Basolateral polymerization of cellular fibronectin (FN) into a meshwork drives endothelial cell (EC) polarity and vascular remodelling. However, mechanisms coordinating α5β1 integrin-mediated extracellular FN endocytosis and exocytosis of newly synthesized FN remain elusive. Here we show that, on Rab21-elicited internalization, FN-bound/active α5β1 is recycled to the EC surface. We identify a pathway, comprising the regulators of post-Golgi carrier formation PI4KB and AP-1A, the small GTPase Rab11B, the surface tyrosine phosphatase receptor PTPRF and its adaptor PPFIA1, which we propose acts as a funnel combining FN secretion and recycling of active α5β1 integrin from the trans-Golgi network (TGN) to the EC surface, thus allowing FN fibrillogenesis. In this framework, PPFIA1 interacts with active α5β1 integrin and localizes close to EC adhesions where post-Golgi carriers are targeted. We show that PPFIA1 is required for FN polymerization-dependent vascular morphogenesis, both in vitro and in the developing zebrafish embryo.

No abstract

We present a novel deep learning approach to reconstruct confocal microscopy stacks from single light field images. To perform the reconstruction, we introduce the LFMNet, a novel neural network architecture inspired by the U-Net design [1]. It is able to reconstruct with high-accuracy a 112 × 112 × 57.6µm 3 volume (1287 × 1287 × 64 voxels) in 50ms given a single light field image of 1287 × 1287 pixels, thus dramatically reducing 720-fold the time for confocal scanning of assays at the same volumetric resolution and 64-fold the required storage. To prove the applicability in life sciences, our approach is evaluated both quantitatively and qualitatively on mouse brain slices with fluorescently labelled blood vessels. Because of the drastic reduction in scan time and storage space, our setup and method are directly applicable to real-time in vivo 3D microscopy. We provide analysis of the optical design, of the network architecture and of our training procedure to optimally reconstruct volumes for a given target depth range. To train our network, we built a data set of 362 light field images of mouse brain blood vessels and the corresponding aligned set of 3D confocal scans, which we use as ground truth. The data set will be made available for research purposes [2].