High throughput T‐DNA insertion mutagenesis in rice: a first step towards <i>in silico</i> reverse genetics

A library of 29,482 T-DNA enhancer trap lines has been generated in rice cv. Nipponbare. The regions flanking the T-DNA left border from the first 12,707 primary transformants were systematically isolated by adapter anchor PCR and sequenced. A survey of the 7480 genomic sequences larger than 30 bp (average length 250 bp), representing 56.4% of the total readable sequences and matching the rice bacterial artificial chromosome/phage artificial chromosome (BAC/PAC) sequences assembled in pseudomolecules allowed the assigning of 6645 (88.8%) T-DNA insertion sites to at least one position in the rice genome of cv. Nipponbare. T-DNA insertions appear to be rather randomly distributed over the 12 rice chromosomes, with a slightly higher insertion frequency in chromosomes 1, 2, 3 and 6. The distribution of 723 independent T-DNA insertions along the chromosome 1 pseudomolecule did not differ significantly from that of the predicted coding sequences in exhibiting a lower insertion density around the centromere region and a higher density in the subtelomeric regions where the gene density is higher. Further establishment of density graphs of T-DNA inserts along the recently released 12 rice pseudomolecules confirmed this non-uniform chromosome distribution. T-DNA appeared less prone to hot spots and cold spots of integration when compared with those revealed by a concurrent assignment of the Tos17 retrotransposon flanking sequences deposited in the National Center for Biotechnology Information (NCBI). T-DNA inserts rarely integrated into repetitive sequences. Based on the predicted gene annotation of chromosome 1, preferential insertion within the first 250 bp from the putative ATG start codon has been observed. Using 4 kb of sequences surrounding the insertion points, 62% of the sequences showed significant similarity to gene encoding known proteins (E-value < 1.00 e(-05)). To illustrate the in silico reverse genetic approach, identification of 83 T-DNA insertions within genes coding for transcription factors (TF) is presented. Based both on the estimated number of members of several large TF gene families (e.g. Myb, WRKY, HD-ZIP, Zinc-finger) and on the frequency of insertions in chromosome 1 predicted genes, we could extrapolate that 7-10% of the rice gene complement is already tagged by T-DNA insertion in the 6116 independent transformant population. This large resource is of high significance while assisting studies unravelling gene function in rice and cereals, notably through in silico reverse genetics.