1. Introduction
As an annual crop and one of the three major food crops, rice (
Oryza sativa) provides staple food for 3 billion people worldwide and plays an essential role in world food security [
1,
2]. Its importance is even more evident in China due to its long history of cultivation, given that it is the most dominant staple food for over 60% of the Chinese [
3]. The traditional cultivation modes require sufficient irrigation water and have caused the serious nutrient loss. Moreover, labor costs, seeds, and other production costs keep rising, which affect people’s enthusiasm and motivation to cultivate perennial rice [
4]. Recently, researchers placed efforts into using the perennial habit of wild rice varieties to breed perennial rice/upland rice and fixing heterosis through vegetative reproduction to increase harvest area and reduce production costs and soil erosion, thereby protecting the environment [
4,
5].
Different perennial characteristics were performed in the species of
Oryza. The common wild rice (
O. rufipogon) grows perennially with stolons. However,
O. longistaminata,
O. rhizomatis, and
O. australiensis grow permanently with rhizomes. Both
O. longistaminata and
O. rufipogon with the typical perenniality have the same AA genome as
O. sativa, which are considered ideal donors for transplanting the perennial habit [
6]. Rhizomes determine the perennial growth of
O. longistaminata and are controlled by one pair of dominant complementary genes,
Rhz2 and
Rhz3, located on chromosomes 3 and 4, respectively [
5]. Further studies have showed that a very complex gene regulatory network mediated the growth and development of rhizomes in
O. longistaminata [
7].
O. rufipogon is the ancestor of
O. sativa [
8], and there are almost no reproductive disorders between them. It is sensitive to photoperiod and has a high outcrossing rate, preferring to live near swamps, lakes, and ditches with perennial habit [
9].
O. rufipogon has a prostrate growth for many years by branching stolons and functional roots to form new individuals and finally achieving perennial survival through vegetative reproduction [
10].
Various important agronomical traits were selected during the domestication of cultivated rice, including seed shattering, erect plant architecture, panicle shape, awn length, grain size and quality, and hull color [
11]. Some key genes controlling rice domestication are well-understood, including the shattering gene
Sb1 and
qSH1 [
12,
13]; awn length gene
An-1,
An-2, and
RAE2 [
14,
15,
16,
17]; hull color gene
Bh4 [
18]; pericarp color gene
Rc [
19]; panicle shape and ligule development gene
OsLG1 [
20]; grain width gene
GW5 [
21,
22]; and tiller angle gene
PROG1 [
23,
24]. The selection and domestication of these genes resulted in dramatic morphological changes in cultivated rice in Asia, from wild varieties adapted to natural environments to cultivated varieties adapted to agricultural environments. In addition, about 40% of wild rice alleles, including perennial alleles in the process of domestication from the wild rice to the cultivated rice, were lost [
25], which resulted in narrowing of the genetic basis of cultivated varieties and further restricted improvement and breakthrough of yield potential of rice varieties. Genetic and molecular analyses of the perennial habit of
O. rufipogon are the basis for breeding utilization and also help reveal the origin and domestication process of cultivated rice. Creeping growth gene
PROG1 and photoperiod sensitive gene
Hd1 and
Ehd1 at the heading stage were associated with the perennial habit of
O. rufipogon. However, the formation of new individuals by stem branches and adventitious roots on stolons and the critical part of the perennial habit, is little known.
Tiller, the shoot branch in rice which is closely related to rice yield, has been extensively studied, and the formation process of tiller has been elucidated [
26,
27], including
MOC1 [
28],
MOC3/
TABLE1 [
29,
30],
TAD1/
TE [
31,
32],
OSH1 [
30],
LAX1, and
LAX2 [
33,
34], which influence tiller bud initiation. Furthermore,
D27 [
35],
OsMAX1 [
36],
OsMADS57 [
37],
D53 [
38],
OsOTUB1 [
39],
IPI1 [
40],
IPA1 [
41],
OsTB1 [
42,
43], and
DLT [
44] have been demonstrated to participate in tiller bud outgrowth. Phytohormones also regulate tiller bud initiation and outgrowth. In rice, strigolactones (SLs), auxin, gibberellins (GAs), and abscisic acid (ABA) can repress bud outgrowth, whereas brassinosteroids (BRs) and cytokinins (CKs) promote bud outgrowth [
45].
The above studies mainly focus on the tiller, a lateral shoot arising from an axillary bud at basal nodes in rice. However, the genetic and molecular mechanism of the stem branch, a lateral shoot arising from an axillary bud at high nodes aboveground, remains elusive. Herein, we investigated the stem branch trait of an introgression line, a cross between cultivated rice (O. sativa) and its wild relative, O. rufipogon. The introgression line formed new individuals from the branches on stem nodes with adventitious roots, which was the same as the O. rufipogon. Furthermore, these individuals can survive by cutting, bear normally, and produce the same yield per plant with the introgression line. This study will provide theoretical basis and practical value for rice breeding using the stem branch trait to fix heterosis and cultivate new rice varieties of vegetative reproduction.
2. Materials and Methods
2.1. Plant Materials and Natural Field Experiment Conditions
The japonica paddy rice varieties, including Yundao1 (YD1), Yunjing37 (YJ37), and Dianjingyou1 (DJY1), were used as the recurrent parents. The common wild rice (YJCWR) collected from Yuanjiang county, Yunnan Province, China was used as the donor parent. IL-J85 was an introgression line from the cross between YD1 and YJCWR. NIL-Y37 and NIL-D1 (BC3F5) were the near-isogenic lines by crossing/backcrossing between IL-J85 and YJ37, respectively. C-J85, C-Y37, and C-D1 were the cuttings from the stem branches of IL-J85, NIL-Y37, and NIL-D1, respectively. S-J85, S-Y37, and S-D1 were the seed seedlings of IL-J85, NIL-Y37, and NIL-D1, respectively. BC3F2 and BC3F3, derived from the cross between YD1 and IL-J85 were cultivated to map the QTLs responsible for the stem branch trait.
All rice materials used for agronomic analysis were cultivated at the breeding base of the Xishuangbanna Botanical Garden, Chinese Academy of Sciences (21°56′ N, 101°15′ E) under natural field conditions twice a year in the spring and fall of 2021. The seeds were soaked for 48 h, and then germinated in a growth chamber at 28 °C for 24 h, subsequently sown on seedbeds. The seed seedlings were transplanted to paddy field at the four-leaf stage with 6 rows per plot and 12 plants per row. The density was 25 cm × 20 cm between each individual. For the breeding utilization of the stem branch trait, about 10 days before harvest of the NILs in the previous season, the seeds of the NILs were sowed normally on the seedling bed. Furthermore, the stem branches were cut and collected in 15 days later after harvest of the NILs, and then the cuttings (C-Y37 and C-D1) were placed in water for 3 days to enable the growth of the adventitious root. The seed seedlings and cuttings of NILs were transplanted to the paddy field simultaneously. The seed seedlings and cuttings of IL-J85 were treated and planted as mentioned above. To better measure phenotypes, all rice materials were planted in random three repetitions. The water in the paddy field was covered 2–5 cm above the soil before the rice seeds matured and supplemented with the compound NPK fertilizer (2 kg/100 m2).
2.2. Phenotypes Analysis
The materials were harvested in the center of each plot in each line to investigate the phenotypes. For the parent YD1 and IL-J85, the agronomic traits, including plant height, primary panicle length, flag leaf length, flag leaf width, flag leaf length–width ratio, the number of panicles, primary branches per panicle, secondary branches per panicle, 1000-grain weight, grain yield per plant, and internode length were investigated for 30 individuals. The number of stem branches and tillers per plant were conducted in 15 days later after harvest. For the analysis in the breeding utilization of the stem branch trait, the agronomic traits, including heading date, plant height, primary panicle length, flag leaf length, flag leaf width, the number of panicles, primary branches per panicle, secondary branches per panicle, 1000-grain weight, and grain yield per plant, were investigated for 30 individuals of the seed seedling and cutting of corresponding ILs/NILs respectively. Each experiment was performed for three biological repetitions. All rice seeds for phenotypes analyses were air-dried. Grain weight was measured with a Yield-Traits Scorer (YTS-5DS).
2.3. DNA Extraction
Rice genomic DNA from leaves was obtained according to the method described by Abdel-Latif et al. [
46].
2.4. Mapping of qSBR1 and qSBR2
We hybridized IL-J85 with YD1 to obtain the BC3F1. The BC3F1 generation all performed the stem branch trait. BC3F2 was obtained from the selfing of BC3F1. We used 210 individuals of the BC3F2 segregating population to map qSBR1 and qSBR5 regions based on the number of stem branches per plant. A total of 100 recombinant individuals from BC3F3 segregating populations were used to fine map the locus of the stem branch trait through the targeted sequencing genotype detection technology of 10 K SNP chip (Shijiazhuang Molbreeding Biotechnology Co., Ltd., Shijiazhuang, China).
2.5. Statistical Analysis
QTL IciMapping 4.1 software was used to construct a genetic linkage map based on the genotypes and phenotypes, and Excel 2007 and SPSS 22 were used for data analysis. All samples were subjected to three replicate measurements. The data were presented as the mean ± standard deviation (SD). Student’s t-test or one/two-way ANOVA was performed to determine the significant difference, as indicated by * and ** at p < 0.05 and p < 0.01, respectively.
4. Discussion
Rice is the staple food for half of the population worldwide. Breeders have placed efforts into improving output per unit area to fulfill the increasing demand of food. Double-cropping rice can be developed in tropical areas to increase rice production. Seed is the main expense of the cost of rice production, especially for hybrid rice. South China and the middle and lower reaches of the Yangtze River are the main areas for rice production. In most rice plant regions, the whole growth duration for a single cropping rice season was surplus but insufficient for double cropping rice. IL-J85 with the stem branch trait discovered in this research can be transplanted to cuttings after harvest and performed with a similar yield compared with YD1, which saved the growth time and production cost. A correlation analysis found that grain yield per plant was negatively correlated with the number of panicles; however, the number of stem branches was significantly correlated with the number of panicles. Furthermore, the number of stem branches affected plant agronomic traits including plant height, flag leaf length, flag leaf width, the number of panicles, and primary panicle length, indicating that there may be one pleiotropic gene or linkage inheritance. A genotypic analysis showed that the phenotype of the stem branch in the heterozygous state was consistent with that in the homozygous state, indicating that it can be broadly used in hybrid rice production. These results suggest that common wild rice has practical value for rice production by using the stem branch trait to fix heterosis in breeding rice varieties of vegetative reproduction.
The novel beneficial alleles and potential genetic diversity have been lost during wild rice domestication in cultivated rice. The wild rice species can better adapt to different environments and various abiotic and biotic stresses owing to the reservoir of novel genes/QTLs. Some agronomically important genes/QTLs for improving abiotic and biotic stresses, resistance, productivity, and grain quality characteristics were identified from AA genome donor wild species and labelled with breeder-friendly molecular markers to elite genetic backgrounds [
47].
O. rufipogon (2n = 24, AA) was considered as the progenitor of
O. sativa. There were several reports of using chromosomal segment substitution lines (CSSLs), backcross inbred lines (BILs), ILs, and NILs to introgress QTLs related to yield and grain quality from different
O. rufipogon accessions into elite
indica and
japonica genetic backgrounds [
48,
49,
50,
51,
52]. In this research, the introgression line IL-J85, constructed with YD1 and Yuanjiang common wild rice (
O. rufipogon), had the stem branch trait and could be cut to form new individuals to be achieved a with similar yield of seed seedlings. The
qSBR1 and
qSBR5 responsible for the stem branch trait were first mapped on chromosomes 1 and 5. However, rice tiller is regulated by complex genes, hormones, and environmental factors. The molecular mechanism and regulatory network of the stem branch trait still need further elucidation.
Although there has been a doubling of major grain crop yields since the 1950s, people subjected to malnutrition prevail worldwide [
53]. However, the crop yield increase cannot fulfill the demand for food with an ever-growing population [
54]. Global food security depends on annual grains:cereals, oilseeds, and legumes, which damage essential ecosystem services, making some beyond sustainable boundaries [
55,
56]. The development of perennial varieties of important grain crops can expand choices of farmers to produce grains under less favorable circumstances, which can ensure food and ecosystem security [
4]. Perennial crops are superior to annuals in sustaining important ecosystem functions, specifically on marginal landscapes or where resources are limited [
57]. Perennial crops tend to have longer growing seasons and deeper rooting depths than annual counterparts. In addition, they intercept, retain, and utilize more precipitation [
4,
58,
59]. Breeders are devoted to pursuing of high yield, high quality, high resistance, and other valuable characteristics of rice varieties. In this research, IL-J85 in YD1 background performed the stem branch trait and can be cut to form new individuals with similar yield of seed seedlings, which save rice seed cost, reduce seedling raising time and water usage. In addition, the cultivated near isogenic lines NIL-Y37 and NIL-D1 in the two backgrounds Yunjing 37 and Dianjingyou 1 perform the same characteristics as IL-J85, indicating the stem branch trait can be used in rice production and breeding utilization of rice varieties of vegetative reproduction.