br Discussion We investigated whether CD

Discussion We investigated whether CD133 serves as a putative epithelial stem/progenitor cell marker in adult murine prostates and explored its function in prostate regenerative capacity. We demonstrated that CD133 is not only expressed in a small fraction of prostate basal rgs protein but is also expressed in some luminal and stromal cells. Moreover, we compared the stem cell potential of CD133-expressing cells and their negative counterparts using both in vitro and in vivo approaches. Consistent with a previous study in human prostates (Richardson et al., 2004), we showed that murine CD133-expressing basal cells exhibited higher stem cell potential in the in vitro prostate sphere assay. However, lineage tracing analysis demonstrated an equal contribution of CD133-expressing cells and their negative counterparts to the maintenance of prostate epithelia. These observations are not necessarily in contradictory with each other because the experimental conditions are distinct in these assays. Instead, these results highlight that cell behaviors can be determined by their biological contexts. Because the experimental condition in the lineage tracing approach is more physiologically relevant, we believe that CD133 does not enrich for the prostate stem cell activities in vivo. Finally, we also demonstrated that CD133 is functionally dispensable for prostate epithelial homeostasis and regeneration under the experimental contexts in this study. Of course, we cannot rule out that loss of CD133 may affect specific cell behaviors in other biological contexts not tested in our study. To this end, it remains unknown why CD133-expressing cells behave distinctly in different stem cell assays. One potential explanation is that the available signaling for cell survival and proliferation is different in these assays. The prostate sphere culture contains a minimal of growth factors such as EGF and FGF, and provides very basic cell signaling that support cell survival and proliferation. CD133 has been shown to potentiate activation of the PI3K/Akt pathway by directly interacting with the p85 subunit of PI3K and promoting phosphorylation of Akt (Wei et al., 2013). Pharmacological inhibition of the PI3K activity in the prostate cancer cell lines DU145 and PC3 decreased the percentage of the CD133+/CD44+ prostate cancer stem cells as well as their sphere-forming activity (Dubrovska et al., 2009). Therefore, CD133-expressing basal cells may have a relatively higher basal level of PI3K-Akt activity, which confers a minor survival advantage on them in the prostate sphere assay. In contrast, additional signaling regulating cell survival, differentiation and proliferation are provided in the organoid assay, such as augmentation of the Wnt signaling and suppression of the TGFβ signaling. These signaling have been shown to suppress apoptosis and anoikis (Yang et al., 2006; Orford et al., 1999; Chen et al., 2001; Ramachandra et al., 2002; Cao et al., 2006). They may override CD133-mediated signaling in the organoid assay and mask the minor survival advantage mediated by CD133. Similarly, in the in vivo lineage tracing assay, the challenge for cell survival does not serve as a prerequisite for assaying stem cell activity, thereby revealing the equal potential of the CD133+ and CD133− cells in maintaining epithelial homeostasis and regeneration. Collectively, our study corroborates that facultative function of specific cell populations induced by experimental conditions may not contribute to the in vivo biology substantially. Although our study shows that CD133+ cells and CD133− cells behave similarly in vivo in maintaining epithelial homeostasis and regeneration, CD133+ and CD133− cells have been shown to play distinct roles in prostate cancer initiation. Taylor RA et al. showed that CD133− cells in an immortalized BPH-1 human prostate epithelial cell line are more proliferative and more susceptible to transformation induced by cancer-associated fibroblasts or hormonal stimulation (Taylor et al., 2012). In addition, CD133+ and CD133− cells are also shown to be functionally distinct in several tumor studies. For example, CD133+ cells in human prostate cancer cell lines proliferated faster in vitro (Reyes et al., 2013) and CD133+ cells in human prostate cancer specimens possessed cancer stem cell activity (Vander Griend et al., 2008; Collins et al., 2005). It remains an open question whether CD133-mediated signaling dictates the functional outcome in these studies, or these studies imply that CD133 expression and putative cancer stem cell activity are co-regulated by the same signaling. Future studies using mouse models for prostate cancer should help address this question.