MdMYB1 Regulates Anthocyanin and Malate Accumulation by Directly Facilitating Their Transport into Vacuoles in Apples  

Abstract

Tonoplast transporters, including proton pumps and secondary transporters, are essential for plant cell function and for quality formation of fleshy fruits and ornamentals. Vacuolar transport of anthocyanins, malate, and other metabolites is directly or indirectly dependent on the H⁺-pumping activities of vacuolar H⁺-ATPase (VHA) and/or vacuolar H⁺-pyrophosphatase, but how these proton pumps are regulated in modulating vacuolar transport is largely unknown. Here, we report a transcription factor, MdMYB1, in apples that binds to the promoters of two genes encoding the B subunits of VHA, MdVHA-B1 and MdVHA-B2, to transcriptionally activate its expression, thereby enhancing VHA activity. A series of transgenic analyses in apples demonstrates that MdMYB1/10 controls cell pH and anthocyanin accumulation partially by regulating MdVHA-B1 and MdVHA-B2. Furthermore, several other direct target genes of MdMYB10 are identified, including MdVHA-E2, MdVHP1, MdMATE-LIKE1, and MdtDT, which are involved in H⁺-pumping or in the transport of anthocyanins and malates into vacuoles. Finally, we show that the mechanism by which MYB controls malate and anthocyanin accumulation in apples also operates in Arabidopsis (Arabidopsis thaliana). These findings provide novel insights into how MYB transcription factors directly modulate the vacuolar transport system in addition to anthocyanin biosynthesis, consequently controlling organ coloration and cell pH in plants.

In plant cells, the vacuole, which occupies most of the cell volume, plays a crucial role in generating turgor, storing metabolites and maintaining cell pH balance (Marty, 1999; Martinoia et al., 2007; Faraco et al., 2014). In the case of fleshy fruits, the function of the vacuole in the peel and flesh cells has direct implications for fruit quality because the vacuole stores a wide variety of compounds, such as sugars, organic acids, and secondary metabolites, the composition and concentrations of which largely determine the appearance (especially color), taste, and flavor of the fruits (Shiratake and Martinoia, 2007; Sweetman et al., 2009; Etienne et al., 2013).

On the vacuolar membrane, two distinct proton pumps, vacuolar H⁺-ATPase (V-ATPase) and H⁺-pyrophosphatase (V-PPase), drive vacuolar acidification by transporting protons across the tonoplast into the vacuole (Shiratake and Martinoia, 2007). Moreover, a large number of secondary transporters and channels on the tonoplast, such as malate transporters (tDT), the vacuolar malate channel ALMT9, MATE-type anthocyanin transporters, and ABC transporters, are responsible for transporting malate and anthocyanins from the cytosol into the vacuole, respectively (Emmerlich et al., 2003; Kovermann et al., 2007; Gomez et al., 2009; Francisco et al., 2013). The activities of these transporters are directly or indirectly dependent on the proton gradients generated by V-ATPase and V-PPase (Shiratake and Martinoia, 2007).

V-ATPase is a complicated complex that is composed of the peripheral subcomplex V₁ and the membrane-bound subcomplex V₀. V₁ has eight different subunits (from A to H) and is mainly responsible for ATP hydrolysis, while V₀ contains six subunits (VHA-a, -c, -c′, -c′′, -d, and -e) and is responsible for proton translocation (Padmanaban et al., 2004). Plant H⁺-PPases are of two distinct types. Type I is K⁺-sensitive, while type II is K⁺-insensitive and Ca²⁺-hypersensitive. These H⁺-PPases generate proton gradients across the vacuolar membrane (Gaxiola et al., 2007). In flesh cells, vacuolar proton pumps and secondary transporters act together to transport a large amount of organic acids into the vacuole, thereby generating a low cell pH value via a so-called “acid trap” mechanism (Martinoia et al., 2007; Etienne et al., 2013), which contributes to the overall organoleptic quality of the fruits.

Organ color, an important economic trait in fruits and ornamental crops, is determined by various pigments. Among these pigments, anthocyanins are often responsible for organ coloration, such as bright red, red, blue, and violet (Młodzińska, 2009). Anthocyanins are biosynthesized via the flavonoid pathway in the cytosol and transported into the vacuole by putative anthocyanin transporters, such as MATE-type and ABC transporters (Debeaujon et al., 2001; Goodman et al., 2004; Marinova et al., 2007; Gomez et al., 2009). This biosynthetic pathway is transcriptionally regulated by an MBW complex containing WD-repeat proteins, bHLH, and MYB transcription factors, which are highly conserved among higher plant species (Koes et al., 2005; Ramsay and Glover, 2005; Ballester et al., 2010; Albert et al., 2011; Xie et al., 2012). Among them, MYB regulators are one of the master candidate genes regulating anthocyanin synthesis (Ban et al., 2007; Lin-Wang et al., 2010). In apples (Malus domestica), MdMYBA, MdMYB1, and MdMYB10 are alleles of a single locus in different genotypes. They differ in a repeated sequence of the promoter region. This repeated sequence ensures high expression levels of MdMYB10 in apple flesh and foliage. In contrast, MdMYB1 and MdMYBA alleles lack such a repeated sequence in their promoters, and their expression is limited to apple peels (Mahmoudi et al., 2012). In fact, the MdMYB1 protein is exactly identical to MdMYBA (Takos et al., 2006) and shares 98% of its homology with MdMYB10 (Espley et al., 2007).

In addition to anthocyanin biosynthesis and transportation, cell pH shifts the anthocyanin absorption spectrum to change the hue of tissues and organs (Yoshida et al., 2009). On the plasma membrane, distinct P-type H⁺-ATPases, such as PM H⁺-ATPases, control the cytoplastic pH (Palmgren, 2001; Palmgren and Nissen, 2011). Meanwhile, vacuolar proton pumps and secondary transporters are responsible for the regulation of vacuolar pH (Verweij et al., 2008). In Japanese morning glory (Ipomoea nil), mutants of the vacuolar Na⁺/H⁺ exchangers, InNhx1 and InNhx2, are unable to increase vacuolar pH and therefore develop purple reddish petals (Fukada-Tanaka et al., 2000; Ohnishi et al., 2005). In Petunia, the flower color generally varies from red to purple; however, seven mutants named ph1 to ph7 exhibit an increased vacuolar pH and produce bluish flowers (de Vlaming et al., 1983). In the past decade, most PHs genes have been identified and characterized. Among these genes, PH1 and PH5 encode vacuolar P_3B-ATPase and P_3A-ATPase, respectively (Verweij et al., 2008; Faraco et al., 2014). Both PH1 and PH5 contribute to vacuolar acidification (Verweij et al., 2008), and PH1 is a putative Mg²⁺ transporter. Although PH1 does not have proton transporter activity, it interacts with PH5 to boost the proton pump activity of PH5, thereby acidifying the cells (Faraco et al., 2014).

PH4 or AN2 (ANTHOCYANIN 2), which encodes an MYB transcription factor, interacts with the WD-repeat protein AN11 and bHLH transcription factor (TF) PH6/AN1 to intensify vacuolar acidification and control flower color in Petunia petal cells (Koes et al., 2005; Quattrocchio et al., 2006; Faraco et al., 2014). The loss-of-function mutant ph4 exhibits a bluish flower color due to the increased pH in the petal cells, while AN2 expression driven by a 35S promoter rescues pigmentation of an2 petals from white to a bluish color (Quattrocchio et al., 2006; Quattrocchio et al., 2013). It has recently been shown that PH1 and PH5 are transcriptionally activated by the AN11-PH6/AN1-PH4 complex (Faraco et al., 2014), suggesting a potential link between vacuolar acidification and anthocyanin accumulation. However, the exact mechanism of this transcriptional regulation has not been identified. In addition, it is not known whether other genes involved in the vacuolar transport of metabolites are transcriptionally regulated by the same complex. We hypothesized that the MYB transcriptional factor for anthocyanin synthesis also directly regulates V-ATPase and possibly other vacuolar transporters in coordinating vacuolar acidification and the transport of anthocyanins and other metabolites.

Remarkably, anthocyanins and malate, which are synthesized in the cytosol, translocate to various cell compartments, especially to the vacuoles, in apples (Fernie and Martinoia, 2009; Li et al., 2012; Etienne et al., 2013). The biosynthesis of anthocyanin is transcriptionally regulated by an MYB-bHLH-WDR (MBW) complex, such as MdTTG1-MdbHLH3-MdMYB1 in apples (An et al., 2012; Li et al., 2012; Xie et al., 2012; Vimolmangkang et al., 2013; An et al., 2015). However, the MdtDT gene is cloned, and preliminary work suggests that it plays a role in malate transport into the vacuole in apples (Yao et al., 2011b). In addition, the Ma locus is a major QTL that controls variations in the flesh and juice acidity levels (Xu et al., 2012). Recent genetic analysis indicates that two ALMT-like genes, Ma1 and Ma2, are strong candidates for the Ma locus (Bai et al., 2012).

In this study, we show that apple MdMYB1/10 TF directly regulates the expression of several genes that encode vacuolar proton pump subunits, including MdVHA-B1, MdVHA-B2, MdVHA-E2, and MdVHP1; an anthocyanin transporter, MdMATE-LIKE1; an ABC transporter; and a malate transporter, MdtDT, in modulating anthocyanin/malate accumulation and cell pH. We also demonstrate that the mechanism through which MYB TF controls cell pH and the vacuolar accumulation of anthocyanins and malate is conserved between apples and Arabidopsis (Arabidopsis thaliana).

RESULTS

Identification of the MYB Transcription Factor Binding to the cis-Element of MdVHA-B1 Promoter

By sequence analysis, we found a typical MYB-binding cis-element present in the promoter region of the V-ATPase subunit gene MdVHA-B1 (Fig. 1A). To identify the potential MYB proteins that recognize and bind to the MYB cis-element, an oligonucleotide probe, ACAATCAACGGTTAAA, from the MdVHA-B1 promoter, which contains the typical MYB cis-element CAACGG, was used for DNA-affinity trapping and electrophoretic mobility shift assays (EMSAs). When the nuclear protein extracts were incubated with biotin-labeled MdVHA-B1 promoter oligonucleotides, a DNA-protein complex with a slower mobility in EMSA was observed (Fig. 1, B and C).

Figure 1.

Identification of the MYB transcription factors binding to the cis-element of the MdVHA-B1 promoter. A, The putative MdMYB1 TF-binding element on the MdVHA-B1 promoter. The black box represents the MdMYB1-binding motif. B, Schematic outline of the method used for the affinity trapping of DNA-binding proteins. C, Identification of the MYB TF proteins binding to the cis-element of the MdVHA-B1 promoter with EMSA. The positions of the free DNA probes and DNA-protein complexes are indicated by arrows. + and − represent with or without the addition of total protein extracted from apple plants, respectively.

Subsequently, a mass spectrometry analysis was performed to identify the proteins that bind to the oligonucleotide. Interestingly, the protein MdMYB1 is a candidate in the LC/MS data (Supplemental Appendix S1). MdMYB1 and its alleles, MdMYB10 and MdMYBA, are master regulators for anthocyanin biosynthesis in apples (Takos et al., 2006; Ban et al., 2007; Li et al., 2012).

MdMYB1 Binds to the Promoters of MdVHA-B1 and MdVHA-B2

To determine whether MdMYB1 binds to the MYB recognition site in the MdVHA-B1 promoter, EMSA was performed using prokaryon-expressed and purified MdMYB1-GST fusion proteins. A specific DNA-MdMYB1 protein complex was detected when the CAACGG-containing oligonucleotide was used as a labeled probe (Fig. 2A). The formation of these complexes was reduced with increasing amounts of the unlabeled MYB competitor probe with the same sequence. This competition was not observed when using a mutated version of the probe (Fig. 2A). The specificity of this competition confirms that the specific binding of MdMYB1 to MdVHA-B1 promoter requires the MYB recognition sequence.

Figure 2.

MdMYB1 specifically binds to the promoters of MdVHA-B1 and MdVHA-B2. A, Interaction of the MdMYB1 protein with labeled DNA probes for the cis-elements of the MdVHA-B1 promoter in the EMSA. pVHA-B1 is used as a negative control without the MdMYB1 protein. B, The cis-element sequences and positions of the oligonucleotides within the MdVHA-B2 promoter that was used for the EMSAs. C, EMSA verification of the interaction between the MdMYB1 protein and the cis-element of the MdVHA-B2 promoter. D, The relative enrichment of the MdVHA-B1, MdVHA-B2 and MdVHA-B3 promoter fragments. The MdMYB1-DNA complex was coimmunoprecipitated from MdMYB1-GFP transgenic apple calli using an anti-GFP antibody. PBIN-GFP was used as a negative control. E, GUS activity in the transgenic apple calli as labeled. The means and standard deviations were calculated from the results of three independent experiments.

In apples, there are three isoforms of V-ATPase subunit B (VHA-B) of the V₁ complex. They are encoded by three genes: MdVHA-B1, MdVHA-B2 and MdVHA-B3 (Supplemental Fig. S1). Our sequence analysis showed that all three MdVHA-Bs contain MYB recognition sites in their promoter regions. EMSA demonstrated that the MdMYB1 protein binds to the promoters of MdVHA-B1 and MdVHA-B2, but not to that of MdVHA-B3 (Fig. 2, A–C; Supplemental Fig. S2, A and B).

To verify in vivo binding of MdMYB1 to the MdVHA-B1 and MdVHA-B2 promoters, ChIP-PCR assays were conducted using 35S::MYB1-GFP and 35S::GFP transgenic apple calli, respectively. The MYB-element-containing promoter regions of MdVHA-B1 and MdVHA-B2, but not MdVHA-B3, were enriched by ChIP in the 35S::MYB1-GFP transgenic calli compared to the 35S::GFP control (Fig. 2D). These results provide in vivo evidence for the binding of MdMYB1 to the MdVHA-B1 and MdVHA-B2 promoters.

GUS assays were then used to confirm the activation of MdMYB1 through the MdVHA-B1 and MdVHA-B2 promoters using the GUS reporter gene. Three recombinant plasmids, P_MdVHA-B1::GUS, P_MdVHA-B2::GUS, and P_MdVHA-B3::GUS, were combined with 35S::MdMYB1 and genetically transformed into apple calli (Fig. 2E; Supplemental Fig. S3, A and B). The transgenic-calli-containing P_MdVHA-B1::GUS (or P_MdVHA-B2::GUS) plus 35S::MdMYB1 exhibited a much higher GUS activity than those harboring P_MdVHA-B1::GUS (or P_MdVHA-B2::GUS) alone. In contrast, the addition of 35S::MdMYB1 had little influence on the GUS activity in P_MdVHA-B3::GUS transgenic calli, indicating that MdMYB1 activates GUS transcription driven by the promoters of MdVHA-B1 (or MdVHA-B2), but not by that of MdVHA-B3.

MdMYB1 Modulates Anthocyanin and Malate Accumulation, As Well As Cell pH Partially via MdVHA-B1 and MdVHA-B2

To characterize the functions of MdMYB1 and MdVHA-Bs, the full-length sense ORFs and antisense cDNA fragments of MdMYB1 and MdVHA-Bs were used to construct expression vectors. A total of five types of 35S-driven vectors of MdMYB1-S (“S” means sense), MdMYB1-AS (“AS” means antisense), MdVHAB1-S and MdVHABs-AS (antisense for both MdVHA-B1 and MdVHA-B2) were made. Subsequently, eight single and combinations of plasmid vectors, i.e. empty vector control, MdMYB1-S, MdVHAB1-S, MdMYB1-S+MdVHAB1-S, MdMYB1-S+MdVHABs-AS, MdMYB1-AS, MdVHABs-AS, and MdMYB1-AS+MdVHABs-AS, were used for genetic transformation. Because it is very difficult to generate transgenic plants in apples, especially for those containing two recombinant plasmids together, apple calli were used for genetic transformation and functional characterization.

As a result, eight transgenic apple calli were obtained, as indicated in Figure 3A. The qPCR assays demonstrated that the overexpression of the MdMYB1 gene enhanced the transcript levels of MdVHA-B1 and MdVHA-B2, but not that of MdVHA-B3, compared with the wild-type control (Fig. 3B), suggesting a positive regulation of MdMYB1 to MdVHA-B1 and MdVHA-B2 genes. Meanwhile, an anti-MdVHAB antibody, which is specific to MdVHA-B1 and MdVHA-B3, but not to MdVHA-B2 (Supplemental Fig. S1B), was used for immunoblotting assays. The abundance of MdVHA-B1 increased with MdMYB1 protein levels, whereas MdVHA-B3 remained at a similar level in the control and transgenic apple calli (Fig. 3C). To examine whether MdMYB1 positively regulates MdVHA-B2 at the protein level, two transgenic apple calli, i.e. P_MdVHA-B2::MdVHA-B2-GUS (transgenic calli containing MdVHA-B2-GUS fusion gene driven by MdVHA-B2 promoter) and 35S::MdMYB1+P_MdVHA-B2::MdVHA-B2-GUS (MdMYB1 overexpressor in a P_MdVHA-B2::MdVHA-B2-GUS background), were obtained. The results showed that the abundances of MdVHA-B1 and MdVHA-B2-GUS proteins were positively regulated by MdMYB1, while that of MdVHA-B3 was not (Supplemental Fig. S4). Therefore, MdMYB1 acts upstream of MdVHA-B1 and MdVHA-B2 to activate their transcript levels and protein abundances in apple calli.

Figure 3.

MdMYB1 controls malate and anthocyanin accumulation and cell pH via MdVHA-B1 and MdVHA-B2 in apple calli. A, Anthocyanin accumulation in wild-type and transgenic apple calli under 17°C plus UVB light conditions. The numbers 1 to 8 represent the wild-type and transgenic apple calli containing different combinations of constructs as indicated. B and C, qRT-PCR (B) and western blotting (C) analyses of MdMYB1 and MdVHA-Bs transcripts/proteins in the wild-type and transgenic apple calli. All protein amounts were normalized based on the protein folds of the wild-type control. Note: In C, protein bands were quantified by scanning densitometry using a Hewlett Packard scanjet scanner and Scanplot software. All protein amounts were normalized based on the protein folds of the wild-type control. D, Anthocyanin content in the wild-type and transgenic apple calli. E, Bafilomycin A₁-sensitive ATP hydrolytic activity of V-ATPase vacuolar membrane vesicles were isolated from plant cells, including wild-type and transgenic apple calli. After preincubation in the presence and absence of bafilomycin A₁, ATP was added as a substrate to start the reaction. Hydrolytic activity was assayed for 30 min. The reaction was stopped by the addition of trichloroacetic acid. The samples were extracted with chloroform to remove lipids and detergents. After centrifugation, the upper aqueous phase was transferred to clean test tubes and incubated with buffers containing ascorbic acid and ammonium molybdate. The concentration of inorganic phosphate was measured by a spectrophotometer at 700 nm and converted to the rate of ATP hydrolysis. F, Proton-pumping activity of V-ATPase was measured by tracking the ATP-dependent quenching of acridine orange using a fluorescence spectrometer with excitation at 493 nm and emissions at 545 nm. Purified vacuolar membrane vesicles were resuspended in 25 mm Tris (pH 7.2), 25 mm KCl, 5 mm acridine orange, and 5 mm MgCl₂ in the presence and absence of bafilomycin A₁. ATP was added at a final concentration of 5 mm to initiate transport. The proton transport rate was determined as the slope of the quench that gave a linear response during the first 10 s immediately after the addition of ATP. G, Malate content in the wild-type and transgenic apple calli. H, The images show emission intensities of protoplast vacuoles in apple calli loaded with BCECF at 488 nm (the first column, red) and 458 nm (the second column, green). The ratio images indicate an increased or decreased vacuolar pH in transgenic apple calli compared to wild type (the third column). Pseudo-color scale on the right indicates the intensity of the fluorescence, in which yellow and red represent the minimum and maximum intensity, respectively. Scale bar = 10 µm. I, Quantification of the luminal pH in wild-type and transgenic apple calli vacuoles. Bars represent SE of 30 measurements from 10 different intact vacuoles. J, pH buffering capacities in the wild-type and transgenic apple calli. In B to G and I to J, data are shown as the mean ± se, which were analyzed based on more than nine replicates. Statistical significance was determined using Student’s t test in different apple calli lines. n.s., P > 0.01; *P < 0.01; **P < 0.001.

In apples, MdbHLH3 binds to the promoter of MdMYB1 to activate its transcription (Xie et al., 2012). As a result, transgenic apple plants overexpressing MdbHLH3 generated more transcripts of the MdMYB1 gene, which subsequently promoted the expression of MdVHA-B1 and MdVHA-B2 genes, thereby exhibiting a much higher malate content and a lower cell pH than the wild-type control (Supplemental Fig. S5, A–C). In addition, MdMYB9 is involved in anthocyanin biosynthesis (An et al., 2015), while MdoMYB121 is involved in abiotic stress tolerance (Cao et al., 2013). Their overexpression did not influence the expression of MdVHA-Bs genes, suggesting that not all MYB TFs affect the regulation of V-ATPase subunits (Supplemental Fig. S5, D and E).

MdMYB1 is a major regulator of anthocyanin biosynthesis (Takos et al., 2006). Its overexpression noticeably influenced anthocyanin accumulation in transgenic apple calli (Fig. 3, A and D), indicating that MdMYB1 successfully functions in these transgenic calli. Subsequently, the V-ATPase hydrolysis activities, H⁺ transport activities, and malate contents were determined. Before the determination of V-ATPase activity, we detected the purity of vacuolar membranes isolated from wild-type apple plants. The results showed that the tonoplast fraction was highly pure, with strong enrichment of vacuolar markers (Supplemental Fig. S6, A and B), and can be used for the further determination of V-ATPase activity. The results demonstrated that MdMYB1 controlled V-ATPase activity and malate contents, which was partially, if not completely, dependent on MdVHA-B1 and MdVHA-B2 in apple calli (Fig. 3, E–G).

The transport of H⁺ into vacuoles contributes to the acidification of vacuolar compartments and the establishment of pH gradients across tonoplasts (Gaxiola et al., 2007). To assess the effects of V-ATPase activities on vacuolar pH, the rationmetric fluorescent pH indicator 2′,7′-bis-(2-carboxyethyl)-5-(6)-carboxyfluorescein (BCECF) was used to measure vacuolar pH (Swanson and Jones, 1996). The pH values were calculated from fluorescence ratios of confocal images (Fig. 3H) based on an in situ calibration curve (Supplemental Fig. S7). As a result, the average vacuolar pH was 4.81 in wild-type apple calli, but the pH of the MdMYB1 and MdVHA-B1 overexpressors shifted to 4.59 and 4.56, respectively (Fig. 3, H and I). Notably, MdMYB1-S+MdVHAB1-S and MdMYB1-AS+MdVHABs-AS apple calli were significantly shifted to 4.23 and 5.21, respectively (Fig. 3, H and I). Hence, a higher luminal H⁺ concentration in MdMYB1 and MdVHA-B1 overexpressors, as indicated by the lower pH values, further supports that MdMYB1 and MdVHA-B1 are required to maintain tonoplast V-ATPase activity.

Furthermore, the pH buffering capacities of wild-type and transgenic apple calli were also determined. The results showed that MdMYB1 and MdVHA-B1 overexpressors have a higher pH buffering capacities compared with WT apple calli (Fig. 3J), further supporting that MdMYB1 controls vacuolar pH, which is dependent on MdVHA-B1 and MdVHA-B2 in apple calli.

Subsequently, a viral vector-based method was used to verify the functions of MdMYB1 and MdVHA-Bs in regulating coloration and acidity in apple fruits and petals. The results showed that MdMYB1 positively activates MdVHA-B1 and MdVHA-B2, thereby increasing the malate contents and decreasing the cell pH in apple fruits and red-flesh apple petals, just as it did in the apple calli (Fig. 3; Supplemental Fig. S8). In addition, the ectopic expression of MdMYB1 or MdVHA-B1 not only promoted anthocyanin accumulation but also enhanced malate contents in transgenic tobacco flowers (Supplemental Fig. S9).

MdMYB10 Influences the Flesh Acidity and Color in Red-Flesh Apples

In apples, MdMYB1, MdMYB10, and MdMYBA are allelic to each other (Ban et al., 2007; Lin-Wang et al., 2010). Compared to MdMYB1 and MdMYBA, the MdMYB10 locus is dominant and constitutively generates an extremely high level of MdMYB10 transcript and protein, thereby making the tree produce red leaves and red flesh fruits (Espley et al., 2007, 2009; Lin-Wang et al., 2010). To verify whether MdMYB10 regulates MdVHA-B1 and MdVHA-B2, in this study, two heterozygous red-flesh apple genotypes were sexually crossed, generating hybrid seeds.

After germination, the hybrid population was divided into red-leaf and green-leaf seedlings (Supplemental Fig. S10A). Subsequently, 10 red-leaf and 10 green-leaf seedlings were randomly chosen for further investigation. Expression and western blotting analyses demonstrated that the red-leaf seedlings produced higher levels of MdVHA-B1 and MdVHA-B2, but not MdVHA-B3, than the green-leaf ones (Supplemental Fig. S10A). At the same time, the red-leaf seedlings produced more anthocyanins and malate while exhibiting a lower pH value than the green-leaf ones (Supplemental Fig. S10, B–D).

The red-leaf trees produced red-flesh fruits, while the green-leaf trees produced non-red-flesh fruits (Fig. 4A). Five trees per flesh color type were chosen. These trees exhibited similar fruit development and ripening stages. The fruits were taken at 30, 90, and 120 d after bloom (DAB) for characterization. qPCR and western blotting analyses demonstrated that the red-flesh fruits produced more MdMYB10 transcript and protein, and consequently more MdVHA-B1 and MdVHA-B2, but not MdVHA-B3, than the non-red-flesh fruits (Fig. 4, B and C). As a result, the red-flesh fruits exhibited higher V-ATPase activity, accumulated more malate, and showed a lower pH than the nonred ones (Fig. 4, D–G).

Figure 4.

MdMYB10 promotes malate accumulation and cell acidification in red-flesh apples. A, Photos of fruits that were harvested from red- and non-red-flesh hybrid trees. B and C, Transcript level (B) and protein abundance (C) of MdMYB10/MdVHA-Bs in the red-flesh and non-red-flesh fruits at 30, 90, and 120 DAB. Note: In C, protein bands were quantified by scanning densitometry using a Hewlett Packard scanjet scanner and Scanplot software. All protein amounts were normalized based on the protein folds of the 30 DAB samples. D and E, The V-ATPase hydrolytic activity (D) and malate content (E) in these two types of apple fruits. F, The images show emission intensities of protoplast vacuoles in apple fruits loaded with BCECF at 488 nm (first column, red) and 458 nm (second column, green). The ratio images indicate an increased or decreased vacuolar pH in red flesh apple fruits compared with non-red flesh apple fruits (the third column). The pseudo-color scale on the right indicates the intensity of the fluorescence, in which yellow and red represent the minimum and maximum intensity, respectively. Bar = 10 µm. G, Quantification of the luminal pH in the vacuoles of red flesh and non-red flesh apple fruits. Error bars represent SE of 30 measurements from 10 different intact vacuoles. In B, D, E, and G, data are shown as the mean ± se, which were analyzed based on more than five trees bearing red- or non-red-flesh fruits. Statistical significance was determined using Student’s t test in the red and non-red-flesh fruits. n.s., P > 0.01; *P < 0.01; **P < 0.001.

To rule out the possible involvement of segregating factors other than MdMYB10 locus to the greatest extent, another hybridization population from a sexual cross between a red flesh genotype, “Jinshanyilamu,” and a non-red flesh genotype, “Yepingguo,” was used. The results further verified the role of the MdMYB10 locus in regulating V-ATPase activity, anthocyanin, and malate content (Supplemental Fig. S11).

MdMYB10 Modulates Malate Accumulation and Cell pH by Activating More Genes Than MdVHA-Bs

To determine whether MdMYB10 regulates genes other than MdVHA-Bs, total RNAs were extracted from fruit samples taken from five trees per flesh color type at the stage of 90 DAB for an RNA-seq analysis. At least 20,000,000 sequence reads were generated, each 100 to 150 bp in length, encompassing 2.2 Gb of sequence data, which was sufficient to quantitatively analyze the gene expression. A total of 40,271 genes were detected. Among these genes, 3938 were up-regulated and 3725 genes were down-regulated in the red-fleshed compared to the non-red-fleshed apple fruits (Fig. 5A).

Figure 5.

Identification of the target genes of MdMYB1/10. A, Number of up-regulated and down-regulated genes in the red-flesh fruits compared with non-red-flesh fruits with RNA-seq. B to D, The relative expression levels of genes involved in anthocyanin transport and biosynthesis (B), V-ATPase subunit genes (C), and genes involved in malate synthesis and transport (D) in red-flesh and non-red-flesh apple fruits. The fruits were sampled at 90 DAB. E, ChIP-qPCR assays of the enrichments of the target gene promoters in the 35S::MdMYB1-GFP transgenic callus compared to the 35S::GFP transgenic calli. Note: MATE-LIKE1-1 and MATE-LIKE1-2 represent cis-elements 1 and 2 in the promoters of MATE-LIKE1 gene, respectively. In B to E, data are shown as the mean ± se, which were analyzed based on more than nine replicates. Statistical significance was determined using Student’s t test in different apple calli lines. n.s., P > 0.01; *P < 0.01; **P < 0 0.001.

To facilitate the global analysis of the gene expression profiles and evaluate the anthocyanin- and malate-related genes that showed significant transcriptional changes in the above samples, a Gene Ontology (GO)-based classification was conducted by subjecting the sequences to InterPro and GO annotations. These annotations revealed that the absolute expression value (log2Ratio) of MdMYB10 in the red-fleshed compared to the non-red-fleshed fruits was 6.45-fold, based on the FDR < 0.01, while the GO functional classification analysis of V-ATPase, V-PPase, malate synthetic, and transporter genes showed that most of these genes were up-regulated (Supplemental Appendix S2 and Supplemental Table S1). To test the accuracy of the RNA-Seq analysis, quantitative real-time PCR (qPCR) was performed on selected genes, including V-ATPase, V-PPase, malate related-genes (MdcyMDH, MdmMDH, MdcyME, MdPEPCK, and MdPEPC), anthocyanin synthesis-related genes (MdANR, MdANS, MdCHI, MdCHS, MdFLS, and MdUFGT) and vacuolar transporters and channels-related genes (MdMATE-LIKE1, MdMATE-LIKE2, MdMATE-LIKE3, ABC transporter, MdALMT9, and MdtDT) (Supplemental Appendix S3). The transcript level of most of these genes was increased, except for that of the malate channel MdALMT9 gene (Fig. 5, B–D), which confirms the results of the RNA-Seq analysis.

To determine whether MdMYB1/10 directly regulates these genes, ChIP-PCR (in vivo) was performed. MdMYB1 directly bound to the cis-element of MdVHA-E2, MdVHP1, MdMATE-LIKE1, ABC transporter, and MdtDT, in addition to MdVHA-B1 and MdVHA-B2, while other genes were the indirect target genes of MdMYB1/10 (Fig. 5E). Therefore, MdMYB1 mediates apple coloration and acidity by directly and indirectly regulating the expression of anthocyanin- and malate-related genes.

AtPAP1, a Master Regulator for Anthocyanin Biosynthesis, Also Modulates Malate Levels and Cell pH in Arabidopsis

In Arabidopsis, PAP1 is homologous to apple MdMYB1 and MdMYB10 (Allan et al., 2008; Ban et al., 2007; Lin-Wang et al., 2010). Here, two 35S::PAP1-GFP transgenic lines (AtPAP1-GFP-1 and AtPAP1-GFP-2) and one T-DNA activation-tagged gain-of-function mutant, AtPAP1-D, were used to determine whether PAP1 functions similarly to MdMYB1/10 in Arabidopsis. The two transgenic lines and the mutant AtPAP1-D accumulated much more anthocyanin than the wild type, and the empty vector pRI-GFP control (Fig. 6, A–C).

Figure 6.

MYB TF AtPAP1 controls malate accumulation and cell acidification. A, Overexpression of PAP1 and its gain-of-function mutant AtPAP1-d promote anthocyanin accumulation. Col., The wild-type background; pRI-GFP, empty vector; AtPAP1-GFP-1 and AtPAP1-GFP-2, two overexpression lines containing the AtPAP1-GFP fusion gene under the control of the 35S promoter; AtPAP1-d, a gain-of-function mutant of AtPAP1 with enhanced anthocyanin accumulation in plants. B, The relative expression level of AtPAP1 in Col., pRI-GFP, AtPAP1-GFP-1, AtPAP1-GFP-2, and AtPAP1-d Arabidopsis. C, Anthocyanin contents in Col., pRI-GFP, AtPAP1-GFP-1, AtPAP1-GFP-2, and AtPAP1-d Arabidopsis. D, ChIP-qPCR assays of the relative enrichment of the target gene promoters in 35S::AtPAP1-GFP transgenic Arabidopsis compared to 35S::GFP transgenic Arabidopsis. E, Malate content in Col., PRI-GFP, AtPAP1-GFP-1, AtPAP1-GFP-2, and AtPAP1-d Arabidopsis. F, The images show emission intensities of protoplast vacuoles in Arabidopsis loaded with BCECF at 488 nm (first column, red) and 458 nm (second column, green). The ratio images indicate an increased or decreased vacuolar pH in AtPAP1 transgenic Arabidopsis than in Col. and PRI-GFP transgenic Arabidopsis (third column). The pseudo-color scale on the right indicates the intensity of the fluorescence, in which yellow and red represent the minimum and maximum intensity, respectively. Scale bar = 10 µm. G, Quantification of the luminal pH in the vacuoles of Col., PRI-GFP, AtPAP1-GFP-1, AtPAP1-GFP-2, and AtPAP1-d Arabidopsis. Error bars represent SE of 30 measurements from 10 different intact vacuoles. In B, C, D, E, and G, data are shown as the mean ± se, which were analyzed based on more than 20 seedlings. Statistical significance was determined using Student’s t test in different Arabidopsis lines. n.s., P > 0.01; *P < 0.01; **P < 0 0.001.

To determine whether in vivo AtPAP1 binds to the promoters of the AtVHA-B1, AtVHA-B2, AtVHA-B3, AtVHA-D, AtAVP1, AtMATE, AtABCB27, AtALMT9, and AttDT genes, 35S::PAP1-GFP (AtPAP1-GFP-1) and 35S::pRI-GFP (GFP control) transgenic Arabidopsis plants were used for immunoprecipitation with the anti-GFP antibody. The enrichment of the promoters was detected with qPCR assays. The promoter regions of AtVHA-B1, AtVHA-B2, AtAVP1, AtABCB27, AtMATE, and AttDT containing the putative MYB cis-elements were enriched by ChIP in AtPAP1-GFP-1 transgenic Arabidopsis compared to the GFP control, while the promoter regions of the other genes did not (Fig. 6D). These results further provide in vivo evidence for the specific binding of AtPAP1 to the AtVHA-B1, AtVHA-B2, AtAVP1 (homologous to apple MdVHP1), AtABCB27, AtMATE, and AttDT promoters, thereby directly mediating the transcriptional activation of these genes. As a result, the malate content was much higher, while the cell pH was lower in the AtPAP1 overexpression lines and the mutant AtPAP1-D than in the control (Fig. 6, E–G). Therefore, the mechanism through which MYB TF controls cell pH and coloration via the regulation of vacuolar proton pumps and malate- and anthocyanin-related transporters is conserved in different species.

DISCUSSION

MdMYB1 Links Vacuolar Acidification and Anthocyanin Accumulation by Activating Vacuolar Proton Pumps and Secondary Transporters

Acidification of the vacuole by proton pumps (V-ATPase and V-PPase) and vacuolar transport of metabolites by tonoplast transporters are important not only for proper cell function (Martinoia et al., 2007; Kovermann et al., 2007; Gomez et al., 2009) but also for quality formation of fleshy fruits and ornamental crops (Lobit et al., 2006; Shiratake and Martinoia, 2007; Etienne et al., 2013). Therefore, it is of considerable interest to understand how the genes encoding these proton pumps and transporters are transcriptionally modulated by TFs. In this study, we found that apple MdMYB1/10 TF and its Arabidopsis counterpart, AtPAP1, which are master regulators of anthocyanin biosynthesis (Espley et al., 2007; Li et al., 2012; Qiu et al., 2014), directly activate the expression of the genes encoding V-ATPase subunits, V-PPase, the anthocyanin transporter MATE-LIKE1, ABC transporter, and the malate transporter tDT. This clearly indicates that MYB transcriptional factors play a key regulatory role in vacuolar acidification and in the transport of anthocyanins and malate.

The direct interaction between MdMYB1 and the promoters of genes encoding V-ATPase subunits MdVHA-B1 and MdVHA-B2 is supported by both in vitro binding demonstrated by EMSA (Fig. 2, A and C) and in vivo binding in ChIP-PCR assays and GUS assays of the transgenic calli (Fig. 2, D and E). This transcriptional activation of both MdVHA-B1 and MdVHA-B2 leads to increases in V-ATPase activity (Fig. 3, E and F) and elevated cell acidification (Fig. 3, H and I). It has been demonstrated in other plant species that the overexpression of a single subunit of V-ATPase improves the expression of other subunits and ultimately increases overall V-ATPase activity (Keenan Curtis and Kane, 2002; Zhao et al., 2009), and our data are in line with these previous findings (Fig. 3, E and F; Supplemental Fig. S12). The higher V-ATPase activities and corresponding lower pH values detected in the red-flesh fruits relative to nonred fruits (Fig. 4, D, F, and G; Supplemental Fig. S11, D, E, and G) and apple fruits transiently overexpressing MdMYB1 (Supplemental Fig. S8, D and F) confirm that MdMYB1 regulates flesh acidification primarily via the transcriptional activation of V-ATPase. In addition, our work shows that MdMYB1 also binds to the promoter of MdVHP1 to regulate its expression (Fig. 5, D and E). It has been shown that ectopic and transient expression of the apple V-PPase gene MdVHP1 promotes malate and anthocyanin accumulation in tomato fruits and grape berries, respectively (Y.X. Yao and Q.L. Dong, unpublished data). These results suggest that MdVHP1 also contributes to cell acidification in MdMYB1 and MdMYB10 transgenic materials.

Although none of the transporters identified for vacuolar transport of anthocyanins (MATE-LIKE and ABC transporter) directly or indirectly use proton gradients in transporting anthocyanins from the cytosol into the vacuole (Gomez et al., 2009; Francisco et al., 2013), our data indicate that vacuolar acidification caused by the overexpression of subunit MdVHA-B1 alone leads to the accumulation of anthocyanins in apple calli (Fig. 3, D, H, and I), apple fruits and petals (Supplemental Fig. S8, B, F, and H), and tobacco flowers (Supplemental Figure S9, C and E). This increased accumulation of anthocyanins, which primarily occurs in the vacuole, can be partially explained by the elevated expression of MdMATE-LIKE1, as demonstrated in the transgenic calli overexpressing MdVHA-B1 (Supplemental Fig. S13). As anthocyanins must be synthesized first before being transported into the vacuole, these results also suggest a close relationship between vacuolar acidification and anthocyanin synthesis. This relationship is supported by previous findings that mutations of the seven pH loci result in a high petal pH and a shift of the flower color to blue in petunia and morning glory (de Vlaming et al., 1983; Fukada-Tanaka et al., 2000; Ohnishi et al., 2005). In grapevines, VvMYB5a and VvMYB5b, which are pH4 orthologs, are involved in the regulation of vacuolar acidification and partially in the flavonoid pathway (Cavallini et al., 2013). Our work also indicates that MdMYB1/10 directly activates MATE-LIKE and ABC transporters for anthocyanin accumulation in the vacuole (Fig. 5, B and E). Therefore, enhanced accumulation of anthocyanins in red-flesh fruits (Fig. 4A; Supplemental Fig. S11C), red-leaf apple seedlings (Supplemental Fig. S10B), and transgenic calli and fruits overexpressing MdMYB1 (Fig. 3D; Supplemental Fig. S8B) results from both the direct regulation of MdMYB1/10 on MATE-LIKE and ABC transporter and indirect effects on the vacuolar transport of anthocyanins via vacuolar acidification.

In addition, It was found in this study that most of the MdMYB1 binding sites are around or < 1 kb far from the transcription start of those target genes such as MdVHP1 (−799 ∼ −721 bp). In contrast, MdMYB1 binds to a MdVHA-B1 cis-element which is > 2 kb far from the transcription start (Fig. 1A), just as several other TFs do (Maruyama-Nakashita et al., 2005; Behnam et al., 2013). In animal, a few TFs have binding sites that even lie beyond the > 10 kb of promoter sequence (Wang et al., 2007). For those TFs and their binding sites, the specific DNA three-dimensional structure may make the TFs close to the transcription start of target genes.

Molecular Mechanisms and Physiological Roles of the Coregulation of the Vacuolar Transport of Malate and Anthocyanins by MdMYB1

In apples, MdMYB1 is a crucial component of the MdMYB1-MdbHLH3-MdTTG1 complex, which acts as the master regulatory machinery for anthocyanin synthesis in controlling the coloration of apple fruit peels and flesh (Takos et al., 2006; Espley et al., 2007; Ban et al., 2007; An et al., 2012; Xie et al., 2012). As the target genes of MdMYB1 TF, anthocyanin structural genes, such as MdDFR and MdUFGT, contain MYB cis-elements in their promoters. Both MdMYB1 and MdbHLH3/33 physically bind to these cis-elements to activate the expression of these genes (An et al., 2012; Xie et al., 2012). Interestingly, our data show that MYB cis-elements are present in the promoters of many genes encoding vacuolar proton pumps and anthocyanin transporters. Among these genes, MdVHA-B1, MdVHA-B2, MdVHA-E2, MdVHP1, MdMATE-LIKE1, and ABC transporter are the direct target genes of MdMYB1 TF (Fig. 5E). In addition, the overexpression of MdbHLH3 enhances the expression of MdVHA-B1 and MdVHA-B2, leading to lower pH values in apple plants (Supplemental Fig. S5, A and C). However, EMSA showed that MdbHLH3 did not directly bind to the promoters of MdVHA-Bs (data not shown), suggesting that MdbHLH3 enhances the expression of the MdVHA-Bs and vacuolar acidification in an indirect way. Therefore, the apple MdTTG1-MdbHLH3-MdMYB1 complex is an important regulatory machinery that modulates not only anthocyanin synthesis but also the activities of V-ATPase and V-PPase and the vacuolar transport of anthocyanins, demonstrating the multiple functions of MdMYB1 in the acidification and coloration of plant organs, such as fruits and flowers. From a physiological perspective, it makes sense that the synthesis and the vacuolar transport of anthocyanins are regulated by the same machinery to prevent any detrimental effect of anthocyanin accumulation in the cytosol on cell metabolism.

In addition to regulating anthocyanin synthesis and vacuolar transport, our data, including RNA-Seq analysis, showed that MdMYB1 directly binds to the promoter of malate transporter MdtDT to transcriptionally activate its expression in apple calli (Fig. 5, D and E; Supplemental Table S1 and Supplemental Appendix S2). We also found that vacuolar acidification resulting from the overexpression of MdVHA-B1 alone leads to elevated malate levels in apple calli (Fig. 3G) and tobacco flowers (Supplemental Fig. S9D). It is interesting that the overexpression of subunit MdVHA-B1 also increases the expression of MdtDT in apple calli (Supplemental Fig. S13). Therefore, both the direct regulation and indirect regulation (via vacuolar acidification) of vacuolar malate transport by MdMYB1 converge at the transcriptional activation of MdtDT, thereby increasing malate levels in red-flesh apples compared to non-red-flesh apples (Fig. 4E; Supplemental Fig. S11F) and in apple calli overexpressing MdMYB1 (Fig. 3G). Considering that malate and anthocyanins are two different types of metabolites stored in the vacuole, it is tempting to ask what advantage, if any, the coregulation of the vacuolar transport of malate and anthocyanins by MdMYB1 confers on cell function. Based on the fundamental role of malate synthesis, degradation and transport in cell pH control (Kurkdjian and Guern, 1989; Hurth et al., 2005; Martinoia et al., 2007), we propose that this coregulation helps to maintain pH homeostasis in the cytosol of plants where MdMYB1 is expressed. When both V-ATPase and V-PPase are up-regulated by MdMYB1, vacuolar acidification is enhanced (Fig. 3, H and I), and corresponding alkalization of the cytosol is expected. It has been shown that the alkalization of the cytosol up-regulates malate synthesis and its subsequent transport into the vacuole, generating protons to maintain the pH of the cytosol (Gout et al., 1992, 1993). In this scenario, the direct transcriptional activation of MdtDT coupled with indirect up-regulation of MdtDT transcript levels by way of vacuolar acidification/cytosol alkalization would be most effective in transporting newly synthesized malate into the vacuole. The fact that the cells in apple calli overexpressing MdMYB1 or MdVHA-B1 (Fig. 3J) have higher buffering capacity but a relatively lower cellular pH (Fig. 3, H and I) supports this hypothesis. However, the mechanism by which MdtDT expression is up-regulated by vacuolar acidification/cytosol alkalinization is not clear. In Arabidopsis, tDT is up-regulated by both malate feeding (Emmerlich et al., 2003) and the acidification of the cytosol in leaf mesophyll cells (Hurth et al., 2005).

The alleles MdMYB10 and AtPAP1-d constitutively express transcripts throughout the entire apple plant and Arabidopsis, thereby producing high levels of anthocyanins in all of the organs (Espley et al., 2007; Li et al., 2012; Qiu et al., 2014). We demonstrated that these alleles also activate vacuolar proton pumps as well as anthocyanin and malate transporters. As a result, plants with MdMYB10 and AtPAP1-d loci have higher malate content and lower pH values (Fig. 4, E–G; Fig. 6, E–G). In apples, however, the differences in malate content between the red-flesh and non-red-flesh fruits are much less obvious in the ripe fruits at 120 d after flowering than in the young ones at 90 d after flowering (Fig. 4E). Espley et al. (2013) also report that the overexpression of MdMYB10 cDNA has little influence on the malate content in the ripe fruits at 130 DAB in the transgenic Royal Gala apple. The cause for the reduced effect of MdMYB10 on malate levels in ripe fruits is not clear, but is likely related to the fruit ripening process. Because MdALMT9, the most likely candidate gene for Ma in apples (Bai et al., 2012), is not a target gene of MdMYB1 (Fig. 5E), nor is its expression up-regulated by vacuolar acidification in apple calli overexpressing MdVHA-B1 (Supplemental Fig. S13) and red-flesh apples (Fig. 5D), it is possible that MdALMT9 plays a larger role in determining malate levels in ripe fruits. This phenomenon, however, is not restricted to red-flesh fruits in apples. Ectopic expression of MdMDH and MdVHP1 remarkably enhances the malate content in young fruits but has little impact on ripe fruits in transgenic tomatoes (Yao et al., 2011a; 2011b). Obviously, further work is needed to understand the regulatory network for controlling malate levels during the fruit-ripening process.

Agricultural Significance and Potential Application of These Findings

Color is one of the most eye-catching traits, not only for ornamental plants (Obón and Rivera, 2006; Młodzińska, 2009) but also for edible crops. In edible plant organs, such as fleshy fruits, anthocyanin and associated polyphenolics are beneficial to human health, while acidity contributes to taste and flavor. These metabolites largely determine the dietary and market values of these economic crops (Butelli et al., 2008; Sweetman et al., 2009; Etienne et al., 2013). Therefore, these characteristics are major targets of breeding programs for these edible crops. Our work clearly demonstrates that the master regulators for anthocyanin synthesis, MdMYB1/10 and AtPAP1, also transcriptionally activate proton pumps and secondary transporters in modulating cellular pH and vacuolar accumulation of anthocyanins and malate (Fig. 7). For apples, this regulation leads to alterations in fruit color and acidity, two important traits for fruit quality. These findings may be useful in developing novel biotechnological strategies, as well as in informing traditional breeding programs in the creation of new cultivars with improved color, taste, and flavor.

Figure 7.

A model of the MYB TF that transcriptionally activates proton pumps and secondary transporters in modulating pH and the accumulation of malate and anthocyanins in the vacuole of plants.

MATERIAL AND METHODS

Plant Materials and Growth Conditions

The apple calli used in this study were induced from the young embryos of Orin apples (Malus domestica Borkh.). They were grown on MS medium supplemented with 0.5 mg L⁻¹ indole-3-acetic acid (IAA) and 1.5 mg L⁻¹ 6-benzylaminopurine (6-BA) at 25°C in the dark. The calli were subcultured three times at 15-d intervals before being used for genetic transformation and in other assays.

The apple fruits used for injection of viral vectors were collected from mature trees of the cultivar Red Delicious grown in a commercial orchard near Tai-An City. Fruits were bagged at 35 DAB. The bagged fruits were harvested at 140 DAB and were de-bagged before injection.

For sexual crossing, a red-fleshed apple B9, which contains a heterozygous MdMYB10 locus, was used as a pollen donor, while another red-flesh apple, YL, having a heterozygous MdMYB10 locus, was the maternal parent. The resultant hybrid population was comprised of 80 red-flesh and 48 non-red-flesh trees. These trees were grown in a field at an experimental farm. Among them, five red hybrid trees and five green ones produced fruits with similar fruit development and ripening stages. The red-flesh fruits from five red hybrid trees and non-red-flesh fruits from five green trees were harvested at different developmental stages. Red-flesh and non-red-flesh fruits were mixed to eliminate genetic differences and used for further investigation. To further eliminate genetic differences to the greatest extent, another hybrid population from a sexual cross between a red-flesh apple, “Jinshanyilamu,” and a non-red-flesh apple, “Yepinguo,” which contains 30 hybrid trees bearing red-flesh fruits and 30 trees bearing non-red-flesh fruits, was used. The fruits were harvested at 90 DAB. Subsequently, the red-flesh and non-red-flesh fruits of 30 trees were mixed for further investigation.

The Arabidopsis (Arabidopsis thaliana) ecotype Columbia, PAP1 gain-of-function mutant PAP1-D, and two 35S::PAP1-GFP transgenic lines, PAP1-GFP-1 and PAP1-GFP-2, were used. After being treated at 4°C for 3 d, the seeds were sown on MS medium. The plants were grown at 21°C in a 16-h light period (200 µmol m⁻² s⁻¹). Tobacco (Nicotiana tabacum) was cultivated in a growth room at 25°C using natural light with a daylight extension of 14 h.

Genetic Transformation and the Construction of the Expression Vectors

To construct MdMYB1 and MdVHA-B1 sense (S) overexpression and antisense (AS) suppression vectors, the full-length cDNAs of MdMYB1 and MdVHA-B1, a specific fragment of MdMYB1 and a conserved fragment of MdVHA-B1 and MdVHA-B2 were isolated from Gala apples using RT-PCR. All cDNAs were digested with EcoRI/BamHI and cloned into pRI plant transformation vectors downstream of CaMV 35S promoters. All primers used are listed in Supplemental Table S2.

Meanwhile, the full-length cDNAs of Arabidopsis PAP1 and PAP2 were amplified from Arabidopsis ecotype Columbia (Col 0) using RT-PCR. The resulting cDNAs were digested with EcoRI/BamHI and cloned into pRI-GFP plant transformation vectors downstream of CaMV 35S promoters. The primers used are listed in Supplemental Table S2.

For apple calli transformation, the recombinant plasmids, including MdMYB1-S, MdVHAB1-S, MdMYB1-S+MdVHAB1-S, MdMYB1-S+MdVHABs-AS, MdMYB1-AS, MdVHABs-AS and MdMYB1-AS+MdVHABs-AS, were introduced into Orin apple calli using an Agrobacterium-mediated method, as described by Horsch et al. (1985) and Li et al. (2012). The resultant transgenic apple calli were grown on MS medium supplemented with 0.5 mg L⁻¹ indole-3-acetic acid (IAA) and 1.5 mg L⁻¹ 6-benzylaminopurine (6-BA) at 25°C in the dark. Subsequently, the wild type and these seven transgenic apple calli were placed under 25°C plus UVB light conditions for two weeks, before for further investigation.

For tobacco transformation, the recombinant plasmids MdMYB1-S and MdVHAB1-S were introduced into Agrobacterium tumefaciens strain LBA4404. Tobacco was transformed with LBA4404 using a leaf disc method (Horsch et al., 1985).

For Arabidopsis transformation, the PAP1-GFP recombinant plasmid was introduced into wild type (Col 0) via Agrobacterium strain GV3101 using a floral dip method (Clough and Bent, 1998). The seeds of the transgenic plants were individually harvested and subsequently shelfed. Homozygous transgenic lines were used for further investigation.

RNA Extraction, RT-PCR, and qRT-PCR Assays

RNA extraction, as well as RT-PCR and qRT-PCR assays, were performed with the methods described by Hu et al. (2015). For all of the analyses, the signal obtained for a gene of interest was normalized against the signal obtained for the 18S gene (Defilippi et al., 2005). All of the samples were tested in three to four biological replicates. All of the primers used for the semiquantitative RT-PCR and qRT-PCR are listed in Supplemental Tables S2 and S3.

RNA-Seq analysis

Total RNA was extracted from the fruits of the red- or non-red-flesh hybrid trees at 90 DAB. Subsequently, the RNAs for the red and non-red-fleshed fruits were used to construct libraries for high-throughput parallel sequencing using an Illumina genome analyzer II. A rigorous algorithm was used to identify the differentially expressed genes in these samples. A 1% false discovery rate was set to measure the threshold of the P-value in our tests and analyses by manipulating the FDR value (Audic and Claverie, 1997). P < 0.001 and the absolute value of log2Ratio > 1 were used as the threshold to determine the significance of the gene expression differences, according to Audic and Claverie (1997). A GO analysis was used to predict gene function and calculate the functional category distribution frequency.

Chromatin Immunoprecipitation qPCR Analysis

35S::MdMYB1-GFP and 35S::GFP transgenic apple calli were used for the ChIP-qPCR analysis. The anti-GFP antibody (Beyotime) was used for chromatin immunoprecipitation (ChIP), as described by Xie et al. (2012). The resultant samples were used as templates for qPCR. The primers used for ChIP-PCR are listed in Supplemental Table S4.

EMSA

EMSA was conducted according to Xie et al. (2012). MdMYB1 was cloned into the expression vector pGEX4T-1. The MdMYB1-GST recombinant protein was expressed in Escherichia coli strain BL21 and purified using glutathione Sepharose beads (Thermo Scientific). An oligonucleotide probe of the MdMYB1 promoter was labeled using an EMSA probe biotin labeling kit (Beyotime) according to the manufacturer’s instructions. The recombined protein of MdMYB1-GST was incubated with 10× binding buffer, 1 μg/μL poly(dI-dC), and 400 fmol of biotin-labeled double-stranded binding consensus oligonucleotides (total volume 20 μL) using a LightShift Chemiluminescent EMSA Kit (Thermo Scientific). The binding reaction was performed at room temperature for 20 min. The DNA-protein complexes were separated on 6.5% nondenaturing polyacrylamide gels, electrotransferred, and detected following the manufacturer’s instructions. The binding specificity was also examined by competition with a fold excess of unlabeled oligonucleotides. The primers used for EMSA are listed in Supplemental Table S5.

Transient Expression Assays

Transient expression assays were conducted using apple calli. The promoters of MdVHA-B1, MdVHA-B2, and MdVHA-B3 were cloned into PBI121-GUS to fuse with the reporter gene GUS. The resultant recombinant plasmids P_MdVHAB1::GUS, P_MdVHAB2::GUS, and P_MdVHAB3::GUS were genetically introduced into Orin apple calli via an agrobacterium-mediated method. Subsequently, 35S::MdMYB1 transformant was cotransformed into the above-mentioned transgenic calli. Finally, histochemical staining was performed to detect GUS activity in the transgenic calli, as described by Xie et al. (2012).

Enzyme Extraction and V-ATPase Activity Assays

The isolation of the tonoplast membranes assays were performed as described by Terrier et al. (2001). Bafilomycin A₁-sensitive ATP hydrolysis of vacuolar membranes was assayed by measuring the production of inorganic phosphate, as described by Lu et al. (2002). Meanwhile, proton transport activity of V-ATPase was measured by ATP-dependent fluorescent quenching of the pH-sensitive fluorescent probe acridine orange (Lu et al., 2002).

Construction of Transient Expression Vectors in Apple Fruits

To construct the antisense expression viral vectors, the MdMYB1 cDNA fragment and conserved MdVHA-Bs fragment were amplified with RT-PCR using apple fruit cDNA as the template. The PCR products were cloned into the tobacco rattle virus (TRV) vector in the antisense orientation under the control of the dual 35S promoter. The resultant vectors were named TRV-MdMYB1 and TRV-MdVHABs. Subsequently, they were introduced into Agrobacterium tumefaciens strain GV3101 by electrical shock. The transformants were then used for apple fruit infiltrations.

To generate the overexpression viral vectors, full-length cDNAs of MdMYB1 and MdVHA-B1 were inserted into the IL-60 vector under the control of the 35S promoter. The resultant vector plasmids were named MdMYB1-IL and MdVHAB1-IL, and then their single or recombined plasmids were used for apple fruit infiltrations. Fruit infiltrations were performed as described by Li et al. (2012).

DNA-Affinity Trapping of DNA-Binding Proteins

DNA promoter fragments of the MdVHA-B1 containing MYB cis-elements were used to isolate the proteins binding to the cis-elements in these promoter fragments. The biotinylated DNA promoter fragments were generated by PCR using the primers that are listed in Supplemental Table S4. Nuclear protein extracts for EMSA were prepared from Red Delicious apple plants that were grown under natural conditions. The biotinylated DNA promoter fragment was immobilized on Dynabeads M-280 streptavidin (Invitrogen) according to the manufacturer’s instructions, in which 2 mg of beads were immobilized using 2× binding buffer (10 mm Tris-HCl, pH 7.5, 1 mm EDTA, and 2 m NaCl). The binding of the protein to the DNA was performed as described by Gabrielsen et al. (1989), with some modifications. A 15-min incubation at 25°C was performed after the beads were resuspended in protein binding buffer (20 mm Tris-HCl, pH 8.0, 1 mm EDTA, 10% (v/v) glycerol, 100 mm NaCl, 0.05% (v/v) Triton X-100, and 1 mm DTT) and mixed with apple nuclear protein extracts. The Dynabeads were washed three times with protein-binding buffer before the proteins were eluted in elution buffer (20 mm Tris-HCl, pH 8.0, 1 mm EDTA, 10% (v/v) glycerol, 1 m NaCl, 0.05% (v/v) Triton X -100, and 1 mm DTT). The protein digestion, mass spectrometry, and data analysis were performed as described by Shaikhali et al. (2012).

Measurement of Vacuolar pH

Vacuolar pH of apple calli protoplasts was monitored with the cell-permeant and pH-sensitive fluorescent dye BCECF-AM (Tang et al., 2012). Vacuolar pH value was quantified by a ratio analysis of the pH-dependent (488 nm) and pH-independent (458 nm) excitation wavelengths from a calibration curve (Supplemental Fig. S7), and ratio images were produced using the ion concentration tool of Zeiss LSM confocal software.

pH Buffering Capacity

The pH buffering capacity of the experimental material was measured using a pH electrode (Rex Instruments; E-201-C) and an ORP electrode (ORP-412, Cany Precision Instruments), as described by Florou-Paneri et al. (2001). A 10-g sample was placed in a beaker, and 100 mL of distilled water was added. The mixture was stirred for approximately 30 min and was subsequently titrated with 0.1 m NaOH, with continuous stirring, to decrease the pH by one unit. The amount of NaOH consumption was used to express the pH buffering capacity of the samples.

Determination of the Total Anthocyanin Content

Total anthocyanins were extracted using a methanol-HCl method and detected as described by Lee and Wicker (1991).

Determination of the Malate Content

Malate content was measured by HPLC, as described by Hu et al. (2015).

Statistical Analysis

Samples were analyzed in triplicate, and the data were expressed as the mean ± sd unless noted otherwise. Statistical significance was determined using Student’s t test. A difference with P ≤ 0.01 was considered significant, and a difference with P ≤ 0.001 was extremely significant.

Supplemental Data

• Supplemental Figure S1. Analysis of the three deduced isoforms of apple V-ATPase subunit B.

• Supplemental Figure S2. The recombinant MdMYB1 does not bind to the promoter region of the MdVHA-B3 gene in the electrophoretic mobility shift assays (EMSAs).

• Supplemental Figure S3. MdMYB1 activates the MdVHA-B1 and MdVHA-B2 but not MdVHA-B3 promoters as detected by GUS assays.

• Supplemental Figure S4. Analysis of protein abundance of MdMYB1, MdVHA-B1, MdVHA-B2, and MdVHA-B3 in transgenic apple calli lines with western blotting.

• Supplemental Figure S5. MdbHLH3 activates MdMYB1, MdVHA-B1, and MdVHA-B2 to increase malate content in transgenic apple plants.

• Supplemental Figure S6. Characterization of the purity of protoplasts, vacuoles, and tonoplast vesicles isolated from WT apple plants.

• Supplemental Figure S7. In situ calibration of BCECF-AM in protoplast vacuoles.

• Supplemental Figure S8. Transient expression of MdMYB1 and MdVHA-B1 via the viral vector-based transformation alters the coloration and acidity of apple fruits and petals.

• Supplemental Figure S9. The ectopic expression of MdMYB1 and MdVHA-B1 promotes acidity and coloration in transgenic tobacco flowers.

• Supplemental Figure S10. The phenotype differences between the red-leaf and green-leaf apple hybrids in a sexually crossed population (YL×B9).

• Supplemental Figure S11. The phenotype differences between red-flesh and non-red-flesh genotypes in a hybridization population from ‘Jinshanyilamu X ‘Yepingguo’.

• Supplemental Figure S12. The relative expression level of the V-ATPase subunit genes in MdVHAB1-Sac and WT apple calli (relative transcripts = MdVHAB1-Sac transcripts/WT transcripts).

• Supplemental Figure S13. The relative expression levels of MdtDT, MdALMT9, MdMATE-LIKE1, MdMATE-LIKE2, MdMATE-LIKE3, and ABC transporter in wild-type (WT) and MdVHA-B1 transgenic apple calli.

• Supplemental Table S1. The transcriptional changes of the anthocyanin- and malate-related genes in red-flesh versus non-red-flesh apple fruits with RNA-seq.

• Supplemental Table S2. List of primers used for real-time PCR.

• Supplemental Table S3. List of primers used for real-time quantitative PCR.

• Supplemental Table S4. List of primers used for ChIP-PCR.

• Supplemental Table S5. List of primers used for EMSA.

• Supplemental Appendix 1. The identified proteins that bind to MYB cis-element of MdVHA-B1 promoter in LC/MS data.

• Supplemental Appendix 2. The relative expression folds of the detected genes in red-flesh versus non-red-flesh apple fruits with RNA-seq assay.

• Supplemental Appendix 3. Amino acid sequence alignment of putative MATE and ABC transporters of Arabidopsis and apple.

ACKNOWLEDGMENTS

We thank Prof. Ilan Sela of Hebrew University of Jerusalem, Israel, for IL-60-BS binary vectors and Prof. Takaya Moriguchi of National Institute of Fruit Tree Science, Japan, for Orin apple calli.

Glossary

 
• V-ATPase

  vacuolar H⁺-ATPase

 
• V-PPase

  vacuolar H⁺-pyrophosphatase

 
• TF

  transcription factor

 
• EMSA

  electrophoretic mobility shift assay

 
• BCECF

  2′,7′-bis-(2-carboxyethyl)-5-(6)-carboxyfluorescein

 
• DAB

  days after bloom

 
• GO

  Gene Ontology

Author notes

1

This work was supported by grants from Ministry of Science and Technology of China (2011AA100204), NSFC (31272142, 31325024, 31471854), Ministry of Education of China (IRT15R42), and Shandong Province (SDAIT-03-022-03).

*

Address correspondence to [email protected].

The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (www.plantphysiol.org) is: Yu-Jin Hao ([email protected]).

Y.-J.H. and D.-G.H. conceived and designed the experiments; D.-G.H., C.-H.S., Q.-J.M., and C.-X.Y. performed the experiments; D.-G.H., Y.-J.H., and L.C. wrote the article.

[OPEN]

Articles can be viewed without a subscription.

www.plantphysiol.org/cgi/doi/10.1104/pp.15.01333