Enabling methodology for high-throughput media screening in the context of multi-cell type co-cultures

Recently, an increasing number of in vitro tissue models have been engineered, incorporating a growing variety of cell types more accurately reproduce native tissue architecture and function. However, optimisation of culture media to support those co-cultures remains challenging due to limited standardised protocols and scalable screening methods. Current co-culture approaches involve either direct cell-cell contact or indirect paracrine signalling through shared media, with each presenting distinct challenges regarding their characterisation. The corresponding commercial co-culture systems are often costly, lack modularity, and are unsuitable for high-throughput screening. Here, we present a semi-automated methodology combining a custom 3D-printed device positioned in standard cell culture well-plates with automated image acquisition and statistical analysis to enable high-throughput screening of media formulations for mono-, bi-, and tri-culture systems. We applied this platform to a respiratory tissue model composed of bronchial epithelial cells, fibroblasts, and endothelial cells, screening 40 culture conditions and analysing 348 growth patterns. Our results highlight the critical influence of media composition and paracrine interactions on cell proliferation, with endothelial cells demonstrating greater sensitivity to media changes compared to epithelial cells. This methodology facilitates reproducible, large-scale evaluation of culture conditions while reducing manual labour through automation, although image curation remains necessary to ensure data quality. Finaly, this study introduces a scalable and robust screening approach for the optimisation of culture conditions in multi-cell co-cultures, providing valuable insights for tissue engineering applications and advancing the development of biomimetic in vitro models.