How can rice genetics benefit from rice-domestication study?

Uncovering the puzzle of the origin and domestication process of cultivated rice has greatly impacted rice genetics, but comprehensively exploiting elite alleles from both wild species and domesticated varieties for modern rice breeding is still a long-term ongoing study.

Crop domestication is one of the greatest innovations in human civilization. Understanding the origin and domestication process of cultivated rice is very important in profiling the history of agriculture. Modern cultivars have many morphological and physiological differences compared with their progenitors. Domesticated cultivars have obtained many beneficial alleles from their progenitors to adapt to local environmental conditions and control domestication alterations.

Asian cultivated rice, Oryza sativa L., has been domesticated from its ancestor, the wild-rice species O. rufipogon. As it is one of the most important food crops, rice cultivation has been deeply influenced by human civilization development [1]. O. sativa varieties and their progenitors are mainly cultivated in temperate, subtropical and tropical zones. O. sativa has two primary subspecies: indica and japonica. According to the analyses of genetic distance and population structure, japonica can be further divided into temperate japonica and tropical japonica subgroups [2]. Uncovering the process of rice domestication could give a better understanding of the nature of artificial selection and propose a model to be applied to the domestication studies of other crop species.

RICE-DOMESTICATION STUDIES BASED ON DOMESTICATED GENES

The process of rice domestication is closely linked with human conscious or unconscious selection and other forces. This process could be described as humans selecting and retaining the favored elite traits according to their selection (Table 1). The resulting domesticated plants therefore have larger seeds, higher resource allocation, more determinate growth and apical dominance, and non-shattering seeds compared with their ancestors. Genetic analysis of the domesticated genes and archaeological records could provide sufficient evidence to uncover the truth of cultivated-rice origin.

Table 1.

The difference between favored domestication traits in O. rufipogon and O. sativa.

Domestication traits	O. rufipogon	O. sativa	Cloned gene(s)
Shattering	Strong	Weak	Sh4, qsh1
Hull color	Most are black	Most are yellow	Bh4
Seed dormancy	High	Low	Sdr4
Grain size	Smaller	Bigger	qsw5
Grain quality	High amylose	Low amylose	Waxy
Awn length	Long	Short or none	An-1, An-2
Tiller angle	Prostrate growth and wider tiller angle	Erected growth and narrower tiller angle	PROG1
Panicle shape	Dispersed	Closed	OsLG1
Germination rate	Low	High	–
Stigma exertion	Yes	No	–

In past studies, efforts ‘from phenotype to genes’ help to identify many domestication-associated genes through quantitative trait locus (QTL) mapping. In maize, Teosinte branched1 (tb1) is the first identified domestication gene controlling the difference in apical dominance compared with its progenitor, teosinte [3]. Higher expression patterns in maize are believed to result from artificial selections. In rice, some favored domestication genes have been characterized. Shattering is a notable domestication trait that could be easily selected by our ancestors and has directly contributed to crop yield. qSH1 is a major QTL controlling shattering, encoding a homeobox-containing transcription factor. The causative mutation is a single nucleotide in a cis-regulatory element, regulating the shattering zone [4]. sh4 is another major QTL controlling shattering that encodes a gene with homology to Myb3 transcription factors, and a single amino change in the DNA-binding domain causes rice to gain the trait of non-shattering [5]. There are many domestication genes relevant to domestication traits, such as Bh4, PROG1, qSW5, An-1, An-2 and OsC1, which have been characterized up to date.

Archaeological evidence has shown that rice domestication began in the Yangtze Valley in China about 8000–9000 years ago and it was cultivated early in the Ganges in India about 4000 years ago [6]. The earlier human-cultivation practice was during the hunter-gatherers’ lives. Human ancestors tried to improve wild cereal crop to meet their needs and make the beneficial domestication genes inherited from its progenitor. Archaeological records provide evidence that seed-size enlargement can occur before the loss of shattering. In wheat and barley, archaeological studies demonstrated that an increase in grain size was followed by the fixation of non-shattering rachises [7]. Base on the genetic and archaeological evidence, many studies addressed the origin of cultivated rice. The most recent genetic studies tended to support the single-origin model through genetic characterization of the key domestication genes. Rice single-origin theory, the snowball model, demonstrated that the earlier critical domestication gene was first introduced to other regions of Asia, and the introgression occurred between the cultivar and local populations of O. nivara and O. rufipogon. The multiple-origin model from their perspective was a combination model. In this model, there were multiple mutations existing in divergent wild populations, and the key domestication genes between indica and japonica were formed by hybridization between the subspecies after their independent domestications [8].

RICE DOMESTICATION: A SINGLE ORIGIN AND MULTIPLE INTROGRESSIONS

Population genetics study in crops showed that the causative mutation in a domestication gene is fixed in the cultivars, causing decreased genetic diversity, which is called ‘selective sweeps’. Recent comprehensive study on rice genome variations of wild-rice O. rufipogon accessions and cultivated-rice O. sativa varieties has provided solid evidence to investigate the phylogenetic relationships between cultivated and wild rice, and identified the signatures of selection in rice domestication. This systematic comparison identified 55 selective sweeps that were involved in domestication [9]. If a gene was related to a domestication trait, then it might show a decrease in nucleotide diversity, increased linkage disequilibrium (LD) and altered population frequencies of polymorphic nucleotides in the gene and linked regions, which can provide sufficient evidence to manifest rice origin and domestication.

Based on rice genome variation studies, cultivated rice is found to undergo a single origin and multiple introgressions. Some useful mutations might have randomly occurred in some populations of wild-rice species, and were then selected and fixed for generating the proto-japonica-like varieties. The proto-japonica-like varieties were farther spread to other places in Asia. The indica varieties were subsequently generated through the crosses between the proto-japonica-like varieties and the O. rufipogon lines in local regions after many cycles of crosses and selections (Fig. 1). The favored mutations that have been fixed with their flanking regions of low genetic diversity (selective sweeps) in cultivated rice provide strong evidence to trace their origins.

Figure 1.

The domestication process of two main O. sativa subspecies: indica and japonica. The data are from the analysis of 55 selective sweeps involved in the Asia rice domestication [9]. The process can be divided into two parts: A and B. (A) Locus A (the red filled triangle and square indicate the two types of allele) and Locus B (the blue filled triangle and circle indicate the two types of allele) represent two genes regarding domestication syndromes, such as reduced shattering, larger seeds, erected growth, etc. The green and black thick lines represent different backgrounds of O. rufipogon III and O. rufipogon I, respectively. Mutations of Locus A and Locus B that randomly occurred in the O. rufipogon III population (from southern China) were favorably selected as domestication genes are indicated by filled triangles. (B) The proto-japonica varieties were domesticated from O. rufipogon III. The indica varieties were subsequently domesticated through crosses between the proto-japonica and O. rufipogon I (from south-east Asia and south Asia) after many cycles of multiple crosses and selections. During this process, the important signatures can lead to the conclusion of a skew in the allele frequency distribution, reduced genetic variation and increased linkage disequilibrium.

RICE DOMESTICATION BENEFITS RICE BREEDING AND FUNCTIONAL GENOMICS STUDIES

Genetic studies on domestication may provide some guidance for future breeding. Improving grain production and quality is an ongoing effort in crop breeding. To meet the challenges of global climate changes, a new expectation of high grain production has arisen in crop production. More and more beneficial gene alleles underlying specific traits have been selected in modern breeding [10]. The high yield and high grain quality of modern cultivars were contributed by the combination of many domestication- and improvement-related genes.

The great diversity in wild-rice populations, which have much more natural allelic variation than domesticated rice (especially the modern cultivars), will facilitate breeding to improve performance in many agronomic traits, including pathogen resistance and abiotic-stress tolerance. Uncovering the origin of cultivated rice has formed a solid foundation for a comprehensive identification of the genetic diversity in wild rice, landraces and elite varieties, and provided useful resources for the characterization of the genetic mechanism of agronomically important traits. The significantly reduced genetic diversity in cultivated rice is a limitation for further rice genetic improvement and the high level of allelic variation in wild rice will be an important resource in designed rice breeding that can be reintroduced into the gene pools of current elite varieties. Taken together, both the rice-domestication studies and the comprehensive analyses of genetic variation in rice will greatly benefit the studies of rice gene function in agronomical traits.