Efficient co-transformation of Nicotiana tabacum by two independent T-DNAs, the effect of T-DNA size and implications for genetic separation.

Abstract

The co-transformation of a single plant genome with two independent T-DNA regions provides opportunities for genetic separation in subsequent generations. In an effective strategy, co-delivery events must form a high proportion of the total transformed population. In this study, using the model plant species tobacco (Nicotiana tabacum), it was shown that the frequency of co-transformation within a given To population could be as high as 100% and this was found to be dependent, at least in part, on designing the plasmid vectors so that the kbp size of the first selected T-DNA region was >2-fold that of the designated T-DNA region for co-transfer. Overall, 40-50% of To lines demonstrated the capacity for segregational separation of co-transformed T-DNA regions. Hence, the estimate of the required number of total transformants for such an independent strategy may seem to be as little as 2-fold that for a conventional, single T-DNA strategy, but we strongly temper such estimates with indications that high co-transformation frequencies may be associated with a higher incidence of linkage. In this co-transformation study we used a single (Agrobacterium) strain system in which a single binary plasmid contained either two or three T-DNA regions, each with a selectable marker. This arrangement could reveal that 'read-through' events within the Agrobacterium cells, resulting in the co-transfer of adjacent T-DNA regions as a single linked unit, accounted for up to 20% of co-transformed plant lines. Such read-through co-delivery appeared to be more frequent from the 'supervirulent' EHA101 A. tumefaciens strain, compared to the 'ordinary' LBA4404 strain. By using the binary plasmid with three selectable T-DNA regions, we have been able to consider the frequency of co-integration of a third independent T-DNA within a T0 subpopulation of co-transformants. This was found to be higher than expected. These observations were applied to the co-transfer of (unwanted) plasmid backbone sequences and showed that screening against such sequences may add a significant factor in achieving the desired, final genotype.