Introduction

The sensory properties of grape (Vitis vinifera L.) berries depend on the accumulation of both primary (sugars and organic acids) and secondary metabolites (CooMbe and McCarthy 2000). Anthocyanins (ATs) and proanthocyanidins (PAs) are important secondary metabolites and accumulate primarily in berry skins, where they are involved in a wide range of biological functions such as antioxidation and protection against biotic and abiotic stresses (Chalker-Scott 1999; Takahama 2004). ATs and PAs are synthesized in grapevines via the flavonoid pathway, with phenylalanine as a precursor. The grape genome contains multiple copies of the genes encoding chalcone synthase (CHS), chalcone isomerase (CHI), and flavanone 3-hydroxylase (F3H), which are involved in the recruitment of flavonoid precursors to the flavonoid biosynthetic pathway. Among them, VvCHI1, VvCHS2, VvCHS3, and VvF3H2 are transcribed early during flavonoid biosynthesis in berry skins for the special synthesis of ATs (Jeong et al. 2008). Genes encoding dihydroflavonol-4-reductase (VvDFR) and leucoanthocyanidin dioxygenase (VvLDOX) mediate the late steps of AT biosynthesis (Soubeyrand et al. 2014), followed by glycosylation and methylation mediated by UDPglucose-flavonoid 3-O-glucosyltransferase (VvUFGT) (Kobayashi et al. 2002) and anthocyanin-O-methyltransferase (VvOMT) (Ageorges et al. 2006; Castellarin et al. 2007), respectively. VvLAR and VvANR are structural genes involved in the regulation of PA biosynthesis (Bogs et al. 2005; Dixon et al. 2005; Marles et al. 2003; Xie and Dixon 2005). Several MYB transcription factors controlling different branches of the flavonoid pathway have also been identified: VvMYB5a, VvMYB5b, VvMYBPA1, and VvMYBPA2 involved in the general PA pathway (Bogs et al. 2007; Deluc et al. 2006, 2008; Terrier et al. 2009) and VvMYBA1 and VvMYBA2 for the AT pathway (Kobayashi et al. 2004; Walker et al. 2007).

AT and PA accumulation and metabolism in grape skins are complex processes of physiological and biochemical changes (CooMbe and McCarthy 2000). Many factors influence the accumulation of ATs and PAs, including grape variety, yield, exogenous stimuli, and climatic variables such as temperature and light (Downey et al. 2006). Ethylene is an important signaling molecule modulating diverse physiological aspects of growth and development. It elicits specific changes in gene expression at various stages of the plant life cycle (De Paepe and Van Der Straeten 2005). Grapes are non-climactic fruits, and even though ethylene concentration and respiration do not dramatically increase at the beginning of ripening (Chervin et al. 2004), ethylene is the dominant regulator of AT accumulation in the berry (Bellincontro et al. 2006; El-Kereamy et al. 2003). The regulatory profiles of other phenolic compounds, such as PAs, are still speculative in the absence of evidence supporting a role for exogenous stimuli. Light is also an important signal affecting AT and PA synthesis. The effects of cluster shading and exposure to sunlight have been extensively documented for table grapes (Wicks and Kliewer 1983; Kliewer and Antcliff 1970) and wine grapes (Bergqvist et al. 2001; Cortell and Kennedy 2006; Crippen and Morrison 1986a, b; Downey et al. 2004; Haselgrove et al. 2000; Morrison and Noble 1990; Price et al. 1995). The variable effect of cluster exposure has been shown to enhance, maintain, or reduce phenolic accumulation. Fully understanding the complex regulation of the AT and PA biosynthetic pathways by various factors, including ethylene and light sensed by the berries, is thus important.

The large size of the vines, the variability of the clusters and berries, the complex environment, and the difficulty of managing hormonal signal transduction provided challenges to the study of the interactive effects of the stimuli on the berries. The regulatory profiles of the AT and PA biosynthetic pathways under the interaction between ethylene and light have also not yet been elucidated. We thus investigated the individual effects of exogenous ethylene and light and their interactive effect on the AT and PA biosynthetic pathways. AT and PA contents and the transcriptional profiles of relevant genes in berry skins across four developmental stages were analyzed. We also analyzed the key genes in endogenous ethylene biosynthesis and the downstream ethylene regulatory pathway. These analyses would provide additional data for a comprehensive understanding of the events underlying the color changes in ethephon-treated berries in dark and light and would broaden our understanding of the response of berries to environmental influences.

Materials and methods

Experimental field site

The field experiment was conducted in a commercial vineyard in Jingyang, Shaanxi, China (34°65′N, 108°75′E) during the 2013 growing season. This region contains hilly and semi-hilly terrain and has a semi-humid, continental monsoon climate with 2195.2 h of sunshine annually and an average frost-free period of 213 days. The mean annual temperature and precipitation are 13 °C and 548.7 mm, respectively. The vineyard soil is classified as sandy loam. All viticultural operations followed standard commercial practices and were identical for all experimental vines.

Berry sampling and experimental design

The experimental vines were five-year-old drip-irrigated grapevines (cv. Cabernet Sauvignon) grafted onto their own rootstocks in a bilateral cordon system orientated east–west with two strong horizontal cordons and vertically trained shoots. Row orientation has a pronounced effect on the amount of photosynthetically active radiation (PAR) received by the two sides of the rows. Grape clusters in the same positions and of the same size were thus selected from the southern sides of rows to maintain a consistent PAR. Three replicates of thirty grape clusters from ten plants, for a total of ninety clusters, received the four treatments: light exposure with (+L+E) and without (+L−E) ethephon, and box-shading with (−L+E) and without (−L−E) ethephon. Ethylene affects the accumulation of regulatory-gene transcripts at the beginning of véraison (Böttcher et al. 2013a, b; Chervin et al. 2008; Coombe and Hale 1973; El-Kereamy et al. 2003). In our study, the clusters at the onset of véraison (softening of 10 % of the berries) were dipped into a 400 mg/L solution of ethephon (>85 % purity; Sangon, Inc., Shanghai, China) containing a buffered wetting agent (1 ml/L Tween 80) for 30 s. Control clusters were dipped into the wetting agent only. The shaded treatments used shade boxes modeled on the design by Downey et al. (2004) to cover the clusters immediately after ethephon treatment, and the shaded clusters remained enclosed until harvest. Disposable USB data loggers (Apresys inc., Duluth, USA) for recording ambient temperature every 20 min were installed beside the grape clusters with and without shade boxes. The similar ambient temperatures indicated that box-shading had little effect on the temperature of the treated clusters (Supplementary Fig. 1).

Grape samples were collected 7, 17, 35, and 49 days after treatment (DAT) (Supplementary Fig. 2). On each sampling date, a total of 250 berries for each replicate were randomly sampled from thirty clusters. The skins of 100 berries were then manually peeled, weighed to determine skin fresh weight, and then stored at −40 °C for analysis of the phenolic compounds, and the skins of another 100 berries were peeled, ground in liquid nitrogen, vacuum freeze-dried, and weighed to determine skin dry weight. The skins of the remaining 50 berries were frozen in liquid nitrogen and stored at −80 °C for RNA extraction and quantification of gene expression by real-time PCR (see below). The remaining treated clusters were sampled 49 DAT for physical and chemical analyses.

Measurement of physiochemical parameters

Two hundred berries were randomly sampled from each experimental clusters at the last sampling date (49 DAT), and the physiological parameters were defined by total-solid and reducing-sugar concentrations, pH, titratable acidity, berry diameter, and berry fresh weight. Twenty randomly sampled berries were measured to determine berry diameter with callipers as previously described (Coombe 1992), and berry fresh weight was obtained by weighing 100 randomly sampled berries. Berry juice was obtained by manual pressing. Total soluble solid content (°Bx) was determined using a PAL-1 digital refractometer (Atago, Tokyo, Japan). The berry pH was measured using a pH meter (model PB-10, Sartorius, Germany). Reducing-sugar concentration and titratable acidity were determined by Fehling reagent titration and sodium hydroxide titration, respectively, according to the national standards of the People’s Republic of China (GB/T, 15038-2006, 2006).

Sample preparation for phenolic analysis

Phenols were extracted from the skins of 100 berries randomly selected from each replicate on each sampling date. Each treatment was subjected to three independent extractions from three replicates. The skins were homogenized in liquid nitrogen with a chilled mortar, and the powder was then immersed in 20 mL of a hydrochloric acid/methanol solution (60 % methanol, 0.1 % hydrochloric acid). The sample was sonicated at 40 % power and 30 °C for 30 min and then centrifuged at 1800 g for 10 min at 4 °C. The supernatant was transferred to 100-mL jars. Twenty milliliters of the hydrochloric acid/methanol solution were added to the pellets, and the sample was extracted twice more under the same conditions. The supernatants of the three extractions were pooled and stored at −80 °C. All extractions were performed in the dark to prevent phenolic decomposition and oxidation.

Phenolic analysis

TP content was determined by the Folin-Ciocalteu method (Jayaprakasha et al. 2001) with some modifications. To a glass cuvette, 2.9 mL of distilled water, phenolic extract (the dose was dependent on the absorbance value, which should be 0.2–0.8; see below), and 0.5 mL of Folin-Ciocalteu reagent were added successively. After extraction for 5 min, 1.5 mL NaCO3 were added. The mixture was allowed to react at room temperature in the dark for 2 h, and absorbance was determined at 765 nm. A control was prepared by replacing the sample with methanol. TP content was expressed as milligrams gallic acid equivalence (GAE) per berry (mg GAE/berry).

TFO content was determined by a previously reported method (Fang et al. 2011; Peinado et al. 2009) with minor modifications. The required amount of phenolic extract was mixed with 0.8 mL of distilled water and 0.06 mL of 5 % NaNO2 in a 10-mL centrifuge tube, and the mixture was incubated at room temperature for 5 min. At the end of the reaction, 0.06 mL of 10 % AlCl3 was added, and the mixture was allowed to stand for 10 min. Finally, 0.4 mL of 10 % Na2CO3 and 0.9 mL of distilled water were added to the reaction mixture, and absorbance was read at 510 nm. A control was prepared by replacing the sample with methanol. TFO content was expressed as milligrams rutin equivalence (RE) per berry (mg RE/berry).

Total PA content was determined by methyl cellulose precipitation (Sarneckis et al. 2006). In a glass cuvette, 3.0 mL of methyl cellulose was added into the required amount of phenolic extract, the solution was incubated at room temperature for 3 min, 2 mL of saturated (NH4)2SO4 was added, and the mixture was then diluted with distilled water to 10 mL and allowed to react at room temperature for 10 min. After centrifugation at 1800g for 5 min, absorbance of the supernatant was determined at 280 nm. The methyl cellulose was replaced by distilled water in the control. Total PA content was expressed as milligrams (+)-catechin equivalence (CE) per berry (mg CE/berry).

Total AT content was determined by the pH-differential method (Stojanovic and Silva 2007). The required volume of extracted sample was added to two tubes, and the samples were diluted to 5 mL with KCl buffer at pH 1.0 and CH3CO2Na·3H2O buffer at pH 4.5, respectively. The absorbances of the two mixtures were measured at 520 nm and 700 nm, respectively, and calculated using the equation A = (A520–A700)pH1.0 − (A520–A700)pH4.5. Each phenolic extract was diluted so that the sample in the buffer at pH 1.0 had an absorbance <1. The AT content was expressed as milligrams of malvidin-3-monoglucoside equivalence per berry (mg ME/berry) and calculated using the equation TA content = (A × MW × DF × Ve × 1000)/(ε × 1 × M), where A is the absorbance, MW is the molecular weight of malvidin-3-glucoside (493.5), DF is the dilution factor, Ve is the extraction volume, ε is the molar extinction coefficient of malvidin-3-glucoside (28,000), and M is the mass of the extracted skins.

RNA extraction and cDNA synthesis

Total RNA was extracted from the skin tissues as described by Reid et al. (2006) using a CTAB-spermidine extraction buffer. The total RNA was treated with RNase-free DNase I (TaKaRa Bio Inc., Dalian, China) to remove any contaminating genomic DNA. The integrity of the RNA was verified by 1 % formaldehyde-agarose gel electrophoresis and subsequent staining with ethidium bromide. RNA purity was verified by absorbance ratios of 1.8–2.0 at 260/280 nm using a NanoDrop ND-1000 Spectrophotometer (NanoDrop Technologies, Rockland, USA). First-strand cDNA was synthesized for each sample using 700 ng of total RNA and Super Script II reverse transcriptase (Takara, Inc., Dalian, China) according to the manufacturer’s instructions. For the following experiment, the products of reverse transcription were diluted 6-fold.

Real-time quantitative PCR analysis

Gene expression in the various treatments was analyzed by quantitative real-time PCR using published gene-specific primers or primers designed by Primer Premier 5. A list of the primer pairs and amplicon lengths of the target genes are provided in Supplementary Table 1. The efficiencies of the primers were tested in preliminary experiments with dilutions of the purified PCR products and maintained an R 2 ≥ 0.90. The levels of expression of the genes were normalized using the grape actin gene (VvActin; GenBank accession EC9699) (Reid et al. 2006). The PCR analysis used an IQ5 Real-Time PCR System (Bio-Rad, Hercules, CA, USA). Each reaction was performed in triplicate in volumes of 20 μl containing 0.8 μl of each gene-specific primer (10 μM), 1.0 μl of cDNA, 10 μl of SYBR premix Ex Taq™ II (Takara, Inc., Dalian, China), and 7.4 μl of sterile distilled water. The reaction was heated to 95 °C for 30 s, followed by 45 cycles of 95 °C for 5 s and 60 °C for 30 s. To determine the specificity of the PCR reaction, a melting curve was obtained at the end of the reaction from 55 to 95 °C at increments of 0.5 °C maintained for 1 s to ensure product homogeneity. The normalized expression of the target gene was calculated by the 2−ΔCT method. Gene expression was presented as the mean and standard error of three replicates.

Statistical analysis

The data for phenolic content and the gene-expression profiles were analyzed using Microsoft Excel 2007. The statistical significance of differences between the four treatments at each sampling date was determined by one-way analyses of variance (ANOVAs) and Duncan’s tests using SAS 9.2 software (SAS Institute Inc., Cary, USA). Two-way ANOVAs tested the significance of the interaction between ethephon and light.

Results

Physiochemical parameters

The ethephon treatments generally did not affect berry weight, berry diameter, or the reducing-sugar content of the juice at 49 DAT (Table 1). The box-shading treatment, however, did affect these indicators. The soluble-solid and reducing-sugar contents, titratable acidity, and the sugar/TA ratio indicated an interactive effect between ethephon and light.

Table 1 Physical and chemical analyses of berries and grape juice sampled at 49 days after treatment
Full size table

Skin fresh and dry weights

Skin fresh and dry weights were significantly influenced by the light and ethylene treatments (Fig. 1). Specifically, the light treatments (+L+E and +L−E) significantly increased the fresh and dry weights for 100 berries at the four sampling times. The ethephon treatment in the dark (−L+E) had no significant effect at 35 and 49 DAT, but the ethephon treatment in the light (+L+E) had a significant influence on skin weight for 100 berries at 49 DAT, with fresh weights of 12.79, 12.26, 11.28, and 10.88 g for +L+E, +L−E, −L+E, and −L−E, respectively (Fig. 1a). Correspondingly, the four treatments had the same effect on the skin dry weight, which increased significantly under the light and ethephon treatments relative to no treatment (Fig. 1b). These results highlighted the essential role of ethylene and light for matter accumulation in berry skins.

Fig. 1
figure 1

Skin dry weight (a) and fresh weight (b) of 100 berries in the four treatments: +L+E, light exposure with ethephon; +L−E, light exposure without ethephon; −L+E, box-shading with ethephon; −L−E, box-shading without ethephon. Different lowercase letters indicate significant differences between treatments calculated by Duncan’s tests (P < 0.05). The bars indicate the means and standard deviations of three replicates

Full size image

Total phenolic (TP) and flavonoid (TFO) contents

The effect of light and ethephon on total phenolic (TP) and flavonoid (TFO) contents in grape skins at different ripening stages are shown in Fig. 2. The light treatments (+L+E and +L−E) significantly increased the TP and TFO contents at the four sampling times. For the ethephon treatments, the TP and TFO contents were consistently higher in +L+E and −L+E (ethephon treatments) than in +L−E and −L−E (ethephon-free treatments), respectively, at all stages except at 17 DAT for TP content. Overall, ethephon and light significantly increased TP and TFO contents. Ethephon and light had a significant interactive effect on the TP and TFO contents at 7, 35, and 49 DAT.

Fig. 2
figure 2

Effects of the four treatments on the amounts of total phenolics (a) and total flavonoids (b) in grape berry skins (cv. Cabernet Sauvignon) during development. The amounts are expressed as milligrams gallic acid equivalence (mg GAE/berry) and as milligrams rutin equivalence (mg RE/berry) per gram of dry berry skins. The treatments are +L+E, light exposure with ethephon; +L−E, light exposure without ethephon; −L+E, box-shading with ethephon; −L−E, box-shading without ethephon. Different lowercase letters indicate significant differences between treatments calculated by Duncan’s tests (P < 0.05). The bars indicate the means and standard deviations of three replicates. The significance of the interactive effect (IE) between ethephon and light by a two-way analysis of variance is represented by asterisks: *P < 0.05, **P < 0.01

Full size image

Analysis of PA content and the key genes regulating its synthesis

The PA content increased significantly under the light and ethephon treatments relative to no treatment, except for the ethephon treatment under dark (−L+E) at 17, 35, and 49 DAT (Fig. 3a). Compared to box-shading without ethephon treatment (−L−E), the PA content increased by 4.01 and 6.97 % at 7 DAT in the light (+L−E) and ethephon (−L+E) treatments, respectively. Until 49 DAT, the PA contents were 67.23 and 33.89 % higher in +L−E and −L+E, respectively, than in −L−E, indicating that both ethephon and light had an effect on the increase in PA content. Ethephon and light also had an interactive effect on PA contents at 17, 35, and 49 DAT.

Fig. 3
figure 3

Analysis of the proanthocyanidin pathway. a Accumulation of proanthocyanidins (PAs) per gram of skin dry weight in grape berries (cv. Cabernet Sauvignon) during development. The amounts are expressed as milligrams (+)-catechin equivalence (CE) per gram of dry berry skin (mg CE/berry) (mean ± SE; n = 3). b Transcript profiles of VvLAR1 as the molar ratio of the mRNA level to that of VvActin in each sample (mean ± SE; n = 3). c Transcript profiles of VvLAR2. d Transcript profiles of VvANR. e Transcript profiles of VvMYB5a. f Transcript profiles of VvMYB5b. g Transcript profiles of VvMYBPA1. See Fig. 2 for a description of the treatments

Full size image

We also analyzed the expression of genes associated with PA biosynthesis (Fig. 3b–g). VvLAR1, VvLAR2, and VvANR mRNA levels were up-regulated by ethephon and light at 7 and 17 DAT and were significantly higher in the ethephon treatment under light (+L+E) at 35 and 49 DAT, when ethephon treatment in the dark (−L+E) inhibited VvLAR2 and VvANR levels. The transcriptional levels of both VvMYB5a and VvMYB5b were lower in the light treatments (+L+E and +L−E) than in the shaded treatments (−L+E and −L−E). Ethephon treatment induced and repressed transcription under light (+L+E) and under dark (−L+E), respectively. Light dominantly increased the transcript levels of VvMYBPA1 at 7 DAT. Notably for the ethephon treatments, the transcript level of VvMYBPA1 was significantly up-regulated by the ethephon treatment under light (+L+E), but ethylene responsiveness was not significantly strengthened in the berries under dark (−L+E). These results indicated that the light and ethephon treatments could increase the transcript levels of VvLAR1, VvLAR2, VvANR, and VvMYBPA1 associated with PA biosynthesis. The ethephon treatment under light (+L+E) had a significant effect, but the ethephon treatment under dark (+L+E) did not.

Analysis of AT content and the key genes regulating its synthesis

Light exposure (+L+E and +L−E) significantly increased AT contents relative to shading (−L+E and −L−E), and the increase was faster than in the ethephon treatments (+L+E and −L+E) (Fig. 4a). The effect of ethephon treatment at 7, 17, and 49 DAT was significant in the berry skins exposed to light (+L+E) but not in the shaded berry skins (−L+E), which led to much larger ranges of increased levels of ATs induced by ethephon under light (+L+E) than those under dark (−L+E). AT content at 49 DAT in the berries treated with ethephon under light (+L+E) increased 0.55-fold, whereas the content increased only 0.09-fold in the shaded berries treated with ethephon (−L+E). Ethephon and light had interactive effects on AT content at the four sampling times.

Fig. 4
figure 4

Analysis of the anthocyanin pathway. a Accumulation of anthocyanins (ATs) per gram of skin dry weight in berry skins (cv. Cabernet Sauvignon) during development. The amounts are expressed as milligrams of cyanidin-3-monoglucoside equivalence (ME) per gram of dry berry skins (mg ME/berry). b Transcript profiles of VvCHI1 as the molar ratio of the mRNA level to that of VvActin in each sample (mean ± SE; n = 3). c Transcript profiles of VvCHS2. d Transcript profiles of VvCHS3. e Transcript profiles of VvF3H2. f Transcript profiles of VvDFR. g Transcript profiles of VvLDOX. h Transcript profiles of VvUFGT. i Transcript profiles of VvMYBA1. See Fig. 2 for a description of the treatments

Full size image

The mRNA levels of VvCHI1, VvCHS2, VvCHS3, VvF3H2, VvDFR, VvLDOX, VvUFGT, and VvMYBA1 during berry development are presented in Fig. 4b–i. VvCHI1, VvCHS2, and VvCHS3 were up-regulated by light at the four sampling times, especially at 7 DAT. The significant increase in the expression of VvCHI1, VvCHS2, and VvCHS3 by ethephon under light (+L+E) continued throughout berry development, but the mRNA levels did not differ significantly and even decreased (for VvCHS2) in the ethephon treatment under dark (−L+E) at 35 and 49 DAT. Light but not ethephon increased the transcript level of VvF3H2 during the berry ripening stage. VvDFR, VvLDOX, and VvUFGT transcription was also induced by light and ethephon in the late steps of AT biosynthesis, and the transcription at 7 DAT confirmed that light increased AT contents markedly more than did ethephon (Fig. 4f–h). Ethephon treatment led to much larger ranges of increased levels of VvDFR, VvLDOX, and VvUFGT under light (+L+E) than under dark (−L+E). The positive effect on the expression of VvUFGT at 7 DAT was 0.025- and 0.195-fold higher in −L+E and +L+E, respectively, than in −L−E and +L−E, allowing the berries to accumulate high contents of AT by exogenous ethephon under light (+L+E). Ethephon and light had also increased the expression of the AT-specific regulatory gene, VvMYBA1, at each sampling date (Fig. 4i). VvMYBA1, which was not significantly transcribed by ethephon treatment under dark (−L+E), was notably up-regulated by ethephon treatment under light (+L+E). Its transcription mimicked the pattern of VvCHI, VvCHS, VvDFR, VvLDOX, and VvUFGT expression, demonstrating that ethephon treatment under light (+L+E) was significantly more effective in up-regulating AT biosynthetic genes than the treatment under dark (−L+E). These results further indicated a potential interactive effect between ethylene and light in the AT biosynthetic pathway.

Endogenous ethylene production and ethylene signal transduction under the two stimuli

The transcript of VvEIN3 is a key transcription factor associated with the ethylene signal transduction pathway (Fig. 5a). The transcriptional levels of VvEIN3 had decreased significantly in the ethephon treatments (+L+E and −L+E) at 7, 17, and 35 DAT. Light exposure (+L+E and +L−E) increased the transcript abundance of VvEIN3 during the same stage, but the effect was not significant. We also analyzed the expression levels of VvERF1 and Vvchit1b, which have been reported to be downstream responsive genes of the ethylene pathway (Alonso et al. 1999), to determine the responses to ethephon in the light and dark (Fig. 5b, c). Transcripts of VvERF1, which were down-regulated by light without ethephon (+L−E), were more significantly enhanced by ethephon under light (+L+E) than under dark (−L+E) at the four sampling dates (Fig. 5b). The transcription level of Vvchit1b increased significantly in +L+E relative to +L−E, but −L+E had little effect relative to −L−E (Fig. 5c). The transcriptional levels of VvERF1 and Vvchit1b supported the differential downstream responsiveness of ethylene under different light conditions, and this difference was inconsistent with the transcriptional level of VvEIN3.

Fig. 5
figure 5

Analysis of the ethylene biosynthetic and signal-transduction pathways. a Transcript profiles of VvEIN3 as the molar ratio of the mRNA level to that of VvActin in each sample (mean ± SE; n = 3). b Transcript profiles of VvERF1. c Transcript profiles of Vvchi1b. d Transcript profiles of VvACO1. e Transcript profiles of VvACO2. See Fig. 2 for a description of the treatments

Full size image

The potential influence of exogenous ethephon on endogenous ethylene responsiveness encouraged us to characterize the transcript abundances of the key genes VvACO1 and VvACO2 in the ethylene biosynthetic pathway during the same four-stage growth pattern (Fig. 5d, e). The expression of VvACO1 increased rapidly at the beginning of véraison, presenting significantly higher levels under light at 7 DAT (Fig. 5d). Ethephon treatment up-regulated the transcriptional levels of VvACO1 and VvACO2 under light (+L−E) but inhibited the expression of VvACO2 in the dark (−L+E), especially at 35 DAT, when ethephon treatment under light increased the expression level of VvACO2 1.4-fold (Fig. 5d). Ethephon thus induced high expression of VvACO1 and VvACO2 under light, similar to the up-regulation of Vvchit1b, VvERF1, and PA and AT biosynthetic genes, further reinforcing the link between light and ethylene.

Hierarchical clustering in the profiles of all genes analyzed

Hierarchical clustering was performed by Genesis (http://genome.tugraz.at) to assess the similarity of the profiles of gene expression in the four treatments. We identified three clusters (Fig. 6). The typical AT biosynthetic genes (VvCHI1, VvCHS3, VvLDOX, VvUFGT, and VvMYBA1) and the PA biosynthetic genes (VvLAR2, and VvMYBPA1) were clustered in a group (cluster A) with Vvchit1b, VvERF1, and VvACO1. The genes in this group were strongly induced by ethephon under light (+L+E) but not by ethephon under dark (−L+E), indicating that the transcriptional enhancement of these genes involving the ethylene biosynthetic and signal-transduction pathway accelerated the ethylene-induced accumulation of PAs and ATs in the presence of light. Additional PA biosynthetic genes, including VvLAR1, VvANR, VvMYB5a, and VvMYB5b, were clustered into another group (cluster B) with VvACO2. The genes in this cluster were up-regulated by ethephon treatment under light (+L+E) but were down-regulated by ethephon treatment under dark (−L+E) at 35 DAT, further supporting the differential regulation of exogenous ethephon under different light conditions in PA biosynthesis associated with endogenous ethylene biosynthesis. VvEIN3, the key transcription factor in ethylene signal transduction pathway, however, was clustered in a separate category, indicating that PA and AT synthesis had no relationship with VvEIN3 at the transcriptional level.

Fig. 6
figure 6

Hierarchical clustering of the transcript profiles of all genes. The relative expression levels of the genes after the four treatments compared to the reference gene, used in the original rank-based algorithm, were used for the hierarchical cluster analysis with Genesis. Red colors represent relatively higher transcript abundances, and yellow colors represent relatively lower transcript abundances. Sampling times and treatments are indicated at the top; the numbers 1, 2, 3, and 4 represent treatments +L+E, +L−E, −L+E, and −L−E, respectively, and the numbers 7, 17, 35, and 49 are days after treatment. See Fig. 2 for a description of the treatments. (Color figure online)

Full size image

Discussion

Grapes are classified as non-climacteric (Coombe and Hale 1973), but climacteric and non-climacteric fruit share similar pathways of ripening through the regulation of ethylene signaling (Barry and Giovannoni 2007). Ethylene does not influence berry weight and cell enlargement (El-Kereamy et al. 2000) and did not contribute to diameter increases in this study (Table 1). Ethephon treatment influenced degrees Brix (°Bx, total soluble solids), reducing-sugar content, titratable acidity, and the sugar/TA ratio, in accordance with the decrease in acidity (Schmidt et al. 2003) and higher sucrose content (Chervin et al. 2006) modulated by ethylene signals. Ethylene generally alters the ripening of berries (Chervin and Deluc 2010; Hale et al. 1970) but not growth (Table 1). Also, light irradiation significantly increased berry weight, diameter, and the maturity parameters at harvest (Table 1), as supported by previous reports (Bergqvist et al. 2001; Dokoozlian and Kliewer 1996; Smart et al. 1988). Ethylene and light signaling played an additional role in the import of sugars, as has also been previously reported (Atanassova et al. 2003; Fillion et al. 1999). Due to a combination of lower titratable acidity and roughly higher Brix levels, ethephon was significantly more effective under light than in the dark. Light was involved in the regulation of ripening by acting as an ethylene synergist, facilitating its role in berry ripening.

Ethylene and light had a consistent influence on the accumulation of phenolic compounds in the skins of the grape berries. Light exposure markedly increased the accumulation of PAs, TP, and TFO. Shading decreased the transcription of VvLAR1, VvLAR2, VvANR, and VvMYBPA1 by at least 5-fold (Fig. 3b–d, g). These results are in line with previous studies showing that artificial shading of clusters can decrease the transcription of some structural genes in the biosynthetic pathways of several phenolics such as PAs and flavonoids in the skins of grape berries (Jeong et al. 2004; Fujita et al. 2006) and that higher cluster exposure can increase phenolic contents (Cortell et al. 2007; Price et al. 1995; Ristic et al. 2007). Ethephon treatment showed similar patterns of accumulation. The total amounts of TP, TFO, and PAs in our study were significantly higher in the ethephon treatments (Figs. 2a, b, 3a), and we inferred that ethephon had a strong effect on the synthesis of PAs in the grapes. To address this point, we analyzed the expression profiles of the PA-relevant genes. Specific genes in the PA pathway were also expressed at significantly higher levels in the ethephon-treated berries, indicated by the 2- to 5-fold increased accumulation of VvLAR1, VvLAR2, VvANR, and VvMYBPA1 transcripts (Fig. 3b–d, g). VvMYB5a and VvMYB5b, however, were not up-regulated in the berries of the light (+L−E) and ethylene (−L+E) treatments (Fig. 3f, g), perhaps due to the timing of VvMYB5a and VvMYB5b expression, which occurs mainly during the early stages of berry development for the metabolism of PAs (Deluc et al. 2006, 2008).

The light and ethephon treatments also had similar patterns of accumulation of anthocyanin compounds. AT contents increased in ethephon- and light-treated berries (Fig. 4a). Increased or decreased light exposure, particularly on fruit clusters, can significantly influence the accumulation of ATs and total phenolics in grape berries (Koyama and Goto-Yamamoto 2008). Light-signaling components such as HY5 and PIF3 bind directly to the promoters of AT structural genes such as CHS, CHI, F3H, DFR, and LDOX (Shin et al. 2007). Light increased the transcription levels of VvCHI1, VvCHS2, VvCHS3, VvF3H2, VvDFR, VvLDOX, VvUFGT, and VvMYBA1, particularly VvCHS2, VvF3H2, VvLDOX, and VvUFGT, at least 5-fold (10-fold for VvCHS2) relative to the shading treatment (+L−E) (Fig. 4). AT biosynthesis is also associated with ethylene (El-Kereamy et al. 2003). Ethylene signaling is involved in the regulation of VvCHS, VvF3H, VvDFR, VvLDOX, and VvUFGT expression (El-Kereamy et al. 2000). The promoter regions of VvUFGT and VvMYBA1 contain putative ethephon-responsive cis-elements (Tira-Umphon et al. 2007). In our study, the transcription of VvCHI1, VvCHS3, VvDFR, VvLDOX, VvUFGT, and VvMYBA1 was up-regulated by ethephon treatment (−L+E) (Fig. 4). Light irradiation increased the transcriptional levels of VvF3H2 approximately 5-fold, but ethephon had little effect (Fig. 4e), which is in line with a previous study reporting that exogenous ethylene induced F3H activity but that ethephon did not affect its activity (El-Kereamy et al. 2003). Ethephon activated PA and AT biosynthesis, supporting our data that the berries treated with ethephon had correspondingly higher levels of TP and TFO due to the increase in the levels of these solutes (Fig. 2).

The interactive effect of light and ethylene on the indicators of grape ripening and PA and AT contents suggested that the activity of ethephon, which was tightly coupled to a cascade of molecular events of downstream responsive genes of the ethylene pathway, may be modulated by light. The interaction between these two signals could affect grape ripening and the accumulation of PAs and ATs in grape berries in two ways.

On the one hand, the more significant effect of ethephon under light was associated with the endogenous production of ethylene caused by light. More ethylene may be produced under light than in the dark (Robert et al. 1975). 1-aminocyclopropane-1-carboxylic acid oxidase (ACO) is a key rate-determining enzyme in the biosynthesis of ethylene. We did not determine the rates of ethylene release or the endogenous ethylene contents, but the expression of the ACO genes, particularly VvACO1, was affected by light at 7 DAT (Fig. 5d, e), suggesting that light acted as a positive regulator of the production of endogenous ethylene. Also, the transcription profiles of the PA- and AT-relevant genes were consistent with VvACO1 and VvACO2 (Fig. 6). The higher endogenous ethylene content in the ethephon-treated berries thus contributed to the significantly higher PA and AT contents under light than in the dark.

On the other hand, the ethylene-insensitive transcription factor EIN3 has emerged as a key player that is a positive regulator of ethylene signals (Chao et al. 1997; Solano et al. 1998). Light signals are involved in the EIN3/EIL1-mediated ethylene signal-transduction pathway (Lee et al. 2006; Zhong et al. 2009, 2012). In our study, the transcription of VvEIN3 was not up-regulated by either light or ethylene (Fig. 5a), and the transcription of AT-relevant genes was concomitant with the expression levels of Vvchit1b and VvERF1 (Fig. 6), which were the direct targets of EIN3 in the ethylene signal transduction pathway. The enhanced ethylene responsiveness and PA and AT contents by ethephon under light may thus be due to the increase in EIN3 levels (Smalle et al. 1997; Knee et al. 2000). Also, the observed interaction between light and ethylene on PA content was associated with EIN3 levels. Ethephon treatment up-regulated the transcription of VvACO2 under light but inhibited the expression of VvACO2 in the dark, especially at 35 DAT (Fig. 5e). Plants overexpressing VvACO2 have a dwarf phenotype, a typical phenotype of the 35S:EIN3 transgene (Chen et al. 2009), indicating a close relationship between VvEIN3 and VvACO2 (An et al. 2010; Lee et al. 2006). Correspondingly, the interaction between ethylene and light on PA content may possibly be concomitant with the up-regulation of VvACO2 and PA-relevant genes (Fig. 6) through a positive feedback by EIN3 on ACO2 (Liu et al. 1985). These reports support an interaction between light and ethylene signaling, so the interruption of ethylene responsiveness by light is partly due to the influence of light on ethylene signal transduction in non-climacteric fruits such as grape berries.

Conclusions

This study provides a comprehensive understanding of the effects of ethephon and light treatment on proanthocyanidin and anthocyanin biosynthesis in the skins of grape berries. Our data suggested that ethephon and light treatment induce the synthesis of proanthocyanidins by up-regulating biosynthetic genes. Light had a larger effect than ethephon on the regulation of anthocyanin biosynthesis. Light, by interacting with exogenous ethephon, can alter the endogenous biosynthesis of ethylene and downstream ethylene signaling and can thus strongly affect the ripening processes and the accumulation of anthocyanin and proanthocyanidin in grape berries.