Biochemical Characterization of Orange-Colored Rice Calli Induced by Target Mutagenesis of OsOr Gene

We generated an orange-colored (OC) rice callus line by targeted mutagenesis of the orange gene (OsOr) using the CRISPR-Cas9 system. The OC line accumulated more lutein, β-carotene, and two β-carotene isomers compared to the WT callus line. We also analyzed the expression levels of carotenoid biosynthesis genes by qRT-PCR. Among the genes encoding carotenoid metabolic pathway enzymes, the number of transcripts of the PSY2, PSY3, PDS, ZDS and β-LCY genes were higher in the OC line than in the WT line. In contrast, transcription of the ε-LCY gene was downregulated in the OC line compared to the WT line. In addition, we detected increases in the transcript levels of two genes involved in carotenoid oxidation in the OC lines. The developed OC lines also showed increased tolerance to salt stress. Collectively, these findings indicate that targeted mutagenesis of the OsOr gene via CRISPR/Cas9-mediated genome editing results in β-carotene accumulation in rice calli. Accordingly, we believe that this type of genome-editing technology could represent an effective alternative approach for enhancing the β-carotene content of plants.


Introduction
In plants, carotenoids play important roles in photosynthesis as auxiliary pigments that harvest energy from the sun and subsequently deliver it to chlorophyll [1,2]. In addition, these pigments not only act as precursors in abscisic acid (ABA) biosynthesis but are also a rich source of provitamin A and thus are of nutritional importance in human diets [3,4]. Carotenoids also play a protective role by interacting with reactive oxygen species to prevent oxidative damage to cells, and serve as useful ecological signals that can be used by pollination and seed dispersal agents, such as insects and birds, to distinguish different plants [5]. These pigments tend to accumulate in specific organs such as fruits, roots, and flowers, thereby conferring their characteristic coloration [6][7][8][9][10]. Carotenoids, which accumulate as red, orange, and yellow pigments, are synthesized in plastids, among which β-carotene, lycopene, and lutein are important constituents exploited by the food and petroleum industries owing to their powerful antioxidant activities [11].
Rice is a staple food crop and a major source of nutrition for more than 40% of the world's population. Numerous studies have been conducted to examine the accumulation of carotenoids in the endosperm rice grains [8,12]. Although it has been established that rice endosperm is not normally involved in carotenoid synthesis, it has been reported that not only geranylgeranyl diphosphate (GGPP), a carotenoid precursor, but also several genes in the carotenoid biosynthetic pathway, including β-LCY, are found in rice endosperm [8,13]. For efficient biosynthesis of β-carotene in rice endosperm, transgenic Golden Rice 2 was developed, which expresses the maize phytoene synthase (PSY) and bacterial CrtI genes under the regulation of the glutelin promoter [8]. Enhanced levels of zeaxanthin, β-carotene, and total carotenoid contents have also been achieved in transgenic potatoes by controlling the expression of LCY-b, CHY-ε, and ZEP [14,15]. In addition, tomato plants overexpressing the PSY, LCY-b, and CHY-b genes were characterized by dwarf phenotypes and fruits with higher contents of lycopene, carotene, and zeaxanthin than the control plants [16]. The orange (Or) gene of plants was first isolated from an orange cauliflower mutant that showed variation in the content of β-carotene [17]. Since this discovery, the gene has been identified in numerous plant species and has been reported to encode cysteine-rich DnaJ proteins confined to plastids and regulate carotenoid accumulation and abiotic stress resistance in various plant species [18]. Recently, overexpression of the Arabidopsis thaliana Or gene in conjunction with that of the PSY and crtl genes was found to enhance carotenoid accumulation in rice calli and maize seeds [13,19]. A single amino acid alteration in wildtype OR significantly enhanced carotenoid accumulation in potato tubes, sweet potato calli, and the melon Cucumis melo [5,17,18,20]. AtOr-His, a variant of the gene that carries golden SNPs in Arabidopsis thaliana, was overexpressed in Arabidopsis callus and tomato (Solanum lycopersicum), leading to a high-level accumulation of carotenoids [21,22]. In addition to its role in carotenoid accumulation, orange gene expression in plants enhances tolerance to environmental stress. For example, in sweet potatoes, overexpression of IbOr has been found to enhance tolerance to heat and oxidative stress [23,24], whereas in alfalfa and potato, overexpression of IbOr has been observed to confer resistance to abiotic stress factors, such as salt, drought, and heat [25,26]. Recently, Endo et al. successfully increased β-carotene accumulation in rice calli by directly modifying the rice OsOr gene via CRISPR/Cas9-mediated genome editing, which disrupted the proper splicing of OsOr, similarly to the cauliflower [17]. In addition, Yu et al. reported that overexpression of the rice orange gene OsOr negatively regulated carotenoid accumulation, leading to higher tiller numbers and decreased stress tolerance in Nipponbare rice [27].
In this study, we generated an orange-colored (OC) rice callus line by targeted mutagenesis of the OsOr gene using the CRISPR/Cas9 system. Our results showed that OC lines contained elevated total carotenoid content, with significant increases in the levels of lycopene, β-carotene, and the two β-carotene isomers. Moreover, we performed gene expression analysis using quantitative RT-PCR to understand the mechanisms associated with the accumulation of orange color.

CRISPR/Cas9 Targeted Mutagenesis of OsOr
To understand the role of OsOr in the mechanisms associated with the accumulation of orange color, we created an orange-colored (OC) rice callus line by targeted mutagenesis of the OsOr using the CRISPR/Cas9 system. OsOr encodes a DnaJ cysteine-rich zincbinding domain consisting of eight exons ( Figure 1A). To obtain OsOr mutants, three single guide RNAs (sgRNAs), sgRNA1, sgRNA2, and sgRNA3, were designed to target the first and second exons, and a region between the third exon and the third intron, respectively ( Figure 1A). Three pBOsC binary vectors containing each sgRNA were used for transformation of the rice cultivar Dongjin, as reported previously by Jung et al. [28] ( Figure 1B). We generated 30 T 0 transformants for constructs containing sgRNA1, sgRNA2, and sgRNA3 and identified the integration of T-DNA into genomic DNA using PCR-based genotyping with bar gene-specific primers (Supplementary Figure S1). Among them, the orange color phenotype was observed in the transgenic callus obtained from the sgRNA3 construct ( Figure 1C). transformation of the rice cultivar Dongjin, as reported previously by Jung et al. [28] (Figure 1B). We generated 30 T0 transformants for constructs containing sgRNA1, sgRNA2, and sgRNA3 and identified the integration of T-DNA into genomic DNA using PCRbased genotyping with bar gene-specific primers (Supplementary Figure S1). Among them, the orange color phenotype was observed in the transgenic callus obtained from the sgRNA3 construct ( Figure 1C). Next, we performed a deep-sequencing analysis of all transgenic calluses to analyze target-site mutations. The mutation rates of sgRNA1, sgRNA2, and sgRNA3 were 36.8%, 40.1%, and 43.5%, respectively (Supplementary Table S1). Deep sequencing results of the target region showed various mutations involving insertions (1-2 bp) and deletions (1-65 bp) of different nucleotides ( Figure 2). The most common types of mutations were 1 bp insertions and deletions (indel) (Figure 2). These mutations were classified as homologous, heterologous, and biallelically edited ( Figure 2). We also investigated sgRNAs using Cas-OFFinder (https://www.rgenome.net/cas-offin der/, accessed on 26 November 2019) [28] and selected two potential off-target sites. However, we were unable to identify offtarget mutations after sequencing of these loci (data not shown). Next, we performed a deep-sequencing analysis of all transgenic calluses to analyze target-site mutations. The mutation rates of sgRNA1, sgRNA2, and sgRNA3 were 36.8%, 40.1%, and 43.5%, respectively (Supplementary Table S1). Deep sequencing results of the target region showed various mutations involving insertions (1-2 bp) and deletions (1-65 bp) of different nucleotides ( Figure 2). The most common types of mutations were 1 bp insertions and deletions (indel) ( Figure 2). These mutations were classified as homologous, heterologous, and biallelically edited ( Figure 2). We also investigated sgRNAs using Cas-OFFinder (https://www.rgenome.net/cas-offinder/, accessed on 26 November 2019) [28] and selected two potential off-target sites. However, we were unable to identify off-target mutations after sequencing of these loci (data not shown).

Analysis of Expression of Carotenoid Pathway Genes
To understand the mechanism of carotenoid accumulation in the orange calli created by OsOr gene editing, we determined the transcriptional levels of genes involved in carotenoid biosynthesis and catabolism by quantitative real-time PCR ( Figure 5). Among the genes encoding the carotenoid metabolic pathway enzymes, the levels of PSY2, PSY3, PDS, ZDS and β-LCY gene transcripts were higher in the OC#1, OC#3, and OC#17 lines than in the WT. On the other hand, the transcription of PSY1, CRTISO and ε-LCY gene was downregulated in OC#1, OC#3, and OC#17 lines compared to that in the WT ( Figure  5, Supplementary Table S6).
We also analyzed the expression levels of CCD1, CCD4a, and CCD4b, which encode carotenoid dioxygenase. In OC#1, OC#3, and OC#17 lines, the expression level of CCD1 was lower than that of WT, whereas the expression levels of CCD4a and CCD4b were higher than those of WT ( Figure 5, Supplementary Table S6).

Analysis of Expression of Carotenoid Pathway Genes
To understand the mechanism of carotenoid accumulation in the orange calli created by OsOr gene editing, we determined the transcriptional levels of genes involved in carotenoid biosynthesis and catabolism by quantitative real-time PCR ( Figure 5). Among the genes encoding the carotenoid metabolic pathway enzymes, the levels of PSY2, PSY3, PDS, ZDS and β-LCY gene transcripts were higher in the OC#1, OC#3, and OC#17 lines than in the WT. On the other hand, the transcription of PSY1, CRTISO and ε-LCY gene was downregulated in OC#1, OC#3, and OC#17 lines compared to that in the WT ( Figure 5, Supplementary Table S6).

Salt-Induces H2O2 Accumulation in OC#1, OC#3, and OC#17 Lines
To evaluate the effects of altered OsOr expression on oxidative stress tolerance, 2week-old transgenic calli were treated with 150 or 200 mM NaCl for 24 h. The OC#1, OC#3, and OC#17 transgenic calli were tolerant of salt-mediated oxidative stress, as determined by qualitative and quantitative analyses of H2O2 ( Figure 6A,B). In addition, exposure to We also analyzed the expression levels of CCD1, CCD4a, and CCD4b, which encode carotenoid dioxygenase. In OC#1, OC#3, and OC#17 lines, the expression level of CCD1 was lower than that of WT, whereas the expression levels of CCD4a and CCD4b were higher than those of WT ( Figure 5, Supplementary Table S6).

Salt-Induces H 2 O 2 Accumulation in OC#1, OC#3, and OC#17 Lines
To evaluate the effects of altered OsOr expression on oxidative stress tolerance, 2week-old transgenic calli were treated with 150 or 200 mM NaCl for 24 h. The OC#1, OC#3, and OC#17 transgenic calli were tolerant of salt-mediated oxidative stress, as determined by qualitative and quantitative analyses of H 2 O 2 ( Figure 6A,B). In addition, exposure to 200 mM NaCl resulted in a higher relative water content (RWC) in the OC#1, OC#3, and OC#17 lines than in the WT lines ( Figure 6C).

Figure 5.
Changes in the expression of carotenoid pathway genes. The Y-axes of graphs indicate relative transcript amounts expressed in terms of 2 −ΔΔCT values determined based on qRT-PCR analyses using the Actin 1 gene as an internal control. Data represent the means ± SD of three independent replicates. Asterisks indicate significant differences between OC callus line and WT callus (** p < 0.01, *** p < 0.001, Tukey's HSD test).

Salt-Induces H2O2 Accumulation in OC#1, OC#3, and OC#17 Lines
To evaluate the effects of altered OsOr expression on oxidative stress tolerance, 2week-old transgenic calli were treated with 150 or 200 mM NaCl for 24 h. The OC#1, OC#3, and OC#17 transgenic calli were tolerant of salt-mediated oxidative stress, as determined by qualitative and quantitative analyses of H2O2 ( Figure 6A,B). In addition, exposure to 200 mM NaCl resulted in a higher relative water content (RWC) in the OC#1, OC#3, and OC#17 lines than in the WT lines ( Figure 6C).

Discussion
Carotenoids can be synthesized in nearly all plastids and stored at different concentrations to produce red, yellow, or orange flowers, fruits, and roots. Although it has been established that rice endosperm does not normally produce carotenoids, they have been shown to contain geranyl geranyl diphosphate (GGPP), a carotenoid precursor, and are known to express several endogenous carotenoid biosynthetic genes, including lycopene β-cyclase (LCY-b) [8,13]. In this regard, the endosperm of Golden rice 2, a transgenic rice variety developed by introducing a corn-derived phytoene synthase gene (PSY) and a soil bacterium-derived phytoene desaturase gene (crtI), has been shown to efficiently synthesize β-carotene [8]. Numerous studies have indicated that both PSY and orange proteins play essential roles in the accumulation of β-carotene [19,20,23,[29][30][31]. Furthermore, it has recently been demonstrated that overexpression of the golden SNP-carrying allele of the Or gene results in elevated carotenoid content and orange-colored fruit or storage root; however, no significant differences in the carotenoid content of transgenic plants overexpressing the Or gene have been reported [20,32]. In a previous study, we successfully developed transgenic rice plants overexpressing OsOr-R115H #20 (leading to Arg to His substitution at position 115 on the OsOr protein) and found that there was no significant difference Plants 2023, 12, 56 7 of 11 in the carotenoid content or calli color of these lines compared with NT plants. However, proline and chlorophyll contents of transgenic lines were significantly higher than those of non-transgenic (NT) plants under heat stress conditions [33]. Recently, Ali et al. reported that OR expression in Arabidopsis thaliana was upregulated by drought treatment and seedlings of OR-overexpressing (OE) lines showed improved growth performance and survival under drought stress [34]. These OE seedlings possessed lower contents of reactive oxygen species, higher activities of both superoxide dismutase and catalase, and a higher level of proline content [34].
Endo et al. achieved a high carotenoid content in orange calli by targeted mutagenesis of the OsOr gene using the CRISPR/Cas9 system [17]. Similar to our results, several sgRNAs were used in these experiments, but only the sgRNA of the exon 2-intron produced the orange calli phenotype. Recently, Yu et al. reported that the carotenoid content in the leaves of the OsOr-KO mutant line produced by CRISPR/Cas9 was similar to that of the control, whereas tillering and biomass were elevated [27].
In this study, we detected significant increases in the orange color intensity and carotenoid content, particularly that of β-carotene, in the OC#1, OC#3, and OC#17 lines, in which the Or gene was edited using the CRISPR/Cas9 system. These callus lines showed similar orange absorption spectra after several subcultures (Figures 3 and 5, and Supplementary Table S4). In carrots, most orange genotypes contain α-and β-carotene, whereas yellow and red genotypes contain predominantly lutein and lycopene, respectively [35]. In tomatoes, the accumulation of carotenoids is closely associated with the expression of genes encoding biosynthetic enzymes, which is in turn influences by the size and number of plastids [36][37][38]. In addition, it has been reported that the carotenoid content in Chrysanthemum is determined by the expression level of carotenoid cleaved dioxygenase (CCD), an enzyme associated with carotenoid degradation [39]. In the present study, qRT-PCR analysis revealed differences among the three experimental rice lines with respect to the expression levels of carotenoid biosynthetic pathway genes (Figure 4), among which, we detected upregulated levels of PSY2, PDS, and ZDS gene transcripts in the OC#1, OC#3, and OC#17 lines. These observations indicate that carotenoid accumulation in rice calli is not directly associated with an increase in the expression of carotenoid biosynthetic genes via expression of the Or gene. Instead, we speculate that carotenoid synthesis in rice calli is correlated with the disruption of an undetermined domain structure in the third and fourth exons of the Or gene (Supplementary Figure S1). This could account for the differences observed in the expression of carotenoid biosynthetic pathway genes in the OC#1, OC#3, and OC#17 lines compared with those in the WT. In addition to an increase in carotenoid content, we also found that the OC#1, OC#3, and OC#17 lines are more tolerant of salt stress than the WT (Figures 5 and 6), which appears to be correlated with the increased level of carotenoids.
We previously reported that carotenoid biosynthetic intermediates function in concert with reactive oxygen species to reduce cellular damage in plants subjected to salt stress [5,40]. Salinity is a major source of environmental stress that reduces crop productivity in arid regions worldwide [41], and evidence obtained to date indicates that enhanced tolerance to abiotic stresses, such as salt stress, is associated with plant resistance to oxidative stress [42]. Accordingly, it is conceivable that transgenic plants with enhanced resistance to different environmental stresses can be generated by manipulating genes associated with carotenoid biosynthesis.

Plasmid Construction and Genetic Transformation of Rice
Target sites and single guide RNAs (sgRNAs) for the first, third, and fifth exons of OsOr (Os02g0651300) adjacent to a protospacer-adjacent motif (PAM) were amplified using specific primer sets (Supplementary Table S1). A CRISPR/Cas9 vector was constructed by selecting three target sites in the OsOr sequence using the CRISPR RGEN software (http://www.rgenome.net/ accessed on 19 November 2019) (Supplementary Table S2). A 20-nt sgRNA scaffold sequence was synthesized by Bioneer Co., Ltd. (Daejeon, Republic of Korea) and the dimer was cloned into an AarI-digested OsU3:pBOsC binary vector, as described by Jung et al. [27]. The constructs obtained were transformed into rice embryogenic calli using Agrobacterium tumefaciens strain EHA105, as previously described [43]. Transformed calli were selected using 6 mg/L phosphinothricin and transformation was confirmed by PCR analysis, as previously reported [43]. To verify target site mutations, PCR amplicons were subjected to MiniSeq paired-end read sequencing (Illumina, San Diego, CA, USA) and analyzed using Cas-Analyzer (https://www.rgeno me.net/cas-analyzer/, accessed on 21 July 2020) [44]. All transgenic callus lines were subcultured several times and maintained in AA medium, as described previously [45] (Supplementary Figure S2 and Supplementary Table S3).

qRT-PCR Analysis
Total RNA was extracted from calli using the RNeasy Plant Mini Kit (Qiagen, Hilden, Germany, www.qiagen.com (accessed on 16 December 2022)), and single-strand cDNA was synthesized with random oligonucleotides using a reverse transcription system (Takara, Kusatsu, Japan, www.takara-bio.com (accessed on 16 December 2022)), based on a previously reported method [25]. To analyze the relative expression levels of genes involved in carotenoid biosynthesis and degradation pathways, qRT-PCR was conducted using cDNA, gene-specific primers, and SYBR Green Real-time PCR Master Mix (Toyobo, http://www.toyobo.co.kr/ (accessed on 16 December 2022)), according to the manufacturer's instructions. The sequences of gene-specific primers used for amplification are listed in Supplementary Table S1. The specificity of the amplicon was verified by dissociation curve analysis (60 to 95 • C) after 40 cycles of PCR and by agarose gel electrophoresis. The OsActin 1 gene was used as an internal standard and relative gene expression levels were calculated using the comparative Ct method [46].

Extraction of Pigments and Spectrophotometric Analysis
Samples of rice calli (0.1 g) frozen in liquid nitrogen were homogenized using a mortar and pestle, followed by the addition of acetone (1 mL), thorough mixing, and centrifugation for 5 min at 13,000× g and 4 • C. Subsequently, the supernatant was collected. This procedure was repeated, and the second supernatant was combined with the first and dried. The obtained residue was dissolved in 500 µL of acetone and diluted 20 times. The absorption of the suspension was measured spectrophotometrically according to a previously reported method [47].

Determination of Carotenoid Content
Carotenoids of calli were extracted with 0.01% solution of butylated hydroxytoluene in acetone and analyzed using an Agilent 1260 high-performance liquid chromatography (HPLC) system (Hewlett-Packard, Waldbronn, Germany), according to the method described by Lim et al. [48]. The carotenoids were quantified using an external calibration method. The standards of β-carotene, β-cryptoxanthin, lutein, violaxanthin, and zeaxanthin were extracted using Carote Nature (Lupsingen, Switzerland). All extraction procedures were performed under low-light conditions to prevent pigment degradation and loss. Carotenoid content was calculated as µg g −1 dry weight of callus tissue.

Salt Stress
To examine salt stress tolerance, 2-week-old calli were incubated for 24 h in liquid AA medium [45] containing 3% sucrose and 2 mg 2,4-D/l, and supplemented with 150 or 200 mM NaCl.

Qualitative and Quantitative Analysis of H 2 O 2
To assess tissue oxidation, calli were placed individually in a 1 mg mL −1 solution of 3,3-diaminobenzidine (DAB)-HCl (pH 3.8) under continuous light, according to the Plants 2023, 12, 56 9 of 11 methods described by Kim et al. [32]. Oxidized DAB was visualized in callus tissues as dark polymerization products generated as a result of the reaction of DAB with H 2 O 2 . The production of H 2 O 2 was quantified by measuring the absorbance of the DAB solution at 460 nm in each callus. Oxidized DAB content was measured using a calibrated DAB standard curve.

Analysis of RWC
Tissue dehydration was analyzed according to the relative water content (RWC) of the transgenic calli after treatment with 150 or 200 mM NaCl. RWC was measured as follows: RWC (%) = (fresh weight − dry weight)/fresh weight [49].

Statistical Analyses
All data were analyzed using one-way analysis of variance (ANOVA). Subsequent multiple comparisons were performed using the least significant difference (LSD) test. According to Tukey's HSD, statistical significance was set at p < 0.05.