Genomic diversity and introgression in O. sativa reveal the impact of domestication and breeding on the rice genome

Abstract

Background: The domestication of Asian rice (Oryza sativa) was a complex process punctuated by episodes of introgressive hybridization among and between subpopulations. Deep genetic divergence between the two main varietal groups (Indica and Japonica) suggests domestication from at least two distinct wild populations. However, genetic uniformity surrounding key domestication genes across divergent subpopulations suggests cultural exchange of genetic material among ancient farmers.

Methodology/principal findings: In this study, we utilize a novel 1,536 SNP panel genotyped across 395 diverse accessions of O. sativa to study genome-wide patterns of polymorphism, to characterize population structure, and to infer the introgression history of domesticated Asian rice. Our population structure analyses support the existence of five major subpopulations (indica, aus, tropical japonica, temperate japonica and GroupV) consistent with previous analyses. Our introgression analysis shows that most accessions exhibit some degree of admixture, with many individuals within a population sharing the same introgressed segment due to artificial selection. Admixture mapping and association analysis of amylose content and grain length illustrate the potential for dissecting the genetic basis of complex traits in domesticated plant populations.

Conclusions/significance: Genes in these regions control a myriad of traits including plant stature, blast resistance, and amylose content. These analyses highlight the power of population genomics in agricultural systems to identify functionally important regions of the genome and to decipher the role of human-directed breeding in refashioning the genomes of a domesticated species.

Figure 1. Population structure in O. sativa estimated from the GoldenGate SNP set.

(A) Population structure estimate from STRUCTURE output for K = 2 to K = 6. (B) Phylogenetic tree. The branch tips of the tree are colored according to the subpopulation assignment in (A) when K = 5.

Figure 2. Average Introgression component in subpopulations.

Each panel represents one chromosome and each SNP is represented as one point in the figure. Here we illustrate introgression from indica (IND) into tropical japonica (TRJ) (red) and from temperate japonica (TEJ) into indica (IND) (dark blue). The dashed lines are the 5^th percentile of all the SNPs for each introgression with the same color. Four genes SD1, GS3, Waxy, Pi-ta located in or near peak regions are indicated along the genome.

Figure 3. Regions of introgression from indica into tropical japonica.

SNP positions are shown as black vertical bars across the bottom; vertical grey lines indicate chromosomes; horizontal grey lines indicate accessions; introgressed regions (defined by ≥5 SNPs) shown in red. The height of the red regions corresponds to the probability of introgression, with a maximum value of 1. Only tropical japonica and admixed accessions (where indica component is less than 25%) are plotted in the figure. SD1 and Pi-ta positions shown as vertical dashed lines. Accession IDs are colored in light blue for TRJ and black for admixed accessions.

Figure 4. Regions of introgression from temperate japonica into indica aligned with admixture mapping and association mapping p-values for amylose content.

Genome position of SNPs and introgressions indicated across the bottom; vertical grey lines indicate chromosomes; position of Waxy gene on chromosome 6 shown as vertical dashed line. (A) Horizontal grey lines indicate accessions; blue-colored regions represent introgressions (defined by ≥5 SNPs) from temperate japonica into indica; variety names (on left) colored dark blue indicate indica accessions carrying a waxy allele introgressed from temperate japonica; those without waxy introgression indicated in red. (B) Admixture mapping p-values for amylose content using the temperate japonica component in the admixed subpopulation. (C) Association mapping p-values for amylose content in all accessions using mixed model approach. Horizontal dotted lines in both (B) and (C) represent significance values of 0.05 after Bonferroni correction.

Figure 5. Admixture mapping and association mapping p-values for grain length.

(A) Admixture mapping p-values for grain length using the tropical japonica component in the admixed subpopulation. (B) Association mapping p-values for grain length in all accessions using the subpopulation component matrix Q (K = 5) as cofactors in the model. Horizontal dotted lines represent significance value of 0.05 after Bonferroni correction. Chromosomes are separated by vertical grey lines; GS3 gene on chromosome 3 shown as vertical dashed line.