Connexin‐mimetic peptides dissociate electrotonic EDHF‐type signalling via myoendothelial and smooth muscle gap junctions in the rabbit iliac artery

Synthetic peptides corresponding to the Gap 26 and Gap 27 domains of the first and second extracellular loops of the major vascular connexins (Cx37, Cx40 and Cx43), designated as 43Gap 26, 40Gap 27, 37,40Gap 26 and 37,43Gap 27 according to Cx homology, were used to investigate the role of gap junctions in the spread of endothelial hyperpolarizations evoked by cyclopiazonic acid (CPA) through the wall of the rabbit iliac artery. Immunostaining and confocal microscopy demonstrated that gap junction plaques constructed from Cx37 and Cx40 were abundant in the endothelium, whereas Cx43 was the dominant Cx visualized in the media. None of the Cx-mimetic peptides affected endothelial hyperpolarizations evoked by CPA directly. When administered individually, 40Gap 27, 37,40Gap 26 and 37,43Gap 27, but not 43Gap 26, attenuated endothelium-dependent subintimal smooth muscle hyperpolarization. By contrast, only 43Gap 26 and 37,43Gap 27 reduced the spread of subintimal hyperpolarization through the media of the rabbit iliac artery. The site of action of the peptides therefore correlated closely with the expression of their target Cxs in detectable gap junction plaques. The findings provide further evidence that the EDHF phenomenon is electrotonic in nature, and highlight the contribution of myoendothelial and homocellular smooth muscle communication via gap junctions to arterial function. Keywords: Gap junctions, connexin, EDHF, hyperpolarization Introduction Synthetic peptides homologous to sequences present in the extracellular loops of connexins (Cxs) 37, 40 and 43 are capable of attenuating endothelium-dependent hyperpolarizations and relaxations that are independent of nitric oxide and prostanoids in isolated arteries and in vivo (Chaytor et al., 1998; 2001; De Vriese et al., 2002; Griffith et al., 2002; Sandow et al., 2002). Since these agents impair intercellular communication via gap junctions, endothelium-dependent smooth muscle hyperpolarization is likely to involve direct cell–cell coupling, rather than extracellular transfer of a freely diffusible endothelium-derived hyperpolarizing factor (EDHF; for a review, see Griffith, 2004). Indeed, in the rabbit iliac artery, Cx-mimetic peptides have been shown to attenuate the transfer of fluorescent tracer dye from the endothelium into the media and transmission of endothelial hyperpolarization to subintimal smooth muscle cells via myoendothelial gap junctions (Griffith et al., 2002; Chaytor et al., 2003). Such peptides are also capable of uncoupling vascular smooth muscle cells, suggesting that a component of their action against the EDHF phenomenon might also reflect an ability to attenuate the electrotonic relay of endothelial hyperpolarization through successive layers of the media (Chaytor et al., 1997; Edwards et al., 2000). In the present study, we have therefore employed four peptides, 37,40Gap 26, 43Gap 26, 37,43Gap 27 and 40Gap 27, homologous to specific domains of the first (Gap 26) and second (Gap 27) extracellular loops of the dominant vascular connexins (Cx37, Cx40 and Cx43), to determine whether such agents can be used to dissociate electrical coupling via myoendothelial and homocellular smooth muscle gap junctions. Cyclopiazonic acid (CPA), a SERCA inhibitor that activates endothelial KCa channels by promoting capacitative Ca2+ entry via store-operated Ca2+ channels, was used to evoke EDHF-type hyperpolarizations of subintimal and subadventitial smooth muscle cells in the rabbit iliac artery (Chaytor et al., 1998; Taylor et al., 1998; Griffith, 2004). Control experiments were performed to confirm that Cx-mimetic peptides do not inhibit endothelial hyperpolarization directly, and their effects against subintimal and subadventitial smooth muscle hyperpolarization correlated with the endothelial and medial localization of gap junction plaques constructed from Cx37, Cx40 and Cx43, as characterized by immunostaining and confocal microscopy. Previous studies have provided evidence that the action of these peptides is Cx-selective. In confluent COS fibroblasts expressing Cx43, for example, intercellular dye transfer of Lucifer yellow is impaired by 37,43Gap 27, but not by 40Gap 27, which differs by just three amino acids (Chaytor et al., 1999). Furthermore, in confluent rat aortic A7r5 myocytes, which are coupled by plaques constructed from Cx40 and Cx43, intercellular transfer of Lucifer yellow can be inhibited by 40Gap 27 and 43Gap 26 in combination, but is essentially unaffected by either peptide individually (Chaytor et al., 2001). This suggests that, in tissues that express multiple Cx subtypes, more than one peptide may be necessary to inhibit intercellular communication because of Cx specificity. It should also be noted that the first and second loop peptides (43Gap 26 and 37,43Gap 27) appear to be equally effective in attenuating dye transfer between confluent HeLa cells transfected to express Cx43 (Berman et al., 2002).