While polycystin-1 (PC1) and polycystin-2 (PC2) are postulated to function as a receptor-channel complex, it is unknown what they sense nor what they signal. Fibrocystin (FPC) is a ciliary protein with Notch-like properties but its functional properties also remain uncertain. The PKD Section is tackling this problem in a variety of ways. In one set of studies, we are seeking to develop cell models that be can used to screen for PKD function. Not only can they be used to determine mechanisms of action, they can be highly useful for screening for compounds that can rescue function of mutant cells, as exemplified by work in the CFTR field. Unfortunately, the PKD field has lacked good methods and high-throughput assay systems that can be used for this purpose. As noted last year, we have generated multiple mouse lines that either express or lack expression of various PKD genes in the hope that we could use them for this purpose. Despite further testing, we have failed to identify any that robustly demonstrate consistent properties determined by the genotype. Under all conditions we have tested, all cells either produce tubules or cysts regardless of genotype. While the results to date have been disappointing, we will continue to tackle this problem because of its importance. In parallel, we have resumed use of the MDCK tubulogenesis system for evaluating PKD gene function. We have expressed multiple mutant forms of PKD1 and multiple tagged versions of WT PKD1 and found a consistent relationship between genotype and morphology. We currently are using this system to further explore PKD1/PKD2 function. In one set of studies, we have tested the relationship between a known ciliary factor and PC1. Knock-out of the ciliary factor in mice results in cystic disease but the mechanism is undefined. Using MDCK cell lines with stable, inducible expression of fluorescently-labeled PC1, we could show co-precipitation, co-trafficking and co-localization of the two proteins. Importantly, we could show that silencing of the ciliary gene in the PKD1-expressing cells resulted in cyst formation and that re-expression of the mouse orthologue of the silenced gene rescued PKD1 function. In contrast, HGF-mediated tubulogenesis was unaffected in the silenced cells. These data suggest that this model may be uniquely qualified to assess how other factors assess PKD1 function since tubule formation in this model is strictly PKD1-dependent. This is an important consideration since many factors, like ciliary proteins, have multiple partners and participate in numerous signaling pathways. Applying this reasoning to the ciliary factor we evaluated, our data suggest that its loss may result in cystic disease in mice predominantly through its effects on PC1 function. In other studies, we have similarly used the MDCK cell system to evaluate the functional relationship between PC1 and a small GTP-binding protein pulled out of a yeast two hybrid library. We have confirmed binding, co-trafficking and initial studies also suggest a functional relationship. We expect to complete the full evaluation in the coming months. We also are pursuing a complementary systems-based strategy to identify gene and protein networks that are either directly or indirectly linked to PKD proteins. This approach offers a number of advantages: it requires no prior knowledge of PKD gene function, it can provide hints about PKD function, and it can be used to identify signaling pathways that might be targeted for future interventions. As noted last year, we have performed a large, systematic study of adult onset mice induced at P40 and followed for up to 200+ days. Several observations are noteworthy: a) the expression differences in normal males and females is about as great as that between WT and mutant kidneys;b) renal cystic disease in female mice develops more slowly and is less severe than that in male mice;c) hepatic disease is worse in female mice than in male mice. These data indicate that this model fully recapitulates the important but under-studied sexually dimorphic features of human ADPKD. This is an exciting observation because the protective effect of female sex on renal function is greater than that of any therapy or other modifier yet reported. The model provides a robust, reproducible system for studying the mechanisms underlying this phenomenon and a unique opportunity to identify gene networks and signaling systems that might be amenable to therapy. While the majority of our system-based analyses have been performed using mouse kidneys, we acknowledge the limitations of using whole organs: heterogeneity of gene inactivation, cellular heterogeneity intrinsic to the kidney, stochastic effects normal in complex biological systems, and other incompletely controlled environmental effects. Our initial plan had been to complement these studies with comprehensive studies of our inducible inactivation Pkd1 mouse renal epithelial cell lines. We reasoned that we could potentially validate our network maps using cells from the same mouse line and more easily switch back and forth to test predictions. However, we have postponed those studies until we can establish a model that reproducibly develops cysts when Pkd1 is inactivated. In the meantime, we have resorted to using the MDCK cell model described above to produce gene and protein network maps. A potent advantage of this approach is that we have already shown that functional interactions can be readily evaluated using this system, providing a practical and cost effective way to overcome one of the biggest challenges associated with network studies. These studies are ongoing. Finally, we have continued to use our mouse models to identify new functions for PKD proteins. In collaboration with investigators at Univ of Maryland who led the effort, we have shown that PC1 and PC2 have novel and essential functions in lymphatic development. This likely accounts for the edema that is a characteristic finding in homozygous mutant nulls. We also have identified a novel phenotype in our Pkhd1 mice that have deletions spanning exon 3 to 67.
Showing the most recent 10 out of 33 publications
