Mutations in WT1 result in different developmental abnormalities, among which is Wilms' tumor. Many different isoforms of the protein have been described, with different functions for at least some of them. However, many of the studies describing these differences have been in vitro experiments using artificial (over expression) situations. In this project we will analyze the phenotypical effects of limiting the number of different isoforms a cell can express through a genetically approach. Our main focus will be on two alternative splicing events in the gene, determining the use of exon 5 and the use of three additional amino acids (KTS), since disturbance of the ratio between expression levels of these variants is a potential inactivation mechanism of WT1 in malignancies. To study the different functions of these variants in a physiological context we will generate a high-efficiency system to manipulate the Wt1 gene in mouse ES cells . We intend to use a RMCF (recombines-mediated cassette exchange) approach, in which F1p-mediated recombination is used to transfer variant Wt1 genes from a BAC to the endogenous Wt1 locus. This system will be used to generate mouse models only expressing Wt1 variants with or without exon 5 for in vivo analysis. Additionally, variants with the four possible combinations of the exon 5 and KTS splice variants will be generated for extensive in vitro analysis in a molecular, biochemical and cell biological context.