Wheeler et al. Abscisic acid and grape berry ripening 195. CSIRO Plant Industry, PO Box 350, Glen Osmond, SA 5064, Australia 2

Transcription

1 Wheeler et al. Abscisic acid and grape berry ripening 195 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 S. WHEELER 1, *, B. LOVEYS 1, C. FORD 2 and C. DAVIES 1 1 CSIRO Plant Industry, PO Box 350, Glen Osmond, SA 5064, Australia 2 School of Agriculture, Food and Wine, PMB1, University of Adelaide, Glen Osmond, SA 5064, Australia *Present address: School of Agriculture, Food and Wine, PMB1, University of Adelaide, Glen Osmond, SA 5064, Australia. Corresponding author: Dr Christopher Davies, fax , christopher.davies@csiro.au Abstract Background and Aims: Grapevines (Vitis vinifera L.) are considered to have non-climacteric fruit, but the trigger initiating ripening (veraison) is poorly understood. This study aimed to further investigate the role of abscisic acid (ABA) during berry ripening. Methods and Results: In field-grown grapes over three seasons, free ABA levels increased at veraison then subsequently declined to low levels. Bound ABA increased as the free ABA level decreased after weeks post-flowering (wpf), but ABA must also be degraded and/or exported. The absence of a large pool of bound ABA before veraison makes it unlikely that the increase in free ABA is due to the mobilization of conjugated ABA. The expression pattern of genes crucial for ABA synthesis, zeaxanthin epoxidase and two 9-cis-epoxycartenoid dioxygenases (NCEDs) indicates that berries may have the potential to synthesize ABA in situ. However, the expression profile of these genes did not correlate well with ABA levels indicating that ABA accumulation is under more complex control. The application of (+)-ABA advanced ripening as measured by colour formation, berry size increase and to a lesser extent sugar accumulation and altered the expression of one of the NCED genes. Conclusions: The changes in berry ABA levels around the time of veraison, which influence the timing of ripening, are under complex developmental control. Significance of the Study: The improved understanding of the control of berry ripening is vital to attempts to successfully manipulate this process. Abbreviations 1-MCP 1-methylcyclopropene; ABA abscisic acid; BRs brassinosteroids; NCED 9-cis-epoxycartenoid dioxygenase; wpf weeks post-flowering; wps weeks post-spray; ZEP zeaxanthin epoxidase Keywords: ABA, berry ripening, biosynthesis, Vitis vinifera Introduction Physiologists commonly divide fleshy fruits into two groups, the so-called climacteric and non-climacteric fruits. Climacteric fruits are those that exhibit an increase in ethylene evolution and respiration at the commencement of ripening (Tucker 1993). There is a large body of knowledge regarding the synthesis and perception of ethylene and the effect it has on ripening-associated changes in gene expression and cell biochemistry (Adams-Phillips et al. 2004). In contrast, the control of ripening of nonclimacteric fruit, such as grapes, whose respiration rates do not exhibit large peaks at around veraison (defined in this paper as the commencement of ripening and judged to be at the last time point before a significant increase in Brix occurred) and that evolve low levels of ethylene compared with climacteric fruit, is less well understood. Although ethylene does not appear to be pivotal to ripening in non-climacteric fruit, as it is in climacteric fruit, it has become increasingly apparent that the division between climacteric and non-climacteric fruit is less clear cut than once thought. The hormonal control of grape berry ripening appears to be relatively complicated. A number of studies where ethylene or the ethylene-releasing compound 2- chloroethylphosphonic acid has been applied to grape berries have produced variable results, but in some cases, an increase in anthocyanin and/or sugar levels was observed (Hale et al. 1970, Coombe and Hale 1973, doi: /j x

2 196 Abscisic acid and grape berry ripening Australian Journal of Grape and Wine Research 15, , 2009 Szyjewicz et al. 1984, Roubelakis-Angelakis and Kliewer 1986, El-Kereamy et al. 2003, Delgado et al. 2004, Tira- Umphon et al. 2007). Ethylene does seem to play a role in the control of some genes expressed during nonclimacteric fruit ripening (Roubelakis-Angelakis and Kliewer 1986, El-Kereamy et al. 2003, Tesniere et al. 2004, Tira-Umphon et al. 2007). The treatment of berries with an inhibitor of ethylene perception, 1-methylcyclopropene (1-MCP), has been reported to inhibit grape berry ripening (Chervin et al. 2004). A decrease in the expression of two sucrose transporters and reduced sugar accumulation were also observed in 1-MCP-treated fruit (Chervin et al. 2006). Recently, it has been shown that grapes and strawberries (also considered to be nonclimacteric) exhibit an increase in ethylene evolution at ripening (Chervin et al. 2004, Iannetta et al. 2006). Furthermore, strawberries also appear to undergo a small rise in respiration at ripening (Iannetta et al. 2006). These results indicate that ethylene may play a role in grape berry ripening, but other factors are likely to be involved. Previous work has shown that auxin application can retard anthocyanin accumulation and, in some cases, ripening (Coombe and Hale 1973, Hale and Coombe 1974, Davies et al. 1997, Ban et al. 2003, Jeong et al. 2004). Brassinosteroids (BRs) also appear to play a role in the control of berry ripening (Symons et al. 2006). The application of epibl appeared to advance the timing of ripening, while the application of brassinazole (an inhibitor of BR synthesis) delayed the onset of ripening. In addition to ethylene and BRs, abscisic acid (ABA) has been implicated in the control of grape berry ripening. Most studies have shown that free ABA levels increase at about the time of veraison, although the reported timing of this increase in relation to veraison somewhat varies (Coombe and Hale 1973, Hale and Coombe 1974, Inaba et al. 1976, Scienza et al. 1978, Cawthon and Morris 1982, Kataoka et al. 1982, Davies et al. 1997, Kondo and Kawai 1998, Zhang et al. 2003, Okamoto et al. 2004, Baydar and Harmankaya 2005, Gagne et al. 2006). The increase in ABA levels at around the time of veraison is consistent with it having a role in the control of berry ripening. Some workers have shown that ABA application can enhance both sugar and anthocyanin accumulation (Hale and Coombe 1974, Düring et al. 1978, Matsushima et al. 1989). However, most reports show that ABA application to grape berries resulted mainly in an increase in colour rather than an advancement of sugar accumulation (Pirie and Mullins 1976, Kataoka et al. 1982, Yakushiji et al. 2001, Delgado et al. 2004, Mori et al. 2005, Cantin et al. 2007, Peppi et al. 2007, Peppi and Fidelibus 2008). The variable effects observed may be due to the timing of application in relation to veraison. To further complicate matters, interactions are likely to occur between the four putative hormonal effectors of berry ripening so far identified. It has been shown in other plants that there is considerable cross-talk between the various hormone signalling pathways as well as the involvement of sugars in signalling (Gazzarrini and McCourt 2003). Before we can understand the more complex interactions between hormones involved in ripening, we need to better understand the role of individual components such as ABA. Our understanding of the role of ABA in a number of plant developmental processes has steadily increased over the years, but the specific role of ABA during grape berry development, particularly in ripening, is still not clear. In this paper, the metabolism of ABA in grape berries was examined in detail throughout the development at both the biochemical and molecular level. The effect of (+)-ABA on the timing and progress of ripening was also examined in field-grown fruit. Materials and methods Berry sampling Berries were sampled over three growing seasons ( , , ) from Vitis vinifera L. cv Cabernet Sauvignon vines grown at the Slate Creek vineyard (35 16 S, E), Willunga, South Australia. Flowering was defined as the time when approximately 50% of the opercula had fallen from the flowers (50% cap-fall). Bunches were collected randomly from vines spread across two to four rows. Samples were taken at fortnightly intervals beginning 2 weeks after flowering (except during the season, when samples were collected weekly) and immediately frozen in liquid nitrogen. In and , all berries were deseeded prior to being frozen. Berries from the seasons were frozen whole, and skin, seed and flesh samples were obtained by peeling the frozen fruit and separating the seeds and flesh prior to extraction. All samples were stored at -80 C until required. To define the stage of berry development, 50 randomly selected berries were weighed and an average berry weight was calculated. The level of total soluble solids ( Brix) was measured in these fruits using a refractometer (model 10430, Reichert, Vienna, Austria). Free ABA analysis Quantitative determination of ABA levels was carried out by stable isotope dilution analysis (Loveys and van Dijk 1988). As an alternative to the ethyl acetate extraction, mg samples of homogenised frozen berries were column purified (Soar et al. 2006). As an internal standard, ng of ( )-3,5,5,7,7,7 -d 6 ABA was added. Gas Chromatography-Mass Spectrometry (GC-MS) was conducted using an Agilent 6890/5973 GC-MS system (Agilent, Santa Clara, CA, USA) essentially as described by Soar et al. (2004). Bound ABA analysis To determine the levels of bound ABA, samples were hydrolysed (Loveys and van Dijk 1988) prior to analysis. To achieve hydrolysis, the aqueous extractions were adjusted to ph 11.0 with 1 M NaOH, heated to 60 C for 1 h, and then adjusted to ph with 1 M HCl. The subsequent analysis was carried out as described above. The level of bound ABA was calculated by subtracting the amount of free ABA from the amount of total ABA (the amount of ABA present in the sample hydrolysed as described).

3 Wheeler et al. Abscisic acid and grape berry ripening 197 Table 1. Oligonucleotide primers for the qreal-time PCR analysis of Vitis vinifera Cabernet Sauvignon NCED 1 and 2 and ZEP gene expression. Primer name GenBank Accession No. Sequence Comment VvNCED1 GF2 AY gagaccccaactctggcagg3 NCED1 Forward primer VvNCED1 GR1 5 aaggtgccgtggaatccatag3 NCED1 Reverse primer VvNCED2 AF AY agttccatacgggtttcatggg 3 NCED2 Forward primer VvNCED2 AR 5 ccattttccaaatccagggtgt3 NCED2 Reverse primer VvZEPF AY taccgggtatttttgggaca3 ZEP Forward primer VvZEPR 5 cttcttcatccgtggcaagt3 ZEP Reverse primer VvUbiqF CF aacctccaatccagtcatctact3 Ubiquitin Forward primer VvUbiqR 5 gtggtattattgagccatcctt3 Ubiquitin Reverse primer Anthocyanin determination Grape berries, frozen in liquid nitrogen, were ground to a powder using an IKA A11 basic analytical mill (IKA, Staufen, Germany). To prepare samples for total anthocyanin determination, 0.1 g of powdered sample was added to 1 ml of methanol containing 1% (v/v) HCl and the anthocyanins were extracted at room temperature in the dark on a rotating mixer for 1 h. The tissue was pelleted by centrifugation at g for 15 min and the supernatant was retained. Total anthocyanins were measured spectrophotometrically by reading absorbance at 520 nm immediately following centrifugation. Preparation of grape RNA and cdna synthesis Grape RNA was isolated essentially as described by Davies and Robinson (1996). Total RNA was quantified using a spectrophotometer (A 260) and 100 mg was further purified and DNase treated using RNeasy mini spin columns (Qiagen, Hilden, Germany) and the RNase-Free DNase set (Qiagen) as described by the manufacturer. One microgram of purified RNA was run on an ethidium bromidestained agarose gel in 20% formaldehyde buffer to check for quality. cdna was synthesised in 20 ml reactions from 2 mg of total RNA using the SuperScript First-strand cdna synthesis system (Invitrogen, Carlsbad, CA, USA) and the 3 -end, (dt) 17-adpater primer (Frohman et al. 1988) according to the manufacturers instructions. cdna reactions were diluted tenfold before use in real time quantitative polymerase chain reaction (PCR) analyses. Quantitative real-time (qreal-time) PCR analysis Expression of VvNCED1, VvNCED2 and VvZEP genes was determined by qreal-time PCR using a Rotor Gene 3000 system (Corbett Research, Australia). The primers used are given in Table 1. Each reaction contained 1 SYBR Green PCR Master Mix (Thermo Scientific), 2 ml of diluted cdna, each primer at 333 nm and water to 20 ml. Cycling conditions were: 95 C, 10 min; 40 cycles: 95 C, 30 s; 58 C, 30 s; 72 C, 30 s, followed by a melt cycle of 1 C increments from 50 C to 96 C (45 s first step, 5 s for each subsequent step). Each primer pair gave a single product of the expected size and sequence, verified by analysis of the melt curve, agarose gel electrophoresis and DNA sequencing. All reactions were performed in triplicate. Standard DNA for the calibration curve of qreal-time PCR was prepared using each specific primer set as follows: first, amplified DNA, which formed a single band, was purified using a QIAquick PCR Purification Kit (Qiagen) according to the manufacturer s instructions; second, to determine the relative expression values, the purified standard DNA was diluted to appropriate concentrations to generate a standard curve for each product. A tenfold dilution series of each purified PCR product was established, qreal-time PCR was performed and arbitrary values were assigned to each standard according to the dilution factor. The expression value of each gene in each cdna tested, which was determined by reference to the standard curve, was normalised to the level of expression of ubiquitin in each cdna. ABA field experiment An experiment to assess the effect of exogenous (+)- ABA on berry development was conducted using Vitis vinifera L. cv Cabernet Sauvignon vines in a vineyard at Charleston, South Australia (34 55 S, E). Six vines in the same row were pruned to 20 bunches on each vine, and all bunches on each vine were sprayed with the same treatment. Three different treatments were applied: (i) 5 mg/l (+)-ABA in 0.5% (v/v) Tween 20; (ii) 400 mg/l (+)-ABA in 0.5% (v/v) Tween 20; and (iii) a control treatment of 0.5% (v/v) Tween 20. Powdered (+)-ABA (90% technical grade, Valent Bio- Sciences) was first dissolved in 1-mL 70% ethanol, which was then diluted to a final volume of 1 L with 0.5% (v/v) Tween 20. The treatments were replicated on separate vines. Bunches were sprayed until runoff with a hand-held sprayer in the late afternoon on 3 days (each a week apart) commencing on 21 January This corresponded to 3 weeks, 2 weeks and 1 week before veraison. For compositional studies, five berries were taken randomly from each of the sprayed bunches at 7, 14, 21, 28, 35, 42, 49, 60 and 75 days after the initial spray. Sampled berries were washed in 70% ethanol and rinsed twice in 0.5% (v/v) Tween 20 solution to reduce the levels of residual ABA on the

4 198 Abscisic acid and grape berry ripening Australian Journal of Grape and Wine Research 15, , 2009 a b Figure 1. Graph showing changes in Brix ( ), A520 nm ( ) and free abscisic acid (ABA) concentration ( ) during the development of Cabernet Sauvignon berries from the 2003/2004 season. Standard errors are indicated. c berry surface then frozen whole in liquid nitrogen. The separate skin and flesh samples were obtained by peeling the frozen fruit prior to extraction. All samples were stored at -80 C until required. Results Changes in free and bound ABA in relation to berry ripening By measuring changes in Brix and absorbance at 520 nm of an acidic methanol extract, a profile of berry development was produced for Cabernet Sauvignon fruit grown in the 2003/2004 season (Figure 1). Free ABA levels were also measured fortnightly throughout development, beginning at 2 wpf, with additional, weekly, measurements being made between 6 and 12 wpf. After being high in young fruit, ABA levels decreased until 7 wpf. From 8 wpf, free ABA levels began to increase sharply, coincident with an increase in berry skin colour due to anthocyanin accumulation (A 520), and the accumulation of sugars ( Brix). Veraison was defined as the time of the last sample prior to an increase in Brix and was at 8 wpf. After peaking at 10 wpf, the level of free ABA decreased to be low at commercial ripeness, 18 wpf. To determine the effect of varying seasonal conditions on the profile of ABA accumulation in berries, free and bound ABA levels were quantified in fruit grown in the 2000/2001, 2002/2003 and 2003/2004 seasons. The timing and pattern of free ABA accumulation was remarkably stable over the 3 years (Figure 2a,b,c). Although the vines used were irrigated, the weather conditions were quite different for these seasons. The fact that the pattern of ABA accumulation was similar in all 3 years indicates that the pattern of free ABA accumulation during berry development is, in the main, developmentally controlled. By plotting the ABA levels on a per berry basis, it can be seen that there is little, if any, nett accumulation of any form of ABA until after veraison at 8 wpf (Figure 2d, only 2003/2004 data are shown but a similar pattern occurred in all 3 years). Bound ABA levels were only a small portion of the total ABA levels in all seasons until the later stages of berry development (Figure 2a,b,c). At harvest, the level of Figure 2. Changes in free ( ) and bound ( ) abscisic acid (ABA) forms throughout Cabernet Sauvignon grape berry development. Veraison (defined as the last sample before the increase in Brix associated with ripening, data not shown) is indicated by an arrow. (a): ABA ng/g for the 2000/2001 season; (b): ABA ng/g for the 2002/2003 season; (c): ABA ng/g for the 2003/2004 season; (d): ABA ng/berry for the 2003/2004 season. bound ABA was equal to, or greater than, the level of free ABA. The sequestration of free ABA into the bound form accounted for nearly 32% of the total decrease in free ABA concentration between weeks 11 and 14 wpf in the 2003/2004 season. In addition, nearly 20% of the decrease in concentration was due to the increase in berry size during this period. In terms of ABA per berry, the amount of free ABA at 11 wpf (582 ng) is reduced to 203 ng at 14 wpf partly due to the conversion of 182 ng of ABA to the bound form (Figure 2d). The remainder of the decrease in free ABA levels must be either due to breakdown via enzymic catabolism, for example to phaseic acid, or due to export from the berry (Cutler and Krochko 1999). ABA biosynthesis pathway gene expression during berry development The transcript level of three ABA biosynthesis genes in deseeded berry tissue was measured throughout development by qreal-time PCR analysis (Figure 3). Southern d

5 Wheeler et al. Abscisic acid and grape berry ripening 199 a b c Figure 4. Concentration of abscisic acid (ABA) in skin, flesh and seed tissues throughout Cabernet Sauvignon berry development during the 2003/2004 season. The dotted line represents the ABA level of deseeded fruit. Veraison is indicated by an arrow. Figure 3. Relative transcript level profiles for abscisic acid biosynthesis pathway genes throughout Cabernet Sauvignon berry development over three seasons (2000/2001 ( ), 2002/2003 ( ), 2003/ 2004 ( )). cdna was made using RNA from deseeded berries. (a) ZEP, (b) NCED1, (c) NCED2. Veraison is indicated by a dashed vertical line. blot analysis (Soar et al. 2004) database searches and PCR using degenerate primers to try to isolate other family members (data not shown) showed that grapevine contains a single ZEP gene (VvZEP) and two NCED genes (VvNCED1 and VvNCED2). In general, the expression patterns for ZEP followed smooth trends across the 3 years studied having a single main peak (Figure 3a). The patterns of ZEP transcript accumulation for the 2000/2001 and 2002/2003 seasons were quite similar to each other, but the peak in ZEP expression appeared to be earlier in the fruit from the 2003/2004 season. ZEP transcript levels were at their lowest in the later stages of ripening. Compared to ZEP, the expression pattern of both NCED genes was more varied within each season (Figure 3b,c). The expression of both the NCED genes was also more variable between seasons, with the only consistent feature being the very low levels of expression found in the 4 weeks prior to harvest. Gene transcription profiling by micro-array analysis (C. Davies, P. Iocco, M. Thomas, unpubl. data, 2005) has confirmed the pattern of gene expression for VvNCED1 and VvZEP (probe sequences representing VvNCED2 are not present on the currently available Grapevine Affymetrix chip) for the 2003/2004 season s fruit and has also shown that there is considerable variation in VvNCED1 and VvZEP transcript levels between fruit grown at different sites. The expression level of VvNCED1 was considerably higher than that of VvNCED2 (data not shown). ABA accumulation in different tissues during grape berry development Subsamples of the berries assayed in Figure 1 were separated into seed, skin and flesh portions and ABA content was measured for each tissue type (Figure 4). As the flesh contributes the greater proportion of mass to the fruit, the level of ABA in the flesh reflected quite closely the levels found in deseeded fruit (dotted line). The skin forms only a relatively small percentage of the total berry weight. The ABA levels in the skin followed a similar pattern to that seen in the flesh. Free ABA levels were higher in the seed than in the skin and flesh except for the sample taken at 10 wpf where the concentration of ABA is similar in all three tissues. The levels in skin and seed were about equal in the 12 wpf sample. As with the skin, the seeds comprise a relatively small portion of the berry volume, particularly later in development when the pericarp expands considerably while the seed volume remains constant. The effect of application of the natural enantiomer of ABA on Cabernet Sauvignon berry ripening (+)-ABA was sprayed onto bunches of Cabernet Sauvignon fruit on field-grown vines at two concentrations (5 and 400 mg/l). The 400-mg/L treatment appeared to induce more rapid anthocyanin accumulation than the 5-mg/L treatment; however, both appeared to enhance colour development. This was reflected in both the percentage of coloured berries each week (Figure 5a) and in the change in absorbance (measured at 520 nm, Figure 5b). These data showed that (+)-ABA treatment, especially the 400-mg/L treatment, increased anthocyanin levels earlier in development compared with the control treatment. The levels of anthocyanins in the (+)-ABAtreated samples did not increase significantly over the period covered by the last three samples before harvest, and the level of anthocyanins in the ABA-treated and control samples were similar during this period (Figure 5b).

6 200 Abscisic acid and grape berry ripening Australian Journal of Grape and Wine Research 15, , 2009 a a b b c c d d Figure 5. Comparison of the progress of berry development in control and (+)-ABA-treated fruit. Cabernet Sauvignon fruit were treated with either solvent alone (Control, ) or treated with 5 mg/l ( ) or 400 mg/l (+)-ABA ( ). (a) colour development as measured by % coloured berries, (b) colour as measured by A520 nm, (c) mean berry weight, (d) total soluble solids as measured by Brix. ABA, abscisic acid. Measurements of berry weight indicated that ABA treatment promoted the typical berry size increase observed after veraison (Figure 5c). The 400-mg/L treatment advanced berry growth such that the expected lag phase observed in the untreated sample was not detectable. The lag phase, during which berry size remains constant, normally occurs between the two phases of rapid berry growth and was only clearly observed in the control fruit (Figure 5c). Brix measurements also indicated that sugar accumulation was enhanced by ABA treatment at the higher ABA concentration (Figure 5d). Again, the difference between the treated and control fruit decreased as ripening progressed. ABA accumulation in ABA-treated fruit and its effect on the expression of ABA biosynthesis genes At the time of sampling, berries were washed with 70% ethanol and a detergent solution to remove any Figure 6. Abscisic acid (ABA) concentrations and ABA biosynthesis gene transcript levels in ABA-treated fruit tissues. Free ABA levels in the skin and flesh of berries treated with 400 mg/l ABA (a). The relative transcript levels of ZEP (b), NCED1 (c) and NCED2 (d) were determined by qreal-time PCR analysis. Berries were sampled two weeks after the first spray, washed then separated before analysis. FA, flesh ABA-treated; FC, flesh control; SA, skin ABA-treated; SC, skin control. unabsorbed ABA resulting from the spray treatment. Two weeks after the first treatment (1 week after the second treatment), there was a significant increase in ABA levels in the ABA-treated tissues compared with the levels in the control tissue samples (Figure 6a). The levels in the ABA-treated skin tissue were the highest, but the difference in ABA levels between the control and treated fruit was greater in the flesh (7.5 ) than in the skin (4.3 ). qreal-time PCR analysis revealed that ZEP transcript levels were reduced in the skin and flesh of ABA-treated samples at 2 weeks post the first spray (2 wps) (Figure 6b). NCED1 transcript levels were much higher in both skin and flesh treated with ABA (Figure 6c). NCED2 transcript levels were increased in treated skin and flesh samples (Figure 6d). The increase in NCED2 transcript level was greater in flesh compared with skin.

7 Wheeler et al. Abscisic acid and grape berry ripening 201 Discussion ABA is most commonly thought of as a hormone associated with processes such as the response to abiotic stress and the maintenance of dormancy. However, it is clear that ABA action is also implicated in many other processes including fruit ripening (Coombe 1976, Finkelstein and Rock 2002). There is evidence in grapes, for example, that ABA application can hasten the initiation of ripening (this work, Hale and Coombe 1974, Düring et al. 1978, Matsushima et al. 1989). However, the mechanism of ABA action and its role during berry development (ripening in particular) are still not well understood. Most previous studies have used a racemic mixture of (+) and (-) enantiomers, which complicates the interpretation of ABA s effects on the induction of ripening. It has been shown that the response to, and metabolism of, (+) and (-) enantiomers of ABA can vary (Loveys and Milborrow 1984, Huang et al. 2007). In the field-based experiments described in this paper, only the natural, (+) enantiomer of ABA has been used. The data presented in Figure 5 show that ABA treatment advanced berry ripening. Given that both the concentration of sugars and the volume of the berry were advanced, it suggests that there is a considerably enhanced flow of carbon into the berry directly after ABA application. However, as development proceeded, values for the colour, size and Brix of the treated fruit tended to asymptote as they approached harvest and the levels in the final samples were similar for the ABA treatments and controls. ABA applications later in development may maintain the difference in levels of some substances, e.g. anthocyanins, between the treatments. However, osmotic considerations and limits to the extent of berry expansion may prevent further size increases and sugar accumulation. ABA-treated fruit may approach this limit earlier than the control fruit due to the ABA-induced advancement of ripening. From a practical point of view, the temporal advancement of ripening may be an asset to viticulturists as it means that fruit may mature earlier, a distinct advantage in cooler areas where an early end to the growing season may prevent adequate fruit maturation. Interestingly, it appears that the ABA treatment used in this study induced a fully coordinated, complete ripening process as the change in all the parameters of ripening measured (anthocyanin accumulation, berry weight increase and the increase in total soluble solids) occurred at the same time (Figure 5). Previous experiments using ethylene show that the response to this hormone, in terms of colour and sugar accumulation, can be variable and may not be coordinately regulated (Szyjewicz et al. 1984). There are reports suggesting that a similar phenomenon can occur when ABA is applied to grape berries. Hale and Coombe (1974) claimed that the timing of hormone application in relation to veraison was a critical factor in determining any effects. Some studies have shown that ABA treatment can induce both skin colouration (anthocyanin accumulation) and sugar accumulation (Matsushima et al. 1989, Hiratsuka et al. 2001). Other workers have reported that ABA application enhanced anthocyanin accumulation but had little effect on sugar accumulation (Kataoka et al. 1982, Yakushiji et al. 2001, Delgado et al. 2004, Mori et al. 2005, Peppi et al. 2007). The timing of hormone application and the timing of measurements of berry parameters are likely to be crucial to the observed outcome. In addition, the experiments mentioned above were conducted using a range of cultivars and there may be genetically determined differences in sensitivity to hormones. It remains to be seen how hastening ripening through ABA application will affect berry composition and the levels of flavour and aroma compounds. The uptake of ABA into the berry through the waxy cuticle is likely to be an inefficient process as this structure forms an effective barrier (Casado and Heredia 1999). However, ABA levels 2 wps were greatly increased in the skin and flesh samples from ABA-treated fruit indicating that either a considerable amount of the sprayed ABA entered into the berry, or the spraying of ABA stimulated ABA accumulation in the treated berries. The analysis of ZEP and NCED gene expression in ABA-treated and untreated fruit showed that VvNCED1 and VvNCED2 transcript levels were elevated in skin and flesh tissues 2 weeks after the first treatment (Figure 6c,d). These tissues also had higher levels of ABA (Figure 6a). This suggests that externally applied ABA may feed forward to elevate the level of ABA biosynthesis pathway gene transcripts and hence its own synthesis. If this is so, then it may only take a small amount of exogenous ABA to enter the berry in order to trigger an increase in ABA synthesis and hence advance berry ripening. As the natural form of ABA was applied to the berries, it was not possible to distinguish between applied ABA and that synthesized within the berry. This apparent feedforward effect may be similar to the autocatalytic production of ethylene in climacteric fruit (Alexander and Grierson 2002). The differential response in the accumulation of ZEP and NCED transcripts may reflect the belief that NCED catalyses the key step in the pathway most responsive to environmental and developmental cues, while ZEP expression is more crucial to the control of carotenoid synthesis (Finkelstein and Rock 2002). Two lines of evidence suggest that not only can exogenous ABA stimulate ripening but ABA may also be crucially involved in controlling ripening during the growth of untreated berries. First, the large increase in free ABA concentration that occurred approximately halfway through the course of berry development coincided with the onset of ripening (Figure 1). Second, a number of genes expressed at veraison (some of which have ripening-related functions) appear to be regulated by ABA (see below). It is not known whether the peak in ABA is a consequence of the initiation of ripening or is part of the mechanism essential for triggering ripening. ABA accumulation may occur as a result of ripening due to the stress caused by the rapid increase in berry sugar levels within the berry. Increased sugar levels may result in stress due to reduced water availability and the consequent osmotic considerations. An alternative explanation is that ABA is the trigger for ripening as ABA application

8 202 Abscisic acid and grape berry ripening Australian Journal of Grape and Wine Research 15, , 2009 certainly seems to advance the onset of ripening (Figure 5). However, although ABA application can trigger berry ripening, it is not necessarily the primary initiator in untreated fruit. Whatever the role of ABA, it appears that high levels are only required transiently and that both the increase in ABA levels from veraison and their subsequent decrease a few weeks later are developmentally controlled. Although sugars and ABA begin to accumulate at about the same time, there is no direct correlation between their levels later in berry development. As can be seen from Figure 1, hexoses are still being accumulated in berries after the peak in free ABA levels. The imported sugars themselves may influence berry ripening as glucose and sucrose are known to be involved in signalling, both in their own and through interaction with hormones, in particular ABA (Cakir et al. 2003, Gibson 2005). It is currently not known where the ABA that accumulates in berries is synthesized. The distribution of free ABA in different berry tissues throughout development generally followed the same pattern (i.e. elevated for a few weeks after veraison, Figure 4). ABA levels were somewhat higher in seed than in either skin or flesh at most time points, and it is possible that the seeds could be the source of at least some of the ABA found in the flesh and skin, although evidence is currently lacking for this. Some authors have suggested that ABA is made in the leaves and transported to the berries along with sucrose after veraison (Antolin et al. 2003). However, Okamoto et al. (2004) showed that berry ABA levels can vary independently of leaf levels during ripening. Although stressing vines may elevate ABA levels throughout the plant, it may still be that ABA is produced within the berry as a response to altered water status rather than imported from the rest of the plant. Free ABA is thought by many to be the active form of the hormone. Opinion is divided on the role of the bound (conjugated) forms of ABA such as the glucosyl ester (Nambara and Marion-Poll 2005). It may be involved in long distance transport within the plant and may, or may not, act as a store of ABA that can be released when required. The release of ABA from a pool of the bound form seems unlikely to occur in grapes as at no time during the preveraison stage of berry development was there a significant pool of bound ABA whose liberation could account for the observed increase in free ABA at veraison, this is best seen in the ABA/ berry plot (Figure 2d). There is no direct evidence to support the release of free ABA from the conjugated form in grapevine. The expression pattern of genes involved in ABA synthesis might be expected to give some indication of the site of synthesis of the ABA that accumulates in berries. ZEP and NCED are two genes known to be crucial to ABA synthesis in other plants. In grapes, a single ZEP gene and two NCED genes were expressed throughout berry development (Figure 3) and, therefore (depending on the expression of the genes encoding the other enzymes), the berries may have the potential to make ABA in situ (Figure 3). However, the patterns of expression of these genes were not closely correlated with the changes in ABA levels that occur during berry development during a season. There is also considerable variation in the gene expression patterns between seasons despite the similar patterns of ABA accumulation (compare Figures 2,3). As the accumulation of ABA was not directly correlated to ZEP and NCED mrna levels, its origin remains unresolved and so the ABA present in berries could arise from multiple sources. The second line of evidence that ABA is involved in the ripening process in untreated berries comes from studies that indicate that some genes expressed at, or after, veraison are either activated by ABA or have homologues that are activated by ABA in other plant species. As mentioned in the introduction, ABA appears to influence the expression of genes in the anthocyanin pathway. The transcription of genes and activity of proteins involved in sugar accumulation and metabolism during ripening are also influenced by ABA (Cakir et al. 2003, Pan et al. 2005, Yu et al. 2006). Sugar metabolism is central to the ripening process not only because of the substantial levels of sugars that are stored but also because of their potential as signal molecules and the indirect changes resulting as a consequence of their osmotic potential. The role of hormones in the control of grape berry ripening appears to be quite complex. Notwithstanding the possible interaction between ABA and sugars, there is evidence that two other growth hormones may play a role in triggering berry ripening. Both ethylene (Szyjewicz et al. 1984) and BRs (Symons et al. 2006) have been shown to have a positive influence on grape berry ripening. The interactions between these hormones and sugars during the initiation and progression of ripening need to be further investigated to determine their relative roles during ripening in this non-climacteric fruit. Acknowledgements The authors would like to thank Ms Katie Spackman for technical assistance, Chalk Hill Wines, Willunga and Nepenthe Wines, Hahndorf for their generosity in providing experimental material and Dr Paul Boss for helpful discussions and comments on the manuscript. The primers used for Real Time PCR analysis of grape ZEP were designed by Dr Jim Speirs. The natural isomer of ABA used in these experiments was a kind gift from Dr Andrew Rath (Valent BioSciences Australia). Ms Wheeler gratefully acknowledges the support by The Grape and Wine Research and Development Corporation in the form of a postgraduate scholarship. References Adams-Phillips, L., Barry, C. and Giovannoni, J. (2004) Signal transduction systems regulating fruit ripening. Trends in Plant Science 9, Alexander, L. and Grierson, D. (2002) Ethylene biosynthesis and action in tomato: A model for climacteric fruit ripening. Journal of Experimental Botany 53, Antolin, M., Baigorri, H., Luis, I., Aguirrezabal, F., Geny, L., Broquedis, M. and Sanchez-Diaz, M. (2003) ABA during reproductive development in non-irrigated grapevines (Vitis vinifera L. cv. Tempranillo). Australian Journal of Grape and Wine Research 9,

9 Wheeler et al. Abscisic acid and grape berry ripening 203 Ban, T., Ishimaru, M., Kobayashi, S., Shiozaki, S., Goto- Yamamoto, N. and Horiuchi, S. (2003) Abscisic acid and 2,4- dichlorophenoxyacetic acid affect the expression of anthocyanin biosynthetic pathway genes in Kyoho grape berries. Journal of Horticultural Science and Biotechnology 78, Baydar, N.G. and Harmankaya, N. (2005) Changes in endogenous hormone levels during the ripening of grape cultivars having different berry set mechanisms. Turkish Journal of Agriculture and Forestry 29, Cakir, B., Agasse, A., Gaillard, C., Saumonneau, A., Delrot, S. and Atanassova, R. (2003) A grape ASR protein involved in sugar and abscisic acid signalling. The Plant Cell 15, Cantin, C.M., Fidelibus, M.W. and Crisostoc, C.H. (2007) Application of abscisic acid (ABA) at veraison advanced red color development and maintained postharvest quality of Crimson Seedless grapes. Postharvest Biology and Technology 46, Casado, C.G. and Heredia, A. (1999) Structure and dynamics of reconstituted cuticular waxes of grape berry cuticles (Vitis vinifera L.). Journal of Experimental Botany 50, Cawthon, D. and Morris, J. (1982) Relationship of number and maturity to berry development, fruit maturation, hormonal changes and uneven ripening of Concord (Vitis labrusca L.) grapes. Journal of the American Society for Horticultural Science 107, Chervin, C., Terrier, N., Ageorges, A., Ribes, F. and Kuapunyakoon, T. (2006) Influence of ethylene on sucrose accumulation in grape berry. American Journal of Enology and Viticulture 57, Chervin, C., El-Kereamy, A., Roustan, J., Latche, A., Lamon, J. and Bouzayen, M. (2004) Ethylene seems required for the berry development and ripening in grape, a non-climacteric fruit. Plant Science 167, Coombe, B. (1976) The development of fleshy fruits. Annual Review of Plant Physiology 27, Coombe, B. and Hale, C. (1973) The hormone content of ripening grape berries and the effect of growth substance treatment. Plant Physiology 51, Cutler, A. and Krochko, J. (1999) Formation and breakdown of ABA. Trends in Plant Science 4, Davies, C. and Robinson, S. (1996) Sugar accumulation in grape berries: cloning of two putative vacuolar invertase cdnas and their expression in grapevine tissues. Plant Physiology 111, Davies, C., Boss, P. and Robinson, S. (1997) Treatment of grape berries, a nonclimacteric fruit with a synthetic auxin, retards ripening and alters the expression of developmentally regulated genes. Plant Physiology 115, Delgado, R., Gallegos, J.I., Martin, P. and Gonzalez, M.R. (2004) Influence of ABA and ethephon treatments on fruit composition of Tempranillo grapevines. Proceedings 26th IHC, viticulture living with limitations. Eds. A.G. Reynolds and P. Bowen. Acta Horticulturae 640, Düring, H., Alleweldt, G. and Koch, R. (1978) Studies on hormonal control of ripening in berries and grape vines. Acta Horticulturae 80, El-Kereamy, A., Chervin, C., Roustan, J.P., Cheynier, V., Souquet, J.M., Moutounet, M., Raynal, J., Ford, C., Latche, A., Pech, J.C. and Bouzayen, M. (2003) Exogenous ethylene stimulates the long-term expression of genes related to anthocyanin biosynthesis in grape berries. Physiologia Plantarum 119, Finkelstein, R.R. and Rock, C.D. (2002) Abscisic acid biosynthesis and response. In: The Arabidopsis book. Eds. C.R. Sommerville and E.M Meyerowitz (American Society of Plant Biologists: Rockville, MD) pp Frohman, M., Dush, M. and Martin, M. (1988) Rapid production of full-length cdnas from rare transcripts: Amplification using a single gene-specific oligonucleotide primer. Proceedings of the National Academy of Sciences of the United States of America 85, Gagne, S., Esteve, K., Deytieux, C., Sauvier, C. and Geny, L. (2006) Influence of abscisic acid in triggering véraison in grape berry skins of Vitis vinifera L. cv. Cabernet Sauvignon. Vitis 40, Gazzarrini, S. and McCourt, P. (2003) Cross-talk in plant hormone signalling: What Arabidopsis mutants are telling us. Annals of Botany 91, Gibson, S.I. (2005) Control of plant development and gene expression by sugar signalling. Current Opinions in Plant Biology 8, Hale, C. and Coombe, B. (1974) Abscisic acid An effect on the ripening of grapes. Royal Society of New Zealand Bulletin 12, Hale, C.R., Coombe, G.B. and Hawker, J.S. (1970) Effects of ethylene and 2-chloroethylphosphonic acid on the ripening of grapes. Plant Physiology 45, Hiratsuka, S., Onodera, H., Kawai, Y., Kubo, T., Itoh, H. and Wada, R. (2001) ABA and sugar effects on anthocyanin formation in grape berry cultured in vitro. Scientia Horticulturae 90, Huang, D., Jaradat, M.R., Wu, S., Ambrose, S.J., Ross, A.R., Abrams, S.R. and Cutler, A.J. (2007) Structural analogs of ABA reveal novel features of ABA perception and signaling in Arabidopsis. The Plant Journal 50, Iannetta, P.P.M., Laarhoven, L.-J., Medina-Escobar, N., James, E.K., McManus, M.T., Davies, H.V. and Harren, F.J.M. (2006) Ethylene and carbon dioxide production by developing strawberries show a correlative pattern that is indicative of ripening climacteric fruit. Physiologia Plantarum 127, Inaba, A., Ishida, M. and Sobajima, Y. (1976) Changes in endogenous hormone concentrations during berry development in relation to ripening of Delaware grapes. Journal of the Japanese Society for Horticultural Science 45, Jeong, S., Goto-Yamamoto, N., Kobayashi, S. and Esaka, M. (2004) Effects of plant hormones and shading on the accumulation of anthocyanins and the expression of anthocyanin biosynthetic genes in grape berry skins. Plant Science 167, Kataoka, I., Sugiura, A., Utsunomiya, N. and Tomana, T. (1982) Effect of abscisic acid and defoliation on anthocyanin accumulation in Kyoho grapes. Vitis 21, Kondo, S. and Kawai, M. (1998) Relationship between free and conjugated ABA levels in seeded and gibberellin-treated seedless, maturing Pione grape berries. Journal of the American Society for Horticultural Science 123, Loveys, B. and van Dijk, H. (1988) Improved extraction of abscisic acid from plant material. Australian Journal of Plant Physiology 15, Loveys, B.R. and Milborrow, B.V. (1984) Metabolism of abscisic acid. In: Biosynthesis and metabolism of plant hormones. Eds. A. Crozier and J.R. Hillman (Cambridge University Press: Cambridge, UK) pp Matsushima, J., Hiratsuka, S., Taniguchi, N., Wada, R. and Suzaki, N. (1989) Anthocyanin accumulation and sugar content in the skin of grape cultivar Olympia treated with ABA. Journal of the Japanese Society for Horticultural Science 58, Mori, K., Saito, H., Goto-Yamamoto, N., Kitayama, M., Kobayashi, S., Sugaya, S., Gemma, H. and Hashizume, J. (2005) Effects of abscisic acid treatment and night temperatures on anthocyanin composition in Pinot noir grapes. Vitis 44, Nambara, E. and Marion-Poll, A. (2005) Abscisic acid biosynthesis and catabolism. Annual Review of Plant Biology 56, Okamoto, G., Kuwamura, T. and Hirano, K. (2004) Effects of water deficit stress on leaf and berry ripening in Chardonnay grapevines (Vitis vinifera). Vitis 43, Pan, Q.-H., Li, M.-J., Peng, C.-C., Zhang, N., Zou, X., Zou, K.-Q., Wang, X.-L., Yu, X.-C., Wang, X.-F. and Zhang, D.-P. (2005) Abscisic acid activates acid invertases in developing grape berry. Physiologia Plantarum 125, Peppi, M.C. and Fidelibus, M.W. (2008) Effects of forchlorfenuron and abscisic acid on the quality of Flame Seedless grapes. Hort- Science 43, Peppi, M.C., Fidelibus, M.W. and Dokoozlian, N.K. (2007) Application timing and concentration of abscisic acid affect the quality

10 204 Abscisic acid and grape berry ripening Australian Journal of Grape and Wine Research 15, , 2009 of Redglobe grapes. Journal of Horticultural Science and Biotechnology 82, Pirie, A. and Mullins, M.G. (1976) Changes in anthocyanin and phenolics content of grapevine leaf and fruit tissues treated with sucrose, nitrate, and abscisic acid. Plant Physiology 58, Roubelakis-Angelakis, K.A. and Kliewer, W.M. (1986) Effects of exogenous factors on phenylalanine ammonia-lyase activity and accumulation of anthocyanins and total phenolics in grape berries. American Journal of Enology and Viticulture 37, Scienza, A., Mieavalle, R., Visai, C. and Fregoni, M. (1978) Relationships between seed number, gibberellin and abscisic acid levels and ripening in Cabernet Sauvignon grape berries. Vitis 17, Soar, C., Speirs, J., Maffei, S., Penrose, A. and Loveys, B. (2004) Gradients in stomatal conductance, xylem sap ABA and bulk leaf ABA along canes of Vitis vinifera cv. Shiraz: Molecular and physiological studies investigating their source. Functional Plant Biology 31, Soar, C., Speirs, J., Maffei, S., Penrose, A., McCarthy, M. and Loveys, B. (2006) Grape vine varieties Shiraz and Grenache differ in their stomatal response to VPD: Apparent links with ABA physiology and gene expression in leaf tissue. Australian Journal of Grape and Wine Research 12, Symons, G., Davies, C., Shavrukov, Y., Dry, I., Reid, J. and Thomas, M. (2006) Grapes on steroids. Brassinosteroids are involved in grape berry ripening. Plant Physiology 140, Szyjewicz, E., Rosner, N. and Kliewer, W.M. (1984) Ethephon ((2-chloroethyl)phosphonic acid, Ethrel, CEPA) in viticulture A review. American Journal of Enology and Viticulture 35, Tesniere, C., Pradal, M., El-Kereamy, A., Torregrosa, L., Chatelet, P., Roustan, J.-P. and Chervin, C. (2004) Involvement of ethylene signalling in a non-climacteric fruit: New elements regarding the regulation of ADH expression in grapevine. Journal of Experimental Botany 55, Tira-Umphon, A., Roustan, J.P. and Chervin, C. (2007) The stimulation by ethylene of the UDP glucose-flavonoid 3-Oglucosyltransferase (UFGT) in grape tissues is independent from the MybA transcription factors. Vitis 46, Tucker, G. (1993) Introduction. In: Biochemistry of fruit ripening. Eds. G.B. Seymour, J.E. Taylor and G.A. Tucker (Chapman and Hall: London) pp Yakushiji, H., Morinaga, K. and Kobayashi, S. (2001) Promotion of berry ripening by 2,3,5-triiodobenzoic acid in Kyoho grapes. Journal of the Japanese Society for Horticultural Science 70, Yu, X.-C., Li, M.-J., Gao, G.-F., Feng, H.-Z., Geng, X.Q., Peng, C.-C., Zhu, S.-Y., Wang, X.-J., Shen, Y.-Y. and Zhang, D.-P. (2006) Abscisic acid stimulates a calcium-dependent protein kinase in grape berry. Plant Physiology 140, Zhang, X.R., Luo, G.G., Wang, R.H., Wang, J. and Himelrick, D.G. (2003) Growth and developmental responses of seeded and seedless grape berries to shoot girdling. Journal of the American Society for Horticultural Science 128, Manuscript received: 31 July 2008 Revised manuscript received: 18 November 2008 Accepted: 25 November 2008