The cell and molecular biological characterization of cell-cell junctions in mammalian heart valves

Abstract

In view of the large proportion of cardiac cells represented by interstitial mesenchymal cells, in particular in the valves, of the frequency and importance of valve surgery, notably replacements, and of the numerous “tissue engineering” projects to provide artificial valve structures it is actually surprising to see how limited the cell and molecular biological knowledge of valvular interstitial cells (VICs) still is. Surprisingly, the molecular components of the special adherens junctions (AJs) of VICs have not yet been elucidated to a satisfying degree. Therefore, the AJs of adult and fetal VICs of human and animal (bovine, ovine and porcine) heart valves have been studied in situ and cell culture preparations, using light and electron microscopy, including immunolocalization techniques, protein as well as glycoprotein analysis by SDS-PAGE, followed by identification of the separated molecules using immunoblotting. In my thesis I could show that adult mammalian VICs in situ possess cell-cell adhering junctions only of the puncta adhaerentia AJ-type, comprising N-cadherin and cadherin-11 as constitutive transmembrane glycoproteins, anchored in cytoplasmic plaques containing α- and β-catenin, plakoglobin, proteins p120 and p0071 which are accompanied by actin-binding proteins such as afadin, vinculin, α-actinin and proteins ZO 1-3. In two-dimensional (2D) cell cultures of adult mammalian VICs of different species, this rather simple molecular AJ ensemble was surprisingly found to be modified by the acquisition of the desmosomal plaque protein plakophilin-2, in the total absence of desmosomal structures and other desmosomal proteins. In three-dimensional (3D) culture constructs mimicking a native valve matrix environment, it could be shown that AJ-plakophilin-2 gradually decreased or was even lost but was able to re-assemble when VICs were re-isolated from 3D constructs and grown in 2D culture again. Human and porcine fetal VICs in situ also presented plakophilin-2 in their AJs, showing that the phenomenon of plakophilin-2 acquisition is not an artefact of cell culture conditions. Primary cultures of porcine fetal VICs also revealed adventitious plakophilin-2. Even more surprisingly fetal endothelial cells of the endocardium – but only those located at the valves and not those at the myocardium – also displayed the addition of this desmosomal protein to their AJ plaques, whereas those of other regions of fetal vascular endothelia remained negative for this protein. Again unexpectedly, pathologically altered cells of adult heart valves, showing an elevated proliferative activity, did not present plakophilin-2 in their AJs. Such as fetal VICs showed only relatively low proliferative activity, these findings may lead to the hypothesis that the acquisition of plakophilin-2 might be rather the result of a general activation- than a proliferation-induced event. Correspondingly, an alternative concept for early valvulogenesis is proposed, differering from the presently prevailing “epithelial- (here better: endothelial-) mesenchymal-transition” (EMT) hypothesis, which is based on the assumption of an ab initio presence of mesenchymal cells of the VIC type, even in the early heart tube “anlage”. The advent of a single desmosome-typical protein, plakophilin-2, in VICs of fetal heart valves and in culture is not an isolated phenomenon as it fits in phenomena observed in recently identified forms of plakophilin-2-containing AJs (coniunctiones adhaerentes) found in other mesenchymal cells such as bone marrow-derived stem cells, malignantly transformed mesenchymal cell lines and certain soft tissue tumors. It is obvious that a more detailed characterization of VICs, notably their AJs, will be needed to provide a safe basis for replacement valve surgery using structures formed by cells grown in valvular cell cultures in vitro.