with Chitinase gene Developed Through Biolistic and Agrobacterium-Mediated Transformation

N. Baisakh, K. Datta, N.P. Oliva and S.K. Datta

Plant Breeding, Genetics and Biochemistry Division, International Rice Research Institute, MCPO Box

3127, 1271, Makati City, Philippines

     Efficient genetic manipulation of rice for agronomically important traits has been made possible by the transfer of foreign gene(s) into the genome through either direct DNA transfer such as biolistic or via Agrobacteri urn-mediated transformation. There has been much debate on these two systems in terms of their efficiency and the pattern of integration, mode of inheritance and stable expression of the transgene(s). However, experiments dealing with both the systems keeping all other factors, like target explants, gene(s) of interest, tissue culture cycle, constant are lacking. We developed a large number of transgenic rice plants from four to six week-old mature seed-derived callus of two indica rice cultivars, vaidehi (suitable for deep water situation in eastern India) and Tulsi (adapted to semi deep lowland condition), with an antifungal PR-protein gene chitinase known to confer resistance against sheath blight fungus. The vectors used were pGL235S- chill (Lin et a!. 1995) for bombardment and LBA-pNol (Datta et al. 1999) for Agrobacteri urn-mediated transformation carrying the same chill and hph (as a selectable marker gene) both driven under the control of CaMV 35S promoter. From three experiments, we found the net transformation frequency for both cultivars was higher in Agrobacterium method as compared to biolistic. The Southern analysis of a total of 179 T0 plants of Vaidehi and Tulsi from Agrobacteriwn-mediated transformation showed a single 1.1-kb size band expected for chill transgene (Fig. la), whereas, multiple and rearranged bands apart from the expected size were common in case of primary transgenics obtained from biolistic experiment (Fig. ib). Further molecular analysis for.copy number showed one to three copies of intact chill in Agrobacterium-derived plants while one to multiple (> 10) with different molecular sizes were observed in the transgenics from biolistic method. In the subsequent segregating generation (Ti), chill was inherited in a typical Mendelian fashion with a ratio ~3:1 in the progenies derived from Agrobacterium transformed plants suggesting single locus integration of chill. On the contrary, both Mendelian (where there was single insertion site) as well as non-Mendelian segregation (in cases of multiple and rearranged bands) of chill were observed in the progenies of primary transgenics developed through the biolistic method. Moreover, differrent banding pattern among the progenies of biolistic-denved transgenics confirms the frequent possibility of integration of multiple and rearranged copies of the foreign gene(s) in more than one locus in the rice genome. This contradicts an earlier report of single locus integration of multiple copies of a single gene or multiple genes from biolistic experiment (Kohli et al. 1998). More details of our study would be published elsewhere.