1. Introduction
According to the fundamental observation made by Müller and Börger [
1], 80 years ago, infection with an incompatible strain of
Phytophthora infestans capable of initiating hypersensitivity reactions in potato tubers (
Solanum tuberosum) significantly prevented the effect of a subsequent infection with another strain (compatible) of
P. infestans. This inhibition phenomenon was linked to the synthesis by the plant of a “chemical principle” making the tissues resistant to infection, which these two researchers named “phytoalexin”. Phytoalexins are produced in a certain number of plant families, mainly and not exhaustively, Leguminosae, Fabaceae, Solanaceae, Malvaceae, Poaceae, Brassicaceae (Cruciferae) and Vitaceae without knowing, due to a gap in knowledge, whether all plants possess phytoalexins [
2,
3]. Phytoalexins are compounds with an antimicrobial activity generally ranging between 10 and 100 μM, with a molecular mass over 1500
Da, high molecular-ordered stilbene oligomers such as pauciflorol D having a mass of 1587.624
Da) [
4,
5,
6]. Their production essentially corresponds to a
de novo synthesis in response to various biotic and abiotic stresses. Their study has given rise over the last thirty years with the progress of structural chemistry, biochemistry and molecular biology, to a ferment of activity on the elucidation of their biosynthetic pathways and the regulation of their syntheses, finding interesting applications in the field of the engineering of these pathways in microbes and plants [
2,
7,
8,
9]. Phytoalexins exhibit a very broad spectrum of biological activity against a number of living organisms, from viruses, bacteria, fungi, plants to animals also taking part in allelochemical processes between plants. Namely, the antimicrobial function of phytoalexins and the role they play in the defense mechanisms of plants have been a central topic of study among the plant pathology community. As a result, a need has arisen to study how to induce the synthesis of these compounds in plants with the aim of elucidating the implementation of their biosynthetic mechanisms and for technological/biotechnological purposes. Of prime concern for plant pathologists is to find out if the stimulation of the synthesis of phytoalexins by different means (chemical substances, biological elicitors, beneficial microorganisms) can be of interest in terms of crop protection [
10].
In the phytoalexin world, stilbene phytoalexins have received a special attention, no doubt linked to their proven role in human health as anti-cancer, antioxidant, cardio-and neuroprotective agents. Although not strictly limited to the Vitaceae, it is nevertheless in this plant family that stilbene-type phytoalexins were first identified in 1976 following the pioneering works of Langcake and Pryce [
11,
12,
13] and that many studies have been carried out there, since. Here, we will unveil the different ways of inducing the synthesis of these compounds in grapevine during experiments carried out in vitro and
in vivo, on isolated leaves or on potted vines, also describing some experiments conducted in the vineyard. The challenge is to verify whether the stimulation of stilbene phytoalexin production can contribute to protecting this plant from pathogens’ attacks with possible applications in the vineyard. More generally, extending the results described in this review may also open the use of stimulating phytoalexin synthesis as a basis for crop protection.
2. Chemistry and antifungal activity of some stilbenes of grapevine: A brief overview
A recent and comprehensive study on these aspects has recently been published [
5]. Stilbenes belong to a restricted chemical group compared to the large family of flavonoids [
14]. But it is in grapevine that this class of compounds presents a spectacular diversity [
5,
15]. In fact, 48 different coding genes have been counted in the stilbene synthase family in grapevine, of which 33 are potentially active [
16,
17]. We will only mention here the elements necessary to the understanding of the data presented in this review.
Building of the resveratrol core begins with phenylalanine, itself deriving from the shikimic acid and arogenic acid pathways [
5,
18]. Phenylalanine undergoes oxidative deamination catalyzed by phenylalanine ammonia lyase (PAL) and leading to
trans-cinnamic acid (
Figure 1). The latter is hydroxylated to
para-coumaric acid (
trans isomer) by a specific cytochrome P450 hydroxylase, C4H,
para-coumaric acid being then converted into its coenzyme A thioester,
para-coumaroyl-CoA, by binding with a Coenzyme A molecule via a cinnamoyl-CoA ligase (4CL). A dichotomous system unique in the plant kingdom and comprising two enzymes, chalcone synthase (CHS) and stilbene synthase (STS) using the same substrates,
p-coumaroyl-CoA and three malonyl-CoA units (formed from glycolysis), affords flavonoids, on the one hand and stilbenes, on the other (
Figure 1). Resveratrol is a hub, seat of an intense metabolism involving in grapevine, methylation, hydroxylation, glycosylation reactions and especially peroxidation reactions leading to the formation of the stilbene oligomers, which makes the originality of this class of compounds [
6].
Certain methylated, glycosylated and hydroxylated stilbene derivatives as well as numerous oligomers have been identified in grapevine [
5,
15]. For all that, few enzymes catalyzing the transformation of resveratrol into its derivatives have been characterized both at the genomic and functional level. Piceatannol, a hydroxylated derivative of resveratrol, was characterized for the first time in grapevine [
19] (not pictured). Though its structural relationship with resveratrol suggests that it could be synthesized directly by hydroxylation of the latter by a flavonoid-3-hydroxylase-like enzyme (F3H) as in flavonoid biosynthesis, no experimental evidence of the genesis of piceatannol from resveratrol has been reported to date. Pterostilbene, which is a dimethylated stilbene identified in 1979 by Langcake’s group [
20], is undoubtedly the best-known resveratrol derivative (
Figure 2). The direct biosynthesis of pterostilbene from resveratrol is catalyzed by a resveratrol-
O-methyl transferase (ROMT), which was cloned from mildew-infected grape leaves and functionally characterized. This enzyme ensured the methylation of resveratrol into pterostilbene in vitro and in vivo [
21]. Isorhapontigenin (not pictured) is a monomethylated stilbene whose probable origin could result from the direct methylation of piceatannol. Glycosylation (addition of one or more sugars to a given compound) is a very common reaction in polyphenol chemistry. Piceid, the 3-
O-resveratrol glucoside, plays an important role in the metabolism of resveratrol where it is considered a storage form (
Figure 2). Glycosylation reactions are catalyzed by glycosyl-transferases generally using uridine diphospho-glucose (UDPG) as a glucose donor and although there are 240 putative genes coding for glycosyltransferases (GTs) in grapevine [
22], only two genes (
VLRSgt and
VvUGT72B27) encoding GTs active on resveratrol and leading to piceid synthesis have been identified in grapevine [
23,
24].
One of the most fascinating aspects in stilbene chemistry is the formation of oligomers; it has been brilliantly demonstrated by Stephenson and his group that resveratrol constitutes the building block of all oligomers whose levels of condensation can reach seven resveratrol units in the case of pauciflorol D. Oligomer formation initially proceeds by the oxidation of resveratrol into several radicals which then condense with each other according to defined coupling modes [
4]. Oxidation of resveratrol is linked to the action of various peroxidases, some of which have previously been described in grapevine, the peroxidases A1, B2 located in the apoplastic compartment and the vacuolar peroxidase B5 [
25,
26,
27]. The horseradish peroxidase (HRP) was employed in Langcake’s group’s pioneering works on stilbene phytoalexins reporting in this case the formation of δ-viniferin [
28]. It is unknown, however, whether peroxidases are able to orientate the polymerization of resveratrol radicals towards specific oligomers, the use of these enzymes such as HRPs in vitro generally affording complex mixtures of stilbene oligomers [
4,
5,
29]. Current hypotheses favor the involvement of chemical reactions between the radicals formed (Friedel-Crafts reactions, oxa-Michael additions) to explain these oligomerization processes [
4].
Many grapevine pathogenic microorganisms can elicit both the synthesis and the accumulation of stilbene phytoalexins. Following the pioneering work of Langcake’s group, other researchers have described the synthesis of resveratrol and its derivatives in response to
B. cinerea [
30,
31],
Plasmopara viticola [
32,
33]
Aspergillus carbonarius, an ochratoxin A-producing fungus [
34] or in response to the Esca complex of diseases responsible for the leaf stripe disease symptoms in grapevine [
35,
36], this list obviously not being exhaustive. Conversely, the produced stilbenes exerted a biocidal activity against these pathogens.
The antifungal activity of stilbenes has recently been described elsewhere [
5]. We will limit ourselves here to presenting the activity of the main stilbene monomers and of some oligomers described in grapevine. Due to their low water-solubility, the antifungal activity of stilbenes has greatly been underestimated. Langcake and Pryce [
11] solubilizing resveratrol in water alone reported concentrations > 870 μM for obtaining 50% inhibition of the spore germination of
B. cinerea (ED
50), for example, though lower values of 438 μM were reported for the inhibition of the development of the mycelium of this fungus when minute quantities of an organic solvent (acetone) was added to the medium and even less (88 μM) for
Fusarium oxysporum ([
11]. Data regarding inhibition of the spore germination of
B. cinerea have since been revisited by incorporating a minimum amount of ethanol into the spore incubation medium to ensure resveratrol solubility. Under these conditions, ED
50 values of 390 μM were noted [
37]. Fairly low inhibition values have been reported concerning the inhibition of zoospore motility of the oomycete
Plasmopara viticola, responsible for grapevine downy mildew, respectively ED
50 of 192 μM and ED
100 500 μM [
38,
39]. Methylated stilbenes such as pterostilbene and isorhapontigenin, respectively derived from resveratrol and piceatannol, exhibited higher biocidal activities, in the order of 2.5 to 10 times compared to the hydroxylated compounds, depending on the type of pathogen. Several studies have reported ED
50 for pterostilbene around 70 μM for the inhibition of the germination of
B. cinerea spores [
28,
37,
40] and 39–390 μM for inhibition of the fungal growth of Esca disease-associated fungi [
41]. The biocidal activity of pterostilbene was more accentuated regarding the inhibition of the motility of the zoospores of
P. viticola with ED
50 of 9 μM [
13] and 14.6 μM [
38]. As confirmation of the greater biocidal activity of methylated stilbenes compared to their non-methylated counterparts, isorhapontigenin has been reported to be 2.5 times more inhibitory to the development of downy mildew (ED
50 of 116 μM) than the highly hydroxylated piceatannol (ED
50 of 254 μM) [
42].
Surprisingly, dimeric stilbenes such as the well-known ε-viniferin showed quite high biocidal activity with ED
50 of 220 and 230 μM for the respective inhibition of spore germination and mycelial development of
B. cinerea [
13] and even more strong activity with regard to sporulation inhibition in
P. viticola (ED
50 of 12.7 and 70 μM) [
38,
43]. Finally, several high molecular-ordered stilbenes, the trimer, miyabenol C and three tetramers, hopeaphenol, vitisin A and vitisin B exhibited remarkable inhibitory activities on the development of downy mildew with very low ED
50 values: 18 μM for hopeaphenol; 40 μM for miyabenol C and, respectively, 20 and 12 μM for the vitisins A and B [
43]. These data are not so surprising because stilbene oligomers possess biological activities, for example, anti-cancer, anti-arthritic properties as well as protective activities against neuroinflammatory processes, superior to resveratrol [
6].
The fact that the synthesis of phytoalexins is induced by a large number of pathogenic fungi, their synthesis being possibly associated with an increase in the resistance of plants to infection, has led researchers to wonder about ways to stimulate the production of these compounds with different types of elicitors or by the use of microorganisms (essentially of bacterial origin) through numerous assays carried out in potted plants or in vivo [
44]. The crucial question of transferring the data acquired in the field has been more rarely addressed [
45,
46]. These questions will be developed in the following sections.
4. Control of grapevine diseases by beneficial organisms involving stimulation of phytoalexin synthesis
The biological control of plant diseases can be defined as the partial or total, direct or indirect, inhibition of the growth and development of a pathogenic agent responsible for a given disease by another living organism (or group of organisms), often antagonistic and considered as a beneficial organism. The living organisms (or group of organisms) responsible for controlling a disease are then called biological control agents (BCAs) [
106,
107,
108,
109,
110,
111]. The fungus,
Trichoderma harzanium strain T39 has been reported, for example, as a biocontrol agent for downy mildew in grapevine showing an 86% reduction in disease symptoms on potted grapevine, cv Pinot Noir. Among the differentially-induced defense responsive genes analyzed,
PAL and
STS genes were found to be up-regulated without mention regarding stilbene phytoalexin synthesis [
112]. There are works describing induction of phytoalexin biosynthesis by BCAs, for example, to name just a few, the accumulation of scoparone and scopoletin in the control of
Penicillium digitatum,
P. italicum and
B. cinerea by the yeast
Rhodotorula glutinis in orange fruit [
113], the synthesis of dianthranilide-type phytoalexins during the biocontrol of fusarium wilt by a
Pseudomonas strain in carnation [
114] or camalexin priming in the induced resistance of Arabidopsis by beneficial bacteria in the control of
B. cinerea and
P. syringae [
115]. To our knowledge, the first study reporting the biocontrol of a grapevine disease, gray mold, by a soil bacterium causing an elicitation of the biosynthesis of the phytoalexin resveratrol dates back 1998 [
116]. Co-inoculation of in vitro grapevine plantlets of
V. vinifera (susceptible) and
V. rupestris (mid-tolerant) with
B. cinerea and the uncharacterized soil bacterium B-781 led to a complete suppression of the disease symptoms. The biocontrol of
B. cinerea could have been linked to an increase in the production of resveratrol in
V. vinifera (6 μg/g FW) and in greater amounts in
V. rupestris (13 μg/g FW). A potentiating effect of the fungus/bacterium co-inoculation was observed on resveratrol production (78 μg/g FW) in
V. vinifera and to a lesser extent in
V. rupestris (31 μg/g FW) [
115] (
Table 4).
Two rhizobacteria of the genus
Pseudomonas,
Pseudomonas fluorescens CHA0 and
Pseudomonas aeruginosa 7NSK2 induced resistance in leaves of in vitro grapevine plantlets of the Chardonnay variety (susceptible) to
B. cinerea [
117], the protection conferred towards grey mold infection being greater than 20% with
P. fluorescens CHA0 and about 35% with and
P. aeruginosa 7NSK2 (
Table 4). The authors ruled out the hypothesis of a direct inhibitory activity of the bacteria towards
B. cinerea. In both cases, induction of the systemic resistance to
B. cinerea could be correlated with a priming effect on stilbene phytoalexin accumulation in the leaves, this effect being initially relatively weak at the pre-contamination stage (before
B. cinerea inoculation), but significant compared to the control (10–25 μg/g FW for resveratrol and 15 to 30 μg/g FW for ε-viniferin). At 3 days post-inoculation with
B. cinerea, these concentrations reached very high values (800 μg/g FW for resveratrol and 250 μg/g FW for ε-viniferin with
P. fluorescens CHA0 and > 700 μg/g FW for resveratrol and 180 μg/g FW for ε-viniferin with
P. aeruginosa 7NSK2). Phytoalexin production was preceded by an early oxidative burst. Very interestingly, two mutants of the 7NSK2 strain, one deficient in pyochelin (Pch) and pyoverdin (Pvd), two bacterial compounds inducers of resistance, the mutant KMPCH, and the other one deficient in Pch, Pvd and SA, the mutant KMPCH-567, showed differential induction profiles in the resistance to
B. cinerea. The KMPCH mutant (Pch- and Pvd-negative) showed a protective effect of 35% against
B. cinerea, that is, an effect comparable to that of the parental strain 7NSK2, the KMPCH-567 mutant (Pch-, Pvd- and SA negative) inducing little or no protection against gray mold infection (< 10%). The protective activity linked to the KMPCH mutant was found to be correlated with a potentiation of the production of resveratrol in the leaves, the KMPCH-567 mutant only inducing weak resveratrol amounts at two days post-inoculation. These data therefore reinforced the role played by phytoalexins in the bacteria-induced disease resistance to
B. cinerea [
117].
Several bacteria isolated in the Champagne vineyard,
Bacillus subtilis,
Pantoea agglomerens,
Acinetobacter lwoffii and
Pseudomonas fluorescens induced a protective effect against
B. cinerea on in vitro grapevine plantlets (cv Chardonnay),
B. subtilis showing a 35% reduction in gray mold symptoms, with
P. fluorescens showing the highest protection rate (70%),
P. agglomerens and
A. lwoffii resulting in a 60% reduction in leaf symptoms [
118]. All the bacteria triggered an early oxidative burst preceding an induction of the biosynthesis of the phytoalexins resveratrol and ε-viniferin of the order of 10 to 20 μg/g FW. A priming effect on phytoalexin accumulation was observed with
P. fluorescens and
A. lwoffii but not in the case of
P. agglomerens and
B. subtilis, where pretreatment with these bacteria had no effect on subsequent
B. cinerea-induced phytoalexin production. Later studies focused on the mechanisms of action of the bacterium
P. fluorescens PTA-CT2 on grapevine [
119,
120]. Induction of the systemic resistance in grapevine against
B. cinerea resulted in a 60% reduction in the symptoms of this fungus on in vitro grapevine plantlets cv Chardonnay, through the differential expression of defensive response genes at the root level (organ colonized by the bacterium) and at the leaf level (lack of colonization). Genes of the phenylpropanoid pathway and resveratrol synthesis (
PAL and
STS) as well as those of the anthocyanin biosynthetic route (
CHS, chalcone synthase,
CHI, chalcone isomerase,
ANS, anthocyanidin synthase) displayed a much higher overexpression level in the leaves compared to the roots [
119]. The PTA-CT2 bacterium was able to prime phytoalexin synthesis (resveratrol, piceid and ε-viniferin) in response to
B. cinerea inoculation, the amounts of stilbenes accumulated in the leaves in response to
P. fluorescens (3 days after the beginning of the infection), being higher in the leaves (piceid: 20 μg/g FW; resveratrol: 4 μg/g FW; ε-viniferin 10 μg/g FW) than in the roots ((piceid: ~ 0 μg/g FW; resveratrol: 1 μg/g FW ε-viniferin 5μg/g FW). Stilbene accumulation was fully consistent with the up-regulation levels of the
PAL and
STS genes, high in the leaves and low in the roots. Other defense genes, particularly those encoding PR proteins, showed a differential response depending on the organs analyzed, leaves or roots (
Table 4).
Although the effectiveness of the PTA-CT2 strain of the bacterium
P. fluorescens has already been demonstrated for the biocontrol of gray mold (
B. cinerea) [
117,
118,
119,
121,
122,
123], that of this bacterium against the oomycete
P. viticola (downy mildew), on the other hand, had not been the subject of any study. In the work of Lakkis et al. [
120], the effectiveness of the PTA-CT2 strain was evaluated on two-year-old potted grapevines of the varieties Pinot Noir (susceptible) and Solaris (tolerant) by soil drenching with the bacterial solution. Application of the bacterium led to a reduction in the growth development of
P. viticola of 80% in Pinot Noir and only 55% in Solaris; this reduction was 73–80% for Pinot Noir and 43% for Solaris following contamination by
B. cinerea. In the absence of any contamination, PTA-CT2 did not induce changes in the basal defenses of the plant, but induced changes at the hormonal level and an improvement in the photosynthetic capacities for the two varieties. In contrast, PTA-CT2 primed defensive pathways including
PAL and
STS gene overexpression, which was correlated with increased phytoalexin levels in both varieties. The two varieties showed quite similar profiles of phytoalexin accumulation after PTA-CT2/
P. viticola or PTA-CT2/
B. cinerea co-inoculation, respectively, resveratrol: 15 μg/g FW, ε-viniferin: 6–8 μg/g FW and δ-viniferin: 12–17 μg/g FW; resveratrol: 40 μg/g FW, ε-viniferin: 12–25 μg/g FW and δ-viniferin: 5–10 μg/g FW. These results clearly demonstrated that the effectiveness of the biocontrol exerted by the bacterium
P. fluorescens was mediated by the upregulation of genes involved in the stilbene biosynthetic pathway resulting in priming of the phytoalexin response and reinforced the role played by these compounds in grapevine/plant pathogens interactions.
There are few experiments describing the use of beneficial bacteria as biocontrol agents of fungal diseases in the vineyard and linking a possible protective effect of these bacteria on the stimulation of grapevine defense systems [
121,
122,
123]. Tests carried out in the vineyard for four consecutive years from 2002 to 2005 regarding the biological control of
B. cinerea by grapevine-associated bacteria, reported interesting results towards the reduction on leaves and berries (reaching 50%) of the symptoms of
B. cinerea by application of bacteria by the soil drenching method or by spray [
121]. This resistance to
B. cinerea was correlated with a stimulation of β-1,3-glucanase and chitinase activities in grape leaves and berries, but no analysis of phytoalexins was described at this time. Very similar work by the same group reported a 22 to 75% decrease in gray mold development in the vineyard following drenching of the grapevine soil with bacterial mixtures, this protection effect being linked to increases in β-1,3-glucanase and chitinase activities without any notification regarding phytoalexin production [
122].
Table 4.
Control of grapevine diseases by beneficial organisms involving stimulation of phytoalexin synthesis.
Table 4.
Control of grapevine diseases by beneficial organisms involving stimulation of phytoalexin synthesis.
Biocontrol Agent |
Plant Material |
Biological Inputs |
References |
Uncharacterized soil bacterium B-781 |
In vitro grapevine plantlets of V. vinifera (susceptible) and V. rupestris (mid-tolerant) |
Complete suppression of gray mold symptoms with increase in resveratrol accumulation |
[116] |
Pseudomonas fluorescens CHA0 and Pseudomonas aeruginosa 7NSK2 |
In vitro grapevine plantlets cv Chardonnay |
Protection towards gray mold > 20% with P. fluorescens CHA0 and about 35% with and P. aeruginosa 7NSK2. At 3 days post-inoculation with B. cinerea, stilbene concentrations reached very high values in the order of several hundred μg/g FW |
[117] |
Bacillus sutilis, Pantoea agglomerens, Acinetobacter lwoffii and Pseudomonas fluorescens
|
In vitro grapevine plantlets cv Chardonnay |
35% reduction in gray mold symptoms with B. subtilis, 70% with P. fluorescens and 60% with P. agglomerens and A. lwoffii. Priming effect on phytoalexin accumulation with P. fluorescens and A. lwoffii but not with P. agglomerens and B. subtilis
|
[118] |
Trichoderma harzanium strain T39 |
Potted grapevine, cv Pinot Noir |
86% reduction in disease symptoms towards downy mildew; induction of PAL and STS genes, no mention of stilbene production |
[112] |
P. fluorescens PTA-CT2 |
In vitro grapevine plantlets cv Chardonnay |
60% reduction in gray mold symptoms. Differential expression of PAL and STS genes (higher in leaves than in roots) correlating with accumulation of stilbenes in the two organs |
[119] |
P. fluorescens PTA-CT2 |
Two-year-old potted grapevins of the varieties Pinot Noir (susceptible) and Solaris (tolerant) |
Reduction in the growth development of P. viticola of 80% in Pinot Noir and only 55% in Solaris, 73–80% for Pinot Noir and 43% for Solaris towards B. cinerea. No induced changes in the basal defenses of the plant with PTA-CT2 alone. PTA-CT2 primed defensive pathways including PAL and STS gene overexpression, which was correlated with increased phytoalexin levels in both varieties |
[120] |
B. subtilis (PTA-271), P. fluorescens (PTA-CT2) and P. agglomerens (PTA-AF2) alone or as binary mixtures |
One-year experiment on grapevine plants (cv Chardonnay) in the vineyard including leaves and berries |
80 to 90% reduction in symptoms towards B. cinerea) with PTA-AF2 + PTA-271 on leaves. 93% reduction of B. cinerea symptoms with CT2 + AF2 on berries well correlating with phytoalexin accumulation |
[123] |
To our knowledge, the only experiment describing use of endophytic and rhizospheric bacteria for the biocontrol of gray mold in vineyard conditions and linking this protective effect to the induced production of phytoalexins both in the leaves and the berries, was that conducted by Aziz et al. [
123] (
Table 4). Three living bacteria isolated from grapevines in the Champagne vineyards, one rhizospheric,
B. subtilis (PTA-271), the other two being endophytic bacteria obtained from tissues of healthy grapevine plants,
P. fluorescens (PTA-CT2) and
P. agglomerens (PTA-AF2), previously studied on grapevine in vitro plantlets [
118], were brought to grapevine plants (cv Chardonnay) in the vineyard by drenching the soil, either individually or in the form of binary mixtures, in two treatments spaced one month apart in June and July. On grape leaves, it was interesting to note that it was the combination PTA-AF2 + PTA-271, which triggered the highest level of systemic resistance (80 to 90% reduction in symptoms towards
B. cinerea), 75 days after the first bacterial application and which also displayed the highest level of total phytoalexins (35 to 40 μg/g FW) as well as the highest values for resveratrol (25 μg/g FW) and ε-viniferin (7 μg/g FW) confirming the good correlation between protection towards contamination by
B. cinerea and rate of accumulation of antifungal phytoalexins. The results were more difficult to interpret regarding bacteria alone, since PTA-271 which provided the best protection level on leaves (75% reduction in
B. cinerea symptoms) was not the one which induced the highest phytoalexin response compared to the PTA-CT2 + PTA-AF2 mixture or PTA-CT2 alone.
B. cinerea is a particularly redoutable parasite that damages the harvested grape quality because of its late attacks on grape berries and due to the fact that the amounts of inducible phytoalexins during stress at this stage of maturity are low [
79,
124]. Specific attention must therefore be paid to the level of protection of the grape clusters towards this fungus. The effectiveness of bacteria alone in terms of reduction of the
B. cinerea symptoms on grape berries was very good, varying from 78% (PTA-CT2) to 87% (PTA-271), which correlated well with the accumulation of total phytoalexins (15 μg/g FW for PTA-271 and around 40 μg/g FW for PTA-CT2) along with a significant production of the antifungal dimer, ε-viniferin (>12 μg/g FW), 75 days after application of the first bacterial treatment. The CT2 + AF2 combination displayed the highest level of protection on grape berries (93% reduction of the
B. cinerea symptoms) correlating with a significant accumulation of total phytoalexins (> 20 μg/g FW). It was interesting to note that, in almost all cases, accumulation levels of phytoalexins in grape berries remained high, even 91 and 99 days after application of the first bacterial treatment, i.e., at dates close to maturity, stages where they are most vulnerable to gray mold attacks [
123].
5. Conclusions
Most of the treatments described in this work (chemicals, phytohormone-derivatives, bio-elicitors, biocontrol agents) led to variable protective effects against various pathogens (powdery mildew, downy mildew, gray mold, esca) which number among the phytopathogenic agents responsible for grapevine major fungal diseases, this protection being most often correlated with overexpression of genes of the phenylpropanoid pathway (
PAL and
C4H) and of the synthesis of resveratrol (
STS) alongside an up-regulation of the expression of other responsive defense genes (
LOX,
GST,
PER,
ERF1 and genes encoding PR proteins).
Induction of PAL and STS genes was generally accompanied by a notable increase in the content of stilbenes (resveratrol, piceid, pterostilbene and dimers) which may explain the protection observed towards pathogens. However, in some cases, the level of accumulation of these compounds regarding the doses required for them to exert their antifungal activity, remained insufficient to support the protective effects observed [
71,
80,
104,
105]
and, sometimes, a decorrelation was described between PAL and STS genes’ up-regulation and stilbene production [
76,
82,
104,
105]
. Even if some data exposed in this review were difficult to interpret, there were compelling facts to support the role of stilbene phytoalexin inducers in controlling various pathogens thereby providing grapevine with relative protection against diseases. Although phytoalexins represent only one component of the plant response to stress, it appeared that their production and the overexpression levels of the genes involved in their biosynthesis (
PAL,
C4H,
STS), were often correlated with those of other responsive defense genes (
PR proteins,
LOX,
GST,
PER, etc...).
Metallic salts possess a significant activity on the stimulation of stilbene synthesis, particularly, aluminum chloride and copper sulfate associated with a certain level of protection of grapevine against gray mold [
50,
63,
64] and, to a lesser extent, towards downy mildew [
50]. The effect of copper sulfate is to be linked to the use since the beginning of the 19th century of the
bouillie bordelaise in the vineyard. The aluminum contained in AlCl
3 was a very effective inducer of phytoalexin synthesis leading to the accumulation of huge amounts of resveratrol in the leaves (> 500 μg/g fresh weight (FW)) [
51]. The mechanisms by which this metal cation triggers the hyperproduction of this stilbene go through an overexpression of two
STS gene sub-families and the transcription factor
MYB14 controlling the
STS gene promoter, have recently been deciphered [
66]. Initiation of the overexpression of these genes, which depends on Al-induced remodeling of intracellular actin, involved the ROBh protein on which elicitor molecules other than aluminum salts whose toxicity is recognized, could be sought representing an interesting path for further technological development.
The different elicitation methods used were able to act either on the stimulation of the basal levels of phytoalexin biosynthesis (action of the treatment at the preinfectional stage), or by a priming effect (synergy of the effect linked to the pre-treatment combined with the phytoalexin response following contamination by the pathogen). The elicitor alone may also have no effect on the basal defenses of the plant, its action being exerted essentially by priming of the phytoalexin response. The level of accumulation of the latter in the plant treated with the elicitor and co-infected with the pathogen is then greater than that of the plant only infected with the pathogen.
Fosetyl-Al, for example, can significantly increase the basal phytoalexin response when applied alone or displays no effect on this response, its activity being mainly due to a priming effect [
61,
62]. Aluminum chloride used alone as well as copper sulfate, chitosan or a combination of both, were capable of inducing the synthesis of high amounts of resveratrol [
50,
51]. Pretreatment with numerous elicitors such as benzothiadiazole, phosphonates, methyljasmonate, ethephon, chitosan or ergosterol also triggered significant changes in the basal levels of stilbene phytoalexins [
50,
73,
78,
81,
82,
85,
92]. Other experiments have shown that benzothiadiazole induced at the same time a rise in the basal syntheses of phytoalexins as well as a phytoalexin priming activity [
73]. β-amino-butyric acid displayed a priming effect on the phytoalexin response but no action on the stimulation of phytoalexin synthesis when applied alone [
95].
Biocontrol agents can stimulate phytoalexin synthesis during preinfectional treatments [
50,
116] or may be able to induce both a basal phytoalexin response and to prime this response [
116,
117]. Finally, inoculation with beneficial bacteria often showed low or no effect on basal phytoalexin levels but, above all, a priming effect [
118,
119,
120].
The tests carried out in the vineyard faced the versatility of the experimental conditions (climate, disease pressure, way of applying the treatments and frequency of applications). Because of these constraints, few trials have been carried out in the vineyard [
64,
76,
81,
123]. Experiments conducted with Synermix (AlCl
3 + seaweed extract) obtained good results in terms of grapevine protection towards
Botrytis cinerea, application trials with Synermix achieving results close to those obtained with Iprodione, the Synermix + Iprodione combination showing a real efficacy, better than the theoretical one [
63,
64]. The protective effects observed with Synermix had then been correlated with the strong eliciting activity of AlCl
3 on resveratrol synthesis [
51]. Surprisingly, this product was withdrawn from the market, probably due to the presence of aluminum in the formulation. Experiments including applications of methyljasmonate in the vineyard every 7–10 days, from bloom to veraison conferred a 73% reduction in the symptom incidence of powdery mildew [
81]. This protective effect was correlated with a large increase in preinfectional accumulation of various stilbene phytoalexins (resveratrol, piceid and dimers).
Input of endophytic or rhizophytic bacteria alone or as mixture to grapevine plants grown in the vineyard by drenching the soil in two treatments spaced one month apart in June and July, showed very promising results in terms of protection of this plant against contamination by gray mold with reduction rates of the disease symptoms reaching 90% in the leaves and 93% in grape berries [
123]. This protection effect was accompanied by a significant increase in the basal levels of phytoalexins (resveratrol and ε-viniferin). Most importantly, phytoalexin amounts in the grape berries remained high, even 91 and 99 days after application of the first bacterial treatment,
i.e., at stages close to fruit maturity when the fruits are particularly susceptible to
B. cinerea attacks.
In sum, all the works presented and discussed in this review, showed that the protective effect observed towards phytopathogenic agents, by the application of different elicitors was correlated with an increase in the biosynthetic basal levels of stilbene phytoalexins as well as phytoalexin priming in grapevine under various conditions, making it possible to validate the concept of using phytoalexin induction as a means for crop protection.