Introduction

Mulberry plants (Morus L.) are well known for their use as forage for silkworms (Bombyx mori L.), and also have important economic and medical uses. Mulberry fruit has been exploited industrially in many countries (Sánchez 2000; Singhal et al. 2010) to take advantages of its delicious taste and beneficial activity for human health (Wu et al. 2013). Anthocyanins are the most abundant nutrient components in mulberry fruits, which also have important biological activities. Liu et al. (2004) measured the total anthocyanin content in 31 mulberry cultivars and revealed that mulberry anthocyanins had broad prospects as natural food colorants in the food industry. Over 11 anthocyanins have been identified in mulberry and the main components are cyanidin-3-O-glucoside and cyanidin-3-O-rutinoside (Du et al. 2008; Dugo et al. 2001; Lee et al. 2004). Characterization of anthocyanin biosynthetic genes will facilitate the genetic improvement of anthocyanins in mulberry. However, little information regarding anthocyanin biosynthetic genes is currently available for this species.

Anthocyanins are a group of water-soluble pigments found in plants. They are found predominantly in the epidermal cells of flowers and fruits and contribute to their brilliant red, purple, and blue colors (Sadilova et al. 2006; Stintzing et al. 2002). Anthocyanins also play important roles in plant growth, development, and defense, acting as major antioxidants in the scavenging of free radicals (Ravindra and Narayan 2003; Lev-Yadun and Gould 2009). In addition, studies reveal that anthocyanins are beneficial to human health in the treatment of inflammation, cancer, and cardiovascular diseases (de Pascual-Teresa and Sanchez-Ballesta 2008).

The biosynthetic pathway of anthocyanins has been well documented in many plants (Holton and Cornish 1995; Koes et al. 2005). A schematic representation is shown in Supplementary Fig. 1. Chalcone synthase (CHS) and chalcone isomerase (CHI) catalyze the formation of the intermediate product naringenin from upstream metabolites. Thereafter, flavanone 3-hydroxylase (F3H), flavonoid 3′-hydroxylase (F3′H) and flavonoid 3′,5′-hydroxylase (F3′5′H) catalyze a series of hydroxylation reactions to produce different kinds of dihydroflavonols. These hydroxylation reactions are essential to the formation of different kinds of anthocyanins, which produce various colors. After the hydroxylation reactions, dihydroflavonol reductase (DFR) catalyzes the reduction of dihydroflavonols to leucoanthocyanins using NADPH as a cofactor. Then, anthocyanin synthase (ANS) catalyzes the conversion of colorless leucoanthocyanins to colored anthocyanins. The connections between anthocyanin contents and the expression of anthocyanin biosynthetic genes have become a topic of intensive research. Boss et al. (1996) investigated the expression profile of genes involved in anthocyanin biosynthesis in five white and three red grape cultivars, and their results revealed that the majority of anthocyanin biosynthesis genes had higher levels of expression in red cultivars than in white ones. In apples, five anthocyanin biosynthesis genes were found to be coordinately expressed during the development of red coloration in the skin and their transcriptional levels showed a positive relationship with anthocyanin concentration (Honda et al. 2002). Salvatierra et al. (2010) investigated the expression of anthocyanin biosynthetic genes in strawberry fruits and found that the transcriptional levels of these genes increased during fruit ripening and the levels of expression were much higher in red fruit than in white fruit.

The complete sequence of the mulberry genome is now available (He et al. 2013). This provides an opportunity for the characterization of genes involved in anthocyanin biosynthesis. Here, nine putative anthocyanin biosynthetic genes were identified in the mulberry genome. The sequence features of these genes not only provide more data for plant comparative analysis but also facilitate future identification of candidate genes from closely related species of the family Moraceae. Their expression patterns were analyzed in two mulberry cultivars with differently colored fruit, and the results indicated that the transcription levels of genes in most anthocyanin biosynthetic steps are associated with anthocyanin content in the mulberry fruit. The information provided here will expand our knowledge of anthocyanin biosynthesis and could be incorporated into anthocyanin improvement strategies for mulberry fruit.

Materials and methods

Mulberry materials

The mulberry Morus notabilis C. K. Schneid. used for the present study was grown in Sichuan province, China. The fruit color of M. notabilis changed from green to white during fruit ripening. Lateral roots, 1-year-old shoots, leaves, male flowers, female flowers, and fruit were collected for RNA extraction. The mulberry cultivars M. atropurpurea Roxb. ‘Da10’ and M. alba ‘Zhenzhubai’ were maintained at the Mulberry Germplasm Nursery in Southwest University. During ripening, ‘Da10’ fruits changed from green to red and finally purple, and ‘Zhenzhubai’ fruits changed from green to white (Supplementary Fig. 2). Fruit samples for RNA and anthocyanin extraction were collected at four stages: I (green, 15 days after anthesis, daa), II (intermediate green, 25 daa), III (semi-mature, 40 daa) and IV (mature, 50 daa). All samples were immediately frozen in liquid nitrogen and stored at −80 °C until use.

Identification and cloning of mulberry anthocyanin biosynthetic genes

The genome sequence of M. notabilis was downloaded from the Morus Genome Database (http://morus.swu.edu.cn/morusdb/). A total of 27,085 high-confidence protein-coding loci were identified in the mulberry genome by combining results from homology-based and de novo gene prediction methods. The amino acid sequences of genes involved in anthocyanin biosynthesis were downloaded from NCBI (http://www.ncbi.nlm.nih.gov/) and then used as queries for blast searches against the mulberry protein database. Gene-specific primers were designed to amplify the anthocyanin biosynthetic genes using mixed cDNA of six mulberry tissues (roots, shoots, leaves, male flowers, female flowers, and fruit) as a template. For a list of primers used in this study see Supplementary Table 1. The purified PCR products were TA cloned into the pMD19-T simple vector (Takara, Otsu, Japan) and the sequences were confirmed by sequencing.

DNA extraction and Southern blotting

Buds of M. notabilis and two mulberry cultivars were powdered in liquid nitrogen and the DNA was extracted using a CTAB method (Lodhi et al. 1994). After digestion with RNase, the DNA was re-extracted with phenol/chloroform/isoamyl alcohol (25:24:1) and chloroform/isoamyl alcohol (24:1). The DNA quality and concentration were measured using a ND-1000 UV spectrophotometer (Nanodrop Technologies, Wilmington, DE, USA). For Southern blotting, each 15 μg of genomic DNA was digested with restriction endonuclease and the resultant fragments were resolved on a 0.8 % (w/v) TAE agarose gel. The bands were transferred to nylon membranes (Roche, Indianapolis, IN, USA). Probes for Southern blotting were designed based on the genome sequence of M. notabilis (For CHS, the sequence identity of MnCHS1 and MnCHS2 was too high to distinguish them, so the same probe was used for both; Supplementary Table 2). The probes were labeled and signals were detected with the DIG High Prime DNA Labeling and Detection Starter Kit II (Roche, Indianapolis, IN, USA) according to the manufacturer’s instructions. After being hybridized with the DIG-labeled probes, membranes were washed twice in 2 × SSC containing 0.1 % SDS for 15 min at room temperature, and then washed twice in 0.1 × SSC containing 0.1 % SDS for 15 min at 65 °C. Ready-to-use CSPD (Roche, Indianapolis, IN, USA) was used to detect the hybridization bands and the signals were observed using a chemiluminescence imaging system (Clinx, Shanghai, China).

RNA extraction and qRT-PCR

Total RNA from six different tissues was extracted using RNAiso Plus (Takara) according to the manufacturer’s instructions. The RNA quality and concentration were measured using a ND-1000 UV spectrophotometer (Nanodrop Technologies). First-strand cDNA was synthesized using 3 μg of total RNA with M-MLV reverse transcriptase (Promega, Madison, WI, USA) in a 25 μl reaction system. For quantitative real-time reverse transcriptional PCR (qRT-PCR), each reaction was prepared according to the manufacturer’s instructions using SYBR® Premix Ex TaqTM II (Takara) and 2 μl of diluted cDNA as a template. The qRT-PCR reactions were conducted on the StepOne Real-Time PCR System (Applied Biosystems, Foster City, CA, USA). The mulberry ribosomal protein L15 (RPL) gene was used as a control to normalize the relative expression of target genes. The relative expression was defined as 2−[Ct (target gene) − Ct (control gene)] × 1,000. For a list of gene-specific primers used for real-time RT-PCR see Supplementary Table 3.

Extraction and quantification of anthocyanins

Anthocyanins were extracted from mulberry fruits as described in Pang et al. (2007) with minor modifications. Briefly, frozen samples were powdered with liquid nitrogen, and 5 ml of extraction buffer (methanol: 0.1 % HCl) was added to 0.5 g ground powder. After 1 h of sonication, the homogenized samples were gently shaken overnight. The samples were centrifuged at 5,000 × g for 10 min and the supernatants were gathered. Thereafter, 1 ml of ddH2O was added to 1 ml of supernatant, followed by the addition of 1 ml chloroform to remove chlorophyll. One milliliter of supernatant was evaporated using a concentrator plus (Eppendorf, Hamburg, Germany) and dissolved in 1 ml of buffer (acetonitrile: ddH2O: 0.1 M HCl = 1:7:2). The extracts were filtered by 0.2 μm PTFE membrane and stored at −80 °C until use. A Waters Acquity UPLC system (Waters, Milford, MA, USA) was used to analyze the anthocyanin levels. Separations were performed on an Acquity UPLC ® BEH C18 column (1.7 μm, 1.0 × 100 mm) at 30 °C. A 7–11 % acetonitrile gradient in 0.07 % H3PO4 was used for elution and the flow rate was 1.4 ml/min. Fractions were monitored at 520 nm. Components were identified by comparison of the retention times of the eluting peaks to those of commercial standards (Sigma, St. Louis, MO, USA) under the same conditions. Dose-dependent calibration curves of the standards were used to determine the concentrations of the components.

Results

Identification and cloning of the anthocyanin biosynthetic genes in mulberry plants

As shown in Supplementary Table 4, we identified nine putative anthocyanin biosynthetic genes in M. notabilis by bioinformatic methods, including two CHS genes, two F3H genes, two F3′H genes, and one gene each of CHI, DFR, and ANS. A F3′5′H gene was not found in the mulberry genome. Prominent among these predicted genes, two CHS genes were located near each other on scaffold508 and two F3′H genes were located near each other on scaffold255. Gene-specific primers were designed based on the genomic sequences and used to clone the cDNAs of these genes. Except for ANS, transcripts of eight mulberry anthocyanin biosynthetic genes were detected and further cloned for sequence analysis. The transcript of ANS was not detected in M. notabilis and its genomic sequence was analyzed. The genomic structures of the cloned genes are illustrated in Supplementary Fig. 3. The structural features of these genes are conserved with their counterparts in other plants (Almeida et al. 2007).

Characterization of anthocyanin biosynthetic genes in mulberry

The deduced amino acid sequences of the nine predicted mulberry anthocyanin biosynthetic genes were blast searched against the SwissProt database. The sequence identities among eight of these mulberry genes ranged from 65 to 93 % (Table 1). The results of multiple sequence alignment indicated that the catalytic domains and active sites were conserved in these mulberry proteins. For example, both MnCHSs showed two conserved phenylalanine residues and the active site cysteine, histidine, and asparagine residues were also conserved. These sites are essential for substrate specificity (Supplementary Fig. 4) (Ferrer et al. 1999). Both the conserved substrate preference residues (Supplementary Fig. 5) and phylogenetic analysis with other plant CHIs (Supplementary Fig. 6) indicated that MnCHI was a type-I CHI, which are capable of isomerizing 6′-hydroxychalcone to 5-hydroxyflavanone (Shimada et al. 2003). MnF3H was found to have conserved histidine, aspartic acid, and arginine residues involved in iron binding (histidine, aspartic acid) and 2OG binding (arginine), which is important for 2-oxoglutarate-dependent dioxygenase activity (Supplementary Fig. 7) (Lukačin and Britsch 1997). Sequence alignment revealed that the mulberry MnDFR contained a conserved motif, VTGASGYIGSWLVMRLLERDY, for NADP binding (Fig. 1); (Johnson et al. 1999). In Fig. 1, underlined amino acids indicate regions that determine substrate specificity. Those labeled aspartic acid/asparagine were found to directly determine the substrate specificity of DFR (Johnson et al. 2001). The MnDFR is an Asp-type DFR, which, like that of petunia, cannot convert dihydrokaempferol to leucopelargonidin (Johnson et al. 2001).

Table 1 Comparison of deduced amino acid sequences of predicted mulberry anthocyanin biosynthetic genes and their counterparts in other plants
Full size table
Fig. 1
figure 1

Multiple sequence alignment of deduced MnDFR and other DFR proteins. The alignment was performed using CLUSTALX and the results were displayed using GENEDOC. Identical residues are highlighted in black. The black box shows the putative NADP-binding domain. The underlined amino acids indicate the region that determines substrate specificity and asterisks indicate residues key to the determination of substrate specificity. Proteins used for alignment are MnDFR from M. notabilis (KF438048), AtDFR from A. thaliana (P51102), VvDFR from V. vinifera (P51110) and PhDFR from P. hybrida (P14720)

Full size image

Southern blotting and transcriptional analyses of mulberry anthocyanin biosynthetic genes

The copy numbers of the anthocyanin biosynthetic genes in M. notabilis and two mulberry cultivars were investigated by Southern blotting. The results indicated that all of the identified genes existed in M. notabilis and two mulberry cultivars. Generally, there were less gene copies in M. notabilis than in ‘Da10’ or ‘Zhenzhubai’ and the hybridization patterns of ‘Da10’ were similar to those of ‘Zhenzhubai’, as shown in Fig. 2. Experiments were carried out to confirm the expression levels of the anthocyanin biosynthetic genes in six tissues of M. notabilis using qRT-PCR. Two MnCHS genes, MnCHS1 and MnCHS2, showed different expression patterns. As shown in Fig. 3a, b, the MnCHS1 gene was expressed exclusively in male flowers and its transcriptional level was much higher than that of MnCHS2, which was mainly expressed in roots and female flowers. The MnCHI gene was expressed at higher levels in roots (Fig. 3c). The expression levels of the two MnF3H genes were tissue-complementary; MnF3H1 was expressed mainly in roots and male flowers, while MnF3H2 was highly expressed in shoots, leaves, and female flowers (Fig. 3d, e). For the two MnF3′H genes, the expression of MnF3′H2 was restricted to roots, while MnF3′H1 was broadly expressed in several tissues (Fig. 3f, g). DFR was found to catalyze the reduction of dihydroflavonols to leucoanthocyanins (Johnson et al. 2001), which are direct precursors of anthocyanins. In M. notabilis, the MnDFR gene was more abundantly expressed in roots than in flowers or fruit (Fig. 3h). ANS is a key enzyme in anthocyanin biosynthesis that catalyzes the conversion of leucoanthocyanin to anthocyanin. No MnANS transcripts were detected in any of the six types of tissues. Also, no anthocyanin was measured in M. notabilis (data not shown).

Fig. 2
figure 2

Southern blotting analysis of anthocyanin biosynthetic genes in M. notabilis and two mulberry cultivars. Letters on the top represent restriction enzymes used to digest genomic DNA. H HindIII, E EcoRI, Bg BglII, Ba BamHI and K KpnI

Full size image
Fig. 3
figure 3

Spatial transcriptional analysis of genes associated with anthocyanin biosynthesis in M. notabilis using qRT-PCR. Six tissues were used: R root, S stem, L leaf, MF male flowers, FF female flowers, and F fruit. Relative gene expression levels were normalized against RPL transcript levels. Values represent the average ± SD of three biological replicates

Full size image

Accumulation of anthocyanin during fruit development in two mulberry cultivars

The types and concentrations of anthocyanins in the fruits of two mulberry cultivars with differently colored fruit were determined using a Waters Acquity UPLC system. By comparing the retention times to standards, the two compounds present in the fruits of ‘Da10’ were determined to be cyanidin-3-O-glucoside and cyanidin-3-O-rutinoside (Fig. 4a). In the ‘Da10’ fruit, the concentrations of the two anthocyanins were increased significantly in stage IV as compared with stage III and the degree of cyanidin-3-O-glucoside increase was much higher than that of cyanidin-3-O-rutinoside (Fig. 4b). No anthocyanin was detected in ‘Zhenzhubai’ fruits at any stage.

Fig. 4
figure 4

Identification and quantification of anthocyanins in the fruits of two mulberry cultivars during fruit ripening. a Identification of the two main anthocyanins in ‘Da10’ fruits using UPLC. b Changes in the concentration of anthocyanin (μg/g fresh weight) during fruit ripening in two mulberry cultivars. D and Z represent ‘Da10’ and ‘Zhenzhubai’, respectively

Full size image

Expression of genes associated with anthocyanin biosynthesis during fruit development in two mulberry cultivars

The expression patterns of the nine mulberry genes associated with anthocyanin biosynthesis were assessed in the fruits of two mulberry cultivars. Based on these expression patterns in two mulberry cultivars, we classified genes with similar pattern into one type. Nine mulberry genes were then divided into four types, as shown in Fig. 5. The genes in type I included CHS1, CHI, F3H1, F3′H1, and ANS. They were expressed increasingly during fruit ripening in the ‘Da10’ cultivar, but only trace expression of these genes was observed in the ‘Zhenzhubai’ cultivar. The gene CHS2, whose expression levels were higher in the ‘Zhenzhubai’ cultivar than in the ‘Da10’ cultivar during fruit ripening, was classified as type II. The F3H2 and F3′H2 genes were classified as type III. The transcriptional levels of these two genes were much lower than those of F3H1 or F3′H1 in both cultivars. The DFR gene was classified as type IV because of its distinct expression pattern. DFR expression was lower in ‘Da10’ fruits than in ‘Zhenzhubai’ fruits during ripening.

Fig. 5
figure 5

Transcriptional analysis of genes associated with anthocyanin biosynthesis during the development of ‘Da10’ and ‘Zhenzhubai’ fruits. The fruits were collected at four developmental stages, as described in the “Materials and methods” section. Black dots represent the transcriptional levels of ‘Da10’ and white dots represent those of ‘Zhenzhubai’. Dashed lines represent branch of delphinidin that was not synthesized in mulberry. The y-axis represents relative gene expression levels normalized against RPL transcript levels. Values represent the average ± SD of three biological replicates

Full size image

Discussion

Mulberry trees are cultivated worldwide. Their leaves are used as forage for silkworms in sericulture. People are also interested in mulberry fruits because of their delicious taste and considerable nutritional value. They are harvested for making jam, juice, wine, and food coloring. Mulberry fruits have been exploited industrially in many countries (Sánchez 2000; Singhal et al. 2010). The improvement of mulberry fruit quality is important for the use of mulberry fruits so genes that control fruit quality are of immediate interest. Anthocyanins are important to fruit quality, and research into anthocyanin biosynthesis in mulberry plants may facilitate improvement of the quality of mulberry fruits.

The completion of mulberry genome sequencing has provided researchers with an opportunity to identify genes involved in anthocyanin biosynthesis. In the present study, we identified nine putative anthocyanin biosynthetic genes that encoded enzymes controlling the committed steps in anthocyanin biosynthesis. However, no F3′5′H gene was identified in the mulberry genome. The F3′5′H gene encodes a cytochrome P450 enzyme that catalyzes the 3′,5′-hydroxylation of dihydroflavonols (Holton et al. 1993). The products of this reaction are the precursors of delphinidin, which is essential for the formation of blue colors in flowers (de Vetten et al. 1999). Previous studies have identified several varieties of anthocyanins in mulberry plants, including cyanidin, pelargonidin, and petunidin derivatives (Dugo et al. 2001). The absence of the F3′5′H gene in mulberry plants may account for the composition of anthocyanins in mulberry plants. The two anthocyanins identified in our experiments were cyanidin derivatives and these results are consistent with earlier reports (Dugo et al. 2001). The substrate specificity of DFR determines the type of anthocyanins produced (Johnson et al. 2001). Sequence analysis indicated that MnDFR belongs to the Asp-type DFRs, which are unable to convert dihydrokaempferol to leucopelargonidin. This would explain why cyanidin derivatives are the most abundant anthocyanins in mulberry plants.

Two CHS, two F3H, and two F3′H genes were identified in the mulberry genome. Each of these genes was found to have different spatial and temporal expression patterns. Similar phenomena have been observed in other plants. For example, in strawberries, one member of the FaCHS family was found to be expressed in petals and transcripts of the other were found to be abundant in fruit (Almeida et al. 2007). Jeong et al. (2006) cloned four F3′H genes in grape (F3′h1 and F3′h2 are alleles of one locus, and F3′h3 and F3′h4 are alleles of another) and found that the transcriptional levels of both sets of alleles (F3′h1–F3′h2; F3′h3–F3′h4) were clearly different in seeds and skins during development. In sorghum, the expression levels of two SbF3′H genes were shown to respond to different stimuli. The expression of SbF3′H1 was shown to be involved in the accumulation of light-specific anthocyanin, while SbF3′H2 expression was involved in pathogen-specific 3-deoxyanthocyanin accumulation (Shih et al. 2006). Differences in the expression of members of the same gene subfamilies might be the cause of functional differences and specialization of anthocyanin biosynthesis. Mulberry F3H1 and F3′H1 may be related to anthocyanin accumulation during fruit ripening, and F3H2 and F3′H2 may have functions that are not yet documented.

The biosynthesis of anthocyanins is part of flavonoid biosynthesis process and most anthocyanin biosynthetic genes participate in the biosynthesis of other flavonoids such as flavonols, isoflavonoids and proanthocyanidins. The transcripts of eight anthocyanin biosynthetic genes were detected in M. notabilis, but anthocyanin was undetectable in this mulberry species. This inconsistency indicated that the genes expressed may contribute to the biosynthesis of other flavonoids rather than anthocyanins.

The connections between the expression patterns of mulberry anthocyanin biosynthetic genes and the anthocyanin concentrations in mulberry fruits were evaluated. During the development of ‘Da10’ fruits, the transcription levels of genes in most anthocyanin biosynthetic steps clearly increased. Conversely, in the white fruit cultivar ‘Zhenzhubai’, the expression levels of anthocyanin biosynthetic genes were relatively low, and no anthocyanin was found in its fruit. Considering the results in these two mulberry cultivars, we found that the transcriptional levels of genes in most anthocyanin biosynthetic steps were positively related with anthocyanin content.

In this study, we identified nine putative anthocyanin biosynthetic genes in M. notabilis genome. Sequence analyses revealed that these genes were conserved with their counterparts of other plants. We speculated that the absence of F3′5′H gene and the substrate specificity of DFR lead to the accumulation of cyanidin derivatives in mulberry. Transcriptional levels of identified genes and the amount of anthocyanins in the fruits of two mulberry cultivars were analyzed. The results indicated that anthocyanin concentrations correlated with the transcriptional levels of anthocyanin biosynthetic genes including CHS1, CHI, F3H1, F3′H1, and ANS during the fruit ripening process. The information provided here will expand our knowledge of the biosynthesis of anthocyanin in mulberry fruits. Artificially changing the gene transcriptional levels by genetic engineering may lead to variations in anthocyanin content. The data from this study of genes associated with anthocyanin biosynthesis will accelerate the agricultural improvement of mulberry fruits.