Taken together these findings indicate that
Taken together, these findings indicate that basal-to-luminal differentiation can occur in prostate organogenesis, pathogenesis, and ex vivo assays, but rarely during normal tissue homeostasis. Thus, it has been unclear to what extent the plasticity of endogenous adult prostate basal hedgehog pathway in ex vivo models and disease states reflects an in vivo activity. In this study, we introduce a mouse model in which a tamoxifen-inducible Cre driver is used to delete E-cadherin in prostatic luminal cells, which are highly susceptible to anoikis (Kwon et al., 2014a). This results in rapid sloughing and death of luminal cells, followed by repair of the damaged epithelium. We show that basal-to-luminal differentiation contributes to tissue repair, providing a new approach for studying prostate stem/progenitor activity and epithelial specification.
Discussion Our studies have shown that E-cadherin plays an essential role in the maintenance of prostate epithelial integrity and are consistent with the requirement for E-cadherin in homeostasis of other epithelial tissues. For example, deletion of E-cadherin in the mouse small intestine and uterus in vivo leads to increased apoptosis and abnormal differentiation, underscoring its essential role in maintenance of epithelial architecture and function (Reardon et al., 2012; Schneider et al., 2010). Similarly, we have found that deletion of E-cadherin in prostate luminal cells results in their sloughing into the prostate lumen followed by anoikis. After loss of luminal cells, the prostate epithelium is repaired by the activity of basal progenitors, likely by differentiation through a transitional “intermediate cell” state, together with activated luminal progenitors. Notably, this process of tissue repair differs from androgen-mediated regeneration of the regressed prostate, which is largely driven by unipotent luminal and basal progenitors, with contributions from bipotential luminal and basal stem/progenitor cells (Choi et al., 2012; Liu et al., 2011; Lu et al., 2013; Wang et al., 2009, 2013, 2015). Such differences in progenitor behavior may reflect the severity of tissue “damage” (Figure 4). Overall, we propose that the identity of stem/progenitor cells depends upon the tissue context, and that there is no single cell population that can account for all stem cell properties within the prostate epithelium. Unlike other known contexts in which adult prostate basal cells can generate luminal cells, the surrounding stromal tissue is not profoundly affected in our epithelial repair model. For example, basal-to-luminal differentiation occurs in sphere formation assays in the absence of stromal tissue, as well as in tissue recombinant grafts with heterologous inductive mesenchyme (Burger et al., 2005; Goldstein et al., 2008, 2010; Hofner et al., 2015; Lawson et al., 2007; Richardson et al., 2004; Wang et al., 2013). Furthermore, a model of bacterial prostatitis also displays acute inflammation and stromal reactivity, which was suggested to be responsible for altered basal cell properties (Kwon et al., 2014b). These previous observations were consistent with the notion that the stroma represents an important niche for basal progenitors. In contrast, basal-to-luminal differentiation occurs in our E-cadherin deletion model when there is no alteration of the surrounding stroma, suggesting that there are signals intrinsic to the prostate epithelium that can promote basal cells to participate in tissue repair. It will be interesting in future studies to determine the identity of such signals that activate basal cell proliferation and differentiation. Our findings potentially reconcile studies of prostate epithelial stem/progenitor activity with work on injury/repair models in other epithelial tissues (Blanpain and Fuchs, 2014). For example, basal cells can form luminal secretory and ciliated cell types in the lung airway epithelium after treatment with noxious chemicals (Hogan et al., 2014), and basal cells in the bladder urothelium can generate intermediate and luminal umbrella cells after bacterial or chemical injury (Shin et al., 2011). Thus, our findings contribute to a unifying perspective for a role of basal cells in injury repair of a range of epithelial tissues. Finally, we suggest that the inherent plasticity of prostate basal cells that can be observed in pathological conditions and in cancer reflects latent progenitor activity that normally functions in tissue repair.