Introduction

Anthocyanins are widely distributed in seed plants and are one of three main pigments for plant coloration, being responsible for orange-to-blue colors found in many flowers, leaves, fruits, seeds and other tissues (Tanaka et al. 2008). A variety of horticultural plants accumulate anthocyanins in ripening fruits, such as grape (Vitis vinifera), apple (Malus domestica), mangosteen (Garcinia mangostana), Chinese bayberry (Myrica rubra), pear (Pyrus pyrifolia), lychee (Litchi chinensis), blood oranges (Citrus sinensis) and strawberry (Fragaria ananassa) (Takos et al. 2006; Espley et al. 2007; Matus et al. 2009; Palapol et al. 2009; Feng et al. 2010; Niu et al. 2010; Wei et al. 2011; Butelli et al. 2012; Schaart et al. 2013). Based on previous studies, it was well recognized that anthocyanin biosynthesis is transcriptionally regulated by the MYB–bHLH–WD40 (MBW) transcription complex (Baudry et al. 2004; Koes et al. 2005; Hichri et al. 2011; Petroni and Tonelli 2011). A ternary complex containing these three members in conjunction with the promoters of anthocyanin structural genes regulates anthocyanin biosynthesis (Koes et al. 2005).

Chinese bayberry (M. rubra Sieb. et. Zucc.) is an economically important subtropical fruit crop belonging to the Myricaceae family and native to southern China and other Asian countries (Chen et al. 2004; Sun et al. 2013). Anthocyanins are the principal color compounds in Chinese bayberry and play an essential role in the fruit quality (Zhang et al. 2008; Niu et al. 2010; Sun et al. 2012a, b). In our previous studies, MrMYB1 and MrbHLH1, the MYB (Niu et al. 2010; Huang et al. 2013) and bHLH (unpublished data) transcription factors regulating anthocyanin biosynthesis in Chinese bayberry, were isolated and their function verified. However, the WD40 partner which could interact with MrMYB1 and MrbHLH1 and then regulate anthocyanin biosynthesis in Chinese bayberry has not been identified.

As a member of the MBW transcription complex, the common and defining feature of WD40s is the WD (Trp–Asp), also known as WD-40 motif, a ~40-amino acid stretch typically ending in amino acid residues WD (Neer et al. 1994). WD40s exist in all eukaryotes and constitute a large family, there are 237 and 200 WD40 members in Arabidopsis and rice, respectively (van Nocker and Ludwig 2003; Ouyang et al. 2012). Commonly, 4–16 copies of conserved WD repeat motifs are present in a single protein, and these motifs act as a site for protein–protein interaction (Smith et al. 1999; van Nocker and Ludwig 2003). WD40s have a wide variety of functions including signal transduction, RNA processing, cytoskeleton dynamics, vesicular trafficking, nuclear export, regulation of cell division, and are especially prevalent in chromatin modification and transcriptional regulation (Smith et al. 1999; van Nocker and Ludwig 2003; Ouyang et al. 2012). However, only a few members of the WD40 gene family have been verified to be involved in transcriptional regulation in plants, such as salt stress response, seed development and anthocyanin biosynthesis regulation (de Vetten et al. 1997; Walker et al. 1999; Sompornpailin et al. 2002; Carey et al. 2004; Pang et al. 2009; Brueggemann et al. 2010; Lee et al. 2010; Matus et al. 2010; Ben-Simhon et al. 2011; An et al. 2012; Bjerkan et al. 2012; Gao et al. 2012).

The first member of the WD40 gene family related to anthocyanin biosynthesis was isolated by transposon tagging and named AN11 which controls flower pigmentation in petunia (de Vetten et al. 1997). Subsequently, in Arabidopsis it was found that a WD40 protein encoded by TTG1 could regulate several developmental and biochemical pathways including the formation of hairs on leaves, stems, and roots, the production of seed mucilage and the biosynthesis of anthocyanins (Walker et al. 1999). To date, WD40 members related to flavonoid and anthocyanin biosynthesis have been isolated in several species, such as PfPFWD in Perilla frutescens (Sompornpailin et al. 2002), ZmPAC1 in Zea mays (Carey et al. 2004), MtWD40-1 in Medicago truncatula (Pang et al. 2009), VvWDR1 in grapevine (Matus et al. 2010), MdTTG1 in apple (Brueggemann et al. 2010), PgWD40 in pomegranate (Ben-Simhon et al. 2011) and FaTTG1 in strawberry (Schaart et al. 2013). As Smith et al. (1999) suggested, the WD-repeat proteins having very similar surfaces are likely to have common binding partners and similar functions, and all of these verified WD40 genes regulating anthocyanin biosynthesis have four highly conserved WD motifs. Moreover, the last two amino acid residues in each repeat domain also retain high conservation among different plant species (Pang et al. 2009; Brueggemann et al. 2010; Ben-Simhon et al. 2011).

For comprehensively elucidating the mechanism of anthocyanin biosynthesis regulation in Chinese bayberry, we isolated and verified MrWD40-1, a WD40 member regulating anthocyanin accumulation in Chinese bayberry.

Materials and Methods

Plant Materials

Fruits at commercial maturity of four different cultivars of Chinese bayberry (M. rubra Sieb. et. Zucc.) — named ‘Shuijing’ (SJ), ‘Fenhong’ (FH), ‘Dongkui’ (DK) and ‘Biqi’ (BQ) — were collected at 86 days after full bloom (DAFB) from commercial orchards in Yuyao County, Zhejiang Province, China, during the 2011 season. Young leaves, stems and fruits at 60, 67, 74, 78 and 82 DAFB of ‘BQ’ were also collected.

All of the samples were healthy and uniformly distributed around the tree. Within 4 h, the plant materials had been collected and transported to the lab, photographs taken and the color measured, then were frozen in liquid nitrogen immediately after being cut into small pieces, and stored at −80 °C. Each sample consisted of three biological replicates (15 fruits for fruit replicates or approximately 30 g for other tissue replicates).

Color Measurement

Color measurements were carried out with a reflectance spectrophotometer (Hunter Lab Mini Scan XE Plus) at harvest and were calculated according to the color index of red grapes (CIRG) as described by Zhang et al. (2008). Based on the method, L*, a* and b* values of fruits were recorded and the color index CIRG = (180 − H)/(L* + C), where H = arc tan b*/a* and C = [(a*)2 + (b*)2]*0.5. Four measurements were made on each fruit at equatorial positions around the fruit.

Anthocyanin Determination

Anthocyanin contents in samples were detected by the pH differential method as described in Zhang et al. (2008). Absorbance of sample extracts at 510 and 700 nm were measured using a UV-2550 spectrophotometer (Shimadzu).

Gene Isolation and Sequence Analysis

Firstly, all WD40 unigenes contained in the Chinese bayberry RNA-Seq database (Feng et al. 2012) were isolated according to annotation information of the unigenes. Then, the phylogenetic relationship between these unigenes and the WD40 genes related to anthocyanin biosynthesis in other plant species was analyzed by means of MEGA v. 5.0 (Tamura et al. 2011). Based on the resulting phylogenetic tree, two closest members, MrWD40-1 and MrWD40-2, were isolated and the sequences were verified by PCR with primers listed in Table 1. Subsequently, the open reading frames (ORF) of these two members were analysed by ORF Finder (http://www.ncbi.nlm.nih.gov/gorf/gorf.html). Sequence homology and alignment were carried out with BLASTn (http://blast.ncbi.nlm.nih.gov/Blast.cgi) and Clustal_X (Thompson et al. 1997), respectively.

Table 1 Primers used in this study
Full size table

Gene Expression Analysis

Total RNA was extracted using a modified cetyltrimethylammonium bromide (CTAB) method (Shan et al. 2008). One gram of DNA-free RNA was used to synthesize first-strand cDNA with Revert Aid™ First Strand cDNA Synthesis kit (Fermentas, USA) in a final 20-μl reaction mixture after removal of genomic DNA by DNase I (Fermentas, USA) according to the manufacturer's instructions. The real-time quantitative PCR (Q-PCR) mixture (25 μl total volume) included 12.5 μl SYBR Green/ROX qPCR Master Mix (Fermentas, USA), 3 μl of each primer (2.5 μM) and 6.5 μl diluted cDNA. Then the Q-PCR was performed on an iCycler iQ real-time PCR instrument (Bio-Rad, USA), initiated by the preliminary steps of 2 min at 50 °C, 10 min at 95 °C, followed by 40 cycles of 95 °C for 15 s, 58 °C for 30 s, and 72 °C for 30 s. All the Q-PCR reactions were normalized using the Ct value corresponding to the actin gene (MrACT, GenBank GQ340770). No-template controls and melting curve analyses were included for each gene and each PCR reaction. Q-PCR primers contained in Table 1 were designed with Primer3 (http://frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi). Gene specificities of the primers were confirmed by melting peaks, dissociation curves, and PCR product sequencing as described in our previous report (Yin et al. 2008).

Tobacco Transient Expression Assay

Tobacco transient expression assay was performed with tobacco (Nicotiana tabacum) leaves as previously reported (Espley et al. 2007; Niu et al. 2010). ORF of either MrWD40-1 or MrWD40-2 were recombined into pGreenII 0029 62-SK vector with the primers described in Table 1. Other recombinant vectors containing either MrMYB1 (Niu et al. 2010; Huang et al. 2013) or MrbHLH1 (submitted elsewhere) were constructed by previous studies. All constructs were electroporated into Agrobacterium tumefaciens GV3101 (MP90) individually, and were infiltrated into the abaxial leaf surface together, either binary or ternary, or alone for the induction of anthocyanins. A strain containing empty pGreenII 0029 62-SK vector served as a negative control. When comparing the effect of anthocyanin biosynthesis regulation by binary complex containing MrMYB1–MrbHLH1 and ternary complex including MrMYB1–MrbHLH1–MrWD40-1 or MrMYB1–MrbHLH1–MrWD40-2, these two groups were infiltrated into either side of the same leaf. Every treatment was carried out with three biological replicates in three different tobacco plants. Digital photographs of anthocyanin development in these patches were taken at 4, 6 and 8 days after infiltration.

Yeast Two-Hybrid Assay

Yeast two-hybrid assay was performed using Matchmaker™ Gold Yeast Two-Hybrid (Clontech, USA) according to the manufacturer's instructions described in Clontech Yeast Protocol Handbook. Briefly, pGADT7, containing GAL4 activation domain, and pGBKT7, containing GAL4 DNA-binding domain, were used. The ORF of MrWD40-1 was inserted into the multiple cloning site (MCS) of pGBKT7 while that of MrMYB1 or MrbHLH1 were inserted into pGADT7. The BD-MrWD40-1 was transformed alone or together with AD-MrMYB1/MrbHLH1 plasmids into the yeast strain Y2HGold using the PEG/LiAc method. The transformed colonies of BD-MrWD40-1 were tested on synthetic dropout (SD) medium with X-α-Gal and Aureobasidin A (AbA) while lacking adenine, histidine, and tryptophan, for their autoactivation. The co-transformed colonies were selected on SD medium lacking leucine and tryptophan (SD/-Leu/-Trp), and screened for growth on quadruple dropout SD medium lacking adenine, histidine, leucine and tryptophan (SD/-Ade/-His/-Leu/-Trp). To further confirm the positive interactions, X-α-Gal was used to assay for β-galactosidase activity.

Results and Discussion

Isolation of Two Putative Anthocyanin Biosynthesis Regulating WD40 Members

A total of 32,805 unigenes have been identified in the Chinese bayberry RNA-Seq database through annotation against public protein databases (Feng et al. 2012). Among these unigenes, 60 WD40 members were identified based on the annotation information. Data from phylogenic analysis suggested that the unigenes 803 and 11128, designated MrWD40-1 and MrWD40-2, respectively, had the highest homology with anthocyanin biosynthesis regulating WD40 genes from other plants (Fig. 1). Through sequence assembling and PCR confirmation, the full length of MrWD40-1 cDNA was 1,654 bp, and the ORF was 996 bp encoding 331 amino acids. At the deduced amino acid level, the homology between MrWD40-1 and WD40 members from other plants ranged from 75.2 % (ZmPAC1) to 91.1 % (VvWDR1). The full length of MrWD40-2 was 1,453 bp, and the ORF was 1,035 bp encoding 345 amino acids. MrWD40-2 showed 68.3 % (ZmPAC1) to 76.7 % (VvWDR1) homology to WD40 member from other plants listed in Fig. 1. The amino acid sequence homology between MrWD40-1 and MrWD40-2 is 75.3 %.

Fig. 1
figure 1

Phylogenetic analysis of WD40 sequences from Chinese bayberry and other plants. Two WD40 genes of Chinese bayberry are marked with black dots, while the genes ID of those from other species are as follows: InWDR1 (Ipomoea nil, GenBank accession number AB232779.1), PhAN11 (Petunia hybrid, AAC18914.1), VvWDR1 (Vitis vinifera, ABF66625.2), RrTTG1 (Rosa rugosa, AFY23208.1), MdTTG1 (Malus domestica, ADI58759.1), PpWD40 (Pyrus pyrifolia, ADU25044.1), MtWD40-1 (Medicago truncatula, ABW08112.1), NtTTG1 (Nicotiana tabacum, ACJ06978.1), AtTTG1 (Arabidopsis thaliana, AED93320.1), ZmPAC1 (Zea mays, AAM76742.1)

Full size image

Genes having similar structures may have similar functions, and vice versa, and four highly conserved WD-repeat motifs were observed in all known WD40 members related to anthocyanin biosynthesis including MdTTG1, AtTTG1 and PfPFWD (Walker et al. 1999; Sompornpailin et al. 2002; Brueggemann et al. 2010) as well as VvWDR1, MtWD40-1, PpWD40, etc. (Fig. 2). These four conserved motifs were also found in MrWD40-1 and MrWD40-2 (Fig. 2). Furthermore, the last two amino acid residues of each WD-repeat motif (WD, FD, LD and WE) also showed high conservation among different species (Fig. 2). This was consistent with previous results showing high conservation of these four WD repeat motifs and the last amino acid residues in each motif of the known WD40 genes related to anthocyanin biosynthesis (Pang et al. 2009; Brueggemann et al. 2010; Ben-Simhon et al. 2011).

Fig. 2
figure 2

Deduced amino acid sequence alignment of MrWD40-1, MrWD40-2 in Chinese bayberry with WD40 sequences known to be related to anthocyanin biosynthesis in other plants. The four WD-repeat domains were marked by four open boxes

Full size image

The sequence similarity between MrWD40-1, MrWD40-2 and anthocyanin biosynthesis regulating WD40 genes from other plants raises the question: is MrWD40-1 or MrWD40-2, or both of them, involved in anthocyanin biosynthesis regulation in Chinese bayberry?

Expression Profile of MrWD40-1 Corresponded Well with Anthocyanin Accumulation in Chinese Bayberry

To test the relationship between expression profile of MrWD40-1, MrWD40-2 and anthocyanin accumulation, ‘BQ’ fruit at six developmental stages were sampled. The sampling date, parameters and CIRG value for fruit color and anthocyanin content of each developmental fruit were described in Table 2, where L*, a* and b* indicated lightness, red-greenness and blue-yellowness, respectively.

Table 2 CIRG values and anthocyanin content in different developmental stages of ‘BQ’ fruits
Full size table

Fruit at early stages of development, i.e., 60, 67 and 74 DAFB (days after full blossoms), accumulated little anthocyanin (no more than 0.84 mg/100 g FW) with green surface color and CIRG values lower than 3.00 (Table 2). Consistent with the low anthocyanin content in fruit, the transcript level of MrWD40-1 remained low during these stages (Fig. 3a). At 78 DAFB, fruit had absolute red surface color with CIRG value increasing to 3.68 and anthocyanin contents to 4.25 mg/100 g FW (Table 2), while expression of MrWD40-1 rose slightly. However, anthocyanin contents increased sharply to 55.61 mg/100 g FW when fruit were at 86 DAFB with purple fruit color (Table 2), and the transcript level of MrWD40-1 increased approximately 3-fold compared to earlier stages (Fig. 3a). The expression of MrWD40-2, on the other hand, was much higher during early stages of fruit development, although it also increased further during ripening (Fig. 3b). The coefficient of determination (R 2) of MrWD40-1 and anthocyanin accumulation was 0.9721, while that of MrWD40-2 was 0.2218 in these developmental stages of fruits.

Fig. 3
figure 3

Analysis of expression patterns of MrWD40-1 and MrWD40-2 in different developmental stages of ‘BQ’ bayberry fruits (a, b) and different tissues (c, d). Error bars indicated SE from three biological replicates

Full size image

Anthocyanin accumulation also varied significantly among different tissues in ‘BQ’ bayberry: anthocyanin was non-detectable in stems; there was a slight accumulation, 15.63 mg/100 g FW, in young leaves, causing them to have a red margin, and sharp accumulation in ripening fruit, 55.61 mg/100 g FW, which appeared dark purple. Like anthocyanin accumulation, the transcript levels of MrWD40-1 showed obvious tissue specificity, and the level for ripe fruit was over 5-fold greater than in young leaves, and 10-fold higher than in stems (Fig. 3c). However, compared with the high positive correlation (R2 = 0.9738) between MrWD40-1 and the anthocyanin biosynthesis, there was a little weaker correlation (R 2 = 0.9321) between the transcript level of MrWD40-2 and anthocyanin accumulation in different tissues, particularly between the young leaves and stems (Fig. 3d).

Different Chinese bayberry cultivars show differences in anthocyanin accumulation in ripe fruits. For example, ‘SJ’ fruit are white colored and do not accumulate anthocyanin, while ‘FH’, ‘DK’ and ‘BQ’ accumulate small, moderate and considerable amounts, respectively (Fig. 4a, b). When the expression profiles of MrWD40-1 were examined in these four different cultivars, a positive correlation (R 2 = 0.9978) was observed between gene transcript abundance and anthocyanin content (Fig. 4b, c). The highest transcript level of MrWD40-1 occurred in purple ‘BQ’ fruits, and it decreased progressively, in line with anthocyanin content, being lowest in ‘SJ’ fruits (Fig. 4c). In contrast, MrWD40-2 was expressed in fluctuating amounts in different cultivars of bayberry, but there was poor correlation (R 2 = 0.5219) with anthocyanin content (Fig. 4d).

Fig. 4
figure 4

Correlation between anthocyanin content and transcription factor expression profile in different cultivars of Chinese bayberry. The images of ripe fruits of four Chinese bayberry cultivars, named ‘SJ’ (Shuijing), ‘FH’ (Fenhong), ‘DK’ (Dongkui) and ‘BQ’ (Biqi) from left to right, indicate the significant different colors (a) due to different anthocyanin accumulation (b). Analysis of expression patterns of MrWD40-1 (c) and MrWD40-2 (d) in these four different cultivars of Chinese bayberry by means of real-time quantitive PCR. Error bars indicated SE from three biological replicates

Full size image

A positive correspondence has been observed between transcript abundance of WD40 genes and function in anthocyanin production. For example, MdTTG1 from apple restores anthocyanin production in young seedlings of ttg1 mutants of Arabidopsis thaliana (Brueggemann et al. 2010), and its highest transcription abundance in apple is detected in fruit skin where anthocyanins accumulate to a high level (An et al. 2012). Anthocyanins in 35S:VvWDR1 transgenic A. thaliana lines were enhanced, and the expression pattern of VvWDR1 was quite well correlated with that of UDP-glucose: flavonoid 3-O-glucosyltransferase (VvUFGT) at different stages of fruit development (Matus et al. 2010). Our results showed good correlations between anthocyanin content and transcript level of MrWD40-1 at different developmental stages of fruits, or in different tissues, or cultivars, but not with MrWD40-2 despite of the high transcript abundance in these samples (Figs. 3 and 4). However, the positive correlation does not always exist between the gene expression and the key function. For example, the effect on the DFR activity of MdbHLH3 was much stronger than the other member, MdbHLH33, despite the fact that it was much less expressed than MdbHLH33 in apple (Espley et al. 2007). These gene expression patterns suggest MrWD40-1, rather than MrWD40-2, has the potential ability to regulate the anthocyanin biosynthesis in Chinese bayberry.

Transient Expression of MrWD40-1 Enhanced MrMYB1–MrbHLH1 Induction of Anthocyanin Accumulation in Tobacco Leaves

Tobacco transient expression assay was used to probe the function of MrWD40-1 and MrWD40-2 in anthocyanin biosynthesis regulation. No anthocyanin accumulation was observed with empty vector or either MrMYB1, MrbHLH1, MrWD40-1, MrWD40-2 singly or the binary complexes containing either MrMYB1–MrWD40-1, MrbHLH1–MrWD40-1, MrMYB1–MrWD40-2 or MrbHLH1–MrWD40-2 infiltrated into the leaves (data not shown). When MrMYB1 and MrbHLH1 were over-expressed together, clear red colored patches developed (Fig. 5), which were due to the significant anthocyanin accumulation (Espley et al. 2007; Niu et al. 2010). Interestingly, when MrWD40-1 was infiltrated into leaves together with MrMYB1 and MrbHLH1, earlier and stronger anthocyanin accumulation was observed (Fig. 5). These results confirmed the suggestion that MrWD40-1 boosted anthocyanin biosynthesis by interacting and most probably stabilizing the MrMYB1–MrbHLH1 complex.

Fig. 5
figure 5

Anthocyanin accumulation activated by binary group of MrMYB1 and MrbHLH1 or ternary groups of MrMYB1, MrbHLH1 and MrWD40-1or MrWD40-2 in tobacco leaves. Pictures were taken at 4 (a), 6 (b) and 8 (c, d) days after infiltration. Three tobacco plants designated 1, 2 and 3, respectively, were analyzed as three biological replicates

Full size image

In contrast, the ternary expression of MrMYB1, MrbHLH1 and MrWD40-2 resulted in less anthocyanin accumulated (Fig. 5d). To further probe the putative function of MrWD40-2, more plant WD40 members were included in phylogenetic analysis, and it was observed that MrWD40-2 clustered into a clade together with VvWDR2 (GenBank accession number ABF66626.2) (Fig. S1). The similarity between MrWD40-2 and VvWDR2 at the amino acid level was 97.7 %, which was much higher than the similarity between MrWD40-2 and MrWD40-1 (75.3 %), as well as between MrWD40-2 and VvWDR1 (76.7 %), a WD40 member regulating anthocyanin biosynthesis in grape (Matus et al. 2010). The function of VvWDR2 is not known; however, it was reported that it could not stimulate anthocyanin biosynthesis in transformed Arabidopsis, although it contained the four conserved WD repeat motifs (Matus et al. 2010). Therefore, it is probable that MrWD40-2 cannot regulate anthocyanin biosynthesis in Chinese bayberry either.

In the MBW complex regulating anthocyanin biosynthesis, WD40 is more likely to enhance gene activation rather than participating in the specific recognition of a target gene promoter because it has no obvious catalytic activity (Baudry et al. 2004; Hichri et al. 2011). However, the presence of a WD40 gene is essential and irreplaceable in the regulation of anthocyanin biosynthesis, as there is no anthocyanin accumulation in the ttg1 mutant Arabidopsis seedling (Brueggemann et al. 2010). Based on the results of transient expression assay, anthocyanin biosynthesis does improve with MrWD40-1, but is also quite strong without MrWD40-1 (Fig. 5). A possible explanation for this phenomenon could point to the homologue of MrWD40-1 in tobacco which has not been identified, yet could interact to some extent in a complex with MrMYB1 and MrbHLH1.

MrWD40-1 Physically Interacts with MYB and bHLH Transcription Factors

As anthocyanin biosynthesis regulation involves three transcription factors (MYB, bHLH and WD40), the physical interaction among MrMYB1, MrbHLH1 and MrWD40-1 was tested by yeast two-hybrid assay. The ORF of MrWD40-1 was inserted into MCS of pGBKT7 vector and no auto-activation was verified on SD medium with X-α-Gal and 125 ng/ml AbA background while lacking adenine, histidine and tryptophan (data not shown). Then the bait construct carrying the BD-MrWD40-1 fusion protein was co-transformed with the individual prey constructs harboring either the AD-MrMYB1 or the AD-MrbHLH1 fusion proteins. It was shown that both MrMYB1 and MrbHLH1 could physically interact with MrWD40-1 (Fig. 6). However, when MrbHLH1 served as the bait, it could not interact with the prey construct carrying MrWD40-1 (Fig. 6). This suggested that the interaction between MrbHLH1 and MrWD40-1 was unidirectional, which has often been encountered in yeast two-hybrid assays (Baudry et al. 2004; Besseau et al. 2012). As the construct harboring BD-MrMYB1 fusion protein had auto-activity on the selection SD medium, the interaction between BD-MrMYB1 and AD-MrWD40-1 was not investigated.

Fig. 6
figure 6

Analysis of physical interaction between MrWD40-1 and MrMYB1, MrbHLH1 in yeast

Full size image

The interaction between WD40s and their binding partners has been studied in several reports on anthocyanin biosynthesis regulation. The existence of physical interaction between AtTT2 (MYB) and AtTTG1 (WD40), as well as between AtTT8 (bHLH) and AtTTG1, has been reported in Arabidopsis (Baudry et al. 2004). Similar phenomena were observed in Chinese bayberry, for both MrMYB1 and MrbHLH1 could physically interact with MrWD40-1 (Fig. 6). However, in maize and apple, the WD40 protein could only interact with bHLH rather than MYB (Grotewold et al. 2000; An et al. 2012). This difference may due to the different species used in these studies.

Conclusion

Two WD40 genes isolated from the Chinese bayberry RNA-Seq database showed high homology with the WD40 genes related to anthocyanin biosynthesis in other species. MrWD40-1 from Chinese bayberry at the deduced amino acid level showed 75.2–91.1 % similarity to other plant WD40 members identified as regulating anthocyanin biosynthesis. MrWD40-2 showed somewhat lower similarity (68.3–76.7 %). However, gene expression analysis indicated a positive correlation between anthocyanin accumulation and the transcript level of MrWD40-1 rather than MrWD40-2. Transient expression assay showed that MrWD40-1 enhanced the anthocyanin biosynthesis induced by MrMYB1-MrbHLH1 in tobacco leaves, while MrWD40-2 could not stimulate anthocyanin accumulation, despite having the four conserved WD repeat motifs known to be located in WD40 genes related to anthocyanin biosynthesis. Furthermore, MrWD40-1 was shown to physically interact with both MrMYB1 and MrbHLH1 in yeast assay, consistent with it being the crucial WD40 gene regulating anthocyanin biosynthesis in Chinese bayberry.