Generation of selectable marker-free transgenic rice using double right-border (DRB) binary vectors

Abstract

Currently employed transformation systems require selectable marker genes encoding antibiotic or herbicide resistance, along with the gene of interest (GOI), to select transformed cells from among a large population of untransformed cells. The continued presence of these selectable markers, especially in food crops such as rice (Oryza sativa L.), is of increasing public concern. Techniques based on DNA recombination and Agrobacterium-mediated co-transformation with two binary vectors in a single or two different Agrobacterium strains, or with super-binary vectors carrying two sets of T-DNA border sequences (twin T-DNA vectors), have been employed by researchers to produce selectable marker-free (SMF) transgenic progeny. We have developed a double right-border (DRB) binary vector carrying two copies of T-DNA right-border (RB) sequences flanking a selectable marker gene, followed by a GOI and one copy of the left border sequence. Two types of T-DNA inserts, one initiated from the first RB containing both the selectable gene and the GOI, and the other from the second RB containing only the GOI, were expected to be produced and integrated into the genome. In the subsequent generation, these inserts could segregate away from each other, allowing the selection of the progeny with only the GOI. We tested this vector using two selectable marker genes and successfully obtained progeny plants in which the second selectable marker gene segregated away from the first. Using the DRB binary vector system, we recovered SMF transgenic lines containing a rice ragged stunt virus (RRSV)-derived synthetic resistance gene in the rice cultivars Jarrah and Xiu Shui. Approximately 36–64% of the primary transformants of these cultivars yielded SMF progeny. Among SMF Jarrah transgenic progeny <50% of plants contained the RRSV transgene. Thus, we have developed an efficient vector for producing SMF plants that allows straightforward cloning of any GOIs in comparison with the published ‘twin T-DNA’ vectors.