Annual rice consumption per person has dropped from 135.6 kg/capita/year in 1979 to 69.0 kg/capita/year in 2013. In conjunction with the drop in consumption, customers’ requirements have diversified, with consumers demanding functional and healthful, as well as tasty rice. Various qualities of rice, such as its physical appearance, cooking properties, eating properties, and, more recently, nutritional value, affect consumers’ acceptance of the product and the product’s market value. Therefore, improving the rice varieties according to consumers needs were main objectives of rice breeding activities (Fitzgerald et al. 2009). The potential of colored rice as a functional food with health benefits (Choi et al. 1994; Kang et al. 1996), as well as its potential as a natural pigment (Choi et al. 1996) and in cosmetics, has been studied.

In the past, the rice bran and germ produced during the milling process was disposed of, without utilizing these by-products of rice production. However, following research showing that rice bran includes a variety of useful components, such as oryzanol, phytic acid, ferulic acid, arabinoxylan, and arabinogalactan, its potential as a new functional ingredient has received increased attention. In particular, pigmented rice bran, rather than the bran of normal rice, is attracting attention because of its high antioxidant activity (Cho et al. 1996; Osawa 1995).

Genetic fortification of rice grain with functional compounds for human health is one of the major objectives in breeding programs for better and diversified grain quality in rice. Colored rice contains a variety of natural colors, ranging from dark purple to maroon and green (Park et al. 1998). Black and dark purple rice are known to contain a large amount of anthocyanin pigments, whereas red and maroon rice contain tannins. The anthocyanin pigment of black-purple rice contains a large amount of cyanidin- 3-glucoside (C3G), up to 80%, as well as peonidin-3- glucoside (P3G), malvidin, pelargonidin, and selphinidin (Reddy et al. 1994; Cho et al. 1996; Ryu et al. 1998).

C3G is the key component of anthocyanin, which is responsible for the color pigmentation in black-purple rice, and is known to have antioxidative potential (Reddy et al. 1995). A Korean blackish-purple rice cultivar, Heugjinju, was reported to contain 470 mg/100g of C3G, which is the highest concentration among cultivated rice in Korea (Kim et al. 2000). The grain color of black-purple rice is known to be controlled by the complementary action of two dominant genes, Pb(Prp-b) and Pp(Prp-a), located on chromosome 4 and 1, respectively (Hsieh & Chang 1964; Yoshimura et al. 1997). The other grain color red also has complementary gene as Rc and Rd (Kinoshita 1984; Nagao & Takahashi 1956).

This study was carried out to investigate the inheritance of the concentration of C3G in segregate population and to explore the relationship between the intensity of pigmentation and the concentration of C3G.

Plant materials

Five black-purple rice varieties, Jilinheimi, Heugjinju, Heidao38, No2, and LK1A-2-12-1-1, and two white rice varieties, Hwacheong super giant embryo and Heugbal, were used as parents of four populations. We performed two crosses each the black-purple variety/black-purple variety and black-purple variety/white variety. The black- purple/black-purple crosses were Jilinheimi/Heidao38 and Jilinheimi/LK1A-2-12-1-1, and the black-purple/white crosses were Heugjinju/Hwacheong-ges and No2/Heugbal.

The colored rices and white rices were crossed in a greenhouse experiment in the summer and subsequent generations were advanced in a greenhouse in the winter. The F₂ population was sown and transplanted in the central region of Korea. In the F₂ population, the plants were harvested one by one, and F₃ seeds were selected from 150 plants. In the analyzed of the F₂ segregation ratio, unhulled seeds of F₃ were used, counted black-purple, red-brown and white seeds by visual, separately.

Phenotypic traits

Color measurements were expressed as tristimulus parameters (L*, a*, and b*) with a color difference meter (Chroma meter CR-200, Minolta, Japan). L* indicate lightness (100 = white, and 0 = black). a* indicates redness- greenness (positive = red, and negative = green), and b* indicates yellowness?blueness (positive = yellow, and negative = blue).

Extraction and detection of C3G

The C3G pigment from 2 g of brown rice was extracted and pulverized with 0.1% Trifluoroacetic acid (TFA) and 20 ml of 95% ethanol solvent, three times every 4 h (Kwon et al. 2011). The extracted solution was filtered with filter paper (Whatman No.2) and concentrated under reduced pressure. The extracted solution was filled up with 2 ml to HPLC analysis. C3G in anthocyanin extracted from the seed coat was analyzed with HPLC (waters 501 pump, millipore gradient controller, waters 480 UV-Vis detector). An ODS-5 column (4.6 mm×250 mm, Nomura Chemical Co., Japan) was used, and detect wavelength was 530 nm. The mobile phase was 0.1% TFA in water, the linear gradient was 0.1% in acetonitrile, and the flow rate was 1.0 ml/min

The segregation ratio of black:brown:white in the F₂ populations of crosses between the black and white rice was 9:3:4, in accordance with previous reports (Table 1). The segregation ratio differed in the F₂ population crosses between the black rice, there were all black. These results indicate that two dominant complementary genes control the color pigmentation in the black-purple pericarp and that the pericarps of Jilinheimi, Heidao38 and LK1A-2-12-1-1 possess the same genes for color pigmentation. However, the depth of pigmentation varied continuously within the color classes, as measured by a color difference meter, implying that other genes might also be involved in the color expression (Fig 1). The C3G concentration of the brown rice exhibited continuous variations, with a tendency toward a lower value in all the F₂ populations (Fig. 2). C3G concentration varied widely, depending on the crosses and also found varying more or less transgressants. The variation in the C3G concentration in F₂ crosses did not correspond with that of color pigmentation. However, there were significant correlations between the C3G concentration and L* (lightness) and b* (yellowness) values (Table 2).

Genetic segregation of grain color

Regardless of the combination, all the F₁ seeds were black, suggesting that a dominant gene affects the grain color. The F₂ seed grain color in the black/black combination, Jilinheimi/Heidao38 and Jilinheimi/LK1A-2-12-1-1, was blackish, demonstrating that the parent of the black/black crosses had the same genes for black grain color.

The results of an χ2 analysis of the segregation of the grain color in the F₂ population of the black/white crosses, Heugjinju/Hwacheong-ges and No2/Heugbal, was 9:3:4 (Table 1), indicating that Heugjinju and No2 has two genes associated with grain color and that these are complementary.

Segregation and heritability of C3G content

The F₁ and F₂ seeds of the black/black crosses were all black-purple, but the seed grain color of the F₃ seeds from the F₂ segregated generation showed a continuous variation, as shown in Figure 1. A colorimeter revealed that the shape of the C3G content in the F₂ population from a combination of the black/black crosses was similar. As expected, transgressive segregation occurred in the Jilinheimi/Heidao38 and Jilinheimi/LK1A-2-12-1-1 crosses (Fig. 2).

Correlations between colorimeter score and C3G content

The C3G content showed a significant negative correlation with L* (lightness) and b* (yellowness), but different aspect showed with a* (redness) by each combination (Table 2). In the Jilinheimi/Heidao38 combination, the lightness, redness, and yellowness had a highly significant negative correlation with the C3G content. The colorimeter values revealed a highly significant positive correlation with each L* (lightness), a* (redness), and b* (yellowness) in theses crosses. In the Jilinheimi/LK1A-2-12-1-1 combination, the C3G content had a significant negative correlation with L* (lightness) and b* (yellowness) but not with a* (redness). The colorimeter values correlation in L* (lightness) with b* (yellowness) and with a* (redness) were highly positive correlation, but negative correlation in a* (redness) with b* (yellowness). The correlation of the C3G content with L* (lightness) and a* (redness) differed, depending on the mating parents.

In particular, the separation aspects of the a*/b* was similar to the separation of the C3G content, which is thought to be due to a*/b* values representing the anthocyanin content (Fig 1 and Fig 2).

The potential of rice as a functional food has attracted attention in recent years, with research focusing, in particular, on dark purple rice. Since the early 1990s, dark purple crop varieties have been introduced from China. Heugjinju (Moon et al. 1998) and Heugnam (Ha et al. 1998) has developed at rural development administration (RDA), but the yield of these varieties has remained at 70% as compared to that of normal varieties (Jung 2000).

The aim of the present study was to obtain basic data on color of pericarp using some pigmented rice varieties and to investigate the genetic underpinning selection for a high C3G content. The C3G content of some lines in the F₂ population was considerably higher or lower than that of the parents. This phenomenon is referred to as transgressive segregation. Transgressive segregation for quantitative traits is very common, but the genetic basis is largely unknown (Liang et al. 2009). Transgressive segregation is a quantitative trait, and the C3G content is controlled by a large number of genes. The Jilinheimi/Heidao38 and Jilinheimi/LK1A-2-12-1-1 combinations showed marked separation transcendence. The objects are separated beyond seems to be due to the cumulative effect or interaction effect of the gene to promote C3G content thought to be useful in the selective breeding. The action of complementary genes may be the primary cause of transgression, although overdominance and epistasis also contribute (Rieseberg et al. 1999). In the present study, standardized heritability of the C3G concentration estimated through selected F₂ and F₃ experiments was significantly high. These are suggest that selection for the C3G concentration in early generations would be effective.

Studies have suggested that the genetic systems conferring color pigmentation and the C3G content might not be the same but inter-related in a specific pathway associated with anthocyanin metabolism. This study can be partly explained by the fact that significant variations exist in the concentration of C3G, even among black-purple rice varieties (Park et al. 1998; Ryu et al. 1998). The results of χ2 analysis of the segregation of F₂ population’s grain color in black/white crosses, Heugjinju/Hwacheong-ges and No2/Heugbal was 9 : 3 : 4. Jung et al. (2000) and Park et al. (2000) reported the same segregation ratio in Heugjinju/IR701078-AC3(white) crosses and Hyangmi (white)/Heugjinju crosses. Shen et al. (2009) reported that the color parameters of rice grain were positively correlated with the total phenolic content, flavonoid content, and antioxidant capacity among a wide collection of rice germplasms, including white, red and black rice, but that the correlations among the white rice accessions were rather weak. Jung et al. (2000) reported that the C3G content of an F₂ population had a highly significant negative correlation with colorimeter a* (redness) with b* (yellowness).

To shed light on the mechanism underlying transgression, it will be necessary to examine the interaction and activity of the dominant allele on heterozygote (Dooner et al. 1991). It will also be necessary to investigate the composition and quantitative differences in anthocyanin in the tissues of other plants, such as corn, snapdragon, and petunia, to shed light on the interaction between the structure and function of regulatory genes involved in the biosynthesis of anthocyanins.