The endothelium controls vascular smooth muscle tone by different dilating mechanisms, including the release of nitric oxide (NO) and prostaglandins, but also by inducing smooth muscle hyperpolarisation (EDH-type dilation). The relative importance of this dilator mechanism varies between arteries of different sizes with a preponderance of NO in conducting arteries and, on the other side, of EDH-type dilations in arterioles. Importantly, the balance may also be shifted during coronary artery disease. The mechanism of the EDH-type dilation is still a matter of discussion but may include a direct transfer of charge from the endothelium into the smooth muscle layer by means of direct cell contact and the formation of heterocellular gap junctions. Such myoendothelial junctions (MEJ) have been demonstrated preferentially in arterioles, in line with the prevailing EDH-type dilation in these vessels. It was previously demonstrated in the laboratory of Brant Isakson that plasminogen activator inhibitor 1 (PAI-1) is required for MEJ formation in arterioles and also accumulates at these sites through specific mechanisms (1,2).
In the present study, Shu and colleagues took this idea further and examined whether a larger conducting artery (A. carotis) in the mouse could be manipulated to acquire anatomical and functional features of an arteriole by prolonged local application of PAI-1 onto the vessel in vivo (3). Seven days after application of PAI-1 they found a substantial increase in the number of holes (3-fold) in the internal elastic lamina that separates endothelium from smooth muscle and endothelial projections protruded through the lamina so that the arteries resembled morphologically arterioles. In addition, functional heterocellular coupling could be demonstrated by means of dye transfer from endothelial cells to smooth muscle cells, a feature that was not observed in untreated or sham treated carotid arteries. EDH-type dilations are initiated by activation of Ca2+-dependent K+-channels (IKCa and/or SKCa) (4,5) which can be activated pharmacologically using NS309. While this activator barely induced a dilation in sham-treated arteries, it dilated PAI-1 treated arteries in an endothelium-dependent fashion indicating the presence of EDH-type dilation. As mentioned above the dilation upon acetylcholine is mediated largely through NO in large arteries and this was also the case in the present study (assessed by blockade of eNOS). However, in PAI-1 treated arteries the acetylcholine-induced dilation was barely sensitive to eNOS inhibition and only abrogated by blockade of KCa channels providing further evidence of an EDH-type dilation in PAI-1 treated vessels. Interestingly, haemoglobin alpha protein was also found at these sites in treated vessels as before also in arterioles at MEJs where it controls NO signalling depending on the oxidation state of its iron atom (6). In these PAI-1 treated arteries haemoglobin was crucially involved in the switch towards EDH-type signalling because in mice deficient for haemoglobin alpha the acetylcholine dilation could still be abrogated by eNOS blockade despite the presence of holes, endothelial projections, and dye transfer.
In summary, the article provides insight into a mechanism by which endothelium-dependent dilation is modified and the type of dilator mechanism shifted, in this case from NO to an EDH-type dilation. This shift was initiated by application of PAI-1 but may likewise occur during disease as is reported also for coronary arterioles (7). It underlines the variability of endothelial dilator pathways and highlights that defects in mechanisms supporting EDH-type dilation may contribute to microvascular coronary disease (8).