Search

The Green Revolution (GR-I) included worldwide adoption of semi-dwarf rice cultivars (SRCs) with mutant alleles at GA20ox2 or SD1 encoding gibberellin 20-oxidase. Two series of experiments were conducted to characterize the pleiotropic effects of SD1 and its relationships with large numbers of QTLs affecting rice growth, development and productivity. The pleiotropic effects of SD1 in the IR64 genetic background for increased height, root length/mass and grain weight, and for reduced spikelet fertility and delayed heading were first demonstrated using large populations derived from near isogenic IR64 lines of SD1. In the second set of experiments, QTLs controlling nine growth and yield traits were characterized using a new molecular quantitative genetics model and the phenotypic data of the well-known IR64/Azucena DH population evaluated across 11 environments, which revealed three genetic systems: the SD1-mediated, SD1-repressed and SD1-independent pathways that control rice growth, development and productivity. The SD1-mediated system comprised 43 functional genetic units (FGUs) controlled by GA. The SD1-repressed system was the alternative one comprising 38 FGUs that were only expressed in the mutant sd1 backgrounds. The SD1-independent one comprised 64 FGUs that were independent of SD1. GR-I resulted from the overall differences between the former two systems in the three aspects: (1) trait/environment-specific contributions; (2) distribution of favorable alleles for increased productivity in the parents; and (3) different responses to (fertilizer) inputs. Our results suggest that at 71.4 % of the detected loci, a QTL resulted from the difference between a functional allele and a loss-of-function mutant, whereas at the remaining 28.6 % of loci, from two functional alleles with differentiated effects. Our results suggest two general strategies to achieve GR-II (1) by further exploiting the genetic potential of the SD1-repressed and SD1-independent pathways and (2) by restoring the SD1-mediated pathways, or ‘back to the nature’ to fully exploit the genetic diversity of those loci in the SD1-mediated pathways which are virtually inaccessible to most rice-breeding programs worldwide that are exclusively based on sd1.

Single nucleotide polymorphisms (SNPs) are the most abundant DNA markers in plant genomes. In this study, based on 54,465 SNPs between the genomes of two Indica varieties, Minghui 63 (MH63) and Zhenshan 97 (ZS97) and additional 20,705 SNPs between the MH63 and Nipponbare genomes, we identified and confirmed 1,633 well-distributed SNPs by PCR and Sanger sequencing. From these, a set of 372 SNPs were further selected to analyze the patterns of genetic diversity in 300 representative rice inbred lines from 22 rice growing countries worldwide. Using this set of SNPs, we were able to uncover the well-known Indica-Japonica subspecific differentiation and geographic differentiations within Indica and Japonica. Furthermore, our SNP results revealed some common and contrasting patterns of the haplotype diversity along different rice chromosomes in the Indica and Japonica accessions, which suggest different evolutionary forces possibly acting in specific regions of the rice genome during domestication and evolution of rice. Our results demonstrated that this set of SNPs can be used as anchor SNPs for large scale genotyping in rice molecular breeding research involving Indica-Japonica and Indica-Indica crosses.

To genetically dissect drought resistance associated with japonica upland rice, we evaluated a doubled haploid (DH) population from a cross between two japonica cultivars for seven root traits under three different growing conditions (upland, lowland and upland in PVC pipe). The traits included basal root thickness (BRT), total root number (RN), maximum root length (MRL), root fresh weight (RFW), root dry weight (RDW), ratio of root fresh weight to shoot fresh weight (RFW/SFW) and ratio of root dry weight to shoot dry weight (RDW/SDW). The BRT was significantly correlated with the index of drought resistance, which was defined as the ratio of yield under the stress of the upland condition to that under the normal lowland condition. A complete genetic linkage map with 165 molecular markers covering 1,535 cM was constructed. Seven additive quantitative trait loci (QTLs) and 15 pairs of epistatic loci for BRT and RN were identified under upland and lowland conditions, and 12 additive QTLs and 17 pairs of epistatic QTLs for BRT, RN, MRL, RFW, RFW/SFW and RDW/SDW were identified under the PVC pipe condition. Four additive QTLs and one pair of epistatic QTLs controlling IDR were also found. These QTLs individually explained up to 25.6% of the phenotypic variance. QTL x environment (Q x E) interactions were detected for all root traits, and the contributions of these interactions ranged from 1.1% to 19.9%. Five co-localized QTLs controlling RFW and RDW, RFW/SFW, RDW/SDW and IDR, BRT and RN, RN, MRL and IDR were found. Four types of QTLs governing BRT and RN were classified by their detection in the upland and lowland conditions. Some common QTLs for root traits across different backgrounds were also revealed. These co-localized QTLs and common QTLs will facilitate marker-assisted selection for root traits in rice breeding programs.