Colloid-patterned surfaces distinguish malignant mechanophenotypes.

Accurate and rapid identification of aggressive cancer cells remains a major clinical challenge. Here, we present a simple, label-free mechanophenotyping platform that integrates controlled colloidal topographies with particle-uptake measurements to reveal biophysical traits associated with metastatic progression. Non-close-packed polystyrene bead arrays were formed on cell culture plates by controlled deposition and stabilized with a thin silicon oxide coating. These arrays display micro- and nano-features with a size range of 0.23-2.3 μm at diverse densities and were used to assess adhesion across cancer cells exhibiting different levels of malignancy. Particle uptake differences were most pronounced for particle diameters above 0.5 μm, whereas adhesion differences emerged predominantly on particles ≥0.7 μm and increased progressively with larger particle sizes. Colloidal topographies were fabricated at particle deposition concentrations of 500 μg/mL and 1000 μg/mL, and adhesion differences were observed under both conditions, with more potent effects at the higher concentration. At the metastatic site, cells exhibited increased particle uptake, stronger adhesion, and a larger morphological engagement on colloid-coated substrates, characterized by extensive actin-rich protrusions wrapping individual particles. AFM force mapping confirmed higher adhesion forces to a colloidal probe, while transcriptomic profiling revealed enrichment of adhesion and ECM-remodeling pathways in the adhesive metastatic state. We also find that lymphatically selected cells exhibit reduced adhesion on colloid-coated surfaces but higher particle uptake compared to the primary tumor cells. These results indicate that after leaving the primary tumor, metastatic cells have reduced adhesive potential, which is only regained upon reaching secondary sites. By exposing adhesion differences that are undetectable on flat substrates and linking them to particle uptake assays, this platform produces functional signatures of metastatic potential. This method is technically accessible, compatible with imaging and molecular workflows, and adaptable for high-throughput or clinical analysis, offering a potential route for label-free detection and classification of cancer cells by their aggressiveness.