The Ge Laboratory, IBCAS

2. Population genetics and speciation of wild rice

Inference of recent evolutionary history of closely related species is one of the most intricate questions for evolutionary biologists. The level and pattern of nucleotide variation in DNA sequences provide important information on the evolutionary history of a species and divergent process of closely related species.
We investigated the nucleotide variation in 10 unlinked nuclear genes for species-wide samples of the C-genome Oryza species (O. officinalis, O. eichingeri and O. rhizomatis). ?The results indicated that three C-genome species harbored low levels of nucleotide variation relative to other plant species partly because of their smaller historic effective population sizes. Two speciation events happened within short time (0.3-0.4 mya), leading to significant lineage sorting among species. A long-distance dispersal from West Africa to Sri Lanka resulted in the disjunct distribution of O. eichingeri (Figure 2-1) (Zhang & Ge. 2007. MBE).

Figure 2-1. A long-distance dispersal from West Africa to Sri Lanka results in the disjunct distribution of O. eichingeri.

Previous studies indicated that the wild progenitors of cultivated rice (O. sativa), O. nivara and O. rufipogon, were two incipient species. They are sympatric but distinct morphologically and ecologically (Figure 2-2), providing a unique opportunity to study speciation. Based on population samples across the geographic range of the two species, we investigated the nucleotide diversity and gene flow using sequences of 7 chloroplast and nuclear loci, and found highly structured populations within species and little nucleotide differentiation between species. Coalescent-based simulations suggested that the two species began to diverge at c. 0.16 MYA and strongly rejected the simple isolation model of zero migration between species, implying recurrent gene flow during the speciation. This study demonstrates that natural selection plays a primary role in driving the divergence of the two Oryza species (Zheng & Ge 2010. ME). An investigation of the molecular basis of the phenotypic differentiation and speciation between the species using genomic SNPs and expression data is currently under way.

Figure 2-2. Geographic distribution, and morphological and ecological difference for two wild rice species, O. nivara and O. rufipogon.

Further investigations based on sequences of 10 nuclear and two chloroplast loci from 26 wild populations across the entire geographic ranges of the two species (Figure 2-3) detected four genetically distinct groups within O. rufipogon and found no correlation between the genetic groups and either species identity or geographical regions (Liu et al. 2015. ME). We suggest that the repeated extinction and colonization of local populations due to multiple glacial-interglacial cycles during the Quaternary was most likely the main factor shaping the confounding population genetic structure of O. rufipogon. Based on the genetic pattern and dynamics of the O. nivara populations, we hypothesize that O. nivara might have independently originated multiple times from different O. rufipogon populations (Figure 2-3).

Figure 2-3. Geographic distribution of O. rufipogon (broken line) and O. nivara (solid line). Blue and red dots represent the O. rufipogon and O. nivara populations sampled, respectively. Three shaded areas are hypothesized origin places of O. nivara.

Recently, by use of digital gene expression technology and the paired-end RNA sequencing method, we conducted a genome-wide gene expression investigation of two Oryza species (Guo et al. 2016. MBE). We found that approximately 8% expressed genes differed significantly in expression levels between the two species and they are randomly distributed across the genome. Moreover, 62% of the differentially expressed genes exhibited a signature of directional selection in at least one species. Importantly, the genes with differential expression between species evolved more rapidly at the 5' flanking sequences than the genes without differential expression relative to coding sequences. Our findings demonstrate that ecological speciation is associated with widespread and adaptive alterations in genome-wide gene expression and provide new insights into the importance of regulatory evolution in ecological speciation in plants (Figure 2-4).

Figure 2-4. Variation patterns of gene expression across the genome. (a) and (b) Venn diagrams showing the numbers of DE and directional selection genes between species among different tissues. (c) Physical distribution of DE genes on 12 rice chromosomes.

Allopolyploidization is one of the major modes of diversification and speciation in plants, and the important source of morphological innovations. In the rice genus (Oryza), about one half of the species are allopolyploids. These species are not only important resources for rice breeding but also provide a unique opportunity for studying evolution of polyploid species. By sequencing four biparentally inherited nuclear loci and three maternally inherited chloroplast fragments from all diploid and tetraploid species with the B- and C-genome types in this genus, we detected at least three independent origins of the BC-genome tetraploid species (Zou et al. 2015. Sci Rep). Specifically, the diploid O. punctata (B-genome) acted as the paternal donor for one tetraploid species and the maternal donor for the other two tetraploid species. We demonstrated that all of the tetraploids originated within the last one million years (Figure 2-5).

Figure 2-5. Schematic of evolutionary relationships among the species with the B-, C-, and BC-genome types. Below is the geographical distribution of these species.