Functional genomics strategies with transposons in rice

Rice is a major staple food crop and a recognizedmonocotylenedousmodel plant from which gene function discovery is projected to contribute to improvements in a variety of cereals like wheat and maize. The recent release of rough drafts of the rice genome sequence for public research provides a vast resource of gene sequences whose functions need to be determined by reverse genetics methods.Characterisation of a mutant phenotype is one of the most promising approaches to link gene to function. Based on this assumption, mutagenesis with transposable elements was chosen as a strategy in the research described in this thesis to address gene function in rice (Chapter 2). The well characterized maize Ac/Ds and En/Spm transposon systems were employed asinsertionalmutagens based on their known ability to transpose inheterologousspecies. Transposon insertions can cause a knockout mutation by blocking the correct expression of a gene, which may result in a mutant phenotype. The mutant gene, thus "tagged" by the inserted transposon, can then be isolated by recovering the DNA flanking the insert and lead to the isolation of the wild-type gene. Constructs for knockout mutagenesis were generated which employed the autonomous Ac element and were tested in rice japonica (Chapter 3) andindica(Chapter 4) varieties. However, the utility of knockout mutations is limited, as the majority of them display no obvious phenotype. This may be due to functional redundancy, where one or more other genes can substitute for the same function, to subtle and conditional phenotypes, or to early lethality caused by the mutation. Gene detection strategies have therefore been developed in addition to classical knockout mutagenesis to address the function of genes that do not directly reveal an obvious phenotype when disrupted (Chapter 5). To utilize gene detection, japonica rice was transformed with advanced two-component enhancer trap vectors, consisting of a mobile transposon element ( Ds or I/dSpm ), and the corresponding stabletransposase( Ac or En/Spm ) source under control of theCaMV35S promoter. The mobile transposons contained in this case a GUS marker gene driven by aCaMV35S minimal promoter that could display the pattern of expression of the adjacent trapped gene and thus provide a clue for its function (Chapters 6 and 7). A large number of rice transformants were produced to test the activity of the different transposon constructs, with the final aim of identifying optimal "starter" lines for the development of tagging populations. Among the factors evaluated were the propensity for continuous transposition through successive generations, the ability to generate large numbers of independent inserts in progeny plants and the target-site specificity of insertion. The usefulness of the selectable markers incorporated in the constructs was also assessed.The results revealed high mobility of the Ac/Ds system in rice (Chapter 3, 4 and 6), although inactivation of Ds was observed in later generations (from T 2 onward). Nevertheless, the high frequency of independent transposition demonstrated to occur in early generations (T 0 and T 1 ) enabled the production of T 2 and T 3 lines with independent "stabilised„ insertions, which can be used directly for reverse genetics screenings without further need for selection against thetransposasesource. The autonomous Ac transposon, in contrast, does not seem to lose mobility and was shown to efficiently transpose in japonica andindicagenotypes as well, supporting its further use in the establishment of a tagging system in this economically important subspecies. Both Ac and Ds displayed amplification of copy number, which enabled the generation of lines containing multiple transposons. Pilot sequencing of genomic sites flanking the Ac and Ds inserts revealed a preferential insertion of these transposons into genes or gene-rich regions and confirmed their tendency to transpose to linked sites, which makes them suitable for targeted tagging. Preliminary testing of the Ac/Ds enhancer trap lines for their ability to function as "detectors" of gene activity, revealed a low frequency of GUS staining patterns in somatic sectors. More thorough screenings are currently under way to fully evaluate the functionality of the system.In contrast to Ac/Ds , the En/Spm system displayed a surprisingly low transposition activity in rice (Chapter 7), restricted to somatic events that were not transmitted to the next generation, in spite of being a well-established mutagenic system inheterologousdicotyledonous species such as Arabidopsis. Transcription analysis of the En/Spm maize element in rice revealed that correct splicing of the element occurs but is not sufficient for transposition ability. Rather, the relative amounts in which the differenttransposaseproducts necessary for transposition are produced seems to be critical and influenced by host factors. In addition, transposition efficiency might be further reduced by the lack of essentialcis-required sequences in the modified I/dSpm version used in this study, although similar constructs were successfully employed in Arabidopsis. Eventually, cross-talk with related endogenous transposable elements may affect the mobility of the maize transposon in rice. Indeed, an interaction of the maizetransposaseswith a rice En/Spm -homologous element was revealed, resulting in the specific suppression of an alternative transcript in the latter (Chapter 8). This finding demonstrates that interference is possible and trans-activation potentially could occur among elements belonging to the same transposon family in different species.Based on the results of these analyses, a core collection of knockout and gene detection Ac/Ds lines with active transposition could be selected as a basis for developing populations for (forward and) reverse genetic screenings. The propagation of lines containing multiple transposons and the preferential insertion into gene-rich regions will help reduce the number of plants that would have to be produced in order to saturate the genome with insertions. At present, over 10,000 stabilised T 2Ac/Ds transposon lines are being analyzed in 5 EU laboratories for transposon flanking sequences that by comparison to the complete and annotated rice sequences will reveal tagged genes of interest that can be used for reverse genetics.