Fine-grained, nonlinear image registration of live cell movies reveals spatiotemporal organization of diffuse molecular processes

Abstract

We present an application of non-linear Image registration that allows spatiotemporal analysis of extremely noisy and diffuse molecular processes across the entire cell. To produce meaningful local tracking of the spatially coherent portion of diffuse protein dynamics, we improved upon existing non-linear image registration to compensate for cell movement and deformation. The registration relies on a subcellular fiducial marker, a cell motion mask, and a topological regularization that enforces diffeomorphism on the registration without significant loss of granularity. We demonstrate the potential of this approach in conjunction with stochastic time-series analysis through the discovery of distinct zones of coherent Profillin dynamics in symmetry-breaking U2OS cells. Further analysis of the resulting Profilin dynamics revealed strong relationships with the underlying actin organization. This study thus provides a framework for extracting functional interactions between cell morphodynamics, protein distributions, and signaling in cells undergoing continuous shape changes. Author SummaryBy adapting optical flow based nonlinear image registration we created a method specific for live cell movies that preserves the dynamics of a signal of interest by remapping using a separate measurable subcellular location fiducial. This is an extension as well on our labs previously published method of cell edge-based referencing of subcellular locations that was incapable of extracting interpretable subcellular time series more than a few microns away from the cell edge. We showed that our method overcomes this key limitation and allows sampling of subcellular time series from every subcellular location through our discovery of organized profilin dynamics in moving cell and that these profilin dynamics are related to actin dynamics due to their ability to bind growing actin structures likely through actin polymerizing factors. Most importantly, our method is applicable to discovering subcellular organization and coordination in a widely used form of live cell microscopy data that hitherto has been largely limited to anecdotal analysis.