The study of hormonal metabolism of Trincadeira and Syrah cultivars indicates new roles of salicylic acid, jasmonates, ABA and IAA during grape ripening and upon infection with Botrytis cinerea

Hormones play an important role in fruit ripening and in response to biotic stress. Nevertheless, analyses of hormonal profiling during plant development and defense are scarce. In this work, changes in hormonal metabolism in grapevine (Vitis vinifera) were compared between a susceptible (Trincadeira) and a tolerant (Syrah) variety during grape ripening and upon infection with Botrytis cinerea. Infection of grapes with the necrotrophic pathogen Botrytis cinerea leads to significant economic losses worldwide. Peppercorn-sized fruits were infected in the field and mock-treated and infected berries were collected at green, veraison and harvest stages for hormone analysis and targeted qPCR analysis of genes involved in hormonal metabolism and signaling. Results indicate a substantial reprogramming of hormonal metabolism during grape ripening and in response to fungal attack. Syrah and Trincadeira presented differences in the metabolism of abscisic acid (ABA), indole-3-acetic acid (IAA) and jasmonates during grape ripening that may be connected to fruit quality. On the other hand, high basal levels of salicylic acid (SA), jasmonates and IAA at an early stage of ripening, together with activated SA, jasmonates and IAA signaling, likely enable a fast defense response leading to grape resistance/ tolerance towards B. cinerea. The balance among the different phytohormones seems to depend on the ripening stage and on the intra-specific genetic background and may be fundamental in providing resistance or susceptibility. In addition, this study indicated the involvement of SA and IAA in defense against necrotrophic pathogens and gains insights into possible strategies for conventional breeding and/or gene editing aiming at improving grape quality and grape resistance against Botrytis cinerea.

On the other hand, the content in anthocyanins at EL38 decreased significantly in the 1 severely infected Trincadeira but not in Syrah. This decrease in anthocyanins was 2 previously observed in berries severely infected with powdery mildew [43] and in grape 3 skins of Botrytis-affected berries [44]. 4 5 3.2 Hormonal metabolism of Trincadeira and Syrah cultivars during grape ripening 6 For jasmonate analyses we focused on the bioactive form, jasmonoyl-isoleucine conjugate 7 (JA-Ile) [45], and its biosynthetic precursor 12-oxo-phytodienoic acid (OPDA). At EL32, 8 both cultivars showed similar amounts of OPDA and started decreasing from the onset of 9 grape ripening until EL38 (Fig. 4). In case of JA-Ile, a decrease from EL32 to EL38 was 1 0 also found in both cultivars; however, at EL32 the JA-Ile concentration in Syrah grapes 1 1 was about three times higher than in Trincadeira and at EL35 it was still higher. The 1 2 expression of a gene coding for allene oxide synthase (AOS) involved in OPDA synthesis 1 3 (Supplementary Figure S1) showed a pronounced increase during ripening of Trincadeira 1 4 grapes. Another biosynthetic gene encoding 12-oxophytodienoate reductase 1 (OPR1) and 1 5 further involved in JA-Ile biosynthesis tended to be slightly higher at the green stage 1 6 (EL32) in both cultivars in accordance with the concentration of JA-Ile. However, at EL35 1 7 Trincadeira presented higher expression of this gene but lower JA-Ile levels. A gene 1 8 coding for JAZ8, a repressor of jasmonate signaling [46] was up-regulated at EL38 in both 1 9 Trincadeira and Syrah grapes. MYC2 expression involved in jasmonate-dependent 2 0 transcriptional activation [31], presented an expression pattern during grape ripening 2 1 similar to JAZ8 but with a lower fold-change level. Previous microarray analysis showed 2 2 that mRNAs involved in the biosynthesis of jasmonates, namely those coding for OPR1 2 3 were less abundant at EL35 and EL36 in Trincadeira grapes [19]. Nevertheless, a gene 2 4 coding for an allene oxide synthase (AOS) involved in jasmonate biosynthesis was 2 5 strongly up-regulated at EL35 and EL36 in this previous study. In the present analyses, 2 6 both cultivars collected in a different terroir from previously one [19] presented an 2 7 increase in the expression of a gene coding for a specific isoenzyme of AOS during grape 2 8 ripening. 2 9 Similar to jasmonates the content in IAA showed a tendency to decrease in Syrah and 3 0 Trincadeira grapes during ripening (Fig. 5). The highest expression levels of the gene coding for Indole-3-acetic acid-amido synthetase, GH3.2, involved in auxin inactivation by 1 conjugation [47] were at stage EL32 in accordance with IAA content. The expression of 2 this gene lowers along grape ripening, and more abruptly in Syrah grapes. Expression of 3 IAA-amino acid hydrolase 6 was higher at EL32 compared with EL35 but tended to 4 increase at EL38. Altogether, these results suggest that free IAA levels are regulated at 5 green stage mainly by conjugation and at harvest stage mainly by hydrolysis of conjugates. 6 In both cultivars, the amount of IAA reached its lowest content at the harvest stage, though 7 higher in Trincadeira. as negative regulators of auxin biosynthesis [48]. The auxin-responsive SAUR29 gene 1 1 presented lower levels of expression at stage EL32 in accordance with the peak of auxin 1 2 content observed in both cultivars at this stage. As the expression of the auxin-responsive 1 3 SAUR29 increased, auxin concentration decreases, until the gene reached its peak of 1 4 expression at stage EL38. The same was verified for Aux1 involved in auxin transport [49]. 1 5 Interestingly, the expression of Aux1 was higher in Trincadeira at EL38 whereas the 1 6 expression of auxin-responsive SAUR29 was higher in Syrah at EL35. 1 7 Regarding SA concentrations they steadily decreased across the various stages of ripening, 1 8 both in Syrah and in Trincadeira grapes (Fig. 6). The expression of genes coding for 1 9 enhanced disease susceptibility (EDS1) and phytoalexin deficient 4 (PAD4) involved in 2 0 SA signaling [50] also followed this trend in both cultivars; the same observation was 2 1 obtained for Trincadeira collected in another terroir and season [19]. Out of the two 2 2 cultivars, Syrah displayed at EL32 higher basal levels of SA than Trincadeira and higher 2 3 expression of PAD4. 2 4 While jasmonates and auxins decreased during ripening, ABA concentration increased 2 5 drastically at veraison in both Trincadeira and Syrah grapes and decreased again at harvest 2 6 ( Fig. 7). The levels of expression of gene coding for 9-cis-epoxycarotenoid dioxygenase 2 7 (NCED) involved in ABA biosynthesis (Supplementary Figure S1) was higher at veraison 2 8 in both cultivars and about two times in Syrah at EL35. These data are in general in 2 9 accordance with the variations in ABA content showed by hormone quantification.

0
Interestingly, the gene coding for ABA receptor PYL4 RCAR10 involved in ABA- 3 1 mediated signaling pathway [51] was more expressed during the green stage, especially in Syrah grapes. In fact, at EL32 when compared with Trincadeira, Syrah grapes presented 1 higher expression of NCED, and ABA receptor PYL4 RCAR10, though levels of ABA were 2 similar for both cultivars. The expression of this ABA receptor was more noticeable in 3 Trincadeira grapes at harvest, possibly related to a higher content in ABA in this cultivar at 4 this stage. Differences in ABA content in the two cultivars during grape ripening seem to 5 be at least partially regulated by ABA catabolism involving ABA 8'-hydroxylase since 6 higher expression of the gene coding for this enzyme was noticed for Syrah at EL35 and 7 EL38. Concerning jasmonates, at EL32, the amount of OPDA in Syrah grapes was maintained 1 2 after infection, while in the susceptible Trincadeira variety, which already showed heavy 1 3 symptoms, OPDA level increased (Fig. 4). The same held true for JA-Ile. However, Syrah 1 4 grapes displayed higher basal levels of JA-Ile comparing to Trincadeira along with higher 1 5 expression of MYC2 and JAZ8. Regarding the expression of genes involved in the 1 6 synthesis of jasmonic acid, it was noted that at green stage expression of AOS was very 1 7 low in infected samples of both cultivars though differences could be noted between them. 1 8 However, at harvest expression of AOS increased in infected berries of Syrah and 1 9 Trincadeira, being highly expressed in the later one. The expression of this gene is in expression peaked at the green stage, which is in accordance with the concentration of JA- 2 2 Ile. Particularly, it illustrates the increase in JA-Ile concentration in Trincadeira grapes 2 3 upon infection accompanied by higher expression of JAZ8 whereas the opposite was 2 4 noticed in Syrah regarding the expression of this repressor of jasmonic acid signaling. 2 5 OPR1 expression decreased at veraison accompanied by a decrease in the content in JA- 2 6 Ile. Interestingly, at EL38 infected Syrah grapes seem to present higher expression of 2 7 OPR1 than Trincadeira but this was not reflected in increased JA-Ile levels. It should be 2 8 noted that only at this stage Syrah grapes presented heavy symptoms of infection. 2 9 Regarding auxins, the content in IAA increased in infected Trincadeira grapes at EL32 and 3 0 EL35 while the expression of IAA-amino acid hydrolase 6 decreased in this cultivar at 3 1 EL32 (Fig. 5). On the other hand, at EL35 levels of IAA seem to be mainly regulated in the two infected cultivars by different expression of Indole-3-acetic acid-amido synthetase, 1 suggesting that free IAA levels are regulated at different stages by active mechanisms of 2 hydrolysis and/ or conjugation. Although not statistically significant (p<0,094), the higher 3 basal levels of expression of IAA-amino acid hydrolase 6 in Syrah grapes at EL32 are 4 accompanied by higher IAA content observed at this stage. Additionally, at EL32 only 5 infected Trincadeira berries showed increased expression of the gene coding for auxin 6 responsive SAUR29 whereas AUX1 increased in Syrah infected grapes at EL38. This 7 suggests different regulation of auxin signaling and transport during ripening of infected 8 berries from different cultivars. 9 On the other hand, SA levels were shown to accumulate more upon infection at EL32 in 1 0 Trincadeira, and at EL35 in Syrah grapes ( Fig. 6). At the harvest stage, SA content was 1 1 relatively the same in both grape cultivars after being infected. Interestingly, the green 1 2 stage was marked by high levels of expression of the genes coding for EDS1 and PAD4 in 1 3 infected Trincadeira grapes. Accordingly, Trincadeira grapes showed a larger increase of 1 4 SA upon infection at this stage. Noteworthy is the higher basal level of expression of 1 5 PAD4 in Syrah grapes at EL32 together with higher basal levels of SA. EDS1 and PR1 1 6 seem to follow the same trend but the differences were not statistically significant. EDS1 1 7 and PAD4 expression levels were lower at veraison and at harvest in both cultivars but 1 8 PR1 increased at EL38 especially in infected Trincadeira berries. 1 9 In what concerns ABA, it displayed a larger increase in concentration in the Trincadeira 2 0 cultivar upon infection, in particular at EL32 (Fig. 7). On the other hand, at EL35 infected 2 1 Syrah berries tended to present higher content in ABA, higher expression of ABA 8'- 2 2 hydroxylase and ABA receptor PYL4 RCAR10 when compared with infected Trincadeira 2 3 berries. 2 4 The expression pattern of NCED at EL32 and EL35 was very similar to the observed 2 5 content in ABA for the variety of samples but not at EL38. The expression of the gene 2 6 coding for ABA receptor PYL4 RCAR10 was higher at green stage, especially in Syrah 2 7 grapes. At EL38, the expression of this gene decreased in Trincadeira berries upon 2 8 infection. At this stage, infected Trincadeira berries exhibiting heavier symptoms of 2 9 infection presented higher content in ABA than infected Syrah berries.  1 2 Hormones play an important role in plant development and stress responses. Disclosing the 3 roles hormones have in grape ripening and grape defense against major fungal pathogens 4 will enable improvement of fruit traits and productivity. In this context, hormonomics 5 which can be considered as a part of metabolomics, may provide invaluable cues 6 concerning hormonal crosstalk occurring in diverse plant processes. However, in general, 7 concentrations of plant hormones are very low. Most plant hormones cannot be identified 8 in common metabolome analysis therefore a specialized and highly sensitive MS detection 9 system is required to analyze them [52]. Hormonomics has been rarely reported in fruit (e.g. IAA, JA, JA-Ile, SA) presented high content in the youngest fruit and then decreased 1 3 dramatically. ABA was among the few hormones that increased in the ripening stage. 1 4 Similar results were obtained in the present study focusing in grape ripening and they were 1 5 confirmed for two cultivars bringing robustness to the study. 4.1 Combined study of two cultivars highlighted differences in the metabolism of ABA, 1 8 auxins and jasmonates during grape ripening that may influence berry quality 1 9 2 0 The important role played by ABA in the onset of grape ripening has been widely referred 2 1 [1, [54][55][56]. Castellarin et al. [57] proposed a timeline of events leading to the onset of 2 2 ripening with increases in ABA occurring early during softening and in the absence of 2 3 significant increases in expression of the V. vinifera 9-cis-epoxycarotenoid dioxygenases. 2 4 These increases in ABA were accompanied by decreases in a product of ABA catabolism, 2 5 diphasic acid, suggesting that initial increases in ABA may be due to decreases in 2 6 catabolism and/or exogenous import. The simultaneous study of grapes from two cultivars 2 7 collected in the same terroir validated the great increase in ABA at initial veraison and a 2 8 putative decrease in ABA catabolism. However, differences between the two cultivars 2 9 were more pronounced in relation to the expression of NCED at EL32 and EL35. Syrah also exhibited a lesser abrupt decrease in ABA catabolism throughout ripening as assessed 3 1 by the expression of ABA 8'-hydroxylase, which was lower in Trincadeira at EL35 and EL38. Furthermore, the metabolism of ABA is also different between Syrah and 1 Trincadeira at harvest stage with Trincadeira accumulating significantly more ABA (likely 2 due to a higher decrease in ABA catabolism) and transcripts of a gene coding for ABA 3 receptor PYL4 RCAR10 but registering no significant difference in NCED expression. 4 Therefore, ABA biosynthesis, catabolism and signaling are dependent on the cultivar and 5 also on the ripening stage. Recently, the short-term effects of ABA on different organs of 6 grapevine were studied and showed that the responses of each organ were unique 7 indicating that ABA signaling varies with the organ [58,59]. It may be that it also varies 8 with the tissue; therefore differences in ABA metabolism between skin and pulp may 9 partially account for the differences noticed here between the whole berries of the two 1 0 cultivars. Additionally, ABA may have an impact in fruit quality [60,61] and Trincadeira 1 1 and Syrah are known to produce grapes and wines with different features. In fact, recent 1 2 studies have shown that exogenous treatment of pre-veraison grape berries with ABA 1 3 activated the expression of genes involved in cell wall modification, lipid, carbohydrate 1 4 and flavonoid metabolisms [56]. 1 5 Contents in auxins and jasmonates are known to decrease during grape ripening while 1 6 some signaling processes of these growth regulators are activated [1,19]. However, these 1 7 aspects have not been previously compared between cultivars collected in the same terroir. 1 8 Interestingly, the contents in IAA and the expression of a gene coding for auxin-responsive 1 9 SAUR29 varied in between the cultivars across the developmental stages. Previously, a 2 0 role of SAUR proteins in promoting cell expansion was described [62] and this process 2 1 occurs during berry ripening. Additionally, differences in the expression of gene coding 2 2 AUX 1 were noticed between the two cultivars at harvest but not in genes coding for 2 3 enzymes involved in auxin glycosylation/ hydrolysis though higher content in IAA was 2 4 observed in Trincadeira at this stage. Since auxin is known to delay increases in berry size, 2 5 sugar accumulation, and anthocyanin content [13,14,16] the fine regulation of the pool of 2 6 free IAA and its conjugates together with auxin transport (e.g. AUX1) and signaling may 2 7 have an impact on final berry quality. Moreover, auxin negatively regulates ABA-induced 2 8 ripening processes [16] and interactions between ethylene and auxin were reported to be 2 9 fundamental to the control of berry ripening [63]. Another level of complexity can be added with the participation of jasmonates in the 3 1 regulation of berry development and ripening as previously suggested [20]. While  The expression of the two genes coding for these receptor and transcription factor followed 6 a similar pattern at EL32 and EL35 for both cultivars though expression levels were 7 generally higher in Syrah. The same was not verified at EL38 when comparing both 8 cultivars indicating that the interplay among hormones is strictly regulated during ripening 9 and depends on the genetic background. In fact, many examples of crosstalk among 1 0 hormones have been described in grape [1]; the present results suggest that this crosstalk is 1 1 likely to occur with specificities associated with each cultivar. 1 2 Additionally, a new molecular mechanism through which SA antagonizes ABA signaling 1 3 has been described [64]. Salicylic acid may delay ripening as suggested by the decrease in 1 4 its content in both cultivars. Additionally, SA treatment has been found to delay the 1 5 ripening of banana, kiwi and grape fruits [65][66][67]. Interestingly, PAD4 and EDS1 were 1 6 shown to regulate cell wall properties in poplar [68] and may eventually play a role in 1 7 regulation of berry softening. Interestingly, at veraison and harvest stage no clear 1 8 differences were noticed between cultivars in what concerns the content in this growth 1 9 regulator as well as in the expression of genes coding for EDS, PAD4 and PR1. Though it 2 0 can be suggested that the role of SA in influencing berry quality characteristics specific of 2 1 each cultivar may be less determinant than other hormones, its function in grape ripening 2 2 may still be uncovered. The analysis of the hormonome showed that Syrah presented at EL32 higher content in 2 8 SA, JA-Ile and IAA, suggesting that these hormones are involved in basal resistance 2 9 against Botrytis cinerea. Salicylic acid is widely known as determinant for the 3 0 establishment of basal defenses, effector-triggered immunity, and both local and systemic 3 1 acquired resistance [69][70][71]. Additionally, SA has been reported to be involved in the activation of plant defenses against biotrophs and hemibiotrophs, and it also appears to 1 enhance susceptibility to necrotrophs by antagonizing the JA signaling pathway and by 2 inhibition of auxin signaling [23,25,72]. However, SA and JA signaling pathways have 3 also been reported to be either antagonistic or synergistic [72][73][74]. Our results suggest that 4 SA may also participate in mechanisms associated with resistance/ tolerance to Botrytis 5 cinerea since the content in SA was higher in infected green but mainly infected veraison 6 Syrah berries comparing to Trincadeira. At this stage only Syrah exhibited mild symptoms 7 due to infection. Previous studies suggest a role of SA in response to infection with B. 8 cinerea [75,76]. In tomato, the balance between SA and JA responses seems to be crucial 9 for resistance of unripe fruit to B. cinerea [26]. The interaction of JA and SA in promoting depend on the accession [77]. It has been also mentioned that depending on the plant 1 2 species, the function of SA in immune responses may vary [78,79]. Previously, azelaic 1 3 acid was identified as a positive marker of infection of Trincadeira green (EL33) and 1 4 veraison (EL35) grapes with B. cinerea [4]. This compound is involved in priming the 1 5 faster and stronger accumulation of salicylic acid in response to pathogen infection in 1 6 Arabidopsis [80]. This accumulation of SA was observed for green Trincadeira grapes but 1 7 not for veraison grapes, highlighting the complex regulation of hormonal metabolism that 1 8 occurs in response to infection at different stages of ripening. It may be that the pathogen 1 9 activates the azelaic acid response, but the priming or another downstream process is 2 0 suppressed or not recognized later during ripening and therefore systemic acquired 2 1 resistance and the salicylic acid response is not activated as previously suggested [4].

2
Altogether, it is therefore tempting to speculate that different cultivars may have a different 2 3 JA and SA balance also in response to necrotrophs and the associated immune response 2 4 may be therefore distinct. Additionally, only at EL38 Syrah presented significant 2 5 symptoms of infection and at this stage SA content is similar between the two cultivars. 2 6 In basal and in effector-triggered immunity, EDS1 with its direct partner Phytoalexin 2 7 Deficient4 (PAD4), promotes SA accumulation, and current models position EDS1/PAD4 2 8 upstream of SA signaling [50,81]. Recently, an early function of EDS1/PAD4 signaling 2 9 independent of generated SA has been described [82]. This may justify the significant 3 0 increase in EDS1 and PAD4 expression in infected Trincadeira berries at EL32 without 3 1 reaching higher content in SA than Syrah. Previously, leaves of the resistant variety Norton 3 2 infected with powdery mildew, a biotrophic pathogen, presented a constitutively high SA 3 3 content as compared to the susceptible Cabernet Sauvignon [83]. Additionally, EDS1 was constitutively expressed to high levels in Norton but not Cabernet Sauvignon, and in 1 Cabernet Sauvignon EDS1 was induced by PM [84]. Similar results were obtained in this 2 study for EDS1 and PAD4 highlighting that some regulatory mechanisms of defense 3 responses may be common to necrotrophic and biotrophic pathogens. On the other hand, 4 EDS1 and PAD4 have been indicated as negative regulators of ethylene/jasmonic acid 5 defense signaling [85], raising the question on whether the high increase in EDS1 and 6 PAD4 expression in infected green Trincadeira berries may lead to impairment in defense 7 associated with JA in this cultivar which is a crucial hormone in response to B. cinerea 8 [79]. 9 The role played by jasmonates in grape response against Botrytis cinerea has been 1 0 previously referred [4,20]. Interestingly, in tomato transcriptional reprogramming of 1 1 important JA-signaling components (e.g., MYC2) was not evident during fruit infection or 1 2 during ripening which may indicate that activation of JA-related defenses in fruit occurs 1 3 via other signaling pathways [26]. However, MYC2 was up-regulated in infected Syrah 1 4 grapes at EL35 and EL38 comparing with Trincadeira highlighting the importance of 1 5 conducting studies in non-vegetative tissues and of comparing climacteric and non- 1 6 climacteric fruits. In addition, Syrah presented at EL32 and EL35 higher basal levels of 1 7 expression of MYC2 than Trincadeira. Mutants of myc2 have been reported to present 1 8 reduced resistance to B. cinerea [86] indicating that MYC2 may be a positive regulator of 1 9 defense.

0
It should be also taken into account that pathogenic fungi can produce plant hormones and 2 1 manipulate hormonal regulation of plant defense [87]. The secretion of fungal 12-OH-JA 2 2 can block JA-mediated signaling to suppress the defense response during host penetration 2 3 [88]. In other cases, pathogen effectors target SA signaling for virulence, by preventing SA 2 4 accumulation [89]. However, there is no direct evidence that fungal SA or JA is required 2 5 for their virulence [90]. 2 6 Regarding auxins, little is known about their role in plant resistance to necrotrophs. 2 7 However, auxin signaling has been reported to be important for innate immunity, the 2 8 activation of the auxin pathway mediates pathogen-associated molecular patterns (PAMP)- 2 9 triggered susceptibility, and auxin opposing regulation mediates PAMP-triggered 3 0 immunity (reviewed by Naseem et al. [91]). Previously, it was suggested an interaction between auxin and jasmonic acid in resistance 1 to necrotrophic pathogens [32]. Considering the higher basal levels in IAA and JA-Ile in 2 Syrah green berries it may be that these growth regulators are interacting in order to 3 provide a fast response to pathogen attack leading to tolerance of this cultivar. 4 However, it has also been reported that several plant pathogens can directly synthesize 5 auxin or induce plant auxin biosynthesis or alternatively modulate auxin signaling to 6 render the host more susceptible to infection [29,[92][93][94]. At EL32 infected Trincadeira 7 berries presented increase in IAA content and up-regulation of auxin-responsive SAUR29 8 comparing to the control indicating that auxin signaling is activated upon the infection. 9 Additionally, at EL35 infected Trincadeira berries presented higher content in IAA than 1 0 infected Syrah berries and tendency for higher expression of AUX1. This gene was also up- 1 1 regulated at EL38 in Syrah berries comparing to the control. At this stage, Syrah already 1 2 presented significant symptoms of infection suggesting that the pathogen may influence 1 3 auxin transport. Furthermore, the aux1 mutant, which is defective in auxin influx, cannot 1 4 develop induced systemic resistance against Botrytis cinerea [49]. Our analysis do not 1 5 seem to corroborate this with increased expression of AUX1 observed first for infected 1 6 Trincadeira grapes at EL35 comparing to infected Syrah and an additional increase noted 1 7 for Syrah at EL38 when they are both exhibiting heavy symptoms of infection. However, 1 8 F. oxysporum, a hemibiotrophic pathogen, requires components of auxin signaling and 1 9 transport to colonize the plant more effectively, suggesting that alteration of polar auxin 2 0 transport may also confer increased pathogen resistance [29]. It should also be taken into 2 1 account that B. cinerea can synthesize auxin [95], but the exact function of the pathogen- 2 2 derived IAA during infection and in interaction with the plant host has not been elucidated. Salicylic acid has been reported to induce the transcription of genes coding for IAA- 2 5 conjugating enzyme GH3.5 which converts free IAA into IAA Asp (inactive auxin) [96]. 2 6 Generally, higher SA levels reduce the pool of active IAA and repress auxin signaling 2 7 leading to enhanced defense and reciprocally, SA-mediated defenses are attenuated by 2 8 auxin (reviewed by Naseem et al. [91]). This was noticed at EL35 when comparing both 2 9 cultivars. On the other hand, both cultivars at EL38 exhibited heavy symptoms of infection 3 0 accompanied by lower levels of SA and increased expression of auxin-responsive SAUR29 3 1 which is in line with previous observations that increased auxin signaling leads to 3 2 significant reduction in SA accumulation after pathogen infection [96,97]. ABA can also influence the outcome of plant-microbe interactions; in particular the effect 1 of ABA in response against necrotrophic pathogens appears to be complex [79,[98][99][100]. 2 Abuqamar and co-workers [101] suggested a link between ABA, the cell wall, and 3 resistance towards B. cinerea in Arabidopsis with ABA treatment inducing the expression 4 of a cell wall loosening gene and contributing to susceptibility. In tomato, increased 5 expression of tomato NCED occurs during early infection of susceptible fruit, which 6 suggests a link between ABA synthesis and fruit susceptibility [26]. Our results suggest 7 that ABA is involved in susceptibility of Trincadeira given the high increase in this growth 8 regulator together with increased expression of NCED in infected grapes at EL32 9 comparing to control, though it can be also due to an acceleration of ripening and/ or 1 0 dehydration promoted by the fungus. This was not noticed in Syrah at this stage; instead 1 1 Syrah present higher basal expression of ABA receptor PYL4 RCAR10, so ABA signaling 1 2 processes can eventually be connected to basal resistance in interaction with other growth 1 3 regulators such as JA and SA [102,103]. In fact, negative and positive roles have been 1 4 described for this hormone depending on the pathosystem, developmental stage of the host, 1 5 and/or the environmental conditions in which the plant-pathogen interaction occurs 1 6 [33,104,105]. The results presented here suggest that it may also depend on intra-species 1 7 genetic variation of hormone networks (Fig. 8). The combined analysis of the hormonal profile and targeted qPCR analysis of genes 2 3 involved in hormonal metabolism during ripening and upon B. cinerea infection suggested 2 4 new roles for SA, IAA and ABA in cultivar fruit specificities and in response to 2 5 necrotrophic pathogens. High basal levels of SA and IAA at an early stage of ripening, 2 6 together with activated SA and IAA metabolism and signaling (Fig. 8), and possibly in 2 7 interaction with JA, seem to be important in providing a fast defense response leading to 2 8 grape tolerance against B. cinerea. 2 9 The results indicate that the role of plant hormones in promoting fruit resistance or 3 0 susceptibility is extremely complex depending not only on the relative content of the 3 1 hormones, but also on the timing of the synthesis and perception of the hormones. 3 2 Additionally, other factors may be important as well, namely the competence of the host 3 3 tissue to respond to active forms of the hormones, the localization of the plant's response 3 4 2 3 to the hormones, the pathogen's infection strategy, including its own production of 1 hormones [26,69]. Moreover, the balance and interaction among hormones may be 2 fundamental in providing either resistance or susceptibility; two genetic backgrounds in the 3 same plant species may present different hormone signaling requirements for resistance 4 [31] as indicated here. Further studies on inter-and intraspecies genetic variation of 5 hormone networks may validate the importance of basal hormonal levels in resistance/ 6 tolerance as suggested here and might provide insights into possible strategies of 7 manipulation using genome-editing technologies. In grapevine, the limitation will be the 8 optimization of transgenesis' protocols for different cultivars enabling functional gene 9 characterization in grapes [106]. Additionally, further hormonomics and RNA sequencing 1 0 applied to the specific study of cells and tissues will elucidate unknown functions of plant 1 1 hormones in fruit growth, development and quality as well as in response against 1 2 pathogens.

Conflict of interest
The authors declare that they have no conflict of interest.

Appendix. Supplementary data
Supplementary Table S1-List of primers used in real-time reverse transcription-PCR.
Primers were designed using Primer Select from DNAstar package.