Due to specific and well characterized genetic tools such as

Due to specific and well-characterized genetic tools such as Villin-Cre mice, which allow for epithelium-specific transgene expression or Cre-mediated genetic excision of conditional urokinase in the intestine, many studies have focused on late development (Madison et al., 2002; El Marjou et al., 2004). Villin-Cre lines efficiently mediate recombination after villus morphogenesis begins, around E14.5, and efficient deletion of conditional alleles is often achieved at mid-gestational stages (Bondow et al., 2012; Walker et al., 2014). Therefore, the goal of the current work was to interrogate a functional role for WNT/β-CATENIN prior to villus morphogenesis. Our results demonstrate that disruption of WNT/β-CATENIN signaling, using Shh-Cre (Harfe et al., 2004) to achieve early epithelium-specific conditional deletion of Ctnn1b (β-catenin) (Brault et al., 2001) or the Frizzled co-receptors Lrp5 and Lrp6 (Lrp5/6) (Zhong et al., 2012), had little effect on the pseudostratified epithelium, indicating that WNT/β-CATENIN signaling was dispensable for proliferation at this time. Significant defects in proliferation and villus formation were only evident at later times, after villus morphogenesis had begun (E15.5). Furthermore, our results show that conditional deletion of Wntless, which is required for proper WNT ligand trafficking and secretion from the cell, from the mesenchymal, but not epithelial compartment, leads to a loss of epithelial proliferation at the time of villus formation. Collectively, our data demonstrate that WNT/β-CATENIN signaling is dispensable for regulating epithelial progenitor cell proliferation in the embryonic gut during the pseudostratified stage of development, whereas active signaling is absolutely required for proliferation and proper villus formation at the time when villus morphogenesis begins.
Results
Discussion Previous embryonic studies have shown that deletion of the β-catenin transcriptional binding partner Tcf7l2 (Tcf4) or the WNT ligand co-receptors Lrp5 and Lrp6 resulted in a loss of proliferation and collapse of the intervillus compartment at late stages of fetal development (E17.5–E18.5), indicating that WNT signaling is critical for proliferation at this developmental time (Korinek et al., 1998; Zhong et al., 2012). In contrast, results from transgenic Wnt reporter mice (TOP-GAL) have suggested that WNT/β-CATENIN activity was absent from the proliferating intervillus domain until postnatal life (Kim et al., 2007). Our results collectively show that WNT/β-CATENIN has biphasic activity, with very low WNT signaling activity during the pseudostratified stages, and with robust WNT signaling activity after the onset of villus morphogenesis. Thus, it is possible that previously published studies have touched on both of these modes of regulation without full appreciation that there are different levels of WNT signaling at different developmental times. In addition, some conclusions in published literature have been drawn from transgenic reporter mice, which may not accurately report signaling activity in certain contexts. For example, while the TOP-GAL mouse has been shown to faithfully report WNT/β-CATENIN signaling in the adult intestine (Davies et al., 2008), side-by-side comparisons of TOP-GAL and Axin2-LacZ reporter activity have indicated that multimerized Tcf/Lef reporter mice may not always be faithful (Al Alam et al., 2011; Barolo, 2006). Here, we presented several lines of evidence that suggest that there are two distinct mechanisms regulating fetal intestinal progenitor cell proliferation. During the pseudostratified stage of development at E13.5 and E14.5, epithelial progenitor cell proliferation occurs normally in the absence of WNT/β-CATENIN signaling, whereas after villus morphogenesis (E15.5), proliferating progenitor cells require WNT/β-CATENIN signaling. Mechanistically, our data point to increased WNT ligand expression in the mesenchyme as a major player in this developmental switch to WNT-dependent proliferation. However, our data do not totally rule out alternative scenarios. For example, it is also possible that ligands that augment WNT signaling, such as R-SPONDIN proteins, also change over developmental time (Kamata et al., 2004; Kim et al., 2008); and yet a second alternative possibility exists whereby an inhibitor of WNT signaling, such as DKK proteins, may be reduced over developmental time (Bafico et al., 2001; Mao et al., 2001; Tamai et al., 2000).