Effect of Grapevine Rootstocks on Anthocyanin Biosynthesis, Sugar Contents, and Endogenous Hormone Contents During the Berry Maturation of ‘Ruby Roman’

Abstract

This study investigated anthocyanin accumulation, sugar contents, and endogenous hormone contents in the berry skin, as well as the expression of genes related to anthocyanin and abscisic acid (ABA) synthesis and metabolism, using grafted ‘Ruby Roman’ berries on the rootstocks of ‘Kober 5BB’ [5BB(2x), a semidwarf rootstock], ‘Hybrid Franc’ [HF(2x), a vigorous rootstock], and their colchicine-induced autotetraploids [5BB(4x) and HF(4x)]. Rootstock had significant effects on the total content, but not on the composition, of anthocyanins. The berries on 5BB(4x) rootstock, where the grapevine showed less vegetative growth, had higher anthocyanin content during the ripening process, and also had higher sugar and ABA contents around véraison. ABA, indole acetic acid (IAA), and cytokinins showed synchronous changes during berry development: they had the lowest levels at pre-véraison, and their metabolic pathways were accelerated after véraison. Furthermore, they all tended to be higher on 5BB(4x) than on the other rootstocks. Since the expression levels of most of the ABA biosynthesis-related genes did not show a corresponding increase with the contents of ABA and ABA-glucosyl ester (ABA-GE), it is considered that the increase in ABA content after véraison may be mainly due to the decrease in catabolism and/or exogenous import from other organs. This study provides an overview showing the dynamic changes and relationship of three phytohormones during the ripening of grape berries grafted on different rootstocks, and explores the mechanisms regulating ripening.

Introduction

Table grapes account for approximately 80% of the total grape yield in Japan, and almost half of the table grapes are large-berry tetraploid cultivars. ‘Ruby Roman’ is a red table cultivar that is selectively bred from seedlings of ‘Fujiminori’ (Vitis vinifera × V. labruscana), a tetraploid black grape. Most tetraploid cultivars, including ‘Ruby Roman’, are interspecific hybrids descended from V. vinifera and V. labruscana. Owing to their low tolerance to grape phylloxera (Daktulosphaira vitifoliae), one of the most destructive grape pests worldwide, tetraploid grapes are generally grafted on resistant rootstocks in practical viticulture. ‘Kober 5BB’ (5BB, V. berlandieri × V. riparia) is the most widely used resistant rootstock in viticulture in Japan. In addition to pest resistance, the 5BB rootstock provides the benefits of high berry quality, early ripening, and wide soil adaptability (Uehara, 1995). Considering that tetraploid grapevines usually show more vigorous cane growth, possibly leading to poor berry coloration and decreased berry quality when they are grafted on a diploid rootstock (Motosugi et al., 2007), some colchicine-induced tetraploid rootstocks have been developed and tested (Motosugi et al., 2002). These tests demonstrated that scions (‘Kyoho’) grafted on the autotetraploid rootstocks had more compact vines and much deeper skin coloration than those grafted on the original diploid rootstocks (Motosugi et al., 2007). In a previous study, we grafted ‘Ruby Roman’ on two rootstock cultivars [5BB and ‘Hybrid Franc’ (HF), a typical vigorous rootstock] and their corresponding tetraploids and investigated the growth of scion vines and roots, as well as the quality of berries (Gao-Takai et al., 2017). ‘Ruby Roman’ grapevines grafted on the autotetraploid of the 5BB rootstock [5BB(4x)] consistently exhibited the lowest vegetative growth, including shoot and root growth, during the four years of investigation. Berries grafted on the 5BB(4x) rootstock had higher sugar content (Brix) and deeper skin coloration than those grafted on the other rootstocks. To better understand the effect of rootstock on berry ripening, this study investigated some physiological changes and analyzed the molecular mechanisms underlying the regulation of berry development.

Grapes, a representative nonclimacteric fruit, have been used to study the mechanism of fruit ripening. Researchers have been looking for a switch that controls ripening initiation in nonclimacteric fruits, similar to the role of ethylene in climacteric fruits. Abscisic acid (ABA) is considered to be a key factor involved in the initiation of the fruit ripening process in grapes because its content increases at the onset of ripening (Pilati et al., 2017; Sun et al., 2010; Wheeler et al., 2009). Simultaneously, ABA is considered to be a promoter of fruit coloring during the fruit ripening process. Some studies reported that ABA can upregulate the expression of MYBA genes encoding myeloblastosis transcription factors (Jeong et al., 2004), which regulate the expression of structural genes in the anthocyanin biosynthesis pathway, including uridine diphosphate-D-glucose: flavonoid 3-O-glucosyltransferase (UFGT), flavonoid 3'-hydroxylase (F3'H), flavonoid 3',5'-hydroxylase (F3'5'H), and flavonoid O-methyltransferase (FAOMT). However, anthocyanin accumulation does not always respond to ABA. For example, Azuma et al. (2012) found that the expression patterns of VlMYBA (V. labrusca myeloblastosis transcription factors) did not resemble the patterns of endogenous ABA contents. In addition, in a previous study, we observed that under elevated temperature conditions, anthocyanin accumulation was inhibited, but ABA content increased (Gao-Takai et al., 2019). It is well documented that the anthocyanin content is affected by the environment and other factors, including rootstock effects. In this study, in order to investigate the effects of rootstock on the expression of anthocyanin-related genes, as well as the mechanism of ABA regulation and the interaction between ABA and other hormones such as auxin and cytokinins, we analyzed and discussed the contents and composition of anthocyanins and sugars, dynamic changes in some endogenous hormones, and expressions of some related genes in ‘Ruby Roman’ berries grafted on four types of rootstock.

Materials and Methods

Plant materials and cultivation management

‘Ruby Roman’ scions were grafted on four types of rootstock: ‘Kober 5BB’ [V. berlandieri × V. riparia, 5BB(2x), a semidwarf rootstock], ‘Hybrid Franc’ [V. rupestris × V. vinifera, HF(2x), a vigorous rootstock], and their colchicine-induced autotetraploids [5BB(4x) and HF(4x)]. Nine grafted vines for each type of rootstock were planted in a vinyl greenhouse located at Ishikawa Prefectural University, Japan (36°30' N: 136°35' E, 38 m). The cultivation management was described in our previous report (Gao-Takai et al., 2017), which investigated the vegetative growth over a period of four years and berry quality in two consecutive seasons. In 2014 and 2015, six new shoots that budded and grew from the spurs (spur-pruned, leaving only two nodes in winter) of each grapevine were trained on a 1.8-m-high trellis and top-pruned once they reached 2 m during the growing season. The clusters and berries were sequentially thinned to finally have two clusters on each grapevine, with 25–30 berries on each cluster. The flower clusters were exposed to gibberellic acid (GA₃) to induce seedlessness. The full bloom date of most clusters was May 25 in 2014 and May 19 in 2015, and the véraison (the day half the clusters became soft) for all types of berries was on July 10 [46 days after full bloom (DAFB)] in 2014 and July 6 (48 DAFB) in 2015.

Analysis of anthocyanin and sugar compositions in berry skin

The total anthocyanin content and total soluble solid content (Brix) in the berry skins grafted on the four types of rootstock were measured in our previous study (Gao-Takai et al., 2017). In the present study, the compositions of anthocyanin and sugar were analyzed using high-performance liquid chromatography (HPLC).

To analyze anthocyanin composition, 15 berries were randomly collected from each type of grapevine on July 24 (60 DAFB), August 5 (72 DAFB), August 15 (82 DAFB), and August 25 (92 DAFB), 2014. Then, they were divided into three biological replicates. The skin of one disk (Φ = 8 mm) was collected from the apex of each berry using a cork borer (five disks per sample, n = 3), and anthocyanins were extracted according to the method described by Gao-Takai et al. (2017). HPLC analysis for anthocyanins was the same as previously described (Gao-Takai et al., 2019). All of the detected peaks were identified by comparing their retention times and order of elution and grouped into cyanidin (Cy), peonidin (Pn), delphinidin (Dp), petunidin (Pt), and malvidin (Mv). The contents of five anthocyanin groups were determined based on the standards of cyanidin 3-glucoside chloride (Cy3G), peonidin 3-glucoside chloride (Pn3G), delphinidin 3-glucoside chloride (Dp3G), petunidin 3-glucoside chloride (Pt3G), and malvidin 3-glucoside chloride (Mv3G) (TOKIWA Phytochemical CO., Ltd., Sakura, Japan), respectively.

For sugar analysis, 15 berries in 2014 and 30 berries in 2015 were collected from nine vines for each type of rootstock at each stage, and were randomly split into three (2014, n = 3) or five (2015, n = 5) biological replicates (5 berries each in 2014 and 6 berries each in 2015). The skins were peeled, lyophilized, and ground into a fine powder using a Multi Beads Shocker (Yasui Kikai Corp., Osaka, Japan). The extraction and detection of fructose, glucose and sucrose were conducted as in our other study (Katayama-Ikegami et al., 2022). An HPLC system equipped with an Asahipak NH2P-50 4E (4.6 mm × 250 mm) column (Showa Denko K. K., Tokyo, Japan), and a charged aerosol detector (Corona Veo; Thermo Fisher Scientific Inc., MA, USA) was used.

Quantification of endogenous plant hormone contents

The lyophilized skin powder samples described above were used to analyze endogenous plant hormones. Endogenous plant hormones including ABA and ABA-derived compounds [dihydrophaseic acid (DPA), phaseic acid (PA), neophaseic acid (neoPA), 7'-hydroxy ABA (7'-OH-ABA), an inactive conjugate of ABA: ABA-glucosyl ester (ABA-GE)], indole-3-acetic acid (IAA) and IAA conjugates [IAA-aspartate (IAA-Asp), IAA-alanine (IAA-Ala), and IAA methyl ester (IAA-ME)], cytokinins [trans-zeatin (tZ) and isopentenyl adenine (iP)] and their sugar conjugates [tZ-riboside (tZR) and iP-riboside (iPR)] in the berry skin were analyzed using a liquid chromatography–triple quadrupole mass spectrometry system (LC: Waters 2695 Separations Module, MS/MS: Waters micromass Quattro micro API Mass Spectrometer; Waters Corp., MA, USA), as described by Gao-Takai et al. (2018, 2019).

Expression analysis of anthocyanin biosynthesis-related genes and ABA metabolism-related genes

Gene expression analysis was conducted using the frozen skin samples of 46 and 60 DAFB berries in 2014. Total RNA extraction and cDNA synthesis were the same as described previously (Gao-Takai et al., 2019). Quantitative PCR was carried out with the StepOnePlus Real-Time PCR System (Applied Biosystems, USA) and SYBR Pre-mix Ex-Taq II (Takara Bio, Japan). The amplification condition was the same as described by Katayama-Ikegami et al. (2016a, b), and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as the internal reference gene. Anthocyanin biosynthesis-related genes, including two genes encoding MYB-related transcription factors, VlMybA1-2 and VlMybA1-3, and four structural genes from V. vinifera (VvUFGT, VvF3'H, VvF3'5'H2, and VvFAOMT), as well as ABA metabolism-related genes, including five genes from the carotenoid cleavage dioxygenase family (CCDs) {VvNCED3 [VIT_19s0093g00550, named VvNCED1 by Sun et al. (2010), also called VvNCED3 by Young et al. (2012)], VvNCED2 (VIT_10s0003g03750), VvNCED4, VvCCD4b, and VvNCED6; some genes involved in ABA biosynthesis: V. vinifera 9-cis-epoxycarotenoid dioxygenase (VvNCED)}, one gene involved in ABA degradation (V. vinifera ABA 8'-hydroxylase, VvCYP707A1), two genes involved in conjugation and reactivation (V. vinifera ABA-glucosyltransferase, VvABAGT and V. vinifera β-glucosidase, VvABABG1), and one responsive gene (V. vinifera ABA-responsive element binding factor2, VvABF2) were analyzed. The primer sequences and the accession number were as our previous study (Gao-Takai et al., 2019; Katayama-Ikegami et al., 2016a, b; Supplementary Table S1), except for three additional genes possibly related to ABA biosynthesis as follows: VvNCED4 (VvCCD4a, VIT_02s0087g00910; XM_002268368, primers used in qRT-PCR for fwd: TCACTGCCCCAAATTCCTTC; for rev: GGTGGTGGAGGTTTGTGGAC), VvCCD4b (VIT_02s0087g00930; XM_002270125, primers for fwd: AAGCCCCTCGTCTCTCCTTC; for rev: TGGGAAGTTGTAGGGGTTGC), and VvNCED6 (VIT_05s0051g00670; XM_002283149, primers for fwd: CCCACTCAGCCTTTGCACTT; for rev: TGGAGGAGAAGCCAGGAGAA).

Statistical analysis

Significant differences among the four rootstocks were determined using Tukey’s test with IBM^® SSPS^® statistics software version 25 (n = 3 in 2014, n = 5 in 2015).

Results

Compositions and contents of anthocyanins and sugars

Seventeen anthocyanin compounds were detected and grouped into Cy (Cy3, 5G, Cy3G, Cy3pG5G, and Cy3pG), Pn (Pn3G, Pn3pG5G, and Pn3pG), Dp (Dp3G, Dp3pG5G, and Dp3pG), Pt (Pt3G, Pt3pG5G, and Pt3pG), and Mv (Mv3G, Mv3aG, Mv3pG5G, and Mv3pG), consistent with the results of our previous study (Gao-Takai et al., 2019). Dihydroxylated anthocyanins (Cy and Pn) were the dominant components at all developmental stages, such that sum the of the two groups accounted for nearly 80% of the total anthocyanin content in most cases, whereas trihydroxylated anthocyanins (Dp, Pt, and Mv) accounted for only the remaining 20% of the content (Fig. 1). The ratios of dihydroxylated anthocyanins to trihydroxylated anthocyanins (the Di/Tri ratio), as well as the ratio of nonmethylated anthocyanins (Cy and Dp) to methylated anthocyanins (Pn, Pt, and Mv) (the NM/M ratio), were generally invariant among the different rootstocks. The contents of all anthocyanin groups in the berry skin were always higher on 5BB(4x) rootstock than on the other rootstocks.

Fig. 1

Effects of rootstocks on the contents and composition ratios of anthocyanin groups in the berry skin of ‘Ruby Roman’ grafted on diploid and autotetraploid of ‘Kober 5BB’ (5BB) and ‘Hybrid Franc’ (HF) rootstocks [5BB(2x), 5BB(4x), HF(2x), and HF(4x)]. The bar graphs on the left show the contents of cyanidin (Cy), peonidin (Pn), delphinidin (Dp), petunidin (Pt), and malvidin (Mv) at 60, 72, 82, and 92 days after full-bloom (DAFB). The graphs on the right show the ratios for each group of anthocyanins. Different letters indicate significant differences among the four rootstock types for each group of anthocyanins followed by Tukey’s test (P < 0.05, n = 3). Error bars represent SE of the mean.

Changes in the contents of fructose, glucose and sucrose during berry maturation are shown in Figure 2. Fructose and glucose showed similar increase patterns, and their contents were higher than sucrose. The contents of the three sugars tended to be higher in the berries on the two 5BB rootstocks [especially on 5BB(4x) rootstock] than on the two HF rootstocks around véraison, although there were no differences among the four types of rootstock near harvest.

Fig. 2

Changes in the contents of fructose, glucose, and sucrose in the berry skin of ‘Ruby Roman’ grafted on 5BB(2x), 5BB(4x), HF(2x), and HF(4x) rootstocks. Different letters indicate significant differences among the four types of rootstock at each stage followed by Tukey’s test (P < 0.05, n = 3 in 2014, n = 5 in 2015). Arrows indicate the date of véraison.

Endogenous levels of ABA, auxin and cytokinins metabolites

Endogenous ABA contents in all types of rootstock decreased from the early fruiting stage to pre-véraison (around 40 DAFB) and then increased until around 70 DAFB (Fig. 3A, B). The berries on 5BB(4x) rootstock had the highest ABA content around véraison (from 40 to 50 DAFB in 2014 and from 48 to 52 DAFB in 2015). The ABA content on the two 5BB rootstocks was higher than that on the two HF rootstocks at 60 DAFB in 2014 and at 70 DAFB in 2015. The berries on HF(4x) had lower ABA content than others at 72 DAFB in 2014 and 59 DAFB in 2015; however, only in the berries on HF(4x), the ABA content continued to increase until 80 DAFB in both years. After 80 DAFB, there were no differences in ABA content among the berries on the four types of rootstock.

Fig. 3

Changes in the contents of ABA (A, B), ABA-glucosyl ester (ABA-GE) (C, D), and total ABA metabolites (E, F) in berry skin of ‘Ruby Roman’ grafted on 5BB(2x), 5BB(4x), HF(2x), and HF(4x) rootstocks. Different letters indicate significant differences among the four types of rootstock at each stage followed by Tukey’s test (P < 0.05, n = 3 in 2014, n = 5 in 2015). Arrows indicate the date of véraison.

The ABA-GE content decreased from the early stage (12 DAFB in 2014 and 20 DAFB in 2015) to 30 DAFB, slowly increased until around 50 DAFB, and then rapidly increased until the harvest stage. Similar to ABA, ABA-GE also tended to have a higher level in the berries on 5BB(4x) rootstock (Fig. 3C, D). The ABA-GE content in the berries on 5BB(2x) was lower than that on other rootstocks after 72 DAFB in 2014.

The contents of the four ABA catabolites, DPA, PA, neoPA, and 7'-OH-ABA, and the total ABA metabolites content are shown in Figure 3E and F. DPA, an ABA catabolite produced via ABA 8'-hydroxylation, showed extremely high levels at the early fruiting stage (e.g., 8897 ng·g⁻¹ DW at 12 DAFB in 2014 in the berries on 5BB(4x) rootstock, which was almost 10 times as much as the ABA content at the same stage). Then, it rapidly decreased and was maintained at a very low level from 40 DAFB until harvest. The rate of decrease in DPA content in the berries on two types of HF rootstocks seemed to be slower than that on the two 5BB rootstocks before véraison in 2015. The other three ABA catabolites, PA, a precursor of DPA, neoPA, an ABA catabolite produced via ABA 9'-hydroxylation and 7'-OH-ABA were also found at very low levels during the entire developmental process. For all rootstocks, the total contents of ABA metabolites in berry skin were the lowest around véraison (40 DAFB in 2014 and 48 DAFB in 2015). The total content of ABA metabolites in the berries on 5BB(4x) tended to be higher than those on the other rootstocks, particularly around véraison.

The content of free IAA in the early fruiting stage was relatively higher and decreased rapidly until 30 DAFB and then remained at a very low level. There were no significant differences in IAA content among the four types of rootstock around véraison (Fig. 4A, B). IAA-Asp content increased rapidly with berry development from 50 to 80 DAFB on all types of rootstock, and the rate of increase was higher on 5BB(4x) (Fig. 4C, D) than on the other rootstocks. The total contents of the four IAA metabolites were low at pre-véraison (30–46 DAFB in 2014 and 43 DAFB in 2015) on all types of rootstock and started to increase after véraison, mainly driven by IAA-Asp. The total IAA metabolites content also tended to be higher in the berries on 5BB(4x) than on the other rootstocks after véraison until 70 DAFB, with the exception of 52 DAFB in 2015 (Fig. 4E, F).

Fig. 4

Changes in the contents of IAA (A, B), IAA-aspartate (IAA-Asp) (C, D), and total IAA metabolites (E, F) in berry skin of ‘Ruby Roman’ grafted on 5BB(2x), 5BB(4x), HF(2x), and HF(4x) rootstocks. Different letters indicate significant differences among the four types of rootstock at each stage followed by Tukey’s test (P < 0.05, n = 3 in 2014, n = 5 in 2015). Arrows indicate the date of véraison.

The tZ (Fig. 5A, B) and iP (Fig. 5C, D) contents were low (several nanograms per gram) compared with ABA (hundreds to thousands of nanograms per gram) and even IAA (several tens of nanograms per gram). Because there were fewer replicates of samples in 2014 (n = 3), the data showed a larger deviation compared with the data for 2015. The contents of two cytokinins, particularly iP, increased with berry development from around 40 DAFB on all types of rootstock. iP content reached its highest value at 82 DAFB in 2014 and at 52 DAFB in 2015. tZ and iP contents in the berries on 5BB(4x) tended to be higher than those on the other rootstocks at véraison in 2015. Total contents of cytokinin metabolites in all types of rootstock were the lowest at 30 DAFB in both years, and they tended to be higher on 5BB(4x) than on the other rootstocks around véraison (Fig. 5E, F).

Fig. 5

Changes in the contents of cytokinins [trans-zeatin (tZ) (A, B) and isopentenyl adenine (iP) (C, D)], and total cytokinin metabolites including tZ, iP, and their sugar conjugates (tZR, iPR) (E, F) in the berry skin of ‘Ruby Roman’ grafted on 5BB(2x), 5BB(4x), HF(2x), and HF(4x) rootstocks. Different letters indicate significant differences among the four types of rootstock at each stage followed by Tukey’s test (P < 0.05, n = 3 in 2014, n = 5 in 2015). Arrows indicate the date of véraison.

Expression of anthocyanin biosynthesis-related genes

The results of gene expression analysis for the anthocyanin biosynthesis-related genes are shown in Figure 6. The expression levels of VlMybA1-2 and VlMybA1-3 in the berry skin on 5BB(4x) were higher than those on the other three rootstocks, although the difference in expression level of VlMybA1-2 at 46 DAFB was not significant. VlMybA1-2 and VlMybA1-3 had similar expression levels among 5BB(2x), HF(2x), and HF(4x) rootstocks. The expression levels of VvUFGT, which encodes a key enzyme for attaching a glucosyl moiety to anthocyanidin aglycones to produce anthocyanin (Ford et al., 1998), were significantly higher on 5BB(4x) than on the two HF rootstocks at both 46 and 60 DAFB and in the berries on 5BB(2x) at 60 DAFB. Moreover, the expression level of this gene in the berries on 5BB(2x) was also higher than that on the two HF rootstocks at 60 DAFB. VvFAOMT, encoding a 3',5'-O-methyltransferase that methylates anthocyanins to produce methylated anthocyanins (Pn, Pt, and Mv) (Hugueney et al., 2009), was found to have the same expression pattern as VvUFGT, with the berries on 5BB(4x) showing the highest expression level, and was higher on the two 5BB rootstocks than on the two HF rootstocks. The expression level of VvF3'5'H2, encoding an enzyme that catalyzes 3' and 5' positions of the B-ring to produce tri-hydroxylated anthocyanidins such as Dp, Pt, and Mv (Bogs et al., 2006; Castellarin et al., 2006), was significantly higher on 5BB(4x) than on the other rootstocks. For all these genes, there was no difference in the expression levels between HF(2x) and HF(4x) rootstocks. The expression level of VvF3'H, encoding a hydroxylase which catalyzes to produce dihydroxylated anthocyanidins (Cy and Pn) (Bogs et al., 2006; Castellarin et al., 2006), did not differ among the four types of rootstock at both 46 and 60 DAFB.

Fig. 6

Comparison of gene expression involved in anthocyanin biosynthesis in berry skin of ‘Ruby Roman’ grafted on 5BB(2x), 5BB(4x), HF(2x), and HF(4x) rootstocks. Expression levels were normalized to those of glyceraldehyde 3-phosphate dehydrogenase (GAPDH). Different letters indicate significant differences among the four types of rootstock at each stage followed by Tukey’s test (P < 0.05, n = 3).

Expression of ABA metabolism-related genes

The results of gene expression analysis for the ABA metabolism-related genes are shown in Figure 7. The expression of VvNCED3 (VvNCED1) tended to be at a higher level in the berries on the two 5BB [particularly on 5BB(4x) rootstock] than on the two HF rootstocks at 46 DAFB (véraison), but there was no difference at 60 DAFB among the different rootstocks. The expression of VvNCED2 tended to be higher on the two HF rootstocks than on 5BB rootstocks at 60 DAFB. A higher expression level of VvNCED6 was observed in the berries on the two 5BB than on the two HF rootstocks at 60 DAFB. There were no significant differences in the expression levels of VvNCED4 and VvCCD4b among the different rootstocks. VvCYP707A1, which encodes ABA 8'-hydroxylase, a predominant enzyme in ABA catabolism (Speirs et al., 2013), was found at a higher expression level on the two HF rootstocks than on the two 5BB rootstocks at 60 DAFB. The expression level of VvABABG1, which encodes a β-glucosidase associated with the release of free (active) ABA from ABA-GE (Sun et al., 2015), significantly increased at 60 DAFB compared with that at 46 DAFB on all types of rootstock; however, there were no differences in its expression among the four rootstocks. The expression level of VvABAGT, which encodes a glucosyltransferase to form the inactive ABA-GE from ABA (Sun et al., 2010), and that of VvABF2, which encodes an ABA-responsive element of binding factors involved in ABA signal transduction (Nicolas et al., 2014), showed no significant differences among the four types of rootstock at any developmental stage.

Fig. 7

Comparison of gene expression involved in ABA metabolism in berry skin of ‘Ruby Roman’ grafted on 5BB(2x), 5BB(4x), HF(2x), and HF(4x) rootstocks. Expression levels were normalized to those of glyceraldehyde 3-phosphate dehydrogenase (GAPDH). Different letters indicate significant differences among the four types of rootstock at each stage followed by Tukey’s test (P < 0.05, n = 3).

Discussion

Previous studies have reported the relationship between sugar content and accumulation of anthocyanins in grape berries, some of these (González-SanJosé, 1992; Pirie and Mullins, 1977) suggested that the sugars in the skin play a role as regulators in the synthesis of anthocyanins. Others (Wicks and Kliewer, 1983) rejected the regulatory role of sugar in anthocyanin synthesis because many external factors, such as light and temperature, can alter anthocyanin accumulation, but not the sugar content. In this study, the contents of sugars and anthocyanins in the berry skin of berries on four types of rootstock did not show the same pattern during berry ripening [for example, the contents of most anthocyanins were highest in the berries on 5BB(4x) rootstock at 72, 82, and 92 DAFB in 2014 (Fig. 1 left), while the contents of fructose, glucose and sucrose did not differ among the four rootstocks at these stages (Fig. 2 left)]; however, the increase in anthocyanin contents may be associated with the content of sugars at véraison. Gambetta et al. (2010) reported that sugar and ABA play a predominant role in regulating the expression of a suite of genes at the onset of ripening, when anthocyanins begin to accumulate.

In the mature stage of ‘Ruby Roman’ berry skins, Cy and Pn groups of anthocyanins (dihydroxylated anthocyanins) are the dominant components (Gao-Takai et al., 2019). In this study, ‘Ruby Roman’ berries grafted on different rootstocks contained similar anthocyanin compositions, and the proportion of each anthocyanin group was generally constant regardless of rootstock type, although the total and individual anthocyanin contents were very different among the four rootstocks (Fig. 1). These results indicate that the effect of the rootstock on anthocyanin composition in the berry skin of the scion is not significant. Suriano et al. (2016) also reported that at a quantitative level, the concentrations of total and individual anthocyanins showed significant differences among five rootstocks studied; however, in percentage terms, the differences in anthocyanin fractions were fairly small.

In the present study, the anthocyanin content and expression of the most closely related genes (VlMybA1-2, VlMybA1-3, VvUFGT, VvFAOMT, and VvF3'5'H2) were highly consistent among four types of rootstock (Figs. 1 and 6). However, the expression level of VvF3'H, which affects the Cy and Pn biosynthetic pathways, showed no variation, although the contents of Cy and Pn, the dominant anthocyanins in the ‘Ruby Roman’ grape, demonstrated considerable differences among the different rootstock types. The lack of change in the expression level of VvF3'H was also observed in our previous research on the effects of temperature (Gao-Takai et al., 2019), suggesting that the key factor(s) determining coloration, even in a red-skinned grape, may be genes other than F3'H.

Although the skin coloration of grape berries mainly depends on their genetic composition, the same grape species can display continuous variations in different environments. Anthocyanin accumulation was affected by environmental factors such as temperature or light (Jeong et al., 2004; Matus et al., 2009), as well as abiotic stresses such as water stress (Castellarin et al., 2007; Mirás-Avalos and Intrigliolo, 2017). The rootstock type can affect the water status and vigor of scions by acting on transpiration, subsequently affecting the composition of berries. In general, vigorous vegetative scion growth on a vigorous rootstock induces poor berry coloration and reduces berry quality (Motosugi et al., 2007). The ‘5BB’ and ‘Hybrid Franc’ rootstocks used in this study belong to a semidwarf rootstock and vigorous rootstock, respectively. However, in our previous study, ‘Ruby Roman’ scions on HF(2x) rootstocks did not show more growth than those on 5BB(2x) rootstocks, and HF(4x) did not show the same pattern of reduced scion vegetative growth and improved the berry quality as 5BB(4x) (Gao-Takai et al., 2017). This may be related to the compatibility between the rootstock and scion, in which hormonal signaling may be important. In fact, the hormone contents in shoots and roots of the grafted vines were different between 5BB and HF rootstocks (Gao-Takai et al., 2018). For example, in the shoot apexes and thin roots, the contents of IAA and tZ on the two 5BB rootstocks were higher than those on the two HF rootstocks, whereas the ABA content tended to be higher on HF rootstocks than on 5BB rootstocks. Further studies are needed to clarify the relationship between scion vegetative growth and rootstock varieties, between diploid and tetraploid rootstocks, as well between vegetative growth and berry quality.

Fruit ripening related processes are thought to be regulated by multiple hormones, including ABA, ethylene, auxins, gibberellins, and cytokinins, through complex interactions (Fortes et al., 2015; Parada et al., 2017). In this study, we analyzed the changes in ABA, IAA, cytokinins, and their metabolites contents in berry skin of scion berries during maturation. The total metabolites contents of ABA, IAA, and cytokinin were present at the lowest level at pre-véraison, with the lowest content for ABA (at 40 DAFB in 2014 and 48 DAFB in 2015) occurring a little later than that for IAA (at 30 DAFB in 2014 and 43 DAFB in 2015) and cytokinin (at 30 DAFB in both years) (Figs. 3, 4, and 5). The metabolic pathways for the three hormones were all promoted after véraison. Interestingly, the total metabolites contents of the three hormones in the berry skin on 5BB(4x) rootstock that had the least scion growth were all the highest.

Several recent studies have reported that a decrease in IAA content is necessary to trigger the onset of ripening (Corso et al., 2016; Fasoli et al., 2018). It has been proposed that high auxin and low ABA levels could slow ripening progression (Gouthu and Deluc, 2015). However, the results of our study indicate that the relationship between ABA and IAA in inducing berry ripening could be related to the timing of content changes, rather than to the contents.

The content of the ABA metabolites pool (mainly ABA and ABA-GE) continuously and substantially increased after véraison, whereas the expression levels of VvNCEDs did not show a corresponding increase. Similar results have also been reported in other studies; for example, the expression pattern of VvNCEDs could not be correlated with changes in ABA levels in grapevine berries (Castellarin et al., 2011; Wheeler et al., 2009). Castellarin et al. (2016) reported that ABA content increases at the onset of berry ripening without induction of ABA biosynthesis genes, suggesting that the ABA increase may reflect an import from other tissues and/or a decrease in ABA catabolism. In the current study, the expression of VvCYP707A1 in the berry skin on two 5BB rootstocks was significantly lower than on the two HF rootstocks at 60 DAFB (Fig. 7). This may be one of the reasons for the higher ABA contents on 5BB rootstocks. In addition, the expression level of VvABABG1 at 60 DAFB was significantly higher than that at véraison (Fig. 7), suggesting that ABA may be derived by cleavage of ABA-GE after véraison. ABA-GE is considered to be a stored form of ABA and is involved in long-distance transport from one plant tissue to another (Jiang and Hartung, 2008; Xu et al., 2014). Based on these results, we proposed that ABA-GE may be imported into berries from the roots or leaves through vascular sap flow, followed by the release of active ABA during berry ripening. Moreover, the amount of ABA-GE imported into the berries varies depending on the rootstock type. For example, it was higher on 5BB(4x) rootstock than on the other three rootstocks. Although the mechanism by which ABA-GE (or ABA) is imported into berries remains unclear, the high amount of ABA and ABA-GE (and perhaps other hormones) in 5BB(4x) rootstock berry skin could be due to a higher concentration in flowing sap, which may be related to lower growth of roots and shoots in 5BB(4x) rootstock. Motosugi and Yamamoto (2000) reported that tetraploid rootstocks showed a lower flow rate of sap bleeding from a stem stump than that observed in corresponding diploid rootstocks.

In summary, the expression levels of anthocyanin biosynthesis-related genes among the different rootstocks were not consistent with the levels of ABA metabolism-related genes. Most genes in the former group generally showed higher expressions on 5BB(4x) rootstock, whereas most of the latter group showed no obvious difference among the four types of rootstock. This indicated that anthocyanin biosynthesis in berry skin is probably mainly associated with the amount of ABA (or ABA-GE) imported rather than with ABA synthesized in the fruit, assuming ABA is a key factor in the induction of fruit ripening and berry coloration.

Conclusions

Rootstock can affect the accumulation of anthocyanins in the berry skin of grafted scion berries, mainly by controlling the amount of anthocyanin synthesis, rather than by changing the anthocyanin composition. The expression levels of anthocyanin biosynthesis-related genes, except for VvF3'H, were highly consistent with the total anthocyanins content among the four types of rootstock in this study.

The higher content of ABA and sugars around véraison in the berry skin on 5BB(4x) rootstock was considered to be the reason for the high accumulation of anthocyanins, which may be achieved by regulating the expression of VlMybA1-2 and VlMybA1-3. However, the expression levels of VvNCEDs did not show a corresponding increase with the contents of ABA and ABA-GE. These results suggest that ABA (as well as other hormones) may be imported into berries from other organs (probably in the form of ABA-GE), rather than synthesized in the fruit. The contents of ABA, IAA, and the cytokinin metabolite pool changed synchronously during berry development; they all had the lowest levels at pre-véraison, and all of them tended to be higher on 5BB(4x) rootstock. The period when all hormone metabolites are at their lowest levels may be the turning point of the phase change to start ripening.

Acknowledgements

We thank our laboratory assistant Mariko Esaki and other technicians at the Agricultural Experiment Station of Ishikawa Prefectural University for their assistance in the experiments and cultivation.

Literature Cited

•  Azuma,  A.,  H.  Yakushiji,  Y.  Koshita and  S.  Kobayashi. 2012. Flavonoid biosynthesis-related genes in grape skin are differentially regulated by temperature and light conditions. Planta 236: 1067–1080.
•  Bogs,  J.,  D.  McDavid and  S. P.  Robinson. 2006. Identification of the flavonoid hydroxylases from grapevine and their regulation during fruit development. Plant Physiol. 140: 279–291.
•  Castellarin,  S. D.,  G. A.  Gambetta,  H.  Wada,  M. N.  Krasnow,  G. R.  Cramer,  E.  Peterlunger,  K. A.  Shackel and  M. A.  Matthews. 2016. Characterization of major ripening events during softening in grape: turgor, sugar accumulation, abscisic acid metabolism, colour development, and their relationship with growth. J. Exp. Bot. 67: 709–722.
•  Castellarin,  S. D.,  G. A.  Gambetta,  H.  Wada,  K. A.  Shackel and  M. A.  Matthews. 2011. Fruit ripening in Vitis vinifera: spatiotemporal relationships among turgor, sugar accumulation, and anthocyanin biosynthesis. J. Exp. Bot. 62: 4345–4354.
•  Castellarin,  S. D.,  G. D.  Gaspero,  R.  Marconi,  A.  Nonis,  E.  Peterlunger,  S.  Paillard,  A. F.  Adam-Blondon and  R.  Testolin. 2006. Colour variation in red grapevines (Vitis vinifera L.): genomic organization, expression of flavonoid 3'-hydroxylase, flavonoid 3',5'-hydroxylase genes and related metabolite profiling of red cyanidin-/blue delphinidin-based anthocyanins in berry skin. BMC Genomics 7: 12. DOI: 10.1186/1471-2164-7-12.
•  Castellarin,  S. D.,  M. A.  Matthews,  G. D.  Gaspere and  G. A.  Gambetta. 2007. Water deficits accelerate ripening and induce changes in gene expression regulating flavonoid biosynthesis in grape berries. Planta 227: 101–112.
•  Corso,  M.,  A.  Vannozzi,  F.  Ziliotto,  M.  Zouine,  E.  Maza,  T.  Nicolato,  N.  Vitulo,  F.  Meggio,  G.  Valle,  M.  Bouzayen,  M.  Müller,  S.  Munné-Bosch,  M.  Lucchin and  C.  Bonghi. 2016. Grapevine rootstocks differentially affect the rate of ripening and modulate auxin-related genes in Cabernet Sauvignon berries. Front Plant Sci. 7: 69. DOI: 10.3389/fpls.2016.00069.
•  Fasoli,  M.,  C. L.  Richter,  S.  Zenoni,  E.  Bertini,  N.  Vitulo,  S. D.  Santo,  N.  Dokoozlian,  M.  Pezzotti and  G. B.  Tornielli. 2018. Timing and order of the molecular events marking the onset of berry ripening in grapevine. Plant Physiol. 178: 1187–1206.
•  Ford,  C. M.,  P. K.  Boss and  P. B.  Høj. 1998. Cloning and characterization of Vitis vinifera UDP-glucose: flavonoid 3-O-glucosyltransferase, a homologue of the enzyme encoded by the maize Bronze-1 locus that may primarily serve to glucosylate anthocyanidins in vivo. J. Biol. Chem. 273: 9224–9233.
•  Fortes,  A. M.,  R. T.  Teixeira and  P.  Agudelo-Romero. 2015. Complex interplay of hormonal signals during grape berry ripening. Molecules 20: 9326–9343.
•  Gambetta,  G. A.,  M. A.  Matthews,  T. H.  Shaghasi,  A. J.  McElrone and  S. D.  Castellarin. 2010. Sugar and abscisic acid signaling orthologs are activated at the onset of ripening in grape. Planta 232: 219–234.
•  Gao-Takai,  M.,  A.  Katayama-Ikegami,  K.  Matsuda,  H.  Shindo,  S.  Uemae and  M.  Oyaizu. 2019. A low temperature promotes anthocyanin biosynthesis but does not accelerate endogenous abscisic acid accumulation in red-skinned grapes. Plant Sci. 283: 165–176.
•  Gao-Takai,  M.,  A.  Katayama-Ikegami,  S.  Nakano,  K.  Matsuda and  H.  Motosugi. 2017. Vegetative growth and fruit quality of ‘Ruby Roman’ grapevines grafted on two species of rootstock and their tetraploids. Hort. J. 86: 171–182.
•  Gao-Takai,  M.,  A.  Shinmura,  H.  Nito and  A.  Katayama-Ikegami. 2018. Comparison of vegetative growth, fruit quality and plant hormone content of ‘Ruby Roman’ grapevines grafted on diploid and tetraploid rootstocks. Acta Hortic. 1208: 207–213.
•  González-SanJosé,  M. L. 1992. Relationship between anthocyanins and sugars during the ripening of grape berries. Food Chem. 43: 193–197.
•  Gouthu,  S. and  L. G.  Deluc. 2015. Timing of ripening initiation in grape berries and its relationship to seed content and pericarp auxin levels. BMC Plant Biol. 15: 46. DOI: 10.1186/s12870-015-0440-6.
•  Hugueney,  P.,  S.  Provenzano,  C.  Verriès,  A.  Ferrandino,  E.  Meudec,  G.  Batelli,  D.  Merdinoglu,  V.  Cheynier,  A.  Schubert and  A.  Ageorges. 2009. A novel cation-dependent O-methyltransferase involved in anthocyanin methylation in grapevine. Plant Physiol. 150: 2057–2070.
•  Jeong,  S. T.,  N.  Goto-Yamamoto,  S.  Kobayashi and  M.  Esaka. 2004. Effects of plant hormones and shading on the accumulation of anthocyanins and the expression of anthocyanin biosynthetic genes in grape berry skins. Plant Sci. 167: 247–252.
•  Jiang,  F. and  W.  Hartung. 2008. Long-distance signaling of abscisic acid (ABA): the factors regulating the intensity of the ABA signal. J. Exp. Bot. 59: 37–43.
•  Katayama-Ikegami,  A.,  T.  Katayama,  M.  Takai and  T.  Sakamoto. 2016a. Reference gene validation for gene expression studies using quantitative RT-PCR during berry development of ‘Aki Queen’ grapes. Vitis 55: 157–160.
•  Katayama-Ikegami,  A.,  T.  Sakamoto,  K.  Shibuya,  T.  Katayama and  M.  Gao-Takai. 2016b. Effects of abscisic acid treatment on berry coloration and expression of flavonoid biosynthesis genes in grape. Am. J. Plant Sci. 7: 1325–1336.
•  Katayama-Ikegami,  A.,  Y.  Sugiyama,  T.  Katayama,  A.  Sakamoto,  R.  Shimada,  C.  Miyazaki and  M.  Gao-Takai. 2022. Difference between coloration and endogenous abscisic acid accumulation patterns in two red grape cultivars, ‘Aki Queen’ and ‘Ruby Roman’ (Vitis labruscana Bailey) berries. Hort. J. 91: 312–321.
•  Matus,  J. T.,  R.  Loyola,  A.  Vega,  A.  Peña-Neira,  E.  Bordeu,  P.  Arce-Johnson and  J. A.  Alcalde. 2009. Post-véraison sunlight exposure induces MYB-mediated transcriptional regulation of anthocyanin and flavonol synthesis in berry skins of Vitis vinifera. J. Exp. Bot. 60: 853–867.
•  Mirás-Avalos,  J. M. and  D. S.  Intrigliolo. 2017. Grape composition under abiotic constrains: water stress and salinity. Front Plant Sci. 8: 851. DOI: 10.3389/fpls.2017.00851.
•  Motosugi,  H. and  Y.  Yamamoto. 2000. The comparison of shoot growth, leaf morphology, and flow rate of bleeding sap between diploids and tetraploids of ‘3309’ and ‘Riparia Gloire’ grape rootstocks. J. Japan Soc. Hort. Sci. 69 (Suppl. 2): 112 (In Japanese).
•  Motosugi,  H.,  K.  Okudo,  D.  Kataoka and  T.  Naruo. 2002. Comparison of growth characteristics between diploid and colchicine-induced tetraploid grape rootstocks. J. Japan Soc. Hort. Sci. 71: 335–341.
•  Motosugi,  H.,  Y.  Yamamoto,  T.  Naruo and  D.  Yamaguchi. 2007. Growth and fruit quality of ‘Kyoho’ grapevines grafted on autotetraploid rootstocks. J. Japan Soc. Hort. Sci. 76: 271–278.
•  Nicolas,  P.,  D.  Lecourieux,  C.  Kappel,  S.  Cluzet,  G.  Cramer,  S.  Delrot and  F.  Lecourieux. 2014. The basic leucine zipper transcription factor ABSCISIC ACID RESPONSE ELEMENTBINDING FACTOR2 is an important transcriptional regulator of abscisic acid-dependent grape berry ripening processes. Plant Physiol. 164: 365–383.
• Parada, F., P. Arce-Johnson and C. Espinoza. 2017. Phytohormonal control over the grapevine berry development. p. 143–159. In: M. El-Esawi (ed.). Phytohormones: Signaling Mechanisms and Crosstalk in Plant Development and Stress Responses. IntechOpen, United Kingdom.
•  Pilati,  S.,  G.  Bagagli,  P.  Sonego,  M.  Moretto,  D.  Brazzale,  G.  Castorina,  L.  Simoni,  C.  Tonelli,  G.  Guella,  K.  Engelen,  M.  Galbiati and  C.  Moser. 2017. Abscisic acid is a major regulator of grape berry ripening onset: new insights into ABA signaling network. Front Plant Sci. 8: 1093. DOI: 10.3389/fpls.2017.01093.
•  Pirie,  A. and  M. G.  Mullins. 1977. Interrelationships of sugars, anthocyanins, total phenols and dry weight in the skin of grape berries during ripening. Am. J. Enol. Vitic. 28: 204–209.
•  Speirs,  J.,  A.  Binney,  M.  Collins,  E.  Edwards and  B.  Loveys. 2013. Expression of ABA synthesis and metabolism genes under different irrigation strategies and atmospheric VPDs is associated with stomatal conductance in grapevine (Vitis vinifera L. cv Cabernet Sauvignon). J. Exp. Bot. 64: 1907–1916.
•  Sun,  J.,  Y.  Dong,  C.  Li and  Y.  Shen. 2015. Transcription and enzymatic analysis of beta-glucosidase VvBG1 in grape berry ripening. Plant Growth Regul. 75: 67–73.
•  Sun,  L.,  M.  Zhang,  J.  Ren,  J.  Qi,  G.  Zhang and  P.  Leng. 2010. Reciprocity between abscisic acid and ethylene at the onset of berry ripening and after harvest. BMC Plant Biol. 10: 257. DOI: 10.1186/1471-2229-10-257.
•  Suriano,  S.,  V.  Alba,  D. D.  Gennaro,  M. S.  Suriano,  M.  Savino and  L.  Tarricone. 2016. Genotype/rootstocks effect on the expression of anthocyanins and flavans in grapes and wines of Greco Nero. (Vitis vinifera L.). Sci. Hortic. 209: 309–315.
• Uehara, S. 1995. Grape. p. 289–334. In: K. Kawase (ed.). Properties and utilization of rootstocks for fruit crops (In Japanese). Nosangkyosonbunkakyokai (Rural Culture Association), Tokyo, Japan.
•  Wheeler,  S.,  B.  Loveys,  C.  Ford and  C.  Davies. 2009. The relationship between the expression of abscisic acid biosynthesis genes, accumulation of abscisic acid and the promotion of Vitis vinifera L. berry ripening by abscisic acid. Aust. J. Grape Wine Res. 15: 195–204.
•  Wicks,  A. S. and  W. M.  Kliewer. 1983. Further investigations into the relationship between anthocyanins, phenolics and soluble carbohydrates in grape berry skins. Am. J. Enol. Vitic. 34: 114–116.
• Xu, Z. Y., Y. J. Yoo and I. Hwang. 2014. ABA conjugates and their physiological roles in plant cells. p. 77–87. In: D. P. Zhang (ed.). Abscisic Acid: Metabolism, Transport and Signaling. Springer, New York.
•  Young,  P. R.,  J. G.  Lashbrooke,  E.  Alexandersson,  D.  Jacobson,  C.  Moser,  R.  Velasco and  M. A.  Vivier. 2012. The genes and enzymes of the carotenoid metabolic pathway in Vitis vinifera L. BMC Genomics 13: 243. DOI: 10.1186/1471-2164-13-243.