The laboratory has continued its investigations into the function of PKD proteins using cell culture and mouse model systems: 1. Characterization of PC1: The PLAT domain of PC1, which lies between the first and second transmembrane spans, is a key feature common to the entire class of PC1-related proteins from worms to humans yet its function is unknown. As a first step in addressing this problem, we had previously screened a mouse testis cDNA library with a homologous sequence from another PC1-related protein using a modified yeast two-hybrid method to identify putative interactors. We have subsequently verified that full-length, recombinant PC1 also binds to the same proteins when co-expressed in cell culture. Because endogenous PC1 is expressed at such low levels and because the antibodies for PC1 and its interactors are of limited quality, we have struggled to demonstrate in vivo interactions. To further validate the relevance of the interactions, we have tested for functional relevance using the MDCK cell culture system that we had previously described. Briefly, MDCK cells cultured in matrix have been used for >20years to study tubulogenesis. Left untreated, they form cysts but in the presence of HGF they form branching tubules. We had previously reported that expression of low levels of human PKD1 in MDCK cells results in spontaneous tubule formation independent of HGF. We presently are using this system to examine functional relationships between interactors, gene network candidates (described below) and PC1/PC2. We have used shRNA constructs to silence/reduce expression of target genes and then determined how this affects PKD1- and HGF-mediated tubulogenesis. We have thus far successfully silenced endogenous expression of one of the putative interactors in the PKD1-MDCK cell line and control MDCK cell line and found that the PKD1-MDCK cells no longer spontaneously formed tubules, though the cells of both cell lines retained their ability to form tubules in response to HGF. This finding suggests that either the two factors (PC1, HGF) induce tubule formation via different mechanisms or that HGF acts downstream of PC1 in this model system. It also suggests that the interaction between PC1 and its partner is functionally relevant. We are presently evaluating the other candidates using the same approach. As noted last year, we have also been seeking to develop alternative cell model systems for studying PKD protein function. We have successfully generated isogenic cell lines both with and without PC1 and compared their functional qualities. Surprisingly, we have found that the cells have remarkably similar properties, independent of PC1 genotype. We have compared morphogenic activity, metabolic status using the Seahorse system, and growth rates. These studies are ongoing, but the lack of an obvious phenotype has thus far limited their utility for study of PC1-regulated pathways. In parallel, we also had obtained a published cell line from an outside group and were so far unable to reproduce the reported phenotypes. We suspect the paucity of reports using cell culture models to study PC1 function may in part reflect the difficulty of establishing a useful cell culture read-out. 2. PKD mouse models: As noted last year, we had completed a comprehensive microarray analysis of the early onset PKD model and published the results in PLoS Genetics. In follow-up to that work, we have now completed a similar comprehensive microarray analysis of the inducible, late-onset model of disease. For each mouse, we have histology, kidney/body weight ratios, and gene array data. We have characterized the temporal course of disease and determined that there is a prolonged period of normal structure followed by an abrupt transition to a cystic state. We have identified sets of differentially expressed genes that precede frank cystic disease and have examined network structure using a variety of bioinformatic tools. We are presently testing for functional relationships between Pkd1 and top targets using in vivo and in vitro approaches. For the in vitro approaches, we are using the PKD1-MDCK cell system described in the preceding section. We have continued to characterize the Pkhd1 mouse lines that we had previously generated. In one line of investigation, we are assessing the role of a putative candidate pathway in the pathogenesis of the disease. We also are working with our collaborator to complete a manuscript describing the novel Pkhd1 mouse line that we had described in last years annual report. This line has an HA-epitope tag knocked into exon 67, the c-terminal end of the protein, along with lox P sites flanking the exon to allow conditional removal. We have used this model to show that most of the C-terminus is not essential for the genes function in mice. We presently are using the HA-tag to study protein-protein interactions in vivo. One final feature of the mouse line is that the lox P sites had been introduced in the same orientation as the lox P sites in the line that we had previously generated. In the prior model, the lox P sites flank exons 3 and 4. Because the Pkhd1 gene is so large (>500kb), we reasoned that we could use meiotic recombination to generate an allele that had lox P sites on both the extreme 5 and 3 ends of the gene, thereby allowing deletion of virtually all of the coding sequence. We have now successfully generated the desired recombinants and are assessing the effects of complete deletion.
Showing the most recent 10 out of 33 publications
