RNAi-mediated suppression of three carotenoid-cleavage dioxygenase genes, OsCCD1, 4a, and 4b, increases carotenoid content in rice

Abstract

Carotenoids of staple food crops have a high nutritional value as provitamin A components in the daily diet. To increase the levels of carotenoids, inhibition of carotenoid-cleavage dioxygenases (CCDs), which degrade carotenoids, has been considered as a promising target in crop biotechnology. In this study, suppression of the OsCCD1, OsCCD4a, and OsCCD4b genes using RNAi was verified in transgenic rice plants by quantitative RT-PCR and small RNA detection. Leaf carotenoids were significantly increased overall in OsCCD4a-RNAi lines of the T1 generation, and the highest accumulation of 1.3-fold relative to non-transgenic plants was found in a line of the T2 generation. The effects on seed carotenoids were determined via cross-fertilization between β-carotene-producing transgenic rice and one of two independent homozygous lines of OsCCD1-RNAi, OsCCD4a-RNAi, or OsCCD4b-RNAi. This showed that carotenoids were increased to a maximum of 1.4- and 1.6-fold in OsCCD1-RNAi and OsCCD4a-RNAi, respectively, with a different preference toward α-ring and β-ring carotenoids; levels could not be established in OsCCD4b-RNAi. In addition, the contents of four carotenoids decreased when OsCCD1, OsCCD4a, and OsCCD4b were overexpressed in E. coli strains accumulating phytoene, lycopene, β-carotene, and zeaxanthin. OsCCD1 and OsCCD4a had a similar high carotenoid degrading activity, followed by OsCCD4b without substrate specificity. Overall, our results suggest that suppresing OsCCD4a activity may have potential as a tool for enhancing the carotenoid content of seed endosperms and leaves in rice.

Fig. 1.

Endogenous transcript levels of OsCCD1, OsCCD4a, and OsCCD4b in various tissues at different developmental stages in rice. (a) The relative expression levels of the genes quantified by qRT-PCRs in leaves (L), roots (R), flowers (F), and seeds (Se) harvested at 40 d after flowering (DAF), at the seedling stage (SS), vegetative stage (VS), and reproductive stage (RS). (b) The relative expression levels of the genes quantified by qRT-PCRs during the development of seeds at 10, 15, 20, 30, and 40 DAF and at an additional stage that was seeds at 40 DAF that had been desiccated for 1 week (40D). The mean (±SE) Ct values of triplicate measurements were used to calculate the expression of the target gene with normalization to an internal control (OsUbi5) using the ΔCt (a) and ΔΔCt (b) methods.

Fig. 2.

Schematic representation of rice OsCCD1, OsCCD4a, and OsCCD4b gene structures and binary vectors used in this study. (a) The conserved position and numbers of key amino acids known for enzymatic action of CCDs and transit peptide regions predicted using the ChloroP program. (b) Diagram of RNAi-mediated suppression vectors of the three genes. The pstPAC vector used to endow the endosperm of stPAC rice with a carotenoid-intensifying golden color trait was developed in a previous study (Jeong et al., 2017) and is presented here because stPAC was used as a male parent in conventional interbreeding in this study. Dotted arrows in (a, b) indicate the gene region used in vector construction for RNAi-mediated suppression. In (b) the bacterial attachment attB sites needed for Gateway cloning are marked with hatched boxes. RB, right border; LB, left border; NPT II, neomycin-resistant gene cassette; BAR, bialaphos-resistant gene cassette; HPT, hygromycin-resistant gene cassette; Ubi-P, maize ubiquitin 1 promoter and 1st intron including splicing acceptor site; Nos-T, 3′-region from the nopaline synthase gene; Glb-P, rice globulin promoter; PinII-T, the 3′-region of the potato proteinase inhibitor II gene; stPsy, rice codon-optimized synthetic gene encoding Capsicum phytoene synthase (PSY); 2A, rice codon-optimized foot-and-mouth disease virus 2A peptide; Tp, transit peptide of rice Rubisco small subunit; stCrtI, rice codon-optimized synthetic gene encoding bacterial desaturase (CRTI).

Fig. 3.

Target gene suppression in the transgenic rice lines OsCCD1-RNAi, OsCCD4a-RNAi, and OsCCD4b-RNAi. (a) Knock-down levels of each target gene were quantified by qRT-PCRs using T₁ leaf tissues of two sibling lines from three independent transgenic plants for each construct. All data are the means (±SE) of triplicate measurements. The mean Ct values were used to calculate the expression of the target gene with normalization to an internal control (OsUbi5) using the ΔΔCt equation. The relative differences to non-transgenic (NT) plants (Oryza sativa cv. Ilmi) were determined using a one-tailed Student’s t-test: ***P<0.001, **P<0.01, *P<0.05. (b) The siRNA detection of each target gene by small RNA gel blot analysis was performed with T₂ leaf tissues of two independent transgenic plants for each construct.

Fig. 4.

Contents and composition of leaf carotenoids in the transgenic rice lines OsCCD1-RNAi, OsCCD4a-RNAi, and OsCCD4b-RNAi. (a) Carotenoid levels by HPLC analysis in T₁ leaf tissues of two sibling lines from three independent transgenic plants for each construct. (b) Levels of carotenoids, measured by HPLC, and chlorophylls, measured spectrophotometrically using absorbance, in T₂ leaf tissues of two independent transgenic plants for each construct. NT, non-transgenic rice (Oryza sativa cv. Ilmi). Data are means (±SD) of three replicates. Differences relative to NT plants were determined using a one-tailed Student’s t-test: ***P<0.001, **P<0.01, *P<0.05.

Fig. 5.

Carotenoid levels and polished color of interbred seeds of rice from two independent lines of OsCCD1-RNAi, OsCCD4a-RNAi, and OsCCD4b-RNAi and a stPAC plant displaying carotenoid-accumulating golden colored seeds. (a) Carotenoid levels, measured by HPLC, of homozygous F₄ seeds containing two transgenes: stPAC and one of the RNAi genes, compared with non-transgenic (NT) plants (Oryza sativa cv. Ilmi). Data are means (±SD) of three replicates. The relative difference compared with each nullizygous (N) rice seed was determined using a one-tailed Student’s t-test. The nullizygous rice line has only a stPAC gene and is without the RNAi gene for each of the OsCCDs after being segregated from interbreeding lines (b) Fold-changes in individual carotenoid components of homozygous F₄ seeds relative to that of each N seed. Differences between groups relative to a value of 1 were also determined using a one-tailed Student’s t-test. All significant differences in (a, b) are indicated as ***P<0.001, * P<0.01, *P<0.05. (c) Phenotypes of endosperm colors in homozygous F₄ seeds after polishing. Line numbers with significantly enhanced carotenoid contents in (a, b) and a slightly more intense golden color compared with each N line are highlighted in bold.

Fig. 6.

Changes in carotenoid levels and colony color according to the individual overexpression of OsCCD1 OsCCD4a, and OsCCD4b in four carotenoid-accumulating E. coli strains. (a) Phytoene contents measured by HPLC in a pPHYT-harboring E. coli strain. (b) Lycopene contents measured by HPLC in a pLYC-harboring E. coli strain. (c) β-Carotene contents measured by HPLC in a pBETA-harboring E. coli strain. (d) Zeaxanthin contents measured by HPLC in a pZEAX-harboring E. coli strain. Null represents each of the four vector-harboring E. coli strains without the rice CCD genes. The relative difference of each null E. coli strain was determined using a one-tailed Student’s t-test: ***P<0.001, **P<0.01, *P<0.05.