A design-of-experiments strategy for engineering 3D topographical features in osteosarcoma modelling

Osteosarcoma is a highly aggressive bone cancer with poor patient outcomes, partly due to the limited predictive power of current preclinical models. Conventional two-dimensional (2D) cultures fail to recapitulate physiologically relevant cell-matrix interactions, while animal models suffer from interspecies variability. To investigate how extracellular matrix (ECM) topographical features influence osteosarcoma behavior, polylactic acid-based microparticles were engineered with bone-mimetic stiffness and defined surface topographies, guided by a Design-of-Experiments (DoE) approach. This enabled systematic variation of microparticle architecture (8-63 μm diameter; 2-13 μm dimple size) for studying the impact of surface topography on osteosarcoma cell behavior in 3D culture, with doxorubicin treatment as a functional readout to evaluate the effects of 3D topographical cues on chemotherapy sensitivity. Topography was found to modulate cell-microparticle aggregation dynamics. At 96 h post-seeding, MG-63 cells displayed significantly reduced metabolic activity on all 3D microparticle designs, with heterogeneously dimpled-topography cultures displaying significantly lower DNA content than conventional 2D cultures. In U2OS cells, metabolic activity was significantly lower on smooth microparticles compared to dimpled designs, with all 3D cultures showing significantly lower DNA content versus 2D. Response to doxorubicin was more strongly influenced by culture dimensionality than surface topography, underscoring the importance of 3D context. Significant metabolic differences between 3D and 2D cultures were also observed, including the enrichment of amino acid related pathways and downregulation of ferroptosis signatures in 3D microparticle cultures. Topography displayed subtler effects on lipid and nucleotide metabolism. This study highlights how topographically-patterned 3D substrates can shape osteosarcoma cell behavior and drug response for disease modelling applications. Our DoE-guided platform enables systematic investigation for dissecting how ECM-inspired physical cues influence osteosarcoma progression and therapeutic resistance.