2015 年 32 巻 2 号 p. 169-173
Although gene targeting (GT) is a useful technology for precise mutagenesis of target sequences, its frequency is quite low. Establishing experimental procedures using a model system will enable us to improve this frequency and apply to GT as a universal system. Here, we propose a convenient system with which to evaluate the frequency of site-directed mutagenesis via GT using a positive-negative selection method. We constructed a GT vector harboring a partial rice acetolactate synthase gene with mutations conferring bispyribac sodium (BS) tolerance and a gene conferring blasticidin-S tolerance as a positive selection marker. In addition, diphtheria toxin A subunit gene was used as a negative selection marker to enrich GT cells. We regenerated GT candidate plants successfully at a frequency of 2.1 putative GT events/gram Agrobacterium-infected callus following dual selection on BS and blasticidin-S. Moreover, molecular analyses confirmed that GT events occurred in >80% of regenerated plants. Existing GT methods using positive-negative selection require that true putative GT events be verified by molecular analysis because of the growth of large numbers of cells in which partial GT vectors containing positive selection marker cassettes, but lacking the negative selection marker, have inserted at random loci. In contrast, the present method with dual selection on both BS and blasticidin-S allows direct enrichment of GT cells at high frequency without the need for further extensive molecular screening.
