Insertion of a Retrotransposon into a Flavonoid 3'-hydroxylase Homolog Confers the Red Root Character in the Radish (Raphanus sativus L. var. longipinnatus L. H. Bailey)

Abstract

Flavonoid 3'-hydroxylase (F3'H) has an important role in determining anthocyanin patterns in plants. Here, we analyzed root coloration characteristics in the radish by investigating anthocyanidin and expression of F3'H genes in cultivars with purple, red, or white roots. Cyanidin was detected in the roots of the purple radish, while pelargonidin was found in the red radish; however, neither of these anthocyanidins was detected in white radish. We isolated the RsF3'H gene of the purple root radish and found that it showed a relatively higher level of expression compared to the RsF3'H of the red radish. Moreover, we identified a retrotransposon, gypsy-Ty3, in the first exon of the F3'H homolog in the red radish. These results suggested that the F3'H enzyme may determine cyanidin-based anthocyanin in the purple radish, and that the lack of F3'H function due to the retrotransposon insertion, contributed to pelargonidin-based anthocyanin accumulation in the red radish.

Introduction

The radish (Raphanus sativus L.) belongs to the Brassicaceae family and is an important vegetable that is grown worldwide. Some varieties of radish (R. sativus L. sativus) and Japanese radish (R. sativus L. var. longipinnatus L. H. Bailey) have red skin and/or flesh due to anthocyanin accumulation. The main anthocyanins in radishes are acylated cyanidin 3-sophoroside-5-glucosides or acylated pelargonidin 3-sophoroside-5-glucosides (Kato et al., 2013). Radish anthocyanins are brightly colored over a wide pH range and their multiple acylation with hydroxycinnamic acids contributes to their remarkable heat stability in acidic environments (Jing et al., 2012). Radish anthocyanins have been widely used as natural colorants and for the health benefits of their antioxidant activities (Matsufuji et al., 2007; Rahman et al., 2006).

The genetics, biochemistry, and molecular biology of the anthocyanin biosynthetic pathway are well characterized (Grotewold, 2006; Koes et al., 2005). This pathway contains seven core enzymes, namely, chalcone synthase (CHS), chalcone isomerase (CHI), flavanone 3-hydroxylase (F3H), flavonoid 3'-hydroxylase (F3'H), dihydroflavonol 4-reductase (DFR), anthocyanidin synthase (ANS), and flavonoid 3-glucosyltransferase (3GT). These enzymes produce cyanidin-based anthocyanins in a chain catalysis process (Fig. 1). In the pathway, F3'H hydroxylates the 3'-position of the B ring of dihydrokaempferol, a pelargonidin precursor, to generate dihydroquercetin, a cyanidin precursor. Dihydroquercetin is an important intermediate in the biosynthesis of anthocyanins (Koes et al., 1994; Werck-Reichhart et al., 2002). Because a F3'H mutant would then result in the production of pelargonidin-based anthocyanin (Hoshino et al., 2003), F3'H plays an important role in production of cyanidin-based anthocyanin.

Fig. 1

Anthocyanin biosynthetic pathway. Enzyme abbreviations: ANS, anthocyanidin synthase; CHI, chalcone isomerase; CHS, chalcone synthase; DFR, dihydroflavonol reductase; F3H, flavonone 3-hydroxylase; F3'H, flavonoid 3'-hydroxylase; 3GT, flavonoid 3-glucosyltransferase.

Recently, the expression of genes encoding enzymes for anthocyanin biosynthesis, including phenylalanine ammonia lyase (PAL), cinnamate 4-hydroxylase (C4H), 4-coumarate-CoA ligase (4CL), CHS, CHI, F3H, DFR, and ANS, was investigated in white and red radishes (Park et al., 2011). However, little attention has been paid to the relationship between F3'H expression and root color in radishes.

In this study, to determine the cause of the red root character, we investigated the types of anthocyanidins in purple, red, and white root radishes. Moreover, to elucidate the molecular basis of the red root character, we investigated the role of the radish F3'H gene in determining radish root color.

Materials and Methods

Plant materials

Radish cultivars (R. sativus L. var. longipinnatus L. H. Bailey) of purple root ‘Karaine-aka’ (Watanabe seed Co., Ltd., Miyagi, Japan), red root ‘Chouan-aomaru-koshin’ (Takii & Co., Ltd., Kyoto, Japan) and ‘Benikururi 521’ (Matsunaga Seed Co., Ltd, Aichi, Japan), and white root ‘Taibyo-soubutori’ (Takii & Co., Ltd.) were used in the present study (Fig. 2A–C, E). White root pungent radish ‘Susanoo’ (Shimane University, Matsue, Japan), which was developed from the Japanese wild radish (R. sativus L. f. raphanistroides Makino) by mass selection, was also used in this study (Fig. 2D). These cultivars were maintained in the experimental field of Shimane University. Their root skin and flesh were collected for High Performance Liquid Chromatography (HPLC) analysis. For molecular analysis, root skin, flesh, and leaves were collected and immediately frozen in liquid nitrogen and stored at −80°C until extraction of total RNA and genomic (g) DNA.

Fig. 2

Photographs of the five radish varieties used in this study: (A) ‘Karaine-aka’, (B) ‘Chouan-aomaru-koshin’, (C) ‘Benikururi 521’, (D) ‘Susanoo’, and (E) ‘Taibyo-soubutori’. Bars indicate 10 cm.

Anthocyanidins by HPLC analysis

The anthocyanidin was extracted using the methods of Mizuta et al. (2009). The raw root skins and flesh (ca 0.5 g) mentioned above were soaked in 50% CH₃COOH in H₂O overnight. The crude extracts were concentrated to small amounts and hydrolyzed with 4 mL of 2N hydrochloric acid at 100°C for 1 h. The hydrolysates were adsorbed on a Sep-pak C18 cartridge (Nihon Waters K. K., Tokyo, Japan). The cartridge was washed to eliminate the water-soluble or hydrophilic contaminants, and then the anthocyanidins were eluted by 50% CH₃COOH in H₂O. A small portion of the eluate was passed through a Millipore filter, and the filtrates were analyzed with some standards by HPLC. As these standards for the assessment of anthocyanidin, commercially available cyanidin and pelargonidin (Extrasynthese, Genay, France) were used.

The analysis for anthocyanidins was performed using a GL science HPLC system (GL-7410, GL-7420, GL-7432, GL-7452A, GL-7480; GL sciences Inc., Tokyo, Japan) equipped with a degasser (BG-240; FLOM Inc., Tokyo, Japan) using an Inertsil ODS-4 column (4.6 × 150 mm; GL sciences Inc.), a column temperature of 40°C, and monitoring at 520 nm. The solvent system had a ratio of 55% solvent A [H₃PO₄-H₂O (1.5 : 98.5, v/v)] to 45% solvent B [H₃PO₄-CH₃COOH-CH₃CN-H₂O (1.5 : 20 : 25 : 53.5, v/v)] run for 25 min with a flow rate of 1.0 mL·min⁻¹.

Isolation of the F3'H gene from mRNA of the purple root radish ‘Karaine-aka’

Total RNA was extracted from root skin of ‘Karaine-aka’ using the hot borate extraction method (Wan and Wilkins, 1994). The RNA extract was treated with DNase I (TaKaRa Bio, Shiga, Japan) to remove contaminating gDNA. cDNA was synthesized from the DNase-treated total RNA (1 μg) using ReverTra-Ace (Toyobo, Osaka, Japan).

To isolate the F3'H gene from the purple radish ‘Karaine-aka’, PCR amplification was carried out using root skin cDNA and with the following primers: forward, 5'-ATGACTAATCTTTACCTCACAATCC-3' and reverse, 5'-TTAAGCCGACCCGAGTCCGTAAGCACTC-3'; these primers were designed to amplify the sequence from the start codon to stop codon in Brassica napus F3'H (DQ324379). The PCR mixture (10 μL) contained 1X Ex-Taq buffer, 200 μM dNTPs, 0.2 μM of each primer, 0.25 U Ex-Taq (TaKaRa Bio), and 2.5 ng cDNA template. Amplification conditions were as follows: preheating at 94°C for 2 min; 35 cycles of denaturation at 94°C for 30 s, annealing at 57°C for 30 s, and extension at 72°C for 1 min 30 s; and final extension at 72°C for 3 min. The amplified fragments were cloned into the pGEM-T easy vector (Promega, Madison, MI, USA) and E. coli HST08 Premium Competent Cells (TaKaRa Bio); they were sequenced using a BigDye Terminator version 3.1 Cycle Sequencing Kit (Applied Biosystems, Foster City, CA, USA) and an ABI 3130 genetic analyzer (Applied Biosystems), following plasmid DNA extraction by a FastGene Plasmid Mini kit (Nippon Genetics Co., Tokyo, Japan). The nucleotide sequences and deduced amino acid sequences of the isolated radish F3'H mRNA were aligned by the BLASTx function in Genetyx ver. 11 (Software Development Co., Tokyo, Japan).

To confirm the sequence in the reverse primer region of the isolated F3'H gene, rapid amplification of cDNA ends (RACE) PCR was performed using the cDNA template produced from the root skin RNA with the 3'-Full RACE Core Set (TaKaRa Bio). To obtain forward primer region sequence information, the 5' upstream region of F3'H was identified by inverse PCR (Ochman et al., 1988) in ‘Karaine-aka’ gDNA extracted from leaves using a modified CTAB method (Kobayashi et al., 1998). We used the restriction enzyme EcoRI to digest 500 ng gDNA and the digested fragments were self-ligated with T4 ligase (Promega). Inverse PCR was performed in 10 μL reaction mixtures containing 1 μL self-ligated gDNA, 1X buffer, 200 μM dNTPs, 0.25 U Ex-Taq (TaKaRa Bio), and the following primers: first PCR, forward 5'-TTCTAACGCTATAGCTCACC-3' and reverse 5'-CTGTCGAACATGTTTAAAATCTTC-3'; second PCR, forward 5'-GTATATTGAGTTATTCGAATTA-3' and reverse 5'-ATCGTGAACTTTCAAGAACT-3'. The reaction conditions of the nested PCR were as follows: preheating at 94°C for 3 min; 35 cycles of denaturation at 94°C for 30 s, annealing at 55°C (1st PCR)/51°C (2nd PCR) for 30 s, and extension at 72°C for 3 min; and final extension at 72°C for 5 min. The amplified fragments were sequenced as described above.

To isolate the complete F3'H sequence from ‘Karaine-aka’, the amplifications were carried out with root skin cDNA and the following primers: forward 5'-ATGACTAATCTCTTCCTCACAATCC-3' and reverse 5'-GTTTAAACAGACCCAAGCCC-3'. These primers were designed to amplify the sequence from the start codon to the stop codon. All amplified fragments were sequenced as described above.

Expression analysis of radish anthocyanin biosynthesis genes

RACE PCR was used to isolate the anthocyanin biosynthesis genes CHS (AF031922), F3H (AB087211), DFR (KF280272), and ANS (KR262954) from the purple radish ‘Karaine-aka’ using a cDNA template produced from root skin RNA with the 3'-Full RACE Core Set (TaKaRa Bio).

Semi quantitative reverse transcription-PCR (semi qRT-PCR) and reverse transcription-quantitative PCR (RT-qPCR) were carried out to determine the relative expression levels of the anthocyanin biosynthesis genes in root skin and flesh from all varieties. In the semi qRT-PCR assay, each 20 μL reaction mixture contained 1X buffer, 200 μM dNTPs, 0.2 μM of each primer for radish CHS (DDBJ accession number LC202030, forward 5'-TCCAAGCGGAGTATCCTGACTAC-3' and reverse 5'-GCACATGTTAGGGTTCTCTTTCAA-3'), F3H (DDBJ accession number LC202031, forward 5'-TGATCTAACCCTCG GACTCA-3' and reverse 5'-TCTGGAACGTGGCTATTGAT-3'), F3'H (forward 5'-GCCGAACAGTTCTTGAAAGT-3' and reverse 5'-CTGTCGAACATGTTTAAAATCTTC-3'), DFR (DDBJ accession number LC202032; forward 5'-ACCGGATGGATGTATTTCATGTC-3' and reverse 5'-ATGATGGAGTAATGTGCCTCGTT-3'), and ANS (DDBJ accession number LC202033; forward 5'-GAGCCTGACCGTCTAGAGAAAGA-3' and reverse 5'-CAAACCAGGAACCATGTTGTGTA-3'), 0.5 U Ex-Taq (TaKaRa Bio), and 1 μL 40X diluted cDNA templates. The PCR conditions were as follows: preheating at 94°C for 2 min; 35 cycles of denaturation at 94°C for 30 s, annealing at 58°C (F3'H), 60°C (radish CHS, F3H, DFR, and ANS), and 61°C (ACTIN) for 30 s, and extension at 72°C for 30 s; and final extension at 72°C for 5 min. The amplified PCR products were separated by electrophoresis on a 1% agarose gel (Nippon Genetics Co.) in 0.5X TBE buffer at 100 V for 30 min. For semi qRT-PCR, the amplified products were normalized against radish ACTIN (FY430005; forward 5'-CGATGGTGAGGACATTCAAC-3' and reverse 5'-TCACCAGAGTCGAGCACAATA-3').

Expression of F3'H in root skin and flesh was assayed by RT-qPCR assay using a 20 μL reaction mix containing 1 μL of 10X diluted cDNA templates, 10 μL of SYBR Premix Ex-Taq II (TaKaRa Bio), and 0.25 μM of 26S ribosomal RNA (26S) primer (Park et al., 2011; forward 5'-AACACCCTTTGTGGGTTCTAGGT-3' and reverse 5'-GCCCTCGACCTATTCTCAAACTT-3') and the above F3'H primers. Each RT-qPCR assay was conducted using a Thermal Cycler Dice Real-Time System (TaKaRa Bio) with an initial denaturation of 30 s at 95°C, followed by 50 cycles of denaturation for 5 s at 95°C, annealing for 10 s at 58°C (F3'H)/56°C (26S), and elongation for 20 s at 72°C. Three technical replicates were analyzed for each cDNA sample. Quantitation was performed by using the difference in the threshold cycle values (ΔCt) between the two samples (the target gene and the reference gene) to calculate the relative amounts of the template present. The value of the relative expression was calculated as the standard using the maximum value for the F3'H gene, and the mean and standard error (SE) were calculated from the value of the relative expression (Cheon et al., 2011).

Isolation of the radish F3'H homolog and insertion sequence from red radish gDNA

The radish F3'H gene in ‘Karaine-aka’ gDNA was isolated by PCR amplification of the region between the 5' upstream to 3' UTR, and was performed in 20 μL reaction mixtures containing 20 ng gDNA, 1X buffer, 200 μM dNTPs, 0.5 U PrimeSTAR GXL polymerase (TaKaRa Bio), and 0.2 μM of each primer. The primers were designed using the 5' upstream and 3' UTR sequences (forward 5'-CTAACCGAGATATGCATGGT-3' and reverse 5'-ATTACACAAACATCACAAGGC-3'). The conditions for PCR amplification were as follows: preheating at 98°C for 5 min; 35 cycles of denaturation at 98°C for 10 s, annealing at 57°C for 5 s, and extension at 68°C for 3 min; and final extension at 72°C for 10 min. A-overhangs were created on the amplified fragments (0.25 U Ex-Taq at 72°C for 30 min) in order to perform TA-cloning. The treated fragments were cloned into E. coli HST08 Premium Competent Cells (TaKaRa Bio) with the pGEM-T easy vector (Promega); plasmid extraction and sequencing analysis were performed as described above.

The F3'H gene of purple, red, and white root radishes was analyzed by PCR amplification of gDNA sequences from the root skin and flesh. The PCR was performed in 20 μL reaction mixtures containing 20 ng gDNA, 1X Ex-Taq buffer, 200 μM dNTPs, 0.25 U Ex-Taq (TaKaRa Bio), and 0.2 μM of each primer. Four primers were used: P1 (5'-CTAACCGAGATATGCATGGT-3'), P2 (5'-CTGTCGAACATGTTTAAAATCTTC-3'), P3 (5'-GAAGAGGTTGGAACACTCATG-3'), and P4 (5'-GGGCTTGGGTCTGTTTAAAC-3') (Fig. 5). The conditions for the PCR amplifications were as follows: preheating at 94°C for 1 min; 35 cycles of denaturation at 94°C for 30 s, annealing at 57°C (P1 and P2)/57°C (P3 and P4) for 30 s, and extension at 72°C for 5 min; and final extension at 72°C for 10 min. The amplified PCR products were confirmed by electrophoresis on 1% agarose gel (Nippon Genetics Co.) in 0.5X TBE buffer at 100 V for 30 min. Subsequently, the fragments amplified using P1 and P2 primers were cloned using a TOPO-XL PCR cloning kit (Invitrogen, Carlsbad, CA, USA) and E. coli HST08 Premium Competent Cells (TaKaRa Bio), and the clones were sequenced as described above. The obtained sequences from ‘Chouan-aomaru-koshin’ and ‘Benikururi 521’ were compared with sequence information of purple colored ‘Karaine-aka’ using Genetyx ver. 11 (Software Development Co.).

Results

HPLC analysis of anthocyanidins

The HPLC analysis of purple, red, and white radishes identified cyanidin in both skin and flesh of the purple radish ‘Karaine-aka’. Pelargonidin was detected in both the skin and flesh of the red radish ‘Chouan-aomaru-koshin’ and ‘Benikururi 521’. However, cyanidin derived from the flesh in ‘Karaine-aka’ and pelargonidin derived from the skin in ‘Chouan-aomaru-koshin’ was at trace amounts compared with other colored tissues, although the precise anthocyanidin contents in each sample were not checked in this analysis method. However, neither of these anthocyanidins was detected in the white radishes ‘Susanoo’ or ‘Taibyo-soubutori’ (Table 1).

Table 1

Anthocyanidin in the skin and flesh of purple, red, and white radishes.

Isolation of the F3'H homolog from mRNA of the purple root radish ‘Karaine-aka’

Two radish F3'H sequences were isolated from the purple radish ‘Karaine-aka’ and termed RsF3'H-a (approximately 1536 bp, DDBJ accession number LC202035), and RsF3'H-b (approximately 1536 bp, DDBJ accession number LC202037) (Fig. 3A). The two gene sequences differed at one position only, with guanine in RsF3'H-a and cytosine in RsF3'H-b. A comparison of the predicted amino acids from these two sequences with published F3'H amino acid sequences showed that both RsF3'H-a and RsF3'H-b showed 97% identity with Brassica rapa F3'H (ABY89687) and 90% identity with Arabidopsis F3'H (AAF73253). The deduced amino acid sequences of the P450 motif P₃₀PGPNPWP₃₇, A₃₀₀GTDTS₃₀₅, F₄₃₆GAGRRICAG₄₄₅, and E₃₅₇-R₃₆₀-R₃₉₉ triad residues, and F3'H-specific motifs V₇₃VVAAS₇₈, G₄₁₈GEK₄₂₁, and V₄₂₄DVKG₄₂₈ encoded by RsF3'H-a and RsF3'H-b were homologous to those in F3'H (Xu et al., 2007) (Fig. 3B).

Fig. 3

Isolation of RsF3'H genes from ‘Karaine-aka’. A: Comparison of RsF3'H-a and RsF3'H-b mRNA sequences from the start codon to stop codon. The symbol M and asterisk indicate the start and stop codon, respectively. A black box indicates different sequences in RsF3'H-a and RsF3'H-b sequences, respectively. Close triangles indicate insertion of retrotransposon and its long terminal region, and open triangles indicate insertion of a poly (T)₂₂ sequence. Horizontal and vertical arrows indicate the primer region for expression analysis and intron regions, respectively. B: Multiple sequence alignment of Arabidopsis F3'H (NP_196416), Brassica rapa F3'H (ABY89687), and the deduced amino acid sequences for RsF3'H-a and -b. Black line boxes are P450 motif PPGPNPWP, AGTDTS, FGAGRRICAG, and E-R-R triad residues, and black dashed-line boxes are the F3'H-specific motifs VVVAAS, GGEK, and VDVKG, respectively.

Analysis of expression of anthocyanin biosynthesis genes

Expressions of RsCHS, RsF3H, RsDFR, and RsANS were analyzed using a semi qRT-PCR assay (Fig. 3). This analysis showed that these genes had the highest expression in the skin of ‘Karaine-aka’, the flesh of ‘Chouan-aomaru-koshin’, and both the skin and flesh of ‘Benikururi 521’. However, expression of RsDFR and RsANS was not found in white radishes. RsF3'H was mainly expressed in the skin of ‘Karaine-aka’, whereas RsF3'H expression was not found in red radishes, and was lower in white radishes (Fig. 4A). An RT-qPCR assay to confirm the patterns of RsF3'H expression in 5 radishes showed that the gene was more highly expressed in ‘Karaine-aka’ than in other radishes (Fig. 4B).

Fig. 4

Expression analysis of the F3'H gene in the purple root radish ‘Karaine-aka’ (KA), red root radishes ‘Chouan-aomaru-koshin’ (CAK) and ‘Benikururi 521’ (BK), and white root radishes ‘Susanoo’ (SU) and ‘Taibyo-soubutori’ (TS). A: Expression analysis using semi qRT-PCR in root skin (S) and flesh (F). Radish ACTIN was used as an internal control. B: Expression analysis of RT-qPCR in skin and flesh. Radish 26S was used as an internal control. The data represent the mean and standard errors obtained from three technical replicates. Different letters indicate significant differences at P < 0.01 (Tukey’s multiple comparison tests).

Analysis of RsF3'H genes from red radish gDNA

The RsF3'H gene was sequenced by isolating RsF3'H-a and RsF3'H-b from ‘Karaine-aka’ gDNA. RsF3'H-a and -b were approximately 3039 bp and 3017 bp in length, respectively (DDBJ accession numbers LC202034 and LC202036, respectively). Analysis of the structure of the gDNA sequences indicated that three introns were present and that the first intron differed in length between the two genes (Fig. 5A).

Fig. 5

Genomic structures of RsF3'H genes in radishes. A: Genomic structure of RsF3'H genes from ‘Karaine-aka’. Closed boxes and E1–E4 indicate exons and first exon-fourth exon, respectively, and lines among exons indicate introns. Numbers of closed boxes (up) and lines (down) indicate length of exons and introns, respectively. Horizontal arrows indicate the primer region and direction for PCR analysis, respectively. B: Amplified PCR products of two primer sets (P1+P2; amplify region between 5' upstream region and first exon and P3+P4; amplify region between second exon and 3' UTR). M: 1 kb DNA Ladder (Nippon Genetics). ‘Karaine-aka’ (KA), ‘Chouan-aomaru-koshin’ (CAK), and ‘Benikururi 521’ (BK). ‘Susanoo’ (SU) and ‘Taibyo-soubutori’ (TS). NC: Negative control. C: Genomic structure of the first exon of RsF3'H genes from five cultivars, I: ‘Karaine-aka’, ‘Susanoo’, and ‘Taibyo-soubutori’, II and IV: ‘Chouan-aomaru-koshin’, II and III: ‘Benikururi 521’. LTR: long terminal repeats, gypsy-Ty3: a retrotransposon of gypsy-Ty3, Poly (T): Poly (T)₂₂ sequence.

The reduced expression of RsF3'H in the red radish was investigated by investigating the genomic gene structures. PCR amplification using primers P3 and P4 generated a band of approximately 1 kb in all radishes. The primers P1 and P2 produced a band of approximately 500 bp (I) in ‘Karaine-aka’, ‘Susanoo’, and ‘Taibyo-soubutori’, two bands of approximately 5 kb (II) and 500 bp (IV) in ‘Chouan-aomaru-koshin’, and two bands of approximately 5 kb (II) and 800 bp (III) in ‘Benikururi 521’ (Fig. 5B). Sequencing of the two PCR products from ‘Chouan-aomaru-koshin’, showed that bands II and IV were 5871 bp (DDBJ accession number LC205688) and 590 bp (DDBJ accession number LC205690) in length, respectively, and those from ‘Benikururi 521’ were 5871 bp (II) and 889 bp (III, DDBJ accession number LC205689). Comparison of these sequences showed that RsF3'H from ‘Chouan-aomaru-koshin’ had two insertion sequences (5322 bp and 27 bp) in the first exon, and that RsF3'H from ‘Benikururi 521’ had two insertion sequences (5322 bp and 340 bp) in the first exon. The 5322 bp insertion sequence identified a target site duplication (TSD; ACCAC) and a long terminal repeat (LTR; 340 bp); moreover, the inserted sequences identified a gypsy-Ty3 retrotransposon that showed 52% identity with Beta vulgaris (AFK13856) in a BLASTx search of GenBank. The 340 bp insertion sequence also identified a TSD (ACCAC) and the inserted sequences showed the LTR of a retrotransposon. The 27 bp insertion sequence identified a TSD (GAAGATTTTAAACATGTT) and a poly(T)₂₂ sequence, but the inserted sequences showed no significant homologies in BLASTx and BLASTn searches of GenBank (Fig. 5C).

Discussion

This study was initiated to elucidate the basis of the red root character in the radish. Mutant type pelargonidin pigments accumulated in plants which accumulated cyanidin pigments in the wild type (Beale, 1941). Analyses of anthocyanidin contents and gene expression were conducted to test the hypotheses that the red root character was caused by accumulation of pelargonidin-based anthocyanin and not of cyanidin-based anthocyanin, and that accumulation of the former was associated with regulation of RsF3'H gene expression.

In anthocyanin analyses, Tatsuzawa et al. (2008, 2010) reported that red and purple radishes contain pelargonidin- and cyanidin-based anthocyanins, and Kato et al. (2013) reported that the aglycones of purple and red radishes contain cyanidin and pelargonidin, respectively. Similarly, in the present study, cyanidin and pelargonidin were detected in purple and red radishes, respectively (Table 1), and pelargonidin-based anthocyanin was accumulated in the roots of ‘Chouan-aomaru-koshin’ and ‘Benikururi 521’.

In addition to the anthocyanidin analysis, when we performed expression analysis of genes related to anthocyanin synthesis, RsF3'H expression differed between purple and red radishes (Fig. 4). Tsuda et al. (2004) reported that in red petunias, the downregulation of the endogenous F3'H gene and over-expression of a rose-derived DFR gene (producing cyanidin) yielded an orange flower containing pelargonidin. Similarly, in Osteospermum, F3'H suppression by RNAi and the introduction of gerbera DFR resulted in pelargonidin accumulation (Seitz et al., 2007). Nakatsuka et al. (2007) also reported that gerbera DFR overexpression and F3'H and FLS suppression in tobacco produced red flowers. In the present study, we found a lower level of RsF3'H expression in the flesh of ‘Chouan-aomaru-koshin’ and both the skin and flesh of ‘Benikururi 521’ than that of the ‘Karaine-aka’ root skin (Fig. 4B), suggesting that the RsF3'H function could be related to cyanidin-based anthocyanin synthesis.

Our gene sequencing analysis identified a retrotransposon, gypsy-Ty3, in the first exon of the F3'H homolog of both ‘Chouan-aomaru-koshin’ and ‘Benikururi 521’ (Figs. 5B and C). Moreover, poly(T)₂₂ and a retrotransposon LTR sequence were found in the same exon in ‘Chouan-aomaru-koshin’ and ‘Benikururi 521’, respectively (Fig. 5C). In Ipomoea purpurea, a pink flower is caused by a mutation (a large insertion in the third exon) of the IpF3'H gene (Zufall and Rausher, 2003). In Dianthus caryophyllus, a change in flower color from purple to deep pink was attributed to an active hAT type transposable element, Tdic101, which was inserted into the second intron of F3'H (Momose et al., 2013). From this, a retrotransposon inserted in the RsF3'H gene indicated that F3'H gene expression is decreased in red root radishes, and that the pelargonidin-based anthocyanin accumulation of red color in radishes is caused by the mutant RsF3'H gene because the red root color is recessive to the purple root color in anthocyanin accumulation in radishes (Hoshi et al., 1963).

The mutant RsF3'H gene and anthocyanidin were not found in roots of white radishes (Fig. 5B; Table 1). In a genetic analysis of radishes (R. sativus L. var. sativus) called ‘Hatsuka Daikon’ in Japan, white radishes were recessive to colored radishes (such as purple and red colors) and the white color was controlled by a single gene because the ratio between colored and white radishes was 3:1 in the F₂ population (Tatebe, 1938). Park et al. (2011) showed that the genes RsDFR and RsANS of the anthocyanin biosynthesis pathway were expressed more highly in the red tissues of red Chinese radishes, and were not expressed in white radishes or white tissues of red Chinese radishes. In the present study, we confirmed the expression patterns of these two genes (Fig. 4A). Anthocyanin biosynthesis is regulated by a transcription factor in a wide range of plant species (Koes et al., 2005). In radishes, overexpression of the R2R3-MYB transcription factor gene, RsMYB1, actively and positively regulates anthocyanin biosynthesis and promotes anthocyanin production by triggering the expression of endogenous basic helix-loop-helix (bHLH) genes as potential binding partners for RsMYB1 (Lim et al., 2016). It is possible that non-anthocyanin accumulation of white radishes was caused by mutations of DFR, ANS, or MYB. Thus, more experiments are needed to clarify the reason for non-anthocyanin accumulation.

In conclusion, we showed that two late genes that are active in the anthocyanin biosynthesis pathway were expressed in purple and red radishes, and that the red root character of radishes is caused by a lack of F3'H function, which is a retrotransposon insertion in the mutant RsF3'H gene. In a future study, we will analyze the mode of inheritance of the mutant RsF3'H. Moreover, to enable selection of the red character for efficient radish breeding, we will develop a gene-based marker using the RsF3'H mutant allele.

Literature Cited

•  Beale,  G. H. 1941. Gene relations and synthetic processes. J. Genet. 42: 197–214.
•  Cheon,  K.-S.,  A.  Nakatsuka and  N.  Kobayashi. 2011. Isolation and expression pattern of genes related to flower initiation in the evergreen azalea, Rhododendron × pulchrum ‘Oomurasaki’. Sci. Hortic. 130: 906–912.
•  Grotewold,  E. 2006. The genetics and biochemistry of floral pigments. Annu. Rev. Plant Biol. 57: 761–780.
•  Hoshi,  T.,  E.  Takemura and  K.  Hayashi. 1963. Genetic modification of hydroxylation pattern in radish anthocyanins. Studies on anthocyanins, XLII. Bot. Mag. Tokyo 76: 431–439.
•  Hoshino,  A.,  Y.  Morita,  J.-D.  Choi,  N.  Saito,  K.  Toki,  Y.  Tanaka and  S.  Iida. 2003. Spontaneous mutations of the flavonoid 3'-hydroxylase gene conferring reddish flowers in the three morning glory species. Plant Cell Physiol. 44: 990–1001.
•  Jing,  P.,  S. J.  Zhao,  S. Y.  Ruan,  Z. H.  Xie,  Y.  Dong and  L.  Yu. 2012. Anthocyanin and glucosinolate occurrences in the roots of Chinese red radish (Raphanus sativus L.), and their stability to heat and pH. Food Chem. 133: 1569–1576.
•  Kato,  K.,  K.  Sato,  T.  Kanazawa,  H.  Shono,  N.  Kobayashi and  F.  Tatsuzawa. 2013. Relationship between root colors and anthocyanins from radishes (Raphanus sativus L.). Hort. Res. (Japan) 12: 229–234 (In Japanese with English abstract).
•  Kobayashi,  N.,  T.  Horikoshi,  H.  Katsuyama,  T.  Handa and  K.  Takayanagi. 1998. A simple and efficient DNA extraction method from the plants, especially from woody plants. Plant Tissue Cult. Biotechnol. 4: 76–80.
•  Koes,  R. E.,  F.  Quattrocchio and  J. N. M.  Mol. 1994. The flavonoid biosynthetic pathway in plants: function and evolution. Bioessays 16: 123–132.
•  Koes,  R.,  W.  Verweij and  F.  Quattrocchio. 2005. Flavonoids: a colorful model for the regulation and evolution of biochemical pathways. Trends Plant Sci. 10: 236–242.
•  Lim,  S. H.,  J. H.  Song,  D. H.  Kim,  J. K.  Kim,  J. Y.  Lee,  Y. M.  Kim and  S. H.  Ha. 2016. Activation of anthocyanin biosynthesis by expression of the radish R2R3-MYB transcription factor gene RsMYB1. Plant Cell Rep. 35: 1–13.
•  Matsufuji,  H.,  H.  Kido,  H.  Misawa,  J.  Yaguchi,  T.  Otsuki,  M.  Chino,  M.  Takeda and  K.  Yamagata. 2007. Stability to light, heat, and hydrogen peroxide at different pH values and DPPH radical scavenging activity of acylated anthocyanins from red radish extract. J. Agric. Food Chem. 55: 3692–3701.
•  Mizuta,  D.,  T.  Ban,  I.  Miyajima,  A.  Nakatsuka and  N.  Kobayashi. 2009. Comparison of flower color with anthocyanin composition patterns in evergreen azalea. Sci. Hortic. 122: 594–602.
•  Momose,  M.,  M.  Nakayama,  Y.  Itoh,  N.  Umemoto,  T.  Toguri and  Y.  Ozeki. 2013. An active hAT transposable element causing bud mutation of carnation by insertion into the flavonoid 3'-hydroxylase gene. Mol. Genet. Genomics 288: 175–184.
•  Nakatsuka,  T.,  Y.  Abe,  Y.  Kakizaki,  S.  Yamamura and  M.  Nishihara. 2007. Production of red-flowered plants by genetic engineering of multiple flavonoid biosynthetic genes. Plant Cell Rep. 26: 1951–1959.
•  Ochman,  H.,  A. S.  Gerber and  D. L.  Hart. 1988. Genetic applications of an inverse polymerase chain reaction. Genetics 120: 621–623.
•  Park,  N. I.,  H.  Xu,  X.  Li,  I. H.  Jang,  S.  Park,  G. H.  Ahn,  Y. P.  Lim,  S. J.  Kim and  S. U.  Park. 2011. Anthocyanin accumulation and expression of anthocyanin biosynthetic genes in radish (Raphanus sativus). J. Agric. Food Chem. 59: 6034–6039.
•  Rahman,  M. M.,  T.  Ichiyanagi,  T.  Komiyama,  Y.  Hatano and  T.  Konishi. 2006. Superoxide radical- and peroxynitrite-scavenging activity of anthocyanins; structure-activity relationship and their synergism. Free Radical Res. 40: 993–1002.
•  Seitz,  C.,  M.  Vitten,  P.  Steinbach,  S.  Hartl,  J.  Hirsche,  W.  Rathje,  D.  Treutter and  G.  Forkmann. 2007. Redirection of anthocyanin synthesis in Osteospermum hybrida by a two-enzyme manipulation strategy. Phytochemistry 68: 824–833.
•  Tatebe,  T. 1938. On inheritance of root color in Raphanus sativus Linn. Jpn. J. Genet. 14: 39–50 (In Japanese with English abstract).
•  Tatsuzawa,  F.,  N.  Saito,  K.  Toki,  K.  Shinoda,  A.  Shigihara and  T.  Honda. 2010. Acylated cyanidin 3-sophoroside-5-glucosides from the purple roots of red radish (Raphanus sativus L.) ‘Benikanmi’. J. Japan. Soc. Hort. Sci. 79: 103–107.
•  Tatsuzawa,  F.,  K.  Toki,  N.  Saito,  K.  Shinoda,  A.  Shigihara and  T.  Honda. 2008. Anthocyanin occurrence in the root peels, petioles and flowers of red radish (Raphanus sativus L.). Dyes Pigments 79: 83–88.
•  Tsuda,  S.,  Y.  Fukui,  N.  Nakamura,  Y.  Katsumoto,  K.  Yonekura-Sakakibara,  M.  Fukuchi-Mizutani,  K.  Ohira,  Y.  Ueyama,  H.  Ohkawa,  T. A.  Holton,  T.  Kusumi and  Y.  Tanaka. 2004. Flower color modification of Petunia hybrida commercial varieties by metabolic engineering. Plant Biotechnol. 21: 377–386.
•  Wan,  C. Y. and  T. A.  Wilkins. 1994. A modified hot borate method significantly enhances the yield of high-quality RNA from cotton (Gossypium hirsutum L.). Anal. Biochem. 223: 7–12.
• Werck-Reichhart, D., S. Bak and S. Paquette. 2002. Cytochromes P450. Arabidopsis Book. <http://dx.doi.org/10.1199/tab.0028>.
•  Xu,  B. B.,  J. N.  Lia,  X. K.  Zhanga,  R.  Wanga,  L. L.  Xiea and  Y. R.  Chai. 2007. Cloning and molecular characterization of a functional flavonoid 3'-hydroxylase gene from Brassica napus. J. Plant Physiol. 164: 350–363.
•  Zufall,  R. A. and  M. D.  Rausher. 2003. The genetic basis of a flower color polymorphism in the common morning glory (Ipomoea purpurea). J. Hered. 94: 442–448.