Surface-guided computing to quantify dynamic interactions between cell morphology and molecular signals in 3D

Abstract

Many signalling circuits are governed by the spatiotemporal organization of membrane-associated molecules. Growing evidence suggests that the mesoscale cell surface geometry is central in modulating this interplay. However, defining the causal hierarchy between geometric and molecular factors that control signals remains challenging. Nonlinearity and redundancy among the components prevent direct experimental perturbation, with shape being the most difficult to independently control. Towards the goal of inferring causality from observational data, we developed u-Unwrap3D as a resource to map arbitrarily complex 3D cell surfaces to diverse representations, each designed to interrogate a different aspect of the dynamic interaction between cell surface geometry and molecular cues. Using u-Unwrap3D, we discover a retrograde protrusion flow on natural killer cells associated with immunological synapse formation with cancer; establish a causal association of K14+ cells with breast tumor organoid invasion; measure the speed of ruffles; and quantify bleb-mediated assembly of septin polymers at the membrane.