The Renal Differentiation and Neoplasia Section studies inductive signaling in tissue development/morphogenesis and, in parallel, its dysregulation in tumorigenesis with emphasis on the ligands that mediate normal tissue interactions and the pathways and targets that are activated in response to signaling. Our focus has been on development of the urogenital tract, which features reciprocal interactions between two distinct mesodermal progenitors, highly coordinated tissue movements, mesenchymal-epithelial transition (MET), integration of structures from different lineages, reiterative cycles of development, and a tumor that caricatures nephrogenesis. More specifically we are interested in the signaling mechanisms that direct metanephric mesenchyme (MM) to convert to the epithelia of the nephron. Wilms tumor (WT) is characterized by an expanded blastemal/progenitor population with a restricted capacity for epithelial conversion (MET). It is our long-term goal to identify targets on which WT cells depend for survival or dysregulated signaling that can be reprogrammed to allow tumor cells to differentiate to a more benign phenotype. Inductive signaling in MM progenitors that results in MET can be mediated by a variety of factors, including Wnt4, which is essential for nephron formation. We previously reported that Wnt4 induces MET by a calcium-dependent mechanism and not by canonical Wnt signaling as thought. We also determined that the cytokine leukemia inhibitory factor (LIF) can similarly induce tubule formation (MET) in renal progenitors, as can small molecule GSK3beta inhibitors. The common thread among these inducers is their ability to activate calcium signaling. LIF, for example, induces phosphorylation of PLCgamma, which results in the downstream activation of calcineurin. Calcineurin then dephosphorylates/activates transcription factors from the NFAT family. LIF also up regulates the expression of NFAT family members. Similarly, the GSK3beta inhibitors block the phosphorylation of NFAT, and induce the expression of MET markers. By ChIP assay, we have found that, following Wnt4 treatment of MM cells, NFAT binds directly to the promoter of a critical target of nephron formation, namely, Lim1/Lhx1. These findings confirm the role of calcium signaling and NFAT activation in MET of MM progenitors. This has led us to speculate that disruption of calcium signaling may be involved in Wilms tumor pathogenesis. Studies by M. Bouchard have demonstrated that calcineurin A-binding protein is overexpressed in more than 70% of WTs, consistent with our hypothesis. This protein interferes with the ability of calcineurin to dephosphorylate and thus activate NFAT, presumably also preventing the induction of MET. Thus regulation of signaling through this pathway may provide a useful target for WT therapy. An examination of this hypothesis is ongoing. In addition to its role in the induction of MET in MM, LIF also appears to have a function in the maintenance and expansion of the MM progenitor population. LIF induces MET in MM at 30-50 ng/ml through activation of calcium signaling in these progenitors. However, LIF also maintains and expands the progenitor population at levels that fail to induce MET (1-5 ng/ml). LIF functions principally through STAT activation, and the lower levels of LIF activate both STAT3 and STAT5 in renal progenitors. At 1 ng/ml, LIF increases the number of cells expressing renal stem cell marker Six2 and this is further increased with the concurrent addition of a Rho kinase inhibitor. The combination in fact causes a massive expansion of Six2-expressing cells in tissue culture. Importantly, these cells retain their ability to undergo MET in culture, thus producing a powerful tool for studying this process. Cells from these cultures have now been passaged multiple times without a loss of ability to differentiate in culture. Similar culture conditions have now also been applied to dissociated tissue fragments from WTs and support the propagation of Six2-expressing tumor cells, which have been shown to populate the blastemal portions of these tumors. We are currently characterizing the lines generated from these primary cultures. If tumorigenic in xenografts, these lines could provide an important new model for drug screening. We have also continued our examination of the role of STATs in the developing metanephros, where we have found STATs 1, 3, 5, and 6 to be highly expressed and phosphorylated/activated. Using conditional loss-of-function (LOF) mouse models, a preliminary assessment of a LOF mutant for Stat3 has revealed extensive defects in the skeletal system but no obvious alterations in the kidney other than possible size differences. In collaboration with colleagues in CDBL, we have found that Stat3 is required for maintenance of the trabecular bone, and the loss of Stat3 results in loss of mineralization in this tissue. We are currently generating conditional double mutants for Stats 1 and 3 and Stats 1 and 5 in order to circumvent the complications of redundancy and further assess the role of these Stats in renal development. Since mesodermal populations are significantly affected in the bone, we are also evaluating the MM progenitors in the metanephros for apoptotic cells. Finally, in collaboration with CDBL PI Terry Yamaguchi, we are investigating the role of Wnt5a in metanephric development. We have found that its specific inactivation in mesoderm using T/Brachyury-Cre results in duplex kidneys and bifid ureter formation bilaterally, a condition that afflicts some 1% of the overall population. Normally the ureteric bud, which forms the collecting ducts and ureter, extends as a single outgrowth from the Wolffian duct (WD) in the intermediate mesoderm (IM) at E10.5 in the mouse. However, Wnt5a is normally not expressed around the site of UB outgrowth at E10.5, suggesting that the cause of duplex kidney precedes this development in the IM. In fact, Wnt5a is expressed in a gradient at both the cranial and caudal ends of the IM at E9.5. To examine the effect of its expression on duplex kidney formation, we ablated Wnt5a in the posterior of the embryo using several tissue-specific Cre lines. T-Cre, Pax3-Cre, and Hoxb6-Cre eliminate Wnt5a expression in the posterior IM, whereas the combination of AP2-Cre and Rarb2-Cre or Dll1-Cre, which are active within the kidney itself, do not. T-Cre, Pax3-Cre and Hoxb6-Cre mediated Wnt5a mutants have a truncated and widened IM and double ureters, while AP2-Cre and Rarb2-Cre or Dll1-Cre-mediated Wnt5a mutants have a normal IM and single ureters. Moreover, tamoxifen induction of Hoxb6-Cre-ERT2 at E6.5 yielded Wnt5a-deficient embryos with an aberrant IM and double ureters;whereas tamoxifen induction at E8.5 resulted in mutants with normal IM extension and single ureters. Further studies using progenitor-specific markers reveal that at E9.5, Wnt5a mutant embryos with an IM disturbance have a striking misorientation of ND cells, typical of disrupted planar cell polarity signaling. Moreover, the ND bifurcates at its caudal terminus, thus forming twin epithelial UB outgrowths in the surrounding MM. These results suggest that Wnt5a regulates IM extension and that its loss in the IM causes double ureter and duplex kidney formation. We are currently collaborating with a computer modeling group to develop hypotheses that might explain this phenotype.
Showing the most recent 10 out of 20 publications
