Dynamic prostate cancer transcriptome analysis delineates the trajectory to disease progression
Marco Bolis, Daniela Bossi, Arianna Vallerga, Valentina Ceserani, Manuela Cavalli, Daniela Impellizzieri, Laura Di Rito, Eugenio Zoni, Simone Mosole, Angela Rita Elia, Andrea Rinaldi, Ricardo Pereira Mestre, Eugenia D’Antonio, Matteo Ferrari, Flavio Stoffel, Fernando Jermini, Silke Gillessen, Lukas Bubendorf, Peter Schraml, Arianna Calcinotto, Eva Corey, Holger Moch, Martin Spahn, George N. Thalmann, Marianna Kruithof‐de Julio, Mark A. Rubin, Jean‐Philippe Theurillat
Abstract
Comprehensive genomic studies have delineated key driver mutations linked to disease progression for most cancers. However, corresponding transcriptional changes remain largely elusive because of the bias associated with cross-study analysis. Here, we overcome these hurdles and generate a comprehensive prostate cancer transcriptome atlas that describes the roadmap to tumor progression in a qualitative and quantitative manner. Most cancers follow a uniform trajectory characterized by upregulation of polycomb-repressive-complex-2, G2-M checkpoints, and M2 macrophage polarization. Using patient-derived xenograft models, we functionally validate our observations and add single-cell resolution. Thereby, we show that tumor progression occurs through transcriptional adaption rather than a selection of pre-existing cancer cell clusters. Moreover, we determine at the single-cell level how inhibition of EZH2 - the top upregulated gene along the trajectory - reverts tumor progression and macrophage polarization. Finally, a user-friendly web-resource is provided enabling the investigation of dynamic transcriptional perturbations linked to disease progression.