Genetic variation architecture of mitochondrial genome reveals the differentiation in Korean landrace and weedy rice

Abstract

Mitochondrial genome variations have been detected despite the overall conservation of this gene content, which has been valuable for plant population genetics and evolutionary studies. Here, we describe mitochondrial variation architecture and our performance of a phylogenetic dissection of Korean landrace and weedy rice. A total of 4,717 variations across the mitochondrial genome were identified adjunct with 10 wild rice. Genetic diversity assessment revealed that wild rice has higher nucleotide diversity than landrace and/or weedy, and landrace rice has higher diversity than weedy rice. Genetic distance was suggestive of a high level of breeding between landrace and weedy rice, and the landrace showing a closer association with wild rice than weedy rice. Population structure and principal component analyses showed no obvious difference in the genetic backgrounds of landrace and weedy rice in mitochondrial genome level. Phylogenetic, population split, and haplotype network evaluations were suggestive of independent origins of the indica and japonica varieties. The origin of weedy rice is supposed to be more likely from cultivated rice rather than from wild rice in mitochondrial genome level.

Figure 1. Overall variation (SNPs and Indels) distribution across the mt genome.

Genes indicating the genes of the mt genome based on the reference of O. sativa japonica. (a–f) Highlights marked on the circle map revealing the SNP/indel positions. (a) Total variations detected in 70 accessions, (b) Variations in Korean landrace rice, (c) Variations in Korean weedy rice, (d) Variations in the rice of indica type, (e) Variations in japonica type, (f) Variations in wild rice. The unit of the outside distance is kb. The number inside the brackets indicated the accession numbers of each subgroup. In case of the space, not all the genes were illustrated in the figure.

Figure 2. Mitochondrial genome nucleotide diversity (pi) and genetic distance (Fst).

(a–c) pi of the whole collection; landrace and weedy rice; wild and the others (landrace and weedy rice). (d) The pi of indica, japonica, O. nivara and O. rufipogon. Values are sorted by ascending. (e) The average pi of indica and japonica in landrace or weedy rice. (f,g) The Fst between different groups. The circles indicated different groups and the circle size indicated the pi value. The Fst value between each two groups were represented by the distance between them. L: landrace rice, W: weedy rice, Wild: wild rice. LI: landrace_indica, WI: weedy_indica, LJ: landrace_japonica, WJ: weedy_japonica, NIV: O. nivara, RUF: O. rufipogon.

Figure 3. Population structure and principal component analysis of the collection.

(a) A neighbor-join tree reveals the phylogenetic of the collection. There main populations were identified. The landrace (indica and japonica type), weedy (indica and japonica type) and wild rice are marked with different color. (b) Population structure clustering of the collection with increasing K value from 2 to 6. (c,d) Principal component analysis of all the accessions (with wild rice) and only landrace and weedy rice accessions. The landrace (indica and japonica type), weedy (indica and japonica type) and wild rice are marked with different color. IND: indica, JAP: japonica, Mix indicated the two accessions in landrace which are mixed with wild rice.

Figure 4. A tanglegram phylogenetic analysis using trees from ML and BI methods to compare the differences from two methods.

(a) Phylogram and radial tree layout of the ML tree. (b) BI-based tree using the same datasets. Best-fit models were evaluated using jModeltest. The tanglegram was implemented in Dendroscope using a Neighbor Net-based heuristic method, which use line connects the same accession in two trees to see the difference phylogenetic structure. indica, japonica, wild and mix groups were marked with different colours. The accessions named with “L” or “W” indicate the landrace and weedy rice.

Figure 5. Haplotype network and population splits of the mitochondrial genome suggested by genome wide variations.

(a) Circle size is proportional to the number of samples within a given haplotype, and dashes between haplotypes represent unobserved, inferred haplotypes. Lines between haplotypes represent mutational steps between alleles. Colors denote rice designation: dark red, O. rufipogon; orange and blue, indica, red and purple, japonica, pink, O. nivara. (b) Population splits based on the six groups. L/W_IND: indica type in landrace or weedy rice, L/W_JAP: japonica type in landrace or weedy rice.