Spheroid Assembly in Microwells of Defined Geometry for Quantitative Assessment of Visco-Capillary Velocity and Shape Control

Three-dimensional (3D) cell spheroids are widely used in biomedical research as in vitro tissue models, yet quantitative understanding of their morphogenesis remains limited. Here, we present an integrated experimental and computational approach to analyze and model the compaction of cell aggregates in agarose microwells with defined cross-sectional geometries. Custom 3D-printed stamps were designed to produce circular, square, and triangular microwells with equal cross-sectional area. Time-lapse imaging and AI-based segmentation were employed to track the evolution of spheroid morphology, with circularity and projected area used as quantitative indicators of compaction dynamics. We show that the compaction process follows predictable exponential trends in both parameters, with mesenchymal spheroids (from human dermal fibroblasts line HDF) compacting faster than epithelial spheroids (from ARPE-19 cells). Spheroid rounding was simulated as a visco-capillary-driven process with a computational fluid dynamics (CFD) model using the Volume of Fluid (VoF) method in OpenFOAM. The visco-capillary velocity extracted from both experimental and simulation data served as a unifying parameter that explained differences in compaction kinetics. Using additional mechanical measurements (AFM and compression), we estimated surface tension and effective viscosity, confirming that surface tension differences predominantly drive the observed kinetics. Pharmacological treatments modulating cytoskeletal tension revealed that contractility inhibition significantly modified spheroid formation dynamics, allowing acquisition of non-spherical cell aggregates. Taken together, our study establishes a robust, geometry-controlled platform for analyzing spheroid formation and quantifying their mechanical properties, as well as provides a framework for creating cellular aggregates of defined shapes.