RESUMO
Most cellular processes in an organism depend on functions of expressed sequences. Thus, efficient large-scale functional assignment of expressed sequences is crucial for understanding cellular processes. Towards this goal in plants, we designed a "random antisense cDNA mutagenesis (RAM)" approach. In a pilot experiment, 1,000 transgenic plants of Arabidopsis thaliana (L.) Heynh. (ecotype Wassilevskija) expressing random antisense cDNA(s) were generated from Agrobacterium cultures harboring an Arabidopsis antisense cDNA library. We identified 104 mutant lines from the transgenic pool by visual screening. Genetic analysis suggested that 37% of the mutations were likely due to antisense effects. When the cDNA inserts were isolated from 11 mutant lines by polymerase chain reaction and reintroduced into plants to express the antisense transcripts, the original mutant phenotypes were reproduced in 7 cDNA clones. One of the cDNA clones did not generate a database match to any sequence with known functions, but did have a dramatic effect on the architecture of the inflorescence in the antisense transgenic plants. Through the RAM approach, it should be possible to assign a large number of expressed sequences to known in vivo functions in plants.
