Yield component QTLs identified by genome-wide association mapping validated in a diverse tropical japonica × tropical japonica rice biparental mapping population

Authors: Eizenga, Georgia; RICE, ADAM - University Of Arkansas; Huggins, Trevis; SHAKIBA, EHSAN - University Of Arkansas; Edwards, Jeremy; Jackson, Aaron; Jia, Melissa; ALI, LIAKAT - University Of Arkansas

Submitted to: Crop Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 4/24/2023
Publication Date: 5/6/2023
Citation: Eizenga, G.C., Rice, A., Huggins, T.D., Shakiba, E., Edwards, J., Jackson, A.K., Jia, M.H., Ali, L. 2023. Yield component QTLs identified by genome-wide association mapping validated in a diverse tropical japonica × tropical japonica rice biparental mapping population. Crop Science. https://doi.org/10.1002/csc2.20999.
DOI: https://doi.org/10.1002/csc2.20999

Interpretive Summary: Rice is the staple food for over half of the world’s 7.6 billion people and population growth is predicted to reach 8.6 billion in 2030 and 9.8 billion in 2050. To meet this growing demand for food, it is essential to understand the plant processes that control grain yield, thus expediting breeding efforts to produce more pounds of rice per acre. Rice yield is determined by several factors including number of panicles per plant, number of panicle branches, number of seeds per panicle, seed size (length and width) and seed weight. To better understand the biological processes controlling these traits, a global collection of 400 diverse rice varieties was phenotyped for the aforementioned traits and mapped with DNA markers for identification of genes controlling these traits. To validate these results, two diverse varieties, L-202, which is short statured with a long grain originating from California, USA and Trembese which is tall, with a medium grain originating from Indonesia, were selected from this diverse collection as parents. From a single cross between these two varieties, 211 progenies were advanced eight generations and evaluated for the same yield-related traits. Using DNA markers we identified 40 chromosome regions across both studies where genes controlling yield-related traits are located. Most significant was a genomic region on chromosome 7 which affects the number of seeds per panicle, panicle branching, seed length and seed weight. These results will be used to develop user friendly DNA markers associated with these yield-related traits that can be used by rice breeders to accelerate selection for the desired panicle architecture, seed size (length and width) and seed weight.

Technical Abstract: Cultivated rice (Oryza sativa L.) is comprised of two subspecies, Indica and Japonica that further divide into five subpopulations (indica, aus, the aromatic, temperate japonica and tropical japonica). The Rice Diversity Panel 1 (RDP1) developed for genome-wide association (GWA) studies includes over 400 accessions with representation from the five subpopulations. Initially, the RDP1 was phenotyped for over 25 yield-related traits and GWA mapping conducted with 44,000 SNPs markers which identified SNPs associated with GWA Quantitative Trait Loci (QTL) that were significant within specific subpopulations, as well as, across the entire panel. Some of these SNPs were co-located with known genes for the trait. QTL mapping with biparental recombinant inbred line (RIL) populations is one method to further validate these GWA QTL. More than half of U.S. cultivars are classified as tropical japonica; thus, it was the focus for validating RDP1 GWA QTL. Two phenotypically and genotypically diverse tropical japonica accessions, L-202 a short stature, long grain originating from the California, USA, and Trembese, a tall, medium grain from Indonesia, which had more seeds per panicle compared to L-202, were selected for population development. The L-202 x Trembese RIL population was phenotyped for 16 yield component traits which were measured in RDP1 including six agronomic traits, six panicle architecture traits and four seed traits plus secondary panicle branching, seed weight per plant and seed weight per plant. Genotyping with 1,489 polymorphic markers and subsequent QTL mapping revealed 43 RIL QTL. Recently, the genotyping density of the RDP1 accessions was increased to 4.8 million SNPs through imputation which enables more precise GWA mapping, thus the GWA analyses were updated for GWA QTL validation with the RIL QTL. Examination of the GWA QTL in the RIL QTL regions identified 175 GWA QTL co-located with 36 RIL QTL. From this comparison, the most significant overlap was on chromosome 7 in the 22.56-29.66 Mb interval where the well-known yield related genes, FRIZZY PANICLE (FZP), DENSE AND ERECT PANICLE 2 (DEP2) and GRAIN WIDTH (GW7) are located. In this region, RIL QTL for six panicle architecture traits and four seed traits were co-located with GWA QTL for 15 panicle architecture traits and 21 seed traits. The allele effects confirmed the increased panicle branching, especially secondary branching, and number of seeds per panicle was attributed to Trembese, whereas the longer seed and increased seed weight to the L-202 parent. The RIL QTL with high R2 values (>15%) which were co-located with GWA QTL are targets for developing user friendly SNP markers that can be utilized in cultivar development.