To evaluate the effect of GA, CPPU, and GA + CPPU treatment at berry set on the volatiles composition, berries of ‘Sable’ were sampled 51 and 70 d after the treatments. GC-MS analysis allowed identification of 58 and 67 volatile compounds with quantitative data calculated relative to the internal standard . Analysis of the differences in volatile content by ANOVA showed that 62 of the 68 compounds present at both time points differed in content by treatments or time . PCA indicated strong separation of the volatile compounds according to the harvest time with PC1 at 49.6% . PC2 contributed 26.5% of the total variance and reflected the effect of CPPU at 51 d, but at 70 d there was convergence of the variance and overall differences among the treatments seemed insignificant. VIP score analysis for the volatile compounds at 51 d pointed to 15 compounds having significant scores . CPPU suppressed the levels of the monoterpenes, α-terpineol, linalool, and limonene,plastic plant containers as well as the carotenoid derivative, β-cyclocitral and the fatty acid derivative, hexanal. However, GA + CPPU increased the levels of these compounds. The levels of 2-hexenal and pentanal were higher in CPPU-treated berries while GA increased the level of methyl geranate.
At 70 d, CPPU increased and GA or GA + CPPU decreased the levels of decanal and 5 other compounds . In contrast, GA increased the levels of trans and cis-rose oxide, β-bourbonene, and methyl geranate. Decanal, methyl geranate, ethyl acetate, and α-terpineol were common among the two time points and the first two compounds showed the same pattern.PGRs may have multiple immediate effects on gene expression, but, in this study, we were interested in understanding the long-term effects of GA and CPPU on the berries at harvest. To this end, RNA-seq was carried out on ‘Sable’ berries at 51 and 70 d after the treatments, the time points that can be defined as early and late ripening with respect to the untreated control.Venn diagram analysis of control vs. GA showed that there were hardly any differences between the control and GA treatments either at the early or late ripening, but there were significant differences in gene expression among the ripening stages . In contrast, the CPPU treatment imposed significant differences on gene expression at either the early or late stages of ripening . Six main clusters representing gene expression patterns were identified . KEGG pathway enrichment showed that only cluster 3 and cluster 4 contained significantly enriched pathways . Six out of the 13 processes related to cluster 3 were assigned to the phenylpropanoid pathway. The first process registered for cluster 3 was ‘circadian rhythms’ but analysis of the relevant KEGG map suggested that all the 20 enriched genes in this bin were actually family members of chalcone synthase which is a central flavonoid pathway.
The ‘ubiquinone and other terpenoid-quinone biosynthesis’ bin included 4CL, trans-cinnamate 4-monooxygenase, and isochorismate synthase. Other bins of cluster 3, excluding the large unspecific bins of ‘Biosynthesis of secondary metabolites’ and ‘Metabolic pathways’, will be discussed later. Cluster 4 contained 2 out of 5 bins assigned to ‘photosynthesis’ or ‘photosynthesis – antenna proteins’. The genes in these bins belonged to photosystem I, photosystem II, cytochrome b6/F complex, and LHC1 and LHC2 complexes. In the ‘Steroid biosynthesis’ bin, genes included squalene synthase, squalene monoxygenase, and other sterol biosynthesis genes upstream of brassinosteroid biosynthesis.Because GA did not contribute a significant number of genes to the DEG pool, it was eliminated from further analysis. There were major differences in ripening between the untreated grapes and grapes treated with CPPU, and this was also expressed in the number of DEGs at both early and late sampling. At the early and late sampling, there were 697 and 382 DEGs between the untreated and the CPPU-treated berries, respectively, and the reduction in the number of DEGs was mainly in the late sampling for genes with lower expression following CPPU treatment . KEGG analysis of these two groups did not show any functional bin enrichment. In a separate analysis, the treatments were compared on the basis of similar Brix; this comparison was possible because the untreated berries reached a Brix of 19.8 at 51 d while the CPPU-treated berries reached Brix of 19.6 at 70 d. Comparing these two groups resulted in 313 and 589 DEGs with higher and lower expression from the untreated and CPPU-treated berries, respectively . KEGG analysis demonstrated that in this ‘similar Brix’ group there was enrichment of 11 functional bins similar to those already present in cluster 3 .
A subset of 42 genes showed significant differences in three combinations: Control vs. CPPU at 51 d; Control vs. CPPU at 70 d; Control at 51 d vs. CPPU at 70 d . Interestingly, very few DEGs had higher expression in the untreated berries atsampling of 51 and 70 d. These genes were a germin-like protein, basic 7 S globulin , protein P21 , polyol transporter 5 , oleosin 1, and DMR6-like oxygenase 2 that functions in response to pathogens. The DEGs that had higher expression in response to CPPU could be assigned mainly to cell wall-related and stress-related functions. Gibberellin has been used to enlarge seedless grape berries for many years. Some cultivars are very sensitive to the treatment while others have a modest response. Sable Seedless can be assigned to the category of ‘modest response to GA’, with a significant effect on size at early sampling that is reduced at late harvest. The transcriptomic data for both 51 and 70 d after the treatment, corresponding to early and late commercial harvest, suggest that GA did not have long-term effects on biological processes and this was also confirmed in the profile of phenylpropanoid metabolites. However, GA did delay maturity as expressed by lower TSS and TA , and also had an effect on volatile compounds . In addition, when combined with CPPU, GA had some additive effects on fruit weight but it mitigated the response to CPPU with respect to acidity and reduction in some phenylpropanoid metabolites . It is likely that higher doses of GA than the concentration used in this study , but still considered in the range of commercial practice for this cultivar, or multiple applications, would further increase the size and delay the ripening as previously reported in many studies. Our data and the literature support the notion that GA can trigger quantitative effects on berry size, but does not have a major impact on ripening processes under standard practice. In contrast to GA, CPPU affected a wide range of phenotypic, biochemical, and molecular responses. Since the introduction of CPPU,blueberry container there have been a number of studies in grapes reporting beneficial effects on fruit size, and delays in maturity, abscission, rachis browning, decay, cuticle formation, volatile profile and gene expression. Our study on the colorless cultivar ‘Thompson Seedless’ pointed to the effect of CPPU on total tannin content resulting in higher astringency. It was therefore of interest to study how the interplay between the anthocyanin branch and the PA branch of the pathway is manifested in colored cultivars. Our data clearly shows that synthesis of PA is at the expense of anthocyanins with no particular effects on the type of anthocyanins produced . The fact that the level of flavonols was also reduced points to the possibility that either PA synthesis was induced by CPPU or that there was multiple repression of both the flavonol and anthocyanidin branches by genes such as MybA1 and MybF1. One interesting trend was the increase of -epicatechin during late ripening, both in the control and CPPU-treated berries , in contrast to previous results showing that flflavan-3-ols in the skins peak at 2–3 weeks after veraison. This increase can result from degradation of PAs, but there was no evidence for this in the total PA levels and the only PA monomer that decreased in content during ripening was ECG . Both flflavan-3-ols, -catechin and -epicatechin, had higher concentration in CPPU-treated grapes and, thus, it can be hypothesized that there was a concomitant induction of their biosynthetic genes, LAR and ANR by TFs such as MybPA1 and MybPA2. CPPU did not affect the expression level of VvMybPA1 at 51 d but the level of this gene was reduced at 70 d post treatment . It is not known if the gene was actually induced by CPPU before veraison. It should however be noted that PAs are synthesized prior to veraison while anthocyanins are synthesized during or after veraison, bringing up the possibility that the decrease in anthocyanins may be due to deficiency in substrates in the berry rather than repression of the ratelimiting genes.
In previous studies, cytokinins were shown to promote anthocyanin synthesis in shoots and leaves of Arabidopsis with concomitant transcriptional activation of the pathway. Likewise, introduction of a bacterial ipt gene to tobacco leaves increased the level of phenolic compounds. The inherent discrepancy of comparing gene expression by treatments that delay maturity was partially compensated by comparing differential expression of phenylpropanoid-related genes at similar Brix, i.e., control at 51 d and CPPU at 70 d after treatment . There is a remarkable reduced expression of 28 STSs in CPPU treated berries . The STS multigene family in grapes includes 48 annotated genes. Expression maps of STS genes suggest induction of gene expression by multiple biotic stresses including ethylene and jasmonate, while abiotic stresses had both suppressive and inductive effects. Therefore, according to our data cytokinin may contribute to long-term inhibition of VvSTS expression. Of the upstream genes in the phenylpropanoid pathway, 6 DEGs corresponded to PAL pointing to a trend of overall reduction in the pathway during ripening. Other genes included C4H, 4CL, CHS, UFGT , and 9 other structural genes of the pathway. Of the related transcription factors, MYB14 and MYB15 are considered as activators of STS gene expression and their low level is in agreement with the low expression of the STS genes. The expression of MYB14 and MYB15 was shown to increase with ripening in Pinot Noir grapes, and therefore delay in ripening mediated by CPPU may reduce their expression. MybA3 was shown to be a truncated non-functional protein. MYB136, MYB137, and MYB139 showed high expression in CPPU-treated berries, and it will be interesting to find out if they are targets for cytokinin signaling in plants and bona fide regulators of VvSTS genes. Of the WRKY genes, WRKY24 and WRKY43 are reported to be co-expressed with STS and MYB genes. For example, MYB14 and MYB15 were coexpressed with STS21, STS29, STS41, and STS48 in agreement with our data . Like MYB14 and MYB15, WRKY24 alone was shown to interact directly with the promoter of STS29. A previous study on the effect of CPPU in ‘Shine Muscat’ grapes focused on transcriptome changes 60 d after the treatment and volatile profile changes 80 d after the treatment. Similar to the previous report, our data showed that some genes of the phenylpropanoid pathway, e.g., PAL, 4CL, and UFGT had lower expression following CPPU treatment as well as fatty acid biosynthesis genes, amino acid biosynthesis, and volatile-related processes. Higher expression was observed for carotenoid-related genes, fatty acid metabolism genes, and auxin-related genes. In hormonal signaling, Wang et al. identified three accessions for TIFY 5A, two of which had higher expression at CPPU application level of 5 mg L−1 but with reduced expression at a level of 10 mg L−1 CPPU. This is consistent with our study where one of the TIFY 5A had increased expression at both 51 and 70 d . Another set of affected hormone-signaling genes in our study was that of the cytokinin receptor genes, of which ARR-5 was repressed by CPPU at the late stage . In contrast, in ‘Shine Muscat’ these genes were induced by CPPU. With respect to the volatile profile, similar reduction in terpenoids was encountered following CPPU treatment with respect to our sampling at 51 d . Several volatile compounds that were reduced after CPPU treatment in our study are consistent with reductions reported in Shine Muscat by Wang et al., i.e., α-terpineol, linalool, 2-hexenal at 51 d post treatment and cis geraniol at 70 d post treatment. Wang et al. also reported an increase in hexanol in ‘Shine Muscat’ following CPPU treatment but this was not the case for ‘Sable’ in our study .