Abstract Neuroendocrine differentiation of prostate cancer (PCa) occurs frequently during the development of castration resistance and rarely in primary tumors. Using the transgenic mouse model of prostate cancer (TRAMP) in a B6/C57 background, we were able to detect, isolate and further characterize basal, luminal and neuroendocrine subsets of cancer cells. We performed allograft experiments in NSG mice to ensure cellular stem and progenitor properties as well as metastatic potential, and thus created a murine model of neuroendocrine prostate cancer. Whilst all three previously described cell populations were present in approximately three quarters of primary TRAMP tumors, some tumors lacked basal stem cells and showed a more aggressive phenotype. These tumors, mainly consisting of small cancer cells, expressed markers of neuroendocrine differentiation such as synaptophysin and chromogranin A as shown by immunohistochemistry and are further referred to as neuroendocrine carcinomas (NECs), as ruled by an experienced uropathologist of our institution. Adenocarcinoma-like tumors (ACs), in contrast, showed high expression of cytokeratins and retained glandular histology. Using fluorescence-activated cell sorting (FACS) against newly discovered NEC markers within the TRAMP model, we found a relative increase of neuroendocrine progenitors in prostate NECs compared to ACs (approximately 75 vs. 33%, respectively). In ACs, on the other hand, luminal progenitors were found to be the predominant drivers of cancer progression. To further evaluate this, we transplanted single-cell-suspensions into NSG mice without androgen supplementation and observed successful engraftment of both non-basal cell populations. Additionally, we were able to passage the resulted tumors for at least two generations and observed maintenance of histology and biological features for tumors of both luminal and neuroendocrine origin. Based on a RNA sequencing, we were able to define gene signatures for neuroendocrine and luminal progenitors, uncovering a number of novel potential therapeutic targets. Ultimately, we were able to evaluate the prognostic value of the signatures obtained from mice in the human disease by in-silico analyses of publically available gene expression profile databases. In conclusion, we created and characterized a murine model of neuroendocrine prostate cancer using flow cytometry and murine allografts. Further, we established gene expression signatures of luminal and neuroendocrine progenitors and translated them to the human disease. Our findings foster the understanding of neuroendocrine differentiation in prostate cancer and may help in developing new targeted approaches in this entity. Citation Format: Maximilian Marhold, Erwin Tomasich, Simon Udovica, Gerwin Heller, Corinna Altenberger, Andreas Spittler, Reinhard Horvat, Peter Horak, Michael Krainer. Neuroendocrine and luminal progenitors drive cancer progression in prostate cancer. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 2407.