Disruption of adherens junctions between endothelial cells results in compromised endothelial barrier function and in altered angiogenesis. Nitric oxide (NO) produced by endothelial NO synthase (eNOS) is essential for increased vascular permeability induced by vascular endothelial growth factor (VEGF). However, the molecular mechanisms by which NO modulates endothelial permeability remain elusive. Here, we show that, within adherens junctions, beta-catenin is a substrate for S-nitrosylation by NO. Stimulation of endothelial cells with VEGF induces S-nitrosylation of beta-catenin, which is dependent on expression and activity of eNOS. Furthermore, VEGF-induced S-nitrosylation of beta-catenin is inhibited in eNOS(-/-) mice. We identify Cys619, located within the VE-cadherin interaction site, as the major S-nitrosylation locus in response to VEGF. Inhibition of S-nitrosylation at Cys619 prevents NO-dependent dissociation of beta-catenin from VE-cadherin and disassembly of adherens junction complexes and inhibits VEGF-stimulated endothelial permeability. Thus, we identify S-nitrosylation of beta-catenin as a modulator of intercellular contacts between endothelial cells.

Abstract-Angiotensin type 2 receptor-mediated effects of angiotensin II appear to counteract many of the effects mediated via the angiotensin type 1 receptor. Compound 21 (C21), a selective angiotensin type 2 receptor agonist, has demonstrated beneficial effects on cardiac function after myocardial infarction in rodents. We hypothesized that C21 alone or in combination with an angiotensin type 1 receptor antagonist would blunt the development of hypertension and vascular damage in stroke-prone spontaneously hypertensive rats. Six-week-old stroke-prone spontaneously hypertensive rats received C21 (1 mg/kg per day), the angiotensin type 1 receptor antagonist losartan (10 mg/kg per day), C21 plus losartan, or vehicle PO for 6 weeks. Systolic blood pressure was lower in losartan and C21-losartan combination groups (PϽ0.001). Endotheliumdependent relaxation was enhanced (PϽ0.001) in the C21-losartan combination group at lower acetylcholine concentrations. C21 or C21-losartan combination reduced vascular stiffness, aortic medial and myocardial interstitial collagen content, and aortic fibronectin (PϽ0.05). C21 and losartan decreased the expression of 2 genes associated with cardiac hypertrophy, myosin heavy chain-␤ (myh7) by 30 to 50%, and ␣-skeletal muscle actin by 30% to 35% (PϽ0.05). C21-losartan combination caused an additional 40% reduction in myh7 compared with C21 (PϽ0.01). Aortic superoxide generation was reduced equally by the 3 treatments (PϽ0.001). Monocyte/macrophage infiltration in the aorta and kidney (PϽ0.001) and T-lymphocyte infiltration in the renal cortex (PϽ0.05) were lowered similarly by the 3 treatments. These data suggest that C21 alone or in combination with losartan may improve endothelial function and vascular composition and mechanics by reducing oxidative stress, collagen content, fibronectin, and inflammatory cell infiltration in stroke-prone spontaneously hypertensive rats. hronic hypertension results in vascular remodeling and inflammation, endothelial dysfunction, and accelerated development of atherosclerosis. 1,2 Of the many factors implicated in hypertensive vascular remodeling, angiotensin (Ang) II, the main effector hormone of the renin-Ang-aldosterone system, seems to be one of the most important. 3 Ang II exerts its effects by binding to 2 membrane-bound, G proteincoupled receptors, Ang II type 1 (AT 1 R) and type 2 receptors (AT 2 R). Most well-known effects of Ang II that contribute to unfavorable vascular remodeling and consequent hypertensive complications are mediated via AT 1 R, whereas those exerted via AT 2 R are less well known and appear to counteract many of its effects exerted via AT 1 R. 4,5 AT 2 Rmediated effects seem to exert vasculoprotective actions via vasodilation, NO production, 6 -8 and apoptosis and inhibition of growth and fibrosis. 5,9,10 Beneficial vascular effects of AT 1 R blockade have been at least partially attributed to unopposed AT 2 R stimulation. 7,11,12

1 The relaxant e ects of isoprenaline may result from activation of another b-adrenoceptor subtype in addition to b 1 and b 2 . This study evaluated the role of a third b-adrenoceptor subtype, b 3 , in b-adrenoceptor-induced relaxation of rat thoracic aorta by isoprenaline. 2 Isoprenaline produced a concentration-dependent relaxation of phenylephrine pre-contracted rings of the thoracic aorta (pD 2 =7.46+0.15; E max =85.9+3.4%), which was partially attenuated by endothelium removal (E max =66.5+6.3%) and administration of the nitric oxide (NO) synthase inhibitor, L-N G -monomethyl arginine (L-NMMA) (E max =61.3+7.9%). 3 In the presence of nadolol, a b 1 -and b 2 -adrenoceptor antagonist, isoprenaline-induced relaxation persisted (E max =55.6+5.3%), but occurred at higher concentrations (pD 2 =6.71+0.10) than in the absence of nadolol and lasted longer. 4 Similar relaxant e ects were obtained with two b 3 -adrenoceptor agonists: SR 58611 (a preferential b 3 -adrenoceptor agonist), and CGP 12177 (a partial b 3 -adrenoceptor with b 1 -and b 2 -adrenoceptor antagonistic properties). SR 58611 caused concentration-dependent relaxation (pD 2 =5.24+0.07; E max =59.5+3.7%), which was not modi®ed by pre-treatment with nadolol but antagonized by SR 59230A, a b 3 -adrenoceptor antagonist. The relaxation induced by SR 58611 was associated with a 1.7 fold increase in tissue cyclic GMP content. 5 Both relaxation and the cyclic GMP increase induced by SR 58611 were greatly reduced by endothelium removal and in the presence of L-NMMA. 6 We conclude that in the rat thoracic aorta, b 3 -adrenoceptors are mainly located on endothelial cells, and act in conjuction with b 1 -and b 2 -adrenoceptors to mediate relaxation through activation of an NO synthase pathway and subsequent increase in cyclic GMP levels.