McMYB12 Transcription Factors Co-regulate Proanthocyanidin and Anthocyanin Biosynthesis in Malus Crabapple

The flavonoid compounds, proanthocyanidins (PAs), protect plants from biotic stresses, contribute to the taste of many fruits, and are beneficial to human health in the form of dietary antioxidants. In this study, we functionally characterized two Malus crabapple R2R3-MYB transcription factors, McMYB12a and McMYB12b, which co-regulate PAs and anthocyanin biosynthesis. McMYB12a was shown to be mainly responsible for upregulating the expression of anthocyanin biosynthetic genes by binding to their promoters, but to be only partially responsible for regulating PAs biosynthetic genes. In contrast, McMYB12b showed preferential binding to the promoters of PAs biosynthetic genes. Overexpression of McMYB12a and McMYB12b in tobacco (Nicotiana tabacum) altered the expression of flavonoid biosynthetic genes and promoted the accumulation of PAs and anthocyanins in tobacco petals. Conversely, transient silencing their expression in crabapple plants, using a conserved gene region, resulted in reduced PAs and anthocyanin production a green leaf phenotype. Meanwhile, transient overexpression of the two genes and silenced McMYB12s in apple (Malus domestica) fruit had a similar effect as overexpression in tobacco and silenced in crabapple. This study reveals a new mechanism for the coordinated regulation of PAs and anthocyanin accumulation in crabapple leaves, which depends on an auto-regulatory balance involving McMYB12a and McMYB12b expression.

resources, due to the high flavonoid levels in its leaves, flowers and fruits 44 . Furthermore, the abundant levels of flavonoids in the leaves and fruits represent an excellent source of antioxidants for use as food nutrition additives 45 , and the consumption of crabapple leaf tea as a health beverage is gaining popularity in Asia. In the current study, we compared PAs biosynthesis in two commercial crabapple cultivars which planted in northern area in China: 'Royalty' with ever-red leaves and 'Flame' with ever-green leaves (Originated in America and have relative closely genetic background) 42 . We also investigated the transcriptional regulation of PAs biosynthesis by functionally characterizing the Malus crabapple TFs, McMYB12a and McMYB12b, which were identified as potential targets because of their similarity to identified PAs regulators in grape.
We present results indicating that both McMYB12 proteins are putative regulators of the PAs and anthocyanins branches of the flavonoid pathway. This reinforces the idea that the biosynthesis of flavonoid compounds is a complex process involving transcriptional network changes, such that TFs serving as main regulators in one branch of the flavonoid pathway may coordinately regulate other flavonoid branches that share an upstream substrate.

Expression of McMYB12 TFs correlates with PAs accumulation during crabapple leaf development.
We previously reported that the anthocyanin concentration in ever-red leaves, was significantly higher than that in ever-green leaves 17 , and that a competitive relationship between anthocyanin and flavonol is important for leaf coloration 46 . We characterized PAs accumulation patterns during five leaf developmental stages (1)(2)(3)(4)(5) of the ever-red leaf cultivar 'Royalty' and the ever-green leaf cultivar 'Flame' , using high pressure liquid chromatography (HPLC) ( Fig. 2A). We observed that in the ever-red 'Royalty' cultivar, the abundance of anthocyanins (cyanidin-3-O-glucoside) and PAs was relatively high in the first three stages, before decreasing to low levels in the last two stages, while epicatechin accumulation gradually increased. In the ever-green leaved 'Flame' cultivar, anthocyanins were almost undetectable, whereas relatively high levels of PAs and epicatechin were measured during leaf development (Fig. 2B). The concentration of flavonols (quercetin-3-O-rhamnoside and rutin) gradually decreased in 'Royalty' and 'Flame' leaves during leaf development, with the exception of stage 4 in 'Flame' . The same trends of decreasing levels were also observed for phlorizin in 'Flame' and for in avicularin in 'Royalty' . In contrast, avicularin abundance increased during the development of 'Flame' leaves, while there was no apparent variation in amount of the phlorizin in 'Royalty' leaves.
We also evaluated the abundance of these flavonoid compounds in the petals, fruit peel and fruit flesh of 'Royalty' and 'Flame' (Supplemental Figs 1A, 2A and 3A). In petals, the concentrations of anthocyanins and avicularin were higher in 'Royalty' than in 'Flame' , and decreased in these two cultivars during petals development. The concentration of (− )-epicatechin significantly increased in the last development stage of petals in 'Royalty' but the compound was not detected in 'Flame' petals. The amounts of quercetin-3-O-rhamnoside and phlorizin were similar in the two cultivars (Supplemental Fig. 1B).
The concentrations of most flavonoids decreased in fruit peels during fruit development in 'Royalty' and 'Flame' . Anthocyanin and avicularin were not detected in 'Flame' peels and the concentrations of PAs ((− )epicatechin and procyanidin B2) in 'Royalty' were higher than in 'Flame' (Supplemental Fig. 2B). We observed a gradual decrease in procyanidin B2 and quercetin-3-O-rhamnoside levels during fruit flesh development, while anthocyanins and quercetin-3-O-rhamnoside were not detected in 'Flame' fruit flesh, and (− )-epicatechin was not detected in 'Royalty' fruit flesh (Supplemental Fig. 3B).
Taken together, these results suggested that the red crabapple color is related to anthocyanin accumulation, and that the levels of PAs show substantial variation during leaf, petal and fruit development.  (Fig. 3A). The two McMYB12 proteins showed sequence homology with known PAs regulatory MYB family proteins from other species 29,30,32,33 , and each was predicted to contain a DNA-binding ID domain motif, a C1 motif and a C3 motif. Phylogenetic analysis showed that McMYB12a and McMYB12b are closely related to the potential PAs regulators MdMYB12, VvMYB5a and VvMYB5b (Fig. 3B). Their expression patterns were found to be positively correlated with the expression of the PAs biosynthetic genes  was used as the reference gene. S1 to S5 represents stage 1, 2, 3, 4 and 5 of leaf development.
Error bars indicate the standard error of the mean ± SE of three replicate measurements. The expression levels and correlation of flavonoid regulatory and biosynthetic genes were calculated using CFX-Manager-3-1 following the manufacturer's instructions (Bio-Rad). Scale bars = 1 cm. Different letters above the bars indicate significantly different values (P < 0.05) calculated using one-way analysis of variance (ANOVA) followed by a Tukey's multiple range test. LacZ activity was detected in yeast cells harboring pJG4-5-McMYB12a together with pLacZi-proMcCHS, pLacZi-proMcANS, pLacZi-proMcANR1, pLacZi-proMcANR2 and pLacZi-proMcLAR2, but not in yeast harboring pLacZi-proMcF3'H, pLacZi-proMcDFR, pLacZi-proMcUFGT, pLacZi-proMcFLS and pLacZi-proMcLAR1 together with pJG4-5-McMYB12a, or in the control. LacZ activity was detected in yeast harboring pJG4-5-McMYB12b together with pLacZi-proMcANS, pLacZi-proMcANR1, pLacZi-proMcANR2, pLacZi-proMcLAR1 and pLacZi-proMcLAR2, but not in yeast harboring pLacZi-proMcCHS, pLacZi-proMcF3'H, pLacZi-proMcDFR, pLacZi-proMcUFGT and pLacZi-proMcFLS together with pJG4-5-McMYB12b, or in the control. From this To further determine whether McMYB12 proteins binds directly to the promoter of the flavonoid biosynthetic genes identified in the yeast one hybrid assay, we performed an electrophoretic mobility shift assay (EMSA). Biotin labeled probes were designed according to the CAACTG element in proMcCHS and proM-cANS, the AACCTAA element in proMcANR1, the TAACTG element in proMcANR2 and the TATCC element in proMcLAR1 and proMcLAR2 (Table S2). We found that McMYB12a bound to the biotin labeled proMcCHS, proMcANS, proMcANR1, proMcANR2 and proMcLAR2 probes, and that McMYB12b bound to the biotin labeled proMcANS, proMcANR1, proMcANR2, proMcLAR1 and McLAR2 probes (Fig. 4B). This binding diminished gradually with an increasing concentration of un-labeled probe, while there was no evidence of competition using a mutated probe (Fig. 4B) In addition, dimethylaminocinnamaldehyde (DMACA) staining revealed that the accumulation of PAs in the transgenic seeds was higher than in control seeds, while HPLC confirmed that the concentration of PAs was at least 2 fold higher in the McMYB12a and McMYB12b transgenic lines than that in control seeds (Supplemental Fig. 7A,B). And there is no obviously variation of flavonoids content in McMYB12s transgenic leaves compared with control leaves (Supplemental Fig. 7C,D).
Overexpression McMYB12s has the same phenotypes with overexpression VvMYB5a and VvMYB5b in tobacco. The flavonoids biosynthesis pathway was only activated in reproductive organs, not in vegetative organs. We speculated that the tobacco leaves maybe lack in the flavonoids biosynthesis substrates. These results also indicate that McMYB12a and McMYB12b can regulate flavonoid metabolism in tobacco flowers and seeds; however, the regulatory modes and function of these two TFs may differ depending on the target genes.  vector, pTRV2-GFP, rapidly accumulated anthocyanins, resulting in red coloration in the upper young leaves (Fig. 7A). In addition, GFP fluorescence was observed in leaf veins and new buds at 14 dpi, indicating that the virus spread through the plants. HPLC analyses confirmed that the anthocyanin concentration in the silenced leaves (12.1 μ g/g FW) was less than that in the leaves of control lines (133.8 μ g/g FW) (Fig. 7B). The PAs content in leaves from plantlets expressing with TRV-GFP-McMYB12s was approximately 70% lower than in the control lines (Fig. 7B). Moreover, levels of rutin, quercetin-3-O-rhamnoside, phlorizin and avicularin were lower than in control leaves.

Overexpression or silencing of
qRT HPLC analysis showed that overexpression of McMYB12a resulted in a significantly increase in anthocyanins accumulation and a small increase in PA contents, while a decrease in quercetin-3-O-rhamnoside in transgenic fruit than that in control lines. In contrast, the anthocyanin content of McMYB12b overexpressing apple fruit was lower than in the control fruit, while epicatechin and procyanidin B2 contents increased almost 2-fold. In pTRV2-McMYB12 infiltrated peels, silencing of McMYB12 caused a decrease in anthocyanin, PA and quercetin-3-O-rhamnoside abundances (Fig. 8B).
The transcript levels of the McMYB12 genes and several flavonoid biosynthetic genes in the transformed apple fruit were evaluated using qRT-PCR. We observed that changes in the PAs and anthocyanin contents were accompanied by increased or reduced transcript levels of the closely related apple genes, MdMYB121 (99% sequence identity to McMYB12a) and MdMYB12 (99% sequence identity to McMYB12b), and the PAs biosynthesis genes (Fig. 8C)

Discussion
Our data suggest that two highly homologous R2R3-MYB TFs (McMYB12a and McMYB12b) can co-regulate the PAs and anthocyanin biosynthesis, thereby revealing a new mechanism for the regulation of PAs biosynthesis.

PAs and anthocyanins in crabapple leaves.
PAs, the products of the flavonoid pathway, accumulate in a wide variety of plant tissues and play roles in many physiological and biochemical processes 47 . Together with anthocyanins and flavonols, which are also antioxidants and have beneficial effects for human health, the regulation of their biosynthesis has been studied in a diverse range of crops, such as apple 11,16,[34][35][36][37][38]40,41 , grape [26][27][28][29][30][31][32][33]48 and strawberry (Fragaria × ananassa) 49,50 , reflecting their importance in fruit and vegetable color and flavor as well as their health promoting attributes.
The biosynthetic pathways of PAs, anthocyanins and flavonols share common steps in the flavonoid pathway (from CHS to F3H), and each class of flavonoid (anthocyanin, flavonol, and flavan-3-ol) is synthesized by a multienzymatic step enzyme reaction branching from the common flavonoid pathway (Fig. 1) 11 . Several studies have shown that the competitive relationship between anthocyanins and flavonols is important for coloration 46,51 ; however, further research is required to understand the association between PAs and anthocyanin accumulation. Our results indicate that crabapple leaf color is closely related to both anthocyanin and PAs abundance. In crabapple, a gradual decrease in anthocyanin content was observed in 'Royalty' leaves during their development, and the levels of PAs compounds showed an increasing trend relative to the decrease in anthocyanin accumulation. In contrast, because of the deficiency in anthocyanin biosynthesis, anthocyanin was almost undetectable in the leaves of the 'Flame' cultivar, and the PAs content was higher than that in 'Royalty' (Fig. 2). In addition, up-regulation in  (Figs 6A and 8). We deduced from these data that there is a metabolic balance between flavonol and anthocyanin biosynthesis 46

Expression of McMYB12 genes correlates with PAs synthesis in Malus crabapple. Several MYB
TFs are known to be involved in the regulation of PAs biosynthesis. In grapevine, the expression of VvMYBPA1 during flower and early berry development, and in seeds before ripening, correlates with PAs accumulation and the expression of PAs-specific biosynthetic genes 29 . The spatiotemporal expression of VvMybPA2, which is restricted to the skin of young berries and leaves, is congruent with a role in the regulation of PAs biosynthesis. Moreover, the expression of VvMybPA2 in berry skins suggests that it is more likely than VvMybPA1 to be involved in PAs synthesis 29,30 . The expression of VvMYB5b, combined with the action of specific regulators, such as VvMYBA1 and VvMYBPA1, controls the biosynthesis of both anthocyanins and PAs throughout grape berry development 52 . Recently, a grape R2R3-MYB transcription factor, VvMYBPAR, was shown to be relatively highly expressed in the skins of young berries, with even higher levels in the seeds, and maximal expression was detected around the veraison developmental stage 31 . In apple, epicatechin and catechin biosynthesis is under the control of the biosynthetic enzymes ANR and LAR, respectively 16 . The expression of MdMYB9 and MdMYB11 followed a similar pattern as the biosynthetic enzyme MdANR, with high relative expression levels detected in flowers and old leaves for all genes 40,41 .
In the current study of the 'Royalty' and 'Flame' crabapple cultivars, we found that the expression patterns of the two McMYB12a and McMYB12b during leaf, petals and fruits development are consistent with them having a role in regulating PAs synthesis. The transcript profiles of McMYB12a and McMYB12b correlated with those of McANRs and McLARs, and with the accumulation of PAs, suggesting that the McMYB12 proteins may regulate PAs synthesis by modulating the expression of genes that are specific to the PAs branch of the phenylpropanoid pathway (Fig. 2C). Interestingly, we found that the expression of McMYB12a partially correlated with the accumulation of anthocyanins, so these TFs may affect not only just one or two branches of the phenylpropanoid pathway, but rather a large portion of the pathway (Fig. 2B,C), as was reported to be the case for VvMYB5a 32 and MdMYB9 and MdMYB11 41 . The patterns of anthocyanin and PAs accumulation showed a similar trend, so we conclude that the biosynthesis of both classes of compounds shares upstream genes/enzymes, and that their biosynthesis occurs concurrently. Variation in the concentrations of phenolic compounds (e.g. rutin, querecetin-3-O-rhamnoside, phlorizin, avicularin) was observed in McMYB12s -overexpressing or silenced organs. In addition, we observed a correlation between variation in expression of the McMYB12 genes and the transcript levels of others MYB transcription factors which involved in flavonol and anthocyanins biosynthesis in crabapple leaves and apple peels. Moreover, the yeast one hybrid results showed that LacZ activities were not detected in yeast harboring pLacZi-proMcMYB4, pLacZi-proMcMYB16 together with pJG4-5-McMYB12a and pJG4-5-McMYB12b. Based on these results, we speculated that variation in levels of these compounds occurs via the McMYB12 genes directly or indirectly modulating the expression of the flavonoid biosynthesis genes. This might involve a variety of mechanisms, including redistribution of secondary metabolites/solutes, or altered secondary metabolism. We conclude that an important function of McMYB12 transcription factors is to regulate the biosynthesis of anthocyanins and PAs by binding to the promoters of downstream anthocyanin and PAs biosynthesis genes.
The extant apple chromosomes homologies derived from a putative nine-chromosome ancestor 53 . Each doublet of the eight apple chromosomes (3-11, 5-10, 9-17 and 13-16) is derived principally from one ancestor, and interestingly, McMYB12a and McMYB12b are located on chromosomes 3 and 11, respectively. In the Rosaceae, including Malus spp., an evolutionary trend toward important trait specialization may have been partially based on gene duplication, resulting in the creation of large families of paralogous genes 53 . McMYB12a and McMYB12b, which share 92% sequence identity, regulate different biosynthetic genes involved in the modulation of anthocyanin and PAs levels in a counterbalanced secondary metabolism. This may account for phenotypic leaf color variation in crabapple. Heterologous and homologous expression analyses results showed that overexpression of McMYB12a affected the expression of various flavonoid structural genes, as well as the amounts of anthocyanins and PAs. In contrast, McMYB12b overexpression promoted the accumulation of PAs, but did not activate anthocyanin synthesis, suggesting it may be specific to the PAs pathway. Conversely, silencing of both McMYB12a and McMYB12b in crabapple plantlets suppressed the accumulation of anthocyanins and PAs and decreased the expression of anthocyanin and PAs biosynthetic genes. We propose that McMYB12b specifically regulates the PAs biosynthetic pathway, as is the case with VvMYBPA1, a transcriptional regulator of PAs synthesis in grape 29 , while McMYB12a is mainly responsible for regulating anthocyanin accumulation and partially responsible for PAs accumulation in crabapple leaves. The different functions of McMYB12a and McMYB12b reflect the regulation of different downstream flavonoid biosynthetic genes (Fig. 9).
In this current study, we observed that several flavonoid biosynthetic genes were activated by McMYB12s, and that this activation was accompanied by the accumulation of anthocyanins and PAs. Promoter binding assays showed that
Nicotiana tabacum cv. W38 was used for overexpression experiments. The plants were grown in a greenhouse at 27 °C under 16-h-light/8-h-dark illumination. Transgenic plants were created using Agrobacterium tumefaciens strain GV3101 by suspension in the Agrobacterium solution. Nicotiana benthamiana was used to promoter activity and transient activation assays. The plants were grown in pots in a growth chamber at 23 °C with a 16 h/8 h photoperiod and ~60% relative humidity.
For the apple studies, bagged fruits were harvested 145 days after full bloom (DAFB) from adult trees of the 'Red Fuji' cultivar (M. pumila Milier). The skin of the fruits was peeled, together with less than 1 mm of cortical tissue, for anthocyanin measurements and other analyses 54 .
Explants of Malus cv. 'Royalty' were harvested from one-year-old branches before spring bud germination, cultured on Murashige and Skoog medium supplemented with 0.1 mg/L 6-Benzylaminopurine (6-BA) and 2 mg/L (2,4-dichlorophenoxy) acetic acid (2,4- were extracted with 10 mL extraction solution (methanol: water: formic acid: trifluoroacetic acid = 70: 27: 2: 1) as previously described at 4 °C in the dark for 72 h, with shaking every 6 h 55 . The liquid was separated from the solid matrix by filtration through sheets of qualitative filter paper. The filtrate was then passed through 0.22-μ m reinforced nylon membrane filters (Billerica, MA, USA). Trifluoroacetic acid: formic acid: water (0.1: 2: 97.9) was used as mobile phase A, and trifluoroacetic acid: formic acid: acetonitrile: water (0.1: 2: 48: 49.9) was used as mobile phase B. PAs were extracted using 1 ml of 70% (v/v) acetone containing 0.1% (w/v) ascorbate, and incubated at room temperature for 24 h in the dark. The extract was centrifuged at 12, 000 g for 15 min at room temperature, and the supernatant was transferred to a new 1.5 ml microfuge tube. An aliquot of 200 μ l of extract was dried at 35 °C and resuspended in 100 μ l of 1% (v/v) HCl/methanol and 100 μ l of 200 mM sodium acetate (pH 7.5). The injection volumes were 20 μ l for anthocyanin and flavonol analysis or 10 μ l for PA analysis. The gradients used were as follows: 0 min, 30% B; 10 min, 40% B; 50 min, 55% B; 70 min, 60% B; 80 min, 30% B. Detection was performed at 520 nm for anthocyanins, 350 nm for flavonols and 280 nm for PAs using an HPLC1100-DAD system (Agilent Technologies, Waldbronn, Germany) 56 . All samples were analyzed in biological triplicate.  the Rapid Amplification of cDNA Ends (RACE) method and nested PCR performed according to the manufacturer's recommendations (Clontech, Palo Alto, CA). All PCR products were sub-cloned into the pGEM T-Easy Vector (Promega, Madison, WI) and transformed into Escherichia coli DH5α cells and sequenced. PCR primer sequences are listed in Table S1. Comparison and analysis of the sequences were performed using the advanced basic local alignment search tool (BLAST) at the National Center for Biotechnological Information (http://www. ncbi.nlm.nih.gov). The full-length DNA and protein sequences were aligned using DANMAN 5.2.2 (Lynnon Biosoft, USA). Phylogenetic and molecular evolutionary analyses were conducted with MEGA version 5.1, using a minimum evolution phylogeny test with a 1,000 bootstrap replicates 57 .

Cloning and GUS activity analysis of the McMYB12s promoters. To analyze the differences in
McMYB12s promoter sequences between the different cultivars, genomic DNA was isolated from leaves using the Plant Genomic DNA Kit (TIANGEN BIOTECH CO., LTD, Beijing, China). Cloning primers PMYB12a-F, PMYB12a-R PMYB12b-F and PMYB12b-R were designed using NCBI Primer BLAST and are listed in Table S1. PCR products were cloned into the pMD-19T vector and sequenced.
The promoter of McMYB12a and McMYB12b were cloned into the modified pBI121 vector 44 using the XhoI and XbaI sites, and Agrobacterium GV3101 carrying McMYB12a promoter-GUS or McMYB12b promoter-GUS were collected by centrifugation at 4000 rpm for 10 min, and suspended in the infiltration buffer (10 mM MgCl 2 , 150 mM acetosyringone, and 10 mM MES, pH 5.6) to a final optical density at 600 nm of 0.8. CaMV 35 S promoter -GUS was as control. 100 μ l bacterial suspension was infiltrated into intercellular spaces of a young tobacco leaf using a 1 ml plastic syringe without needle 58 . 2-4 pieces fully expanded young leaves were infiltrated per plant in three plants. After agroinfiltration, tobacco plants continued growing at 23 °C for 3 days as usual. Six technical replicates of 3 mm diameter leaf discs (about 1/3 of infiltrated area) were excised from each plant using a leaf hole-punch and buffered in Phosphate Buffer Saline (PBS). Subsequently, discs were respectively used for the histochemical GUS analysis and GUS activity assay 59 . qRT-PCR analysis. Total RNA from crabapple leaves, tobacco petals and apple peel were extracted as described above. DNase I (TaKaRa, Ohtsu, Japan) was added to remove genomic DNA, and the samples were then subjected to cDNA synthesis using the Access RT-PCR System (Promega, USA), according to the manufacturer's instructions. The expression levels of flavonoid biosynthetic genes in crabapple and tobacco were analyzed using qRT-PCR and the SYBR Green qPCR Mix (TaKaRa, Ohtsu, Japan) and the Bio-Rad CFX96 Real-Time PCR System (BIO-RAD, USA), according to the manufacturers' instructions. The PCR primers were designed using NCBI Primer BLAST and are listed in Table S1.
qRT-PCR analysis was carried out in a total volume of 20 μ l containing 9 μ l of 2 × SYBR Green qPCR Mix (TaKaRa, Ohtsu, Japan), 0.1 μ M specific primers (each), and 100 ng of template cDNA. The reaction mixtures were heated to 95 °C for 30 s, followed by 39 cycles at 95 °C for 10 s, 50-59 °C for 15 s, and 72 °C for 30 s. A melting curve was generated for each sample at the end of each run to ensure the purity of the amplified products. The transcript levels were normalized using the Malus 18 S ribosomal RNA gene (DQ341382, for apple and crabapple) or the NtActin gene (GQ339768, for tobacco) as the internal controls and calculated using the 2 (−∆∆Ct) analysis method 60 .

Overexpression in tobacco. The full length McMYB12a and McMYB12b open reading frames (ORFs)
were cloned into the pBI121 vector 59 using the BamHI and SacI sites, and A. tumefaciens strains carrying these constructs were used to transform Nicotiana tabacum 'W38' using the leaf disk method 61 . The PCR primers used are listed in Table S1. Transgenic plants were selected by kanamycin resistance. Two independent lines from the T2 progeny were used to compare color changes with the untransformed wild-type line.
DMACA Staining of PAs. The presence of PAs in plant tissue was detected by staining the tissues with DMACA (dimethylaminocinnamaldehyde) solution (1% DMACA, 1% 6 N HCl in methanol). Dried seeds were stained for 6 h and the seedlings for 20 min. After staining, the samples were transferred to distilled water and blue staining was visualized and photographed with a Zeiss Discovery V20 stereomicroscope. Transient expression assays in crabapple plantlets and apple fruit. Fragments for the pTRV2-GFP-McMYB12s (458 bp) constructs were amplified by PCR with gene-specific primers, from a cDNA library derived from Malus crabapple leaves (cv. 'Royalty'), using Taq DNA polymerase (TaKaRa, Ohtsu, Japan), according to the manufacturer's instructions. The PCR primers used are shown in Table S1.
A. tumefaciens cells were grown, collected, and resuspended in a 10 mM MES, 10 mM MgCl 2 , and 200 mM acetosyringone solution to a final optical density of 1.5 at 600 nm and then incubated at room temperature for 3 to 4 h without shaking. Before infiltration, A. tumefaciens cultures containing combinations of pTRV1 (acts as an assistant vector and is responsible for virus replication and for allowing systemic movement throughout the host) 62 , pTRV2-GFP or its derivatives were mixed in a 1:1 ratio.
The infiltration protocol and culture methods for transient expression assays in crabapple plantlets were adapted as previously described 63,64 . The infiltration protocol and culture methods for apple fruits were adapted as previously described 46 .
For visualization of the GFP fluorescence in whole crabapple plants, the samples were illuminated with a 100 W hand-held long-wave ultraviolet lamp (UV products, Upland, CA, USA; Black Ray model B 100AP/R) and were photographed using a Kodak wratten filter 15.
Yeast one-hybrid assay. A yeast one-hybrid system was used to assay the transcriptional activation by and McLAR-2 promoter sequences were inserted upstream of the reporter LacZ gene in the pLacZi vector. The effector and reporter or control constructs were transformed into competent cells of the yeast strain EGY48, resulting in the following yeast strains: pJG4-5-McMYB12s/pLacZi-promoter of flavonoid biosynthetic genes, pJG4-5/pLacZi-pro of flavonoid biosynthetic genes, pJG4-5-McMYB12s/pLacZi, pJG4-5/pLacZi. The cells were selected on synthetic drop-out media lacking tryptophan and uracil, and positive colonies were spotted onto glucose plates (2%) containing X-gal at 28 °C for 2 days to confirm blue color development 17 11 . N. benthamiana plant growing conditions, A. tumefaciens infiltration processes, and luminescence measurements were as described previously 65 . For each TF-promoter interaction, three independent experiments were performed (at least six replicates in each experiments). Bimolecular fluorescence complementation assay. Bimolecular fluorescence complementation (BiFC) assay in vivo was performed as described 66 . McMYB12a, McMYB12b and McbHLH3 full-length sequence were amplified and cloned into the plasmid pSPYCE-35S/pUC-SPYCE and pSPYNE-35S/pUC-SPYNE, respectively. The pairs of the resulting fusion protein with corresponding empty vector were used as negative controls. These vectors were introduced into the Agrobacterium strain GV3101. The Agrobacterium were grown overnight in YEB media with appropriate antibiotic selection. Agrobacterium cultures containing the BiFC constructs plasmid were mixed at OD600 of 0.5: 0.5. Cells were pelleted by centrifugation and resuspended in infiltration medium (10 mM MgCl2, 10 mM MES, and 100 mM acetosyringone). After incubation for at least 2 h at room temperature, the suspension was infiltrated into leaves of 1-month old tobacco (Nicotiana benthamiana) plants. The photographs of YFP fluorescence were taken after 24-72 h of transformation using a confocal laser-scanning microscope (FV1000, Olympus, Tokyo, Japan).