Endothelial cells (ECs) respond to shear stress by aligning in the direction of flow. However, how ECs respond to flow in complex in vivo environments is less clear. Here we describe an endothelial-specific transgenic zebrafish line, whereby the Golgi apparatus is labelled to allow for in vivo analysis of endothelial polarization. We find that most ECs polarize within 4.5 h after the onset of vigorous blood flow and, by manipulating cardiac function, observe that flow-induced EC polarization is a dynamic and reversible process. Based on its role in EC migration, we analyse the role of Apelin signalling in EC polarization and find that it is critical for this process. Knocking down Apelin receptor function in human primary ECs also affects their polarization. Our study provides new tools to analyse the mechanisms of EC polarization in vivo and reveals an important role in this process for a signalling pathway implicated in cardiovascular disease.

Trabeculation is crucial for cardiac muscle growth in vertebrates. This process requires the Erbb2/4 ligand Neuregulin (Nrg), secreted by the endocardium, as well as blood flow/cardiac contractility. Here, we address two fundamental, yet unresolved, questions about cardiac trabeculation: why does it initially occur in the ventricle and not the atrium, and how is it modulated by blood flow/contractility. Using loss-of-function approaches, we first show that zebrafish Nrg2a is required for trabeculation, and using a protein-trap line, find that it is expressed in both cardiac chambers albeit with different spatiotemporal patterns. Through gain-of-function experiments, we show that atrial cardiomyocytes can also respond to Nrg2a signalling, suggesting that the cardiac jelly, which remains prominent in the atrium, represents a barrier to Erbb2/4 activation. Furthermore, we find that blood flow/contractility is required for Nrg2a expression, and that while non-contractile hearts fail to trabeculate, non-contractile cardiomyocytes are also competent to respond to Nrg2a/Erbb2 signalling.

Despite great strides in understanding cardiac trabeculation, many mechanistic aspects remain unclear. To elucidate how cardiomyocyte shape changes are regulated during this process, we engineered transgenes to label their apical and basolateral membranes. Using these tools, we observed that compact-layer cardiomyocytes are clearly polarized while delaminating cardiomyocytes have lost their polarity. The apical transgene also enabled the imaging of cardiomyocyte apical constriction in real time. Furthermore, we found that Neuregulin signaling and blood flow/cardiac contractility are required for cardiomyocyte apical constriction and depolarization. Notably, we observed the activation of Notch signaling in cardiomyocytes adjacent to those undergoing apical constriction, and we showed that this activation is positively regulated by Neuregulin signaling. Inhibition of Notch signaling did not increase the percentage of cardiomyocytes undergoing apical constriction or of trabecular cardiomyocytes. These studies provide information about cardiomyocyte polarization and enhance our understanding of the complex mechanisms underlying ventricular morphogenesis and maturation.

Complex interplay between cardiac tissues is crucial for their integrity. The flow responsive transcription factor KLF2, which is expressed in the endocardium, is vital for cardiovascular development but its exact role remains to be defined. To this end, we mutated both klf2 paralogues in zebrafish, and while single mutants exhibit no obvious phenotype, double mutants display a novel phenotype of cardiomyocyte extrusion towards the abluminal side. This extrusion requires cardiac contractility and correlates with the mislocalization of N-cadherin from the lateral to the apical side of cardiomyocytes. Transgenic rescue data show that klf2 expression in endothelium, but not myocardium, prevents this cardiomyocyte extrusion phenotype. Transcriptome analysis of klf2 mutant hearts reveals that Fgf signaling is affected, and accordingly, we find that inhibition of Fgf signaling in wild-type animals can lead to abluminal cardiomyocyte extrusion. These studies provide new insights into how Klf2 regulates cardiovascular development and specifically myocardial wall integrity.

Zebrafish regenerate damaged myocardial tissue very effectively. Hence, insights into the molecular networks underlying zebrafish heart regeneration might help develop alternative strategies to restore human cardiac performance. While TGF-β signaling has been implicated in zebrafish cardiac regeneration, the role of its individual ligands remains unclear. Here, we report the opposing expression response during zebrafish heart regeneration of two genes, mstnb and inhbaa, which encode TGF-β family ligands. Using gain-of-function (GOF) and loss-of-function (LOF) approaches, we show that these ligands mediate inverse effects on cardiac regeneration and specifically on cardiomyocyte (CM) proliferation. Notably, we find that Inhbaa functions as a CM mitogen and that its overexpression leads to accelerated cardiac recovery and scar clearance after injury. In contrast, mstnb GOF and inhbaa LOF both lead to unresolved scarring after cardiac injury. We further show that Mstnb and Inhbaa inversely control Smad2 and Smad3 transcription factor activities through alternate Activin type 2 receptors.