Verification of the inheritance pattern of rice pericarp color
The Rc-controlled pigments accumulate in the lower-epidermal cell layer of the pericarp, which is maternal tissue (Gu et al. 2011), suggesting that the pericarp color is dependent on the maternal genotype (Sweeney et al. 2006). To verify the inheritance pattern of rice pericarp color, we conducted a forward cross and reverse cross between H5Rc (brown pericarp) and H5rc (white pericarp) (Fig. 2). If the mother plant had a white pericarp (rcrc), all the F1 seeds produced white pericarp; if the mother plant had a brown pericarp (RcRc), all the F1 seeds produced brown pericarp. Thus, pericarp color of the F1 seeds is dependent on the mother’s genotype rather than on the genotype of the F1 seeds.
All of the F2 seeds that the F1 plants (Rcrc) produced were brown, showing no genetic segregation, irrespective of whether the female parent’s pericarp was white or brown, indicating that the pericarp color of the F2 seeds was dependent on the maternal (F1 plant) genotype rather than on that of the F2 seeds. The F2 plants produced two types of F3 seeds: brown and white, showing that the pericarp color of the F3 seeds was dependent on the maternal (F2 plants) genotype rather than that of the F3 seeds. Therefore, Rc has a maternal effect on pericarp color.
An investigation of 576 F2 plants (that produced two types of F3 seeds: brown and white) showed that the brown:white segregation ratio of 437:139 (≈ 3.14:1) corresponded to the 3:1 Mendelian expectation (χ2 = 0.1875; P > 0.833), further confirming that pericarp color is controlled by a single completely dominant gene, Rc.
These observations indicate that Rc has a completely dominant maternal effect on pericarp color; thus, we speculate that Rc has a completely dominant maternal effect on PHS resistance.
The PHS resistance effect of Rc and release ofRc-controlled dormancy in japonica rice
To further ameliorate the effects of genetic background and environment, we developed a NIL-derived population. From this population, we randomly sampled the tagged panicles and evaluated PHS resistance immediately after each harvest. To ensure random sampling, we did not detect sample genotype before evaluating PHS resistance for each harvest. During the seed maturation period (from 21–49 DAH), the SPs of seeds produced by maternal plants of the RcRc, Rcrc, and rcrc genotypes all showed an increasing trend, with ranges of 25.8%–74.3%, 0–26.9%, and 0–23.8%, respectively (Fig. 3a). A Chi-square test showed that the SPs of seeds from maternal plants of the RcRc and Rcrc genotypes were both significantly lower than that of the rcrc genotype at each DAH by an average of ~43%, peaking at 55.0% (SP value of rcrc minus that of the RcRc genotype) and 53.3% (SP value of rcrc minus that of the Rcrc genotype), respectively, at 42 DAHs. By contrast, there was no significant difference between the Rcrc and RcRc genotypes at any DAH, indicating that Rc had a completely dominant effect over rc on PHS resistance and that the seeds from maternal plants of the Rcrc genotype showed no genetic segregation in PHS resistance. Therefore, Rc has a completely dominant maternal effect on PHS resistance. As Rc-controlled dormancy may reduce the germination percentage (GP) of seeds in the next agricultural production cycle, causing reductions in quality of seeds for planting, we thus conducted dormancy-releasing experiments (Fig. 3b). In the control experiment, freshly harvested seeds were air-dried to a moisture content of ~13.5%. The GPs of seeds from maternal plants of the RcRc and Rcrc genotypes were significantly (~50%) lower than that of seeds from maternal plants of the rcrc genotype, further verifying that the Rc gene controls dormancy. There was no significant difference in GP between the Rcrc and RcRc genotypes, indicating that seeds from Rcrc plants showed no genetic segregation in dormancy.
For the con1 (treatment at 25 °C with 65% RH for 4 weeks) and the con2 (storage under ambient conditions in Hangzhou for 12 weeks from November 2018 to February 2019), the GPs of seeds from maternal plants of the RcRc, Rcrc, and rcrc genotypes were all greater than ~ 93%, with no significant difference, indicating that RcRc and Rcrc genotypes seeds can completely release dormancy within 12 weeks by after-ripening. Thus, releasing the dormancy retains the quality of the seed for planting, and Rc-controlled dormancy does not negatively affect the next agricultural production cycle. This is indirectly supported by the observation that red rice is cultivated widely around the world.
Effect of Rc on PHS resistance and release of Rc-controlled dormancy in indica hybrid rice
According to our results obtained in inbred japonica rice, all F1 seeds that are produced with brown-pericarp rice as the female parent are brown, i.e., resistant to PHS; all F2 seeds that F1 plants (Rcrc) produced are brown, i.e., resistant to PHS, with no genetic segregation, which confers an advantage to hybrid rice. Thus, we investigated further the effect of the Rc gene on PHS resistance and the release of Rc-controlled dormancy in indica hybrid rice.
In our hybrid seed (F1) production field where GA was applied, we randomly sampled tagged panicles (containing two types, brown- and white-pericarp rice) and evaluated PHS resistance for each harvest. During the seed maturation period (15–35 DAP), the SPs of brown and white hybrid seed both showed an increasing trend, ranging from 50.8–84.2% and 0–0.3%, respectively (Fig. 4a). A Chi-square test showed that the SP of brown hybrid seed was significantly lower than that of white hybrid seed at each DAP by an average of ~ 50%, peaking at 58.5% (the SP value of white hybrid seed minus that of brown hybrid seed) at 20 DAP. This result confirmed that GA application during hybrid seed production failed to counteract the effect of the Rc gene on PHS resistance (Gu et al. 2011).
The freshly harvested seeds were air-dried to a moisture content of ~ 13.5% and then stored under ambient conditions for 12 weeks in Hangzhou from November 2018 to February 2019. Seedlings established from brown and white hybrid seed showed no visible difference (Fig. 4b), further confirming that Rc-controlled dormancy does not negatively affect the next agricultural production cycle.
In our hybrid rice production (F2 seeds as food) field, we randomly sampled tagged panicles (containing two types, brown and white) and evaluated PHS resistance for each harvest. During the seed maturation period (20–40 DAHs), the SPs of brown and white hybrid rice both showed an increasing trend, ranging from 35.4–78.3% and 0–27.4%, respectively (Fig. 4c). A Chi-square test showed that the SP of brown hybrid rice was significantly lower than that of white hybrid rice by an average of ~ 48%, peaking at 59.1% (the SP value of white hybrid rice minus that of brown hybrid rice) at 30 DAHs. Therefore, the Rc gene has a significant effect on PHS resistance in indica hybrid rice.
Differences in milled rice qualities between the RcRc, Rcrc, and rcrc genotypes
Rice is generally consumed in the form of milled rice, from which the bran (seed coat and aleurone layer) has been removed. There is no visible difference between the appearances of milled rice of the RcRc, Rcrc, and rcrc genotypes (Fig. 5a), consistent with a report by (Gu et al. 2011) showing that the pigments accumulate in the lower epidermal cell layer of the pericarp tissue.
Starch, composed of amylose and amylopectin, is the major constituent of milled rice and accounts for 80–90% of its dry weight (Butardo et al. 2019). The cooking and eating quality of rice is routinely estimated, primarily on the basis of starch properties, using three main physicochemical tests: total amylose content (TAC), gel consistency (GC), and alkali spreading value (ASV) (Butardo et al. 2019). The results of our study showed no significant difference between the RcRc, Rcrc, and rcrc genotypes in TAC (Fig. 5b) or GC (Fig. 5d). The ASVs of the RcRc, Rcrc, and rcrc genotypes were all 7, meaning that all samples were completely dispersed and intermingled with no differences. The hot-water-insoluble amylose content (WIAC) differs markedly among different varieties with TAC in the same range (Bhattacharya et al. 1978). WIAC correlates more strongly with textural quality than does TAC (Bhattacharya 2009). Therefore, we determined the WIACs of the RcRc, Rcrc, and rcrc genotypes. There was no significant difference in WIAC values between the genotypes (Fig. 5c). Thus, the starch properties, as quantified by TAC, WIAC, GC, and ASV, did not significantly differ between the RcRc, Rcrc, and rcrc genotypes.
The rapid viscosity analyzer (RVA) profile of rice flour is an important predictor of rice cooking, eating, and processing quality characteristics (Champagne et al. 2004; Champagne et al. 1999). The viscosity curves are the most useful tool available for assessing cooking and eating quality of rice, with the results associated with sensory or processing attributes (Butardo et al. 2019). The viscosity curves can derive seven characteristic parameters (CPs): peak, trough, final viscosity (FinalVisc), breakdown, setback, initial gelatinization temperature (PastingTemp), and peaktime (Champagne et al. 2004; Champagne et al. 1999). We conducted an RVA profile analysis of the RcRc, Rcrc, and rcrc genotypes in triplicate; viscosity curves expressed as the mean of three replicates showed no obvious difference among the genotypes (Fig. 5e). Further analysis showed that the seven CPs derived from the viscosity curves also had no significant difference among the genotypes (Table 2).
CPs
|
Genotypes
|
Mean
|
SD
|
95% CI
|
Upper
|
Lower
|
Peak
(cP)
|
RcRc
|
3789.00 A*
|
140.82
|
3439.18
|
4138.82
|
Rcrc
|
3671.67 A
|
251.96
|
3045.75
|
4297.58
|
rcrc
|
3741.00 A
|
49.11
|
3619.00
|
3863.00
|
Trough
(cP)
|
RcRc
|
2232.33 A
|
113.46
|
1950.49
|
2514.17
|
Rcrc
|
2224.67 A
|
67.17
|
2057.80
|
2391.54
|
rcrc
|
2276.67 A
|
96.26
|
2037.54
|
2515.79
|
Breakdown
(cP)
|
RcRc
|
1556.67 A
|
47.93
|
1437.60
|
1675.73
|
Rcrc
|
1447.00 A
|
213.70
|
916.14
|
1977.86
|
rcrc
|
1464.33 A
|
120.56
|
1164.85
|
1763.82
|
FinalVisc
(cP)
|
RcRc
|
3770.00 A
|
110.18
|
3496.30
|
4043.70
|
Rcrc
|
3722.00 A
|
53.86
|
3588.20
|
3855.80
|
rcrc
|
3772.00 A
|
75.15
|
3585.33
|
3958.67
|
Setback
(cP)
|
RcRc
|
-19.00 A
|
45.43
|
-131.86
|
93.86
|
Rcrc
|
50.33 A
|
222.27
|
-501.81
|
602.47
|
rcrc
|
31.00 A
|
102.80
|
-224.36
|
286.36
|
PeakTime
(min)
|
RcRc
|
5.78 A
|
0.04
|
5.68
|
5.88
|
Rcrc
|
5.87 A
|
0.14
|
5.53
|
6.20
|
rcrc
|
5.84 A
|
0.15
|
5.47
|
6.22
|
PastingTemp
(°C)
|
RcRc
|
86.55 A
|
1.00
|
84.07
|
89.03
|
Rcrc
|
86.55 A
|
0.98
|
84.11
|
88.99
|
rcrc
|
87.23 A
|
0.06
|
87.09
|
87.38
|
Table 2: Summary of 7 characteristic parameters (CPs) derived from the viscosity curves, including the genotypic mean and its standard deviation (SD), and 95% credible intervals (CI). * The same characters (A and A) mean that their genotypic means have no significant difference at the P > 0.1. Definition of CPs refers to Champagne et al. (1999; 2004).
Aside from starch, the main macronutrients in rice are storage proteins and crude fat. The former, accounting for 6–8% of milled rice on a dry-weight basis, is a major source of food protein (Shih 2004) and plays an important role in nutritional quality, as well as in textural and sensory traits (Butardo and Sreenivasulu 2016). The latter, accounting for < 1% of milled rice on a dry-weight basis, plays a minor role in influencing pasting properties (Godber and Juliano 2004). We detected no significant difference in the crude protein content (CPC) (Fig. 5f) or crude fat content (CFC) (Fig. 5g) between the RcRc, Rcrc, and rcrc genotypes.
In conclusion, we detected no significant differences in milled rice qualities, including TAC, WIAC, GC, ASV, RVA profile properties, CPC, and CFC, between the RcRc, Rcrc, and rcrc genotypes. Thus, the Rc gene retains the same milled rice qualities in brown-pericarp rice as those in white-pericarp rice, indicating that the pigments accumulated in the pericarp do not negatively affect milled rice quality.
Differences between the RcRc, Rcrc, and rcrc genotypes in the antioxidative capacity of the bran
Red rice possesses a powerful antioxidative capacity that is derived mainly from its proanthocyanidins. Proanthocyanidin synthesis involves the Rc and Rd genes in red rice (RcRd); thus, proanthocyanidins are not detectable in brown-pericarp rice (Rcrd), which lack the Rd allele (Furukawa et al. 2006). We determined the antioxidative capacity of the RcRc, Rcrc, and rcrc genotypes using a 2,2-diphenyl-1-picrylhydrazyl (DPPH•) assay (Fig. 5h). The DPPH• inhibition percentage (DIP) of the samples was calculated as an indicator of antioxidative capacity. DIPs of RcRc and Rcrc rice were significantly higher (nearly double) than that of rcrc rice, whereas there was no significant difference in DIP between Rcrc and RcRc rice. Although proanthocyanidins were not detected in the brown-pericarp rice (Furukawa et al. 2006), the brown-pericarp rice possessed a higher antioxidative capacity than white-pericarp rice.