Our data demonstrate that activated RAAS induces FGF23 expression in cardiac myocytes and thereby stimulates a pro-fibrotic crosstalk between cardiac myocytes and fibroblasts, which may contribute to myocardial fibrosis in CKD.

The focal adhesion kinase (FAK) regulates the dynamics of integrin-based cell adhesions important for motility. FAK’s activity regulation is involved in stress-sensing and focal-adhesion turnover. The effect of FAK on 3D migration and cellular mechanics is unclear. We analyzed FAK knock-out mouse embryonic fibroblasts and cells expressing a kinase-dead FAK mutant, R454-FAK, in comparison to FAK wild-type cells. FAK knock-out and FAKR454/R454 cells invade dense 3D matrices less efficiently. These results are supported by FAK knock-down in wild-type fibroblasts and MDA-MB-231 human breast cancer cells showing reduced invasiveness. Pharmacological interventions indicate that in 3D matrices, cells deficient in FAK or kinase-activity behave similarly to wild-type cells treated with inhibitors of Src-activity or actomyosin-contractility. Using magnetic tweezers experiments, FAKR454/R454 cells are shown to be softer and exhibit impaired adhesion to fibronectin and collagen, which is consistent with their reduced 3D invasiveness. In line with this, FAKR454/R454 cells cannot contract the matrix in contrast to FAK wild-type cells. Finally, our findings demonstrate that active FAK facilitates 3D matrix invasion through increased cellular stiffness and transmission of actomyosin-dependent contractile force in dense 3D extracellular matrices.

Mutations of the Pkhd1 gene cause autosomal recessive polycystic kidney disease (ARPKD). Pkhd1 encodes fibrocystin/polyductin (FPC), a ciliary type I membrane protein of largely unknown function, suggested to affect adhesion signaling of cells. Contributions of epithelial cell adhesion and contractility to the disease process are elusive. Here, we link loss of FPC to defective epithelial morphogenesis in 3D cell culture and altered cell contact formation. We study Pkhd1-silenced Madin-Darby Canine Kidney II (MDCKII) cells using an epithelial morphogenesis assay based on micropatterned glass coverslips. The assay allows analysis of cell adhesion, polarity and lumen formation of epithelial spheroids. Pkhd1 silencing critically affects the initial phase of the morphogenesis assay, leading to a reduction of correctly polarized spheroids by two thirds. Defects are characterized by altered cell adhesion and centrosome positioning of FPC-deficient cells in their 1-/2-cell stages. When myosin II inhibitor is applied to reduce cellular tension during the critical early phase of the assay, Pkhd1 silencing no longer inhibits formation of correctly polarized epithelia. We propose that altered sensing and cell interaction of FPC-deficient epithelial cells promote progressive epithelial defects in ARPKD.

Three-dimensional culture systems that allow generation of monolayered epithelial cell spheroids are widely used to study epithelial function in vitro. Epithelial spheroid formation is applied to address cellular consequences of (mono)-genetic disorders, that is, ciliopathies, in toxicity testing, or to develop treatment options aimed to restore proper epithelial cell characteristics and function. With the potential of a highthroughput method, the main obstacle to efficient application of the spheroid formation assay so far is the laborious, time-consuming, and bias-prone analysis of spheroid images by individuals. Hundredths of multidimensional fluorescence images are blinded, rated by three persons, and subsequently, differences in ratings are compared and discussed. Here, we apply supervised learning and compare strategies based on machine learning versus deep learning. While deep learning approaches can directly process raw image data, machine learning requires transformed data of features extracted from fluorescence images. We verify the accuracy of both strategies on a validation data set, analyse an experimental data set, and observe that different strategies can be very accurate. Deep learning, however, is less sensitive to overfitting and experimental batch-to-batch variations, thus providing a rather powerful and easily adjustable classification tool.

Scientific Reports 7: Article number: 42780; published online: 16 February 2017; updated: 03 May 2017. This Article contains an error in the legend of Figure 5F. “(F) The Young’s modulus of suspended FAKwt/wt cells (blue, n = 20) is higher compared to suspended FAKR454/R454 cells (red, n = 20)”. should read: