CRAC-TENSOR-FLOWS

Large-scale tissue deformations during development or tumor progression are governed by rheological properties, with cell rearrangements relaxing cell deformations. Current methods for characterizing these properties rely on invasive techniques such as laser ablation or deformable bead injection. However, living tissues can exert forces on themselves, which opens the possibility of a non-invasive, image-sequence-based rheological inference of tissues. Towards this aim, we propose to estimate the relationship between the mesoscopic tensor fields of cell shapes and rearrangements in various types of tissue flows. We will connect the relationship between mesoscopic tensors and the microscopic processes regulating cell-cell junctions. Our project has 3 primary objectives. Aim 1 involves quantifying cell-cell rearrangements by generating a comprehensive database using fluorescent labeling and advanced tracking tools across 2D and 3D experiments in cancer and embryonic vertebrate tissues. Sub-aims include studying cell migration on diverse adhesive patterns, applying shear flow in microfluidic setups, and analyzing natural flows in neural epithelia. We will explore the relationships between kinematic tensors and their temporal fluctuations, investigating phenomena like shear thinning and avalanche-like behaviors. Aim 2 focuses on mapping adhesion turnover to rheological models by examining junction tension and adhesion dynamics using techniques like FRAP, FRET sensors, and force inference. It incorporates four-way vertices into vertex models to simulate tissue mechanics and active feedback mechanisms. Aim 3 seeks to innovate quantification methods for mesenchymal and 3D tissues by leveraging generative AI to create realistic simulations for training deep-learning models, enhancing the detection and analysis of cell rearrangements. This integrated approach combines experimental and computational strategies to advance the understanding of tissue biomechanics and dynamics.