VERTEBRATE GENE TARGETING--A MODEL SYSTEM

点击翻译按钮获取中文摘要

When DNA molecules are introduced into mammalian cells or into the nuclei of Xenopus laevis oocytes, the major pathway available for homologous recombination of this exogenous DNA is a nonconservative one. In the case of Xenopus oocytes, it has been possible to describe this pathway in some detail by characterizing recombination intermediates and by examining the effects of injecting modified substrates. Because the mechanism of recombination involves resection of the introduced DNA by a 5'->3' exonuclease prior to annealing of complementary strands, it is essential that the substrate DNA be linear. Extrapolation of this model to gene targeting protocols in mammalian cells suggests that the reason that homologous interactions with chromosomal sites are so infrequent is that the target is not broken. Experiments are proposed to demonstrate that circular DNAs, which are inert for recombination in oocytes, become activated by cleavage in vivo. Then approaches will be explored that will allow the introduction of a double-strand break into circular DNA in oocytes at a specific, but arbitrarily chosen site. These reagents will be based on homologous interactions between single-stranded oligonucleotides and double-stranded DNA promoted by E. coli RecA protein. Light-activated DNA modifying agents will be tethered to such oligonucleotides and introduced into oocytes that already contain a circular target DNA. The prototype reagent will carry a psoralen, and the fate of psoralen crosslinks will be explored in detail to determine whether they are recombinagenic. In addition to attempting to achieve targeted cleavages, other parameters of the oocyte recombination reaction that may be relevant to gene manipulation experiments will be investigated. These include effects of the length and quality of sequence homology on the efficiency of recombination and the fate of sequence mismatches in recombination products. Finally, critical experiments will be performed in Xenopus eggs, which have a more complex range of recombination capabilities, similar to mammalian cells. If successful, these studies will permit targeted manipulations of mammalian genomes with high efficiency. This would be useful for genetic investigation of gene function in experimental animals, like mice, and would improve the feasibility of genetic intervention in the treatment of inherited diseases in humans.