The large-scale Munich ENU-mouse-mutagenesis screen

Within the next few years the complete sequence of the human genome will be available (Schuler et al. 1996), and the postgenome era will start with the systematic analysis of gene function and its role in human pathogenesis and disease. Characterization of spontaneous and induced mutants, the analysis of transgenic and gene-targeted phenotypes in animals, i.e., fruitfly, zebrafish, or rodents, are common tools to obtain insight into the biological function of genes. With respect to the genetics and pathogenesis of human diseases, animal models are essential for further investigations; and in particular, the mouse has had a major role as a model system owing to the similarity of its genome, developmental and biochemical pathways, and physiology to humans. Two different strategies can be attempted for the systematic production of mutant phenotypes in the mouse: the gene-driven and the phenotype-driven approach. The gene-driven approach is based on the mouse embryonic stem cell technology, in which mouse mutants can be generated for any targeted mutation engineered through homologous recombination (Thomas and Capecchi 1987; Ramirez-Solis et al. 1993). The limiting factor is the production of the targeting construct, which involves an engineered DNA fragment containing the desired mutation. To date, automation of this technique on a large scale could not be carried out because of the heterogeneity of gene structures. In most cases, the engineered construct replaces the targeted DNA region of the wildtype gene and interrupts the gene’s function, which results in a null allele or a ‘knocked-out’ gene. Another gene-driven approach is the gene trap strategy (Evans et al. 1997). In this approach, a selectable insert is introduced into the ES cell DNA, and by building a large number of cell lines carrying disrupted units, an ES cell mutation bank is established for generating mouse mutants (Wiles et al. 2000). The disadvantage of the gene-driven strategy is the production of mostly null alleles, which often do not reveal all biological functions of a gene. With respect to clinically relevant diseases in humans, the focus is on the result of a partial, but not complete loss of gene function. Thus, for a genetic analysis of inherited diseases, it is necessary to generate multiple alleles of a single gene, which results in hypomorphs, alleles of different strength or gain-of-function alleles. Complementary to such a gene-driven approach, in which mutants are produced for those genes that are already known, the phenotype-driven approach identifies new genes, gene products, and their relevant biological pathways by recovering mouse mutants with a novel phenotype (Brown and Peters 1996). Random mutagenesis from ionizing radiation induces large deletions, chemical mutagenesis with ENU (ethylnitrosourea) (Russell et al. 1979, 1990; Russell 1982; Peters 1985; Dove 1987; Bode et al. 1988; Favor et al. 1990a, 1990b; Rinchik 1991), causes point mutations (Popp et al. 1983; Harbach et al. 1992), while chlorambucil induces small deletions. These approaches have a long tradition in classical genetics and have generated a large number of mutant phenotypes. The responsible genes are then identified through positional cloning or other strategies. The validity and success of these approaches were demonstrated in the course of the genetic and molecular dissection of the pathway that set up the Drosophila body pattern (Nüsslein-Vollhard and Wieschaus 1980; Ashburner 1989). The main interest in such a phenotype-driven strategy is the establishment of appropriate procedures to assess the mutant phenotypes of interest, to obtain animal models of human diseases and insight into the gene functions. To date, only a few protocols have met these demanding requirements (Rogers et al. 1997). A screening and phenotyping protocol for pathophysiological abnormalities—the Munich protocol—has been established in the last 3 years in the ENU-mouse-mutagenesis project of the German Human Genome Project (Hrabé de Angelis and Balling 1998) to assess mutant phenotypes for specific, postnatal abnormalities comprising congenital malformations, clinical chemical, biochemical, hematological, immunological defects and complex traits such as allergies and behavior.