The use of particle gun for the production of marker-free plants is scant in published literature. Perhaps this is a reflection of the widely held notion that the events generated through bombardment tend to have multiple copies of transgenes, usually integrated at a single locus, features which precludes segregating away the selectable marker gene. However, our previous studies have shown that single-copy integrants are obtained at a high frequency if limited quantity of DNA is used for bombardment. Also, the concatemerized insertion of transgenes has been demonstrated to be greatly reduced if "cassette DNA" is employed in place of whole plasmid DNA for bombardment. Based on the above findings, in the present study the feasibility of co-bombardment was evaluated for the production of marker-free plants in corn, employing a combination of limited quantity DNA and cassette DNA approaches for bombardment. Transgenic events were generated after co-bombardment of a selectable marker cassette containing the nptII gene (2.5 ng per shot) and a GUS gene cassette (15 ng per shot). Among these events single-copy integrants for nptII gene occurred at an average frequency of 68% within which the co-expression frequency of GUS and nptII genes ranged from 41% to 80%. Marker-free corn plants could be identified from the progeny of 28 out of the 103 R0 co-expressing events screened. The results demonstrate that by using cassette DNA and low quantities of DNA for bombardment, marker-free plants are produced at efficiencies comparable to that of Agrobacterium-based co-transformation methods.

Some late embryogeny abundant (LEA) proteins, which are developmentally regulated in embryos, are also known to be expressed in meophytic tissues in response to osmotic stress. Here we report the extent of genetic variability in the level of expression of lea2 and lea3, under stress, in fingermillet and rice seedlings. In both species, the expression of lea genes was seen in the mesophytic tissue in response to salinity, partial dehydration and abscisic acid. Tolerant genotypes exhibited higher expression of rab16A and M3 that code for LEA2 proteins, than susceptible genotypes. A novel approach, that of raising antibodies against the conserved peptides of these proteins was used to study genetic variability in LEA protein levels. Since stress proteins are known to be expressed in response to mild, non-lethal induction-stress (Uma, Prasad and Udayakumar, Annals of Botany 76 : 43-49, 1995), we developed an optimum induction protocol for salinity stress in rice and fingermillet. We studied the quantitative differences in expression of these proteins by western blot and ELISA techniques in different genotypes. A positive correlation was found between LEA2 and LEA3 protein levels and the growth of seedlings during stress and recovery in both rice and fingermillet, indicating a possible relevance of these proteins in stress tolerance.

Identification of an appropriate selection agent and its corresponding selectable marker gene is one of the first steps in establishing a transformation protocol for a given plant species. As the promoter controls expression level of the genes, the promoter driving the selectable marker gene can affect transformation. However, investigations into the direct effect of promoters driving selectable marker on transformation are lacking in the literature though many reports of relative strengths of promoters driving reporter genes like GUS or CAT or GFP are available. In the present study, we have compared rice Actin1 and CaMV.35S (commonly used promoters in monocotyledonous plant transformation) promoters driving nptII for their effectiveness in paromomycin selection of transgenic corn events. To enable statistically meaningful analysis of the results, a large sample size of nearly 5,000 immature embryos (explants) was employed producing approximately 1,250 independent events from each of the two constructs in four independent experiments. The rate of appearance of resistant calli and percentage of resistant calli recovered was higher with P-Os.Actin1/nptII/nos3' as compared to P-CaMV.35S/nptII/nos3' in all four experiments. There was no appreciable difference either in the frequency of plant regeneration or in the morphological characteristics of plants recovered from the two constructs. Although the escape rate trended lower with P-Os.Actin1 as compared to P-CaMV.35S, the recovery of low copy events was significantly higher with P-CaMV.35S. The higher transformation frequency with P-Os.Actin1 could be related to the strength of this promoter as compared to P-CaMV.35S in the explants and/or calli. Based on these results, we infer that the promoter driving the selectable marker is an important factor to be considered while establishing a high throughput transformation protocol as it could not only influence the transformation frequency but also the copy number of the transgene in the recovered transgenics.