Multi-omics resources for targeted agronomic improvement of pigmented rice

Pigmented rice (Oryza sativa L.) is a rich source of nutrients, but pigmented lines typically have long life cycles and limited productivity. Here we generated genome assemblies of 5 pigmented rice varieties and evaluated the genetic variation among 51 pigmented rice varieties by resequencing an additional 46 varieties. Phylogenetic analyses divided the pigmented varieties into four varietal groups: Geng-japonica, Xian-indica, circum-Aus and circum-Basmati. Metabolomics and ionomics profiling revealed that black rice varieties are rich in aromatic secondary metabolites. We established a regeneration and transformation system and used CRISPR–Cas9 to knock out three flowering time repressors (Hd2, Hd4 and Hd5) in the black Indonesian rice Cempo Ireng, resulting in an early maturing variety with shorter stature. Our study thus provides a multi-omics resource for understanding and improving Asian pigmented rice.

- Supplementary Fig. 1. Dot plot comparing the genome assemblies of the five varieties and the reference Nipponbare genome. - Supplementary Fig. 2. Genomic landscapes of the 12 chromosomes in the five genome assemblies of pigmented rice. - Supplementary Table 1. Summary of whole-genome sequence assembly, annotation, gene prediction, and structural variants. Table 2. BUSCO assessment of the five genome sequence assemblies.

Relationship between grain pigments and metal ions
To investigate relationship between pigment enrichment in the rice grain and metal ion concentration, we performed a correlation test between anthocyanin or proanthocyanidins and the 22 trace metal ion concentrations using GraphPad Prism 9.3.1. Pearson's correlation coefficients (R) and P-values (P) were calculated. The correlation was considered to be significant at 0.05 level. The abundance of anthocyanin was positively correlated with Fe, Mn, V, Cr, Mo, and B and negatively correlated with Mg. However, the abundance of proanthocyanidins was negatively correlated with Fe, Cr, and, Mo concentrations (Supplementary File 5). These inverse relationships detected for Fe, Cr, and Mo and anthocyanin/ proanthocyanidins were not previously reported for whole grain rice varieties and merits further mechanistic investigation.

Supplementary S4
Methods Embryogenic callus induction and regeneration. Mature dry seeds of Cempo Ireng (local to the Yogyakarta region, Indonesia) were used in this study. The seeds were dehusked and sterilized with 70% ethanol for 1 min and 30% (v/v) commercial bleach (5.25% sodium hypochlorite) for 40 min with continuous shaking, and then rinsed five times with sterilized distilled water and dried on Whatman paper for 5 min. For callus induction, 36 seeds were inoculated per Petri dish on the callus-induction media 2N6 or 2NBK (modified from the protocol developed by Hiei and Komari 11 and incubated at 30°C and 32°C in the dark or under continuous light for 7 d. All culture media components and preparations are detailed in Supplementary S6. The use of 2NBK medium at 32°C under continuous light were selected as the best conditions for callus induction based on the induced callus mass and quality. The scutella were subcultured onto fresh 2NBK medium for another seven days under the same conditions. The embryogenic calli were induced by subculturing the calli on nNBKC for five days at 32°C. Ten different regeneration media were developed and tested for shoot induction based on MS medium and different growth regulator combinations (Supplementary S6). The embryogenic calli were selected and placed on the tested media at 32°C under continuous light for 14 days. The regeneration frequency was calculated based on the number of regenerated shoots compared with the total number of calli. For the development of roots, the regenerated shoots were transferred into magenta boxes containing MSRO rooting medium under the same temperature and light conditions.
Agrobacterium-mediated transformation. The binary vector (pRGEB32) used in this study contains Hygromycin phosphotransferase (Hpt) as a selectable marker under the control of the CaMV 35S promoter, and the Cas9 gene under the control of the rice Ubiquitin promoter. The vector was transferred into Agrobacterium tumefaciens strain EHA105 by electroporation in a Bio-Rad Laboratories Escherichia coli pulser. Bacterial culture and transformation were conducted according to Hiei and Komari 11 , with some modifications. After the co-cultivation step, the calli were placed on the first selection medium (NBKCH20) for 14 days and were then moved to the second selection medium (nNBKCH40) for five days. The calli were incubated in all steps from callus induction to the rooting stage at 32°C and under continuous light, except for the co-cultivation step, which took place at 25°C in the dark. The embryogenic calli were placed on the R8H5 medium containing 5 mg hygromycin for selection. The regenerated shoots were grown on the rooting medium (MSROH5), after which the seedlings were acclimatized in the soil and greenhouse conditions.
Genomic DNA extraction and genotyping. Total genomic DNA was extracted from approximately 0.5 g of fresh leaves from the transformed plants using the DNAquick Plant System (Tiangen Biotech), according to the manufacturer's protocol. A PCR screening was performed to amplify 240 bp of the CRISPR transgene using Phusion High-Fidelity DNA Polymerase (Thermo Fisher Scientific) and the Cas9-F7/Nos-R7 primer pair (Supplementary Table 8). The PCR products were analyzed by gel electrophoresis on 1% agarose gels.
Phenotyping of CRISPR/Cas9-targeted Cempo Ireng mutants. The genome-engineered plants and the wild-type controls were grown in the KAUST greenhouse rooms at 28°C under natural sunlight. Fifteen random homozygous mutants were selected (5 plants per target) and 5 wild-type controls for detailed agronomic trait analysis. The heading date was recorded as the first day on which the first panicle emerged. The plant height was measured on the same day using the measuring scale. The plants were grown until fully mature, at which point the plant tillers were counted, the panicles were harvested, the number of filled and unfilled spikelet were counted in each panicle and based on that seed setting rate (the ratio of number of filled grains to total number of spikelets) was calculated. The rough, paddy rice grain length and width were measured by ImageJ software, and 1000 paddy grains had husk removed and were cleaned prior to weight measurements with a sensitive electronic balance (0.001 g sensitivity). The total yield per plant was measured as the total grams of grain collected per plant. All data were statistically analyzed by unpaired t-test using GraphPad Prism 9.3.1. and significant difference were indicated by asterisks: *, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001.

Establishment of a regeneration and transformation system for BR
To establish a regeneration protocol for Cempo Ireng, we tested callus inducibility using the established protocol for japonica and indica rice developed by Hiei and Komari 11 with some modification. We used the mature grains of Cempo Ireng for in vitro callus induction. We tested the medium we developed (CIM57) and another two media developed by Hiei and Komari 11 with our modifications (see Supplementary S6, below), called 2N6 and 2NBK. We also tested two different incubation temperatures (30°C and 32°C) and two different light conditions (complete darkness and continuous light).
No callus was induced using CIM57 supplemented with dicamba and 1-naphthaleneacetic acid (NAA) growth regulators; however, a typical yellowish and friable callus mass emerged from the grain using media containing 2,4dichlorophenoxyacetic acid (2N6 and 2NBK) under all test conditions ( Supplementary Fig. 7e). We observed an increase in the number of induced calli when the temperature was increased to 32°C and in the presence of light. Callus induction frequency varied between 26.3% and 30.6% across the tested conditions, with the highest callus induction frequency obtained using 2NBK at 32°C in light ( Supplementary Fig. 7e). Based on these results, we selected 2NBK as the best callus-induction medium and 32°C and continuous light as the conditions for the further regeneration and rooting stages ( Supplementary Fig. 7).
Somatic embryos were successfully induced on the nNBKC medium containing a high concentration of sorbitol (55 g/L) and sucrose (20 g/L) (Supplementary S6). We tested ten different Murashige and Skoog (MS) basal media supplemented with different concentrations of various growth regulators (R1-R10) for their shoot induction abilities ( Supplementary Fig. 7f, Supplementary S6). We did not achieve regeneration using R1, R5, and R7; however, the other seven media induced green shoots after 12-16 days of cultivation, with frequencies ranging from 33.3% to 83.3%. The highest regeneration frequency (83.3%) was achieved using R8 medium, which contained a 1:2 combination of NAA and BAP ( Supplementary Fig. 7f). The shoots were healthy and developed a good root system on MSRO rooting medium for maturation into a seedling. For acclimatization, the seedlings were transferred into a greenhouse room at 28°C.
Next, we used our optimized regeneration protocol to establish a genetic transformation protocol. We tested the Agrobacterium tumefaciens-mediated protocol developed by Hiei and Komari 11 to transform proliferating Cempo Ireng calli. We used the pRGEB32 vector harboring hygromycin as a selectable marker and delivered it into Agrobacterium tumefaciens strain EHA105. We transfected the induced calli and selected them using 20 mg/L hygromycin for 14 days and 40 mg/L hygromycin for five days (Supplementary Fig. 8b-c). We observed no shoot regeneration of the transformed calli in the presence of a high concentration of hygromycin (50 mg/L). To determine the optimal hygromycin concentration for selection of transformed calli, we tested different levels of hygromycin (5-50 mg/L), and found that 5 mg/L was sufficient for selection of transformed shoots ( Supplementary Fig. 8d). Using these conditions, we found that 78.7% of the regenerated plants had successfully integrated T-DNA sequence into their genome.

Supplementary S6
Media Setup: CIM57: Add 4.3 g Murashige and Skoog (MS) basal medium, 30 g sucrose, 1 g casein hydrolysate. Adjust pH to 5.8 and add 4 g agar and 3 g Gelzan. Autoclave at 121°C for 20 min, cool to 60°C, then add 2 mg of filter-sterilized dicamba and 10 mg of filter-sterilized AgNO3.
NBKCH20: Add 0.4 mL of 50 mg/mL hygromycin B, and 1 mL of 200 mg/mL timentin to the autoclaved 1L 2NBK medium after cooling to 60°C.