) of stimulated neutrophil adherence to fibrinogen. These observations not only confirm the presence of A1 and A2 receptors on PMN but also suggest two opposing roles for adenosine in inflammation. Occupancy of A1 receptors promotes neutrophil adherence to endothelium and chemotaxis (a proinflammatory role) whereas occupancy of A2 receptors inhibits adherence and generation of toxic oxygen metabolites (an antiinflammatory role)

Corticosteroids are the preeminent antiinflammatory agents although the molecular mechanisms that impart their efficacy have not been defined. The endothelium plays a critical role in inflammation by directing circulating leukocytes into extravascular tissues by expressing adhesive molecules for leukocytes [e.g., endothelial-leukocyte adhesion molecule 1 (ELAM-1) and intercellular adhesion molecule 1 (ICAM-1)]. We therefore determined whether corticosteroids suppress inflammation by inhibiting endothelial expression of adhesion molecules for neutrophils (polymorphonuclear leukocytes). Preincubation of endothelial cells with endotoxin [lipopolysaccharide (LPS), 1 microgram/ml] led to a 4-fold increase in subsequent adherence of polymorphonuclear leukocytes (P < 0.0001, n = 10) to endothelial cells, an increase that was markedly attenuated when endothelial cells were treated with dexamethasone (IC50 < 1 nM, P < 0.0001, n = 6 or 7) during preincubation with LPS. Moreover, the steroid receptor agonist cortisol (10 microM), but not its inactive metabolite tetrahydrocortisol (10 microM), diminished LPS-induced endothelial cell adhesiveness. Further evidence that the action of dexamethasone was mediated through ligation of corticosteroid receptors [human glucocorticoid receptors (hGRs)] was provided by experiments utilizing the steroid antagonist RU-486. RU-486 (10 microM), which prevents translocation of ligated hGR to the nucleus by inhibiting dissociation of hGR from heat shock protein 90, completely aborted the effect of dexamethasone on adhesiveness of endothelial cells (P < 0.0005, n = 3). Treatment of endothelial cells with LPS (1 microgram/ml) stimulated transcription of ELAM-1, as shown by Northern blot analysis, and expression of membrane-associated ELAM-1 and ICAM-1, as shown by quantitative immunofluorescence (both P < 0.001, n = 9). Dexamethasone markedly inhibited LPS-stimulated accumulation of mRNA for ELAM-1 and expression of ELAM-1 and ICAM-1 (IC50 < 10 nM, both P < 0.001, n = 4-9); inhibition of expression by dexamethasone was reversed by RU-486 (both P < 0.005, n = 4-6). As in the adhesion studies, cortisol but not tetrahydrocortisol inhibited expression of ELAM-1 and ICAM-1 (both P < 0.005, n = 3 or 4). In contrast, sodium salicylate (1 mM) inhibited neither adhesion nor expression of these adhesion molecules. These studies suggest that antagonism by dexamethasone of endotoxin-induced inflammation is a specific instance of the general biological principle that the glucocorticoid receptor is a hormone-dependent regulator of transcription

Although commonly used to control a variety of inflammatory diseases, the mechanism of action of a low dose of methotrexate remains a mystery. Methotrexate accumulates intracellularly where it may interfere with purine metabolism. Therefore, we determined whether a 48-hr pretreatment with methotrexate affected adenosine release from [14C]adenine-labeled human fibroblasts and umbilical vein endothelial cells. Methotrexate significantly increased adenosine release by fibroblasts from 4 +/- 1% to 31 +/- 6% of total purine released (EC50, 1 nM) and by endothelial cells from 24 +/- 4% to 42 +/- 7%. Methotrexate-enhanced adenosine release from fibroblasts was further increased to 51 +/- 4% (EC50, 6 nM) and from endothelial cells was increased to 58 +/- 5% of total purine released by exposure to stimulated (fMet-Leu-Phe at 0.1 microM) neutrophils. The effect of methotrexate on adenosine release was not due to cytotoxicity since cells treated with maximal concentrations of methotrexate took up [14C]adenine and released 14C-labeled purine (a measure of cell injury) in a manner identical to control cells. Methotrexate treatment of fibroblasts dramatically inhibited adherence to fibroblasts by both unstimulated neutrophils (IC50, 9 nM) and stimulated neutrophils (IC50, 13 nM). Methotrexate treatment inhibited neutrophil adherence by enhancing adenosine release from fibroblasts since digestion of extracellular adenosine by added adenosine deaminase completely abrogated the effect of methotrexate on neutrophil adherence without, itself, affecting adherence. One hypothesis that explains the effect of methotrexate on adenosine release is that, by inhibition of 5-aminoimidazole-4-carboxamide ribonucleotide (AICAR) transformylase, methotrexate induces the accumulation of AICAR, the nucleoside precursor of which (5-aminoimidazole-4-carboxamide ribonucleoside referred to hereafter as acadesine) has previously been shown to cause adenosine release from ischemic cardiac tissue. We found that acadesine also promotes adenosine release from and inhibits neutrophil adherence to connective tissue cells. The observation that the antiinflammatory actions of methotrexate are due to the capacity of methotrexate to induce adenosine release may form the basis for the development of an additional class of antiinflammatory drugs

Patients with chronic renal failure who undergo hemodialysis experience accelerated atherosclerosis and premature death. Since the end-metabolite, oxalic acid, accumulates in plasma in proportion to the severity of renal failure, we studied whether sodium oxalate (0 to 300 microM) is an endothelial toxin and, therefore, might enhance atherogenesis. Exposure to uremic levels of oxalate (greater than 30 microM) for 9 to 28 days depressed endothelial cell replication by 33% to 84% (mean +/- SD, 54% +/- 15.7%, n = 17 experiments, p = 0.002). In contrast, replication of fibroblasts exposed to 200 microM oxalate for 45 days was not inhibited. The inhibitory effect of oxalate on endothelial cell replication was both dose- and time-dependent (both p less than 0.0001) and was first detected 3 to 7 days after the initial exposure to oxalate. Further, the inhibitory effect was fully reversible upon removal of oxalate, but only if exposure was limited to 5 days or less. Sodium salts of other carboxylic acids (citric, succinic, glyoxylic, and malonic; 200 microM) as well as HCl (200 microM) did not suppress endothelial cell replication. Oxalate also inhibited endothelial cell migration but had no effect on basal, thrombin-induced, or arachidonate-induced prostacyclin production by endothelial cells. Exposure of endothelial cells to sodium oxalate (200 microM) for as little as 24 hours-a time period sufficient to induce delayed, transient inhibition of replication not detectable until approximately 1 week after exposure-inhibited incorporation of 3H-leucine into protein by 40% (p = 0.009). We conclude that sodium oxalate acts as a uremic toxin, inhibiting endothelial cell replication and migration, functions which may be important for constitutive inhibition of atherosclerosis.