Somaclonal variation does not preclude the use of rice transformants for genetic screening

Rice (Oryza sativa) is one of the world's most important crops. Rice researchers make extensive use of insertional mutants for the study of gene function. Approximately half a million flanking sequence tags from rice insertional mutant libraries are publicly available. However, the relationship between genotype and phenotype is very weak. Transgenic plant assays have been used frequently for complementation, overexpression or antisense analysis, but sequence changes caused by callus growth, Agrobacterium incubation medium, virulence genes, transformation and selection conditions are unknown. We used high-throughput sequencing of DNA from rice lines derived from Tainung 67 to analyze non-transformed and transgenic rice plants for mutations caused by these parameters. For comparison, we also analyzed sequence changes for two additional rice varieties and four T-DNA tagged transformants from the Taiwan Rice Insertional Mutant resource. We identified single-nucleotide polymorphisms, small indels, large deletions, chromosome doubling and chromosome translocations in these lines. Using standard rice regeneration/transformation procedures, the mutation rates of regenerants and transformants were relatively low, with no significant differences among eight tested treatments in the Tainung 67 background and in the cultivars Taikeng 9 and IR64. Thus, we could not conclusively detect sequence changes resulting from Agrobacterium-mediated transformation in addition to those caused by tissue culture-induced somaclonal variation. However, the mutation frequencies within the two publically available tagged mutant populations, including TRIM transformants or Tos17 lines, were about 10-fold higher than the frequency of standard transformants, probably because mass production of embryogenic calli and longer callus growth periods were required to generate these large libraries. Significance Statement Large-scale insertional mutant libraries in rice have a much lower tagging efficiency than those in Arabidopsis, where the generation of insertional mutants does not rely on tissue culture methods. Here we used next generation sequencing to assess sequence changes in large-scale rice insertional mutant libraries and in more recently generated transformants. The incidence of sequence changes in recently generated transformants was no different than that for somaclonal variation, but the large-scale libraries had 10-fold more sequence changes, perhaps due to the longer tissue culture procedures used to generate these large libraries.