Efforts to obtain the complete DNA sequences of mouse and man will yield an enormous amount of information regarding the content of mammalian genomes. However, the most profound biological insights will come only upon linkage of sequences with a description of the role of each gene in the organism. In this application, we describe a methodology that permits the generation of random insertional mutations in the mouse genome. Inherent in the design of the proposed system are methods for detecting the presence of an insertion event by visual inspection. In fact, coat color can be used to distinguish whether animals are heterozygous or homozygous for the insertion event, allowing genetic crosses to be followed without the need for molecular analysis. We have used a previous incarnation of this general approach to generate and analyze approximately 250 strains of randomly mutated mice. Of these, approximately 10% showed overtly interesting phenotypes. Several displayed craniofacial defects such as cleft palate. However, the current design of the approach has several deficiencies that ultimately result in it taking several years to clone genes responsible for observed phenotypes. In this application, we propose to create a vastly improved version of the system that incorporates the successful aspects of the current design but that provides simplified methods for isolating the genes that have been disrupted by insertion. This will be accomplished through the use of recombinant retroviruses that are packaged in the sertoli cells of the testes and that are transmitted as new integration sites only through the male germ line. This methodology will initially be applied to the generation of mouse models with craniofacial defects; however, if successful, the approach will also find much broader application.
This proposal describes the creation of a methodology that can be used to create libraries of random insertional mutations in mouse and provides ways to screen those libraries for mice having specific desired phenotypes. Such a library could be used to identify mouse models of disease, the use of which would be widespread in industry.
