Identification of a link between prostate stem cell self-renewal and carcinogenesis

Growing evidence suggests that the self-renewal pathways critical for normal tissue stem cell maintenance are also important for cancer progression in the same tissue. Identification and characterization of the genes and pathways that control prostate stem cell self-renewal are essential to better understand normal prostate homeostasis as well as carcinogenesis. An essential first step to this goal is to be able to isolate and maintain prostate stem cells in vitro. Fluorescence Activated Cell Sorting (FACS) technology was used to identify a defined population of prostate cells enriched for stem cell activity, as defined by growth in an in vitro colony forming assay and an in vivo dissociated cell prostate regeneration assay. The Lineage (CD45, CD31, Ter119) -Sca-1+ CD49f+ cells, named the 'LSC' fraction, contained all of the colony and tubule forming activities. A semi-solid prostate sphere assay was developed to quantify and maintain these cells, and study their unique properties. Only LSC cells form clonally derived spheres, which develop spontaneous lineage hierarchies, and could be driven to terminal differentiation in a 2D colony assay. The spheres can be passaged individually or in bulk for up to 12 generations, which makes this a powerful tool to study self-renewal in vitro. Using these tools, several self-renewal pathways were tested to identify ones expressed in and crucial for the maintenance of prostate stem cells. The polycomb group protein Bmi-1 is enriched in the LSC population. Using an shRNA based knock-down strategy, we found that loss of Bmi-1 reduces sphere forming and self-renewal activities of primary prostate cells, and diminishes their in vivo regeneration capacity. This is partially accomplished by the induction of pl6Ink4a and p19ARF, which control senescence and apoptosis, respectively. Inhibition of Bmi-1 could reverse the added self-renewal activity driven by Wnt/beta-catenin, suggesting that Bmi-1 is required for the functional activity of at least one other self-renewal pathway. Finally, Bmi-1 loss protected primary prostate cells from paracrine growth factor signal induced hyperplasia, and attenuated the growth of prostate cancer induced by more potent oncogenic stimuli. Using novel isolation and culturing techniques, we conclude that Bmi-1 is a crucial regulator of prostate stem cell self-renewal, and it plays important roles in prostate cancer initiation and progression as well.