Spheroid assembly in microwells of defined geometry for quantitative assessment of aggregation kinetics and shape...

Three-dimensional (3D) cell spheroids are widely used asin vitrotissue models, yet quantitative understanding of their morphogenesis remains limited. We present an integrated experimental-computational framework to analyze, model, and modulate the compaction of cell aggregates in agarose microwells of defined geometries. Custom 3D-printed stamps produced circular, square, and triangular microwells of equal cross-sectional area. Time-lapse imaging combined with AI-based segmentation enabled tracking of spheroid morphology, with circularity and projected area serving as quantitative descriptors of compaction. The process followed predictable exponential kinetics, with mesenchymal (HDF) spheroids compacting faster than epithelial (ARPE-19) ones. Computational fluid dynamics (CFD) simulations modeled spheroid rounding as a visco-capillary-driven process, where the extracted visco-capillary velocity unified experimental and simulated dynamics. Mechanical measurements by atomic force microscopy and compression confirmed that differences in surface tension predominantly governed the observed kinetics. Pharmacological modulation of cytoskeletal tension revealed that inhibition of contractility markedly altered spheroid formation dynamics, enabling the generation of stable, non-spherical aggregates. Using this principle as a shape-engineering strategy, we produced aggregates with distinct geometries (brick-like, prismatic, and star-shaped), characterized by an increased surface-to-volume ratio compared to conventional spheroids. Limitations of the approach include the use of pharmacological cytoskeletal modulation and constraints in geometric fidelity arising from printing resolution, agarose casting, cell filling, and intrinsic smoothing of sharp features during cell aggregation. Collectively, this work establishes a geometry-controlled platform for quantitative analysis of spheroid formation and mechanical behavior, and provides a versatile framework for designing cell aggregates with defined shapes.