Keywords

22.1 Introduction

Rice (Oryza sativa L.) is the most important food crop in the world, being a staple for over half of the world’s population. Rice cultivars must produce high yields and have strong stress resistance. However, the market price and consumer acceptance of rice are largely determined by rice grain quality, which is a primary consideration of consumers, the food industry, farmers, and seed producers (Hori and Yano 2013; Bao 2014). Therefore, rice grain quality is a major target in rice breeding programs (Champagne et al. 1999; Fitzgerald et al. 2009).

Rice grain quality comprises a set of complex traits encompassing a wide range of physical, chemical, and physiological characteristics (Hori and Yano 2013). Grain quality consists of four major components, concerning grain appearance, milling qualities, nutrition, and cooking and eating qualities (Bao 2014). This chapter focuses on grain chalkiness, one of the most important traits in grain appearance. Recent progress in the genetic analysis of grain chalkiness has revealed many QTLs and genes, which can be used in breeding selection for the development of novel cultivars with high grain appearance quality.

22.2 Components of Grain Appearance Quality

Pericarp color, grain shape, grain size, and grain chalkiness (translucency) contribute to grain appearance. Several rice cultivars have red, brown, purple, or black pericarps (Wang and Shu 2007). Proanthocyanidins and anthocyanins accumulate in the grains of red and black rice, respectively. Both chemicals have antioxidant activity with health benefits for humans (Maeda et al. 2014). Grain size and shape are described by the kernels’ length, width, and thickness. Medium- and short-grain cultivars are generally preferred in Japan, northern China, and Korea, where japonica rice is cultivated. Long and slender grain even with aroma cultivars are preferred in many countries where indica rice is cultured and traded (Bergman et al. 2004). Previous molecular genetic studies have identified many QTLs and genes involved in the control of grain size and shape and pericarp color (Furukawa et al. 2007; Huang et al. 2013) (See also Chap. 11).

22.3 Grain Chalkiness

Chalky grains have opaque spots in various regions of the endosperm (Hori and Yano 2013; Sreenivasulu et al. 2015; Ishimaru et al. 2016; Lo et al. 2016). Generally, increasing chalkiness reduces grain quality and retail prices. The physicochemical properties involved in grain chalkiness have been investigated for over a hundred years (Inagaki 1899; Goto 1904). Normal translucent endosperm consists of large, tightly packed, polyhedral, single starch granules. On the other hand, chalky endosperm consists of small, loosely packed, round, compound starch granules, and therefore its cause is inferred by the defects in genes affecting starch biosynthesis, starch granule structure, or grain-filling manner (Bao 2014; Lo et al. 2016). In indica cultivars, chalkiness is measured visually on a scale of 0–9, where 0, none; 1, minor (<10% of the endosperm); 5, moderate (10–20%); and 9, extensive (>20%) (Fig. 22.1a). In japonica cultivars, chalkiness is classified according to the location of the chalkiness: white-back chalkiness on the dorsal side of the grain, white-belly chalkiness on the ventral side, white-base chalkiness around the embryo, white-core chalkiness in the center, and milky-white chalkiness of the whole grain (Fig. 22.1b).

Fig. 22.1
figure 1

Grain appearance in rice cultivars. (a) Chalky grains of indica cultivars. (b) Chalky grains of japonica cultivars: (i ) normal (no chalkiness), (ii) white-back, (iii) white-base, (iv) white-belly, (v) white-core, (vi) milky-white, and (vii) abortive in normal color (upper) and color-negative (lower) images. Scale bars, 5 mm

Full size image

However, consumer preferences necessitate white-core or whole-grain chalkiness in some japonica cultivars (Juliano and Hicks 1996; Hori and Yano 2013). For example, arborio-style and Japanese “sake” brewing cultivars need white-core chalkiness, and glutinous rice cultivars have a completely opaque endosperm. Therefore, genetic control of grain chalkiness is necessary for both increasing and decreasing chalkiness.

22.4 Alteration of Milling Quality by Grain Chalkiness

Chalky grains show significantly different physicochemical, morphological, thermal, cooking, and textural properties from completely translucent grains. In addition, grain chalkiness greatly influences milling quality, especially in indica cultivars, which have slender grains (Del Rosario et al. 1968; Bao 2014). Although annual world rice production is 750 Mt a year, the final milled rice yield is 490 Mt a year. Thus, as much as 260 Mt a year is lost after milling, although this includes hulls and bran (Sreenivasulu et al. 2015). Chalky grain is brittle and is easily broken during milling and so decreases the final yield of polished grains.

Grain size and shape are highly correlated with degree of chalkiness and milling quality (Zheng et al. 2007). Grain length is negatively associated with milled-grain yield, whereas grain width and thickness are positively correlated with milled-grain yield. Many QTLs for grain size and shape are co-localized on the same chromosome regions as QTLs for chalkiness (He et al. 1999; Tan et al. 2000; Wan et al. 2005; Ebitani et al. 2008; Wang et al. 2016a). Thus, optimization of grain size and shape would be an effective strategy for decreasing chalkiness and improve milling quality. In particular, because many indica cultivars have long and slender grains, the genetic control of grain size and shape in these cultivars needs to be improved to increase milled-grain yield while keeping chalkiness low.

22.5 Effect of High Temperature at Grain-Filling Stage on Grain Chalkiness

Rice growth is seriously affected by climate change, and global warming has become a major constraint on rice production. In the past three decades, Earth’s surface temperature has warmed faster than at any time preceding 1850. The global mean surface temperature may rise by up to 4.8 °C by the end of this century relative to the period from 1986 to 2005 (IPCC 2013; Ishimaru et al. 2016). Rice is highly sensitive to heat stress, particularly during the reproductive and grain-filling stages (Mitsui et al. 2013; Jagadish et al. 2015). Heat stress induces abnormal grain formation at all stages from pollen development through endosperm filling to harvest. High temperatures during grain filling reduce starch accumulation and shorten the grain-filling period, resulting in low grain weight and increased chalkiness characterized by both irregular and round starch granules (Arshad et al. 2017).

In japonica cultivars, grain chalkiness is induced under conditions of average daily mean temperatures above 27 °C during the first 20 days after heading (Wakamatsu et al. 2009; Ishimaru et al. 2016). High nighttime temperatures significantly increase grain chalkiness. Insolation strongly influences the emergence of grain chalkiness under high temperatures or high humidity (Wakamatsu et al. 2009; Tanaka et al. 2010): the prevalence of white-back grains increases under high insolation and high humidity, while that of milky-white grains increases under low insolation and high humidity. Nitrogen application also influences the frequency of chalkiness (Tanaka et al. 2010): the prevalence of milky-white, white-base, and white-back grains decreased with increased nitrogen uptake and content. This association suggests that maintaining a high nitrogen content at the panicle formation stage is important for optimizing protein content and reducing chalkiness.

Crop management can enhance heat resistance in rice plants (Bita and Gerats 2013; Sreenivasulu et al. 2015). Increasing the water depth in paddy fields is one of the most effective methods to manage heat stress and to minimize damage to panicles and grains. A water depth of 20–25 cm significantly shields panicles and grains from injury during the grain-filling period. Shifting earlier or later the planting time can also avoid grain filling during hot weather. Such management methods prove successful at reducing grain chalkiness. Collaboration among farmers, breeders, physiologists, and meteorologists is necessary to realize this.

22.6 Genetic Dissection of Grain Chalkiness

Understanding of the genetic factors associated with rice grain quality is necessary for the efficient development of new cultivars producing high-quality grain required and preferred by consumers. Shumiya et al. (1972) reported that the occurrence of white-core grains varied from 0% to 49% in 49 japonica cultivars. Ebata and Tashiro (1973) reported that the occurrence of white-belly grains ranged from 0.3% to 99.4% in 88 indica and japonica cultivars. As the occurrence of chalky grains in F2 populations showed continuous frequency distributions (Kamijima et al. 1981; Takeda and Saito 1983), grain chalkiness represents a set of complex traits controlled by multiple genes or QTLs.

Many QTLs for grain chalkiness have been detected in segregating populations such as F2 populations and recombinant inbred lines (RILs) derived from crosses between indica and japonica cultivars (Yamakawa et al. 2008; Bao 2014; Ishimaru et al. 2016). Genetic studies have detected over 140 QTLs for grain chalkiness across all 12 rice chromosomes in the Gramene QTL database (http://archive.gramene.org/qtl/) (Monaco et al. 2014; Sreenivasulu et al. 2015). He et al. (1999) reported the first detection of QTLs for grain chalkiness, detecting three QTLs for the percentage of white-core grains and the square of the white-core in doubled haploid lines (DHLs) of ZYQ8 × JX17. Tan et al. (2000) found 14 QTLs for white-belly, white-back, and white-core grains in RILs of Zhenshan 97 × Minghui 63. Ebitani et al. (2008) reported 12 QTLs for 6 types of grain chalkiness in chromosome segment substitution lines (CSSLs) of Nipponbare × Koshihikari.

A few QTLs for grain chalkiness have so far been fine mapped and isolated. qPGWC8 is a major-genetic-effect QTL for this phenotype on chromosome 8 in CSSLs of Asominori × IR24 (Wan et al. 2005). Guo et al. (2011) narrowed down its chromosomal region within a 142-kbp region by using 1801 BC4F2 plants of Asominori × IR24. qPGWC7, on chromosome 7, was identified using a set of CSSLs of PA64s × 9311: fine mapping in a population of 3221 F2 plants delimited it to a 44-kbp region including 13 predicted genes (Zhou et al. 2009). qACE9 was also detected in CSSLs of PA64s × 9311 (Gao et al. 2016): fine mapping in 920 BC4F2 plants narrowed it down to within a 22-kbp region including five predicted genes. Map-based cloning of the chalkiness 5 (CHALK5) QTL, a major-effect QTL in DHLs of H94 × Zhenshan 97, isolated a gene encoding a vacuolar H+-translocating pyrophosphatase with activity of inorganic pyrophosphate hydrolysis and H+-translocation (Li et al. 2014) (Table 22.1).

Table 22.1 Grain chalkiness genes in rice isolated by map-based cloning strategies
Full size table

Several rice mutants show grain chalkiness phenotypes due to disruption of genes associated with starch biosynthesis and metabolism pathways (Table 22.1). Tos17 transposon and T-DNA insertion mutants of starch synthase IIIa (SSIIIa) showed loose starch granule packing and a chalky phenotype (Fujita et al. 2007; Ryoo et al. 2007). A single nucleotide substitution at the splice junction of UDP-glucose pyrophosphorylase 1 (UGPase1) induced chalky endosperm (Woo et al. 2008); overexpression of UGPase1 protein produced normal grains in transgenic plants of the UDPase1 mutant. A gene named floury endosperm 4 ( flo4) was isolated from a mutant with a floury white-core endosperm caused by T-DNA insertion into the rice pyruvate orthophosphate dikinase B (OsPPDKB) gene (Kang et al. 2005). Chalkiness was produced by downregulation of starch-branching enzyme IIb (SBEIIb) in rice endosperm (Butardo et al. 2011; Yang et al. 2012). Likewise, a mutant of the substandard starch grain 4 (SSG4) gene had altered starch granules and a chalky phenotype (Matsushima et al. 2014). Recently, Tang et al. (2016) and Wei et al. (2017) identified grain incomplete filling 2 (gif2), which encodes the rice ADP-glucose pyrophosphorylase large subunit 2 (OsAGPL2) protein. The rice brittle 1 (OsBT1) gene encodes an ADP-glucose transporter, and a mutant had white-core grains (Li et al. 2017). The ethyl methane sulfonate (EMS) mutants of pyrophosphate: the fructose-6-phosphate 1-phosphotransferase (PFP1) gene exhibited floury endosperm (Duan et al. 2016). Alteration of the expression of starch metabolism genes also influences the degree of chalkiness: downregulation of α-amylase genes Amy1A and Amy3E produced less chalky grains in RNAi transgenic plants (Hakata et al. 2012). Fu and Xue (2010) reported that rice starch regulator 1 (RSR1) gene, which encodes an APETALA2/ethylene-responsive element binding protein family transcription factor protein, negatively regulates the expression of starch synthesis genes. And OsbZIP58 is a key transcriptional regulator required for starch synthesis through directly binding to the promoters of OsAGPL3, GBSSI, OsSSIIa, SBE1, OsBEIIb, and ISA2 to promote their expression (Wang et al. 2013).

The formation of chalky grain is also triggered by abnormal accumulation of seed storage protein. Many storage protein mutants preferentially accumulate a large amount of glutelin precursor within chalky grains (Table 22.1). The endosperm storage protein mutant 2 (esp2) gene encodes disulfide isomerase-like 1-1 (PDIL1-1) protein (Takemoto et al. 2002) (Fig. 22.2). The glutelin precursor mutant 4 (glup4) gene encodes a rice ortholog of the small GTP-binding and GTP-hydrolyzing protein Rab5a (Fukuda et al. 2011). The glutelin precursor mutant 6 (glup6) gene encodes guanine nucleotide exchange factor, which activates Rab5a protein (Fukuda et al. 2013). The glutelin precursor accumulation 3 (gpa3) gene encodes a plant-specific kelch-repeat protein, which is associated with the trans-Golgi network (Ren et al. 2014). The glutelin precursor mutant 2 (glup2) gene encodes Golgi transport 1B (GOT1B) protein, which has membrane-spanning domains (Fukuda et al. 2016; Wang et al. 2016b). These results indicate that disruption of the storage protein trafficking system during endosperm development elevates the amount of small-vesicle-like abnormal protein bodies, decreases the number and size of protein bodies, and ultimately increases grain chalkiness.

Fig. 22.2
figure 2

Phenotypes of rice grains in wild type (WT) and esp2 mutant. (a) Appearance of polished rice grains. Scale bar, 5 mm. (b) SDS-PAGE image of storage proteins in mature seeds. Arrow indicates accumulation of glutelin precursor in the esp2 mutant. (c) Scanning electron microscopy images of transverse sections of mature seed. Scale bars, 10 μm

Full size image

Mutations of genes not involved in starch biosynthesis and metabolism or protein biosynthesis can also lead to phenotypes similar to grain chalkiness (Table 22.1). The grain incomplete filling 1 (gif1) mutant has a disrupted allele of a gene encoding a cell-wall invertase protein required for carbon partitioning during early grain filling (Wang et al. 2008). The floury endosperm 2 ( flo2) mutation is caused by a disrupted allele of a gene encoding the rice tetratricopeptide repeat domain-containing (OsTPR) protein (She et al. 2010). The floury endosperm 6 (FLO6) gene encodes an unknown protein with a C-terminal carbohydrate-binding module (Peng et al. 2014). The substandard starch grain 6 (ssg6) gene encodes a protein homologous to aminotransferase (Matsushima et al. 2016). The mutation of OsAlaAT1 gene, which encodes cytosolic aminotransferase associated with the interconversion of pyruvate to alanine, decreased the expression of starch biosynthetic genes (Yang et al. 2015). The floury endosperm7 (FLO7) gene encodes a protein of unknown function; however, FLO7 is necessary for starch synthesis and amyloplast development (Zhang et al. 2016). RNAi knockdown of the expression of the rice nuclear factor Y B (OsNF-YB) transcription factor gene, which is co-expressed with starch biosynthesis genes in rice endosperm, leads to small grains with chalky endosperm (Xu et al. 2016).

A comprehensive atlas of transcriptomic changes under high temperatures reveals downregulation of starch biosynthesis and upregulation of starch degradation (Jagadish et al. 2015; Yamakawa et al. 2007). The incidence of chalkiness is increased as a result of high expression of genes for starch degradation enzymes such as α-amylases (Amy1A, Amy1C, Amy3A, Amy3D, and Amy3E; Yamakawa et al. 2007, Hakata et al. 2012). Heat stress also downregulated the expression of starch synthesis genes of SBEIIb and OsPPDKB (Li et al. 2011). To decrease grain chalkiness, it would be effective to adjust the balance of gene expression between biosynthesis and metabolism of starch and storage proteins.

22.7 Improvement of Grain Quality in Breeding Selection of Rice Cultivars

DNA markers tightly linked with grain chalkiness QTLs and genes facilitate marker-assisted selection for the development of novel cultivars with appearance of high grain quality. Hori et al. (2012) found two QTLs for grain chalkiness on chromosomes 8 and 11, qDWK8 and qDWK11. Two CSSLs with Koshihikari alleles of these QTLs in the Nipponbare background had lower proportions of white-base and white-back grains (Hori et al. 2017). In addition, CSSL SL844, with qDWK11, had fewer abortive and cracked grains after milling (Fig. 22.3). Kobayashi et al. (2007) detected a major-effect QTL for white-back grains on chromosome 6 (qWB6). A near-isogenic line (NIL) carrying the Hanaechizen qWB6 allele of this QTL in the Niigatawase background had significantly less white-back chalkiness during grain filling at >27 °C (Kobayashi et al. 2013). Shirasawa et al. (2013) detected another QTL on chromosome 6 very close to qWB6. NILs with the Kokoromachi allele of this QTL in the Tohoku 168 background had fewer white-back grains. Two NILs carrying a QTL on chromosome 8 from Chikushi 52 in the Tsukushiroman background had less white-back chalkiness (Wada et al. 2015). Murata et al. (2014) detected an appearance quality QTL Apq1 on chromosome 7. A NIL with the Habataki Apq1 allele in the Koshihikari background had a significantly higher rate of translucent grains. Kobayashi et al. (2016) reported that a NIL carrying the Kasalath allele of the seed dormancy QTL Sdr4 in the Koshihikari background had a lower rate of white-back grains. Although these reports concern white-back chalkiness mostly, they clearly indicate that introgression of these QTLs can significantly decrease grain chalkiness. These NILs are good candidates for novel cultivars with high grain quality.

Fig. 22.3
figure 3

Milling quality of Nipponbare and SL844, with a Koshihikari chromosome fragment (red) in the Nipponbare genetic background (white). (a) Graphical genotypes representing the 12 rice chromosomes. (b) Abortive and cracked grains after milling 1.5 kg of brown rice

Full size image

It is also important to find additional QTLs and to develop novel NILs by screening many cultivars for new genetic variation. Rice cultivars and wild relatives have a broad range of adaptations to heat stress and can be used in breeding to reduce grain chalkiness (Shumiya et al. 1972; Ebata and Tashiro 1973; Arshad et al. 2017). The indica cultivars Takanari, Habataki, and Kasalath have low chalkiness (Ebitani et al. 2008; Murata et al. 2014). Several japonica cultivars, including Fusaotome, Koshijiwase, Hanaechizen, Kokoromachi, and Tentakaku, have low chalkiness (Kobayashi et al. 2007; Ishimaru et al. 2016). Thus, both indica and japonica cultivars could be used as donors for reducing grain chalkiness. The use of additional genetic resources would provide advantages in the development of novel rice cultivars with high grain quality.

22.8 Conclusions and Prospects

Many QTLs for grain chalkiness have been detected on all rice chromosomes. Several have been fine mapped, and the responsible genes have been identified by map-based cloning strategies. Further work has led to the development of NILs by marker-assisted selection with DNA markers linked to major-effect QTLs and genes.

However, we are still far from clearly understanding how grain chalkiness forms. To uncover all genetic factors involved in its control, it is necessary to perform further analyses with new techniques. So far, QTL analysis has detected a few QTLs associated with phenotypic differences between pairs of cultivars. Genome-wide association study (GWAS) can collect many QTLs and identify the genes responsible directly without the need for segregating populations (Zhao et al. 2011; Yano et al. 2016). GWAS has recently been used in the study of grain quality traits in rice (Qiu et al. 2016; Wang et al. 2016a). GWAS will be a powerful tool to collect additional QTLs and to isolate new genes in multiple cultivars. Further, novel analytical methods such as genome sequencing by next-generation sequencer, genotype-by-sequencing (GBS), RNA-seq, and metabolome analysis can enhance the resolution of genetic analysis. Chen et al. (2016) combined GBS and RNA-seq to detect grain chalkiness QTLs and to identify candidate genes highly expressed in chalky endosperm. Metabolome analyses in rice cultivars and mutants revealed that the metabolism of carbon, nitrogen, and phospholipids is important in the formation of chalky grains (Yamakawa and Hakata 2010; Kusano et al. 2012; Lin et al. 2017). Such trials could open an opportunity to identify in detail the molecular mechanisms that regulate grain chalkiness.

In addition to the development of NILs carrying single major-effect QTLs and genes, it is possible to develop NILs by introducing multiple QTLs and genes simultaneously. Combining mutant alleles of starch synthase and branching enzyme genes altered amylose content and the chain-length distribution of amylopectin (Fujita et al. 2011; Zhang et al. 2011; Abe et al. 2014; Toyosawa et al. 2016). Such double-mutant lines could reveal how starch biosynthesis genes interact with each other and with the interrelationship between starches and storage proteins and phospholipids. NILs with multiple QTLs and genes would also be more effective at developing higher-grain-quality cultivars than NILs with single ones. For instance, combining a grain chalkiness QTL and a functional impaired allele of granule-bound starch synthase I (GBSSI) gene could produce and generate novel rice cultivars with low grain chalkiness and high eating quality.

Because the development of grain chalkiness is complex and is determined by multiple QTLs and genes, the development of new cultivars with high grain quality might be challenging for traditional breeding approaches. Other breeding concepts would be more appropriate for improving grain quality controlled by multiple QTLs with small genetic effects. Breeding by genomic selection can estimate the genetic effects (breeding value) of each locus on the whole genome by simultaneously accounting for all SNP markers (Meuwissen et al. 2001). Although genomic selection has been used mainly in dairy cattle breeding (Schaeffer 2006), methods that consider multiple genetic loci simultaneously offer powerful tools to reduce grain chalkiness in the development of novel rice cultivars.

Recent studies have identified many QTLs and genes involved in grain chalkiness, but many more loci are needed for the improvement of current rice cultivars. Advances in research are still needed to reveal the molecular mechanisms of the development of grain chalkiness in rice.