Transcriptome-Guided Reverse Engineering of Human Prostate Cancer

Over many decades ex vivo culture methods for human cells have emerged with the hope of replacing animal experiments. However, how well these culture conditions reflect animal models or even humans has not been systematically analyzed. As a result, the choice of a specific culture method and its predictive power remains largely a matter of belief rather than an informed decision. Using prostate cancer as an example, we propose transcriptional profiling to assess qualitative and quantitative differences between ex vivo and in vivo conditions. More specifically, we will take advantage of established and tuneable culture matrices to evaluate the influence of biophysical/-chemical properties on the transcriptional output program of prostate cancer cells. Subsequently, we will use this knowledge and artificial intelligence to refine existing ex vivo culture models to better match the transcriptional fingerprint of animal models or prostate cancer in humans. Finally, selected culture methods will be validated using complementary functional approaches, such as comparing drug responses ex and in vivo and the ability to culture patient-derived human cells ex vivo directly. Cost-effective and manageable protocols for everyday laboratory research will be prioritized during this circular optimization process. The universally applicable framework may enable the identification and subsequent optimization of more predictive ex vivo culture models and thereby help replace current animal experiments and increase the chances that promising laboratory-based findings validate in a clinical setting.