Live-Cell Labeling and Artificial Intelligence Approaches for High-Resolution XYZT Imaging of Cytoskeletal Dynamics...

Collective cell migration is crucial for a variety of pathophysiological processes including embryonic development, wound healing, carcinoma invasion, and sprouting angiogenesis. The behavior of leading and following cells during migration is highly dynamic and involves extensive cellular morphological changes mediated by the actin cytoskeleton. Imaging these rapid and dynamic changes over time requires expression of fluorescent proteins and/or live labeling with fluorescent probes, followed by acquiring series of image stacks at short intervals. This presents significant challenges related to dye cytotoxicity, signal loss, and in particular phototoxicity resulting from repeated irradiation, especially when using separate channels for multiple dyes and when imaging large z-stacks at short time intervals. In this chapter, we present methods for multicolor live-cell labeling of primary human endothelial cell populations, followed by multi-position time-lapse imaging in 2D and in 3D protein matrices. These approaches can be performed in combination with RNA interference to suppress the expression of specific proteins, as well as in mosaic assays using mixtures of differentially labeled cell populations. Finally, we present a protocol for long-term imaging at low laser intensity to minimize laser-induced cell damage, followed by post-imaging signal enhancement using artificial intelligence.