Integration of Forward and Reverse Genomic Approaches to Identify and Clone QTLs for Agronomic Traits in Rice

QTL cloning of genes controlling rice agronomic traits has elucidated the contributions of such genes to phenotypes. However, we face difficulties in the isolation of genes with relatively small effects on phenotypes. Next-generation sequencing technology allows us to explore the allelic differences among diverse populations by rapid full-genome sequencing. The concomitant use of information from QTL mapping and genome sequencing could facilitate the allele mining of natural genetic variants affecting agronomic traits in rice. To this end, we have been developing chromosomal segment substitution lines (CSSLs) by using diverse rice accessions and high-depth short-read re-sequencing of diverse donor cultivars. Phenotyping of CSSLs provides us with accurate information on the genetic effects of a single chromosomal segment and the position of the candidate genes. Sequence comparisons of relevant small regions between multiple donor cultivars and a common recurrent parent subsequently reduce the number of candidate genes. Additional information about the candidate genes, such as functional annotations and expression profiles (for example, expression levels at different stages), also accelerates allele mining. The combination of QTL mapping for heading date (Hd) by using CSSLs and sequence information can serve as an example of the efficient allele mining of genes associated with Hd. Additional information has helped us to reduce the number of candidates. To verify the phenotypic effects of sequence changes in a few genes, the development of overexpressing or RNAi lines and the selection of gene disruption lines for candidate genes is underway. The effective strategies for allele mining will be discussed.