Valve morphogenesis : directing cardiac form and function

Abstract

MicroRNAs regulate endocardial cushion differentiation To investigate more specifically the function of miRNAs in both the myocardium and endocardium, I generated MZdicer mutant lines carrying either a cmlc2:GFP (myocardium) or a kdr-l:GFP (endothelial cells, endocardium) transgene. We performed detailed protein and mRNA expression analysis and found that the endocardium was most affected by a loss of miRNAs. Endocardial cushion (EC/ valve) specific markers like hyaluronic acid synthase 2 (has2) and the cell surface protein Alcama/Dm-grasp, were ectopically expressed in endocardial cells lining the chambers of MZdicer mutant hearts indicating an expansion of the EC primordium. We subsequently identified miR-23 to be functionally responsible for restricting valve formation in the embryo. MiR-23 knockdown resulted in a similar EC expansion and reintroducing miR-23 rescued the MZdicer endocardial phenotype. We could detect miR-23 expression in the ECs from 53hpf, 20 hours after the onset of cardiac cushion differentiation. Interestingly, when we overexpressed miR-23 in mouse embryonic endothelial cells (MEECs), we could inhibit a TGF-β induced EMT response. We next searched for miR-23 target genes. We selected a group of candidate genes that were specifically expressed in mouse embryonic hearts at E10.5 and contain a putative miR-23 3’UTR target sequence (www.targetscan.org). We subsequently selected candidate genes that were shown to be up-regulated during EMT in MEECs and identified functional interactions with miR-23 using zebrafish 3’UTR sequences. This study identified three novel miR-23 targets; has2, transmembrane protein 2 (tmem2) and β-catenin interacting protein 1 (ctnnbip1/icat). Having established that miR-23 regulates these genes and EC formation in the embryo, we finally showed that extracellular HA production is directly responsible for the excessive endocardial cushion cell differentiation in dicermutants.HA can bind the CD44 receptor and thereby stimulate the PI3K pathway, which results in increased Has2 expression. In addition, the HA/CD44/ErbB2 interaction induces β-catenin activation, which results in the induction of Has2 expression. Therefore based on my studies we have proposed that excessive HA production in the cardiac jelly due to a loss of miR-23 function, inducesan auto regulatory loop in normally non-responsive cells which start to express has2 resulting in more HA secretion. With this study we were the first to report a role for miRNAs during cardiac valve development.