Replacement of connexin 43 by connexin 32 in a knock-in mice model attenuates aortic endothelium-derived hyperpolarizing factor-mediated relaxation.

Previous studies demonstrated that intercellular communication through endothelial, smooth muscle or myoendothelial connexin channels contributes to the control of vascular tone. At least four connexin types are present in the arterial wall. The aim of the present work was to assess the role played by connexin 43 (Cx43)-formed gap junctions on vessel function. Aortic reactivity to noradrenaline, acetylcholine and sodium nitroprusside, and endothelium-derived hyperpolarizing factor (EDHF)-mediated relaxations, were analysed in a Cx43KI32 mouse model in which the coding region of Cx43 was replaced by that of connexin 32 (Cx32). Aortic rings were placed in organ baths containing a Krebs solution oxygenated at 37 degrees C (pH 7.4). Confocal images of aortic rings confirmed connexin substitution in mutant mice. In control conditions, replacement of Cx43 by Cx32 in homozygous mutant mice did not modify endothelium-independent contractile responses to noradrenaline, or relaxations in response to sodium nitroprusside (endothelium independent) or acetylcholine (endothelium dependent). However, residual endothelium-dependent relaxations in response to acetylcholine after nitric oxide synthase and cyclooxygenase inhibition (EDHF type) were significantly reduced in homozygous Cx43KI32 mice (maximal effect values: 4.86 +/- 0.37% of noradrenaline precontraction versus 7.06 +/- 0.31% in wild-type, n = 8, P < 0.05). This attenuation was mimicked by treatment of rings from wild-type animals with the connexin-mimetic peptide (37,43)Gap27 (5 x 10(-6)m). In conclusion, replacement of Cx43 by Cx32 attenuates EDHF-mediated relaxations in mice aortic rings, suggesting that they are, at least in part, dependent on Cx43-formed gap junctions. In contrast, aortic responses to tested endothelium-independent agonists were not modified in knock-in animals.