Abstract

Background

Plasmopara viticola (Berk. et Curt.) Berlese and De Toni is an obligate parasite, able to biotrophically grow inside the susceptible tissues of species belonging to the genus Vitis and particularly of the extremely susceptible Vitis vinifera L., which is, unfortunately, the only species extensively used in the global wine industry [1]. P. viticola infects all green parts of the vine [2], leaves and bunches in particular, penetrating through stomata and extensively forming an intercellular mycelium with haustoria in the mesophyll cells [3]. In favourable climatic conditions, the pathogen causes numerous infection cycles, which are responsible for both quantitative and qualitative yield reductions. Since agricultural practices are almost ineffective in reducing disease incidence [4], chemical control is, at the moment, the most efficient mean to prevent severe downy mildew epidemics [5] and obtain good quality grapes.

Damages due to P. viticola could be reduced by using less susceptible grapevine varieties obtained by crossing V. vinifera with resistant species and in particular with American Vitaceae co-evolved with the pathogen such as V. labrusca L., V. aestivalis Michx., V. rupestris Scheele and V. riparia Michx. [6]. The resulting hybrids, however, are often unsuitable for the production of high quality wines due to their unpleasant flavours and aromas.

Both monogenic and polygenic resistance to P. viticola have been described in the Vitaceae family [7,8]. More recently, single quantitative trait loci (QTL) have been associated with resistance to P. viticola in V. amurensis Ruprecht [9], V. riparia[10] and in V. vinifera cv Regent [11]. Monogenic resistance is determined by a single dominant gene and has been associated with an hypersensitive response causing stomatal death at the point of infection and was found in the genera Tetrastigma and Cissus[7], whereas polygenic resistance induces a reduction of the pathogen growth after infection and is polygenically determined. The resistance imposed by a single gene is usually easily circumvented by pathogens characterized by a great evolutionary potential [12] such as P. viticola which has a high asexual sporulation efficiency, a polycyclic behaviour and sexual reproduction through the formation of oospores. However, also in polygenic resistance, that leads to a disease reduction rather than the absence of disease, host genotypes can respond differently to different pathogen strains [13] upon pathogen adaptation [14]. In the coevolutionary arms race between host and pathogen, the pathogen possesses the advantage of relatively short generation times and higher abundance in number compared to the host [15]. As a consequence, the pathogen can overcome the host defence mechanisms.

The variety Bianca, achieved by crossing the susceptible V. vinifera cv Bouvier with the resistant interspecific hybrid Villard Blanc [16], is together with Regent and Solaris one of the few cultivated hybrids resistant to P. viticola. Due to the multiple crosses involved in the breeding program, the genetic background of Bianca is mainly composed by V. vinifera (78.09%), and also by the Northern American species V. rupestris (14.58%), V. berlandieri Planch. (3.13%), V. lincecumii Buckl. (2.64%) and V. labrusca (1.56%) [17]. Recently, QTL mapping showed that resistance in cv Bianca is controlled by a single dominant allele at the Rpv3 locus on chr18, that explains most of the phenotypic variance in downy mildew resistance [18]. The resistance phenotypic traits observed in Bianca are localized necrosis and reduction of the colonization extent and sporulation capacity of P. viticola. Information on the interactions between P. viticola, cv Bianca and its siblings obtained from a cross with Chardonnay mainly derives from experimental inoculations carried out with isolates coming from naturally infected leaves collected in vineyards [18-20]. Since in single leaves different P. viticola genotypes can coexist, host reactions can not be attributed to a specific interaction with a pathogen individual. Therefore both the possible range and effectiveness of resistance should be ascertained only by using single spore isolates, i.e. distinct pathogen genotypes, and by quantitatively assessing disease parameters such as the area under disease progress curve (AUDPC). Host responses to individual P. viticola genotypes provide not only a profile of resistance but also insights on the evolutionary potential of the pathogen strains through the investigation of their virulence, defined as degree of damage caused to the host [21], and their fitness i.e.their ability of surviving and reproducing [22]. A rapid increase of genotypes characterized by the ability to breakdown monogenic/polygenic resistance, as already assessed on cv Bianca in Czech Republic where it is cultivated on a limited acreage [19], affects both effectiveness and durability of resistance [13].

Detailed investigations carried out on various host-pathogen binomials pointed out that resistance reactions to biotrophic fungal infections depend on the activation of several defence mechanisms, including the production of antimicrobial metabolites and proteins (pathogenesis related proteins, PRs) and, at the cell wall level, thickenings, callose appositions in the paramural space and accumulation of phenolic compounds and reactive oxygen species [23,24]. While defence reactions against P. viticola have been investigated at the morphological and histochemical level mainly on resistant species and only on two resistant hybrids, Solaris [25] and Seyval [26], incomplete information on the resistance mechanisms and on their histochemical features are available on cv Bianca and its siblings. Moreover, since the pathogen colonization pattern and the host histochemical reactions have often been independently examined, it is not clear if the observed structural or histochemical changes in the grapevine tissues are actually associated with the pathogen presence.

The aims of the present work are: (i) to quantitatively evaluate the phenotypic responses of both the parental and the hybrid lines obtained by crossing the susceptible cv ‘Chardonnay’x‘Bianca’ to 5 P. viticola strains each of which derived from a single oospore and assess both the host resistance level and the virulence of the pathogen isolates; (ii) to estimate the fitness of the most virulent isolate of P. viticola; (iii) to characterize at different infection stages the histochemical reactions of the host plants associated with the colonizing structures of the most virulent strain of P. viticola.

Results and discussion

Histological analyses

Colonization pattern

Encysted and germinating zoospores were observed in proximity of the stomata 24 hours after inoculation on all the examined grapevine lines. In most of the cases the substomatal vesicle, originating from the infection peg, was visible (Figure ​(Figure1b),1b), whereas just in a few infected stomata the primary hyphae with the first haustorium were also recognizable (Figure ​(Figure1a).1a). In some cases (18103 and 18110), callose depositions were observed on the cells surrounding the substomatal cavity immediately underneath the germinating zoospore (Figure ​(Figure1c).1c). At 2 dai, the primary hypha with haustoria, recognizable from the brightly fluorescent neck covered by callose, started branching inside the leaf tissues of all the plants except for 18096 and 18106 (Figure ​(Figure1d-f).1d-f). However, some differences in the pathogen structures could be detected. Due to the great variability observed in the length of the primary hypha, the different cv/hybrids could be divided into three groups: the first characterized by high values, ranging between 100–200 μm, of the pathogen vegetative structures, the second by an intermediate range (45–100 μm) and the third by low values (44–19 μm). Members of the different resistant classes could be found in each group (Figure ​(Figure2).2). For example, the first group includes 18048, 18088 (IS), 18036, 18103 (IR), 18100 and 18110 (V), while the second contains the parental cultivars and 18006, 18017 (IS), 18120 (IR) and 18001 (V) and the third 18053 (IS), 18099 (IR) and 18090 (V). Due to the extensive growth inside the tissues of some grapevine plants, the hyphal growth could not be measured from 3 dai onwards. The hyphal diameter ranged between 5 and 10 μm on most of the investigated cv/hybrids except for a few cases including the IR Bianca at 2 and 3 dai (3 and 1.8 μm), 18120 at 2 dai (2.7 μm) and 18096 at 3 dai (3.1 μm) and the V 18090 at 2 dai (2.7 μm).

Open in a separate window

Figure 1

Time course of colonization of grapevine cv/hybrids by P. viticola strain G. The pathogen structures (red colour) inside the leaf tissues (green colour) of susceptible (a, d, g, j, m) and resistant (b, c, e, f, h, i, k, l, n, o) grapevine plants following inoculation with G strain and aniline blue staining are represented. Pathogen diffusion was monitored 1 (a-c), 2 (d-f), 3 (g-i), 4 (j-l) and 7 (m-o) days after inoculation. (a) Substomatal vesicle (sv) with primary hypha (h) and first haustorium (harrow head) in the substomatal cavity of the leaf tissues of hybrid 18006. (b) Substomatal vesicle under an infected stoma of grapevine line 18096. (c) Callose depositions (asterisks) on the spongy mesophyll cells under an infected stoma. (d,e) Branched hyphae with haustoria (arrowhead) developing from the substomatal vesicles inside susceptible cv Chardonnay and resistant hybrid 18120. (f) Low diameter hypha (arrow) in the resistant cv Bianca. (g) Regular diffusion of the mycelium inside an infected area of Chardonnay. (h) Reduced colonization of the leaf tissues of the resistant hybrid 18103. (i) Callose on the hyphal wall inside the resistant hybrid 18120. (j) Complete colonization of the infected areas of cv Chardonnay. (k) Limited diffusion of P. viticola mycelium that in some cases was covered by callose (asterisk). (l) Degenerated mycelium (left) under a necrotic area and regular hyphae (arrows) inside the same infected leaf of the intermediate hybrid 18090. (m) Sporangiophores emerging from Chardonnay. (n) Mycelium diffusion inside Bianca. (o) Detail of an hypha covered by callose (asterisk). C = chloroplast, LV = leaf vein. Scale bar = 50 μm.

Open in a separate window

Figure 2

Early development of P. viticola inside the leaf tissues. Hyphal length (a) and diameter (b) measured at 2 and 2–4 dai respectively. Bars represent confidence interval (95%).

At 3 dai the differences between susceptible and resistant plants were clearly visible. While in the first case P. viticola started spreading inside the infected tissues and was mainly limited by the leaf veins (Figure ​(Figure1g),1g), in the resistant individuals the mycelium growth was restricted to the area close to the penetration stoma (Figure ​(Figure1h)1h) and in some cases brightly fluorescent appositions of callose were observed around the hyphae (Figure ​(Figure1i,o)1i,o) and in the surrounding stomata (Figure ​(Figure3d).3d). At 4 dai, the mycelium diffusion inside the susceptible individuals almost completely covered the infected area (Figure ​(Figure1j),1j), while on the resistant ones the pathogen growth was still restricted (Figure ​(Figure1k)1k) and degenerated hyphae were visible near the penetration point. However, close to the collapsed pathogen structures, it was still possible to observe some normal hyphae characterized by regular diameters (Figure ​(Figure1l).1l). At 7 dai the mycelium growth increased (Figure ​(Figure1n)1n) and sporulation occurred (Figure ​(Figure1m)1m) in all the samples even if at a variable rate.

Open in a separate window

Figure 3

Localization of peroxidase acitivity in infected leaf tissues. POX (blue) have been visualized by transmission light microscopy following TMB staining (a, c, e, g, h). The corresponding P. viticola vegetative structures (red) inside the leaf tissues (green) have been visualized in epifluorescence after aniline blue staining (b, d, f). (a) Stoma of the susceptible hybrid 18048 showing POX at 2 dai. (b) P. viticola mycelium (m) with haustoria (arrowhead) close to the stoma visualized in (a). Autofluorescent red chloroplasts inside the stomata are visible. The picture was obtained by overlapping a surface section (epidermal cells) and a lower section (mesophyll). (c) Epidermal cells of the resistant cv Bianca showing POX on the plasma membrane, that is detached from the wall, and browning under the stoma at 3 dai. (d) Lower section of (c) showing an altered mycelium (m₁) intensely coloured, probably due to callose deposition (asterisk), and no haustoria immediately under the browning area and regular vegetative structures of the pathogen (m₂) in a surrounding area showing no peroxidase activity. Asterisk indicates callose, S callose in a closed stoma. (e) Leaf tissues of cv 18048 lacking POX at 5 dai. (f) Regular mycelium (m) with haustoria (arrowheads) in the same area. (g) POX in the epidermal walls of the cells surrounding the stomata with sporangiophores (arrows). Hybrid 18100, 5 dai. (h) Detail of a sporangiophore with POX at the exit point. Scale bar = 50 μm.

The amount of P. viticola DNA inside each of the 12 inoculated leaf discs was estimated quantitatively by real-time PCR (calibration curve: y = −3.4x + 25.4; r² = 0.996; PCR efficiency = 0.97). The average DNA amount ranged between 20–70 ng/leaf discs in most of the samples apart from Bianca, 18001, 18036, 18103, 18110, 18120, showing lower values (0.5-14.2 ng), and 18099, characterized by a higher value (117 ng) (Figure ​(Figure4).4). A positive linear correlation was found among the amount of P. viticola DNA, the dimension of the infected areas and the number of sporangia differentiated by the pathogen on the susceptible parental line Chardonnay, suggesting that a high hyphal density and an abundant sporulation correspond to an extensive mycelium colonization (Table ​(Table4).4). A positive correlation between the dimension of the infected areas and sporulation was also found on the resistant cv Bianca. However, a negative correlation characterized the amount of P. viticola DNA and both dimension of the colonized areas and amount of sporangia, suggesting that in this case growth and sporulation are associated with a low hyphal density. In most of the hybrids examined, no significant correlation was found among the parameters except for 18036 (DNA-infected areas), 18053 (DNA-sporangia) and 18106 (infected areas-sporangia).

Open in a separate window

Figure 4

Average P. viticola DNA amount (ng) in the leaf discs. Bars represent confidence interval (95%).

Table 4

Correlation among DNA amount, infected area and number of sporangia

Cv/hybrid	Variable	Infected area	Sporangia
Chardonnay	DNA	0.55 (0.07)	0.700 (0.01)
	Infected area	1	0.721 (0.01)
Bianca	DNA	−0.811 (0.001)	−0.748 (0.005)
	Infected area	1	0.723 (0.008)
18036	DNA	0.664 (0.018)	-
	Infected area	1	-
18053	DNA	-	0.599 (0.04)
	Infected area	1	-

Open in a separate window

Spearman’s rho correlation coefficients for DNA amount (ng), dimension of infected areas (mm²) and number of sporangia differentiated and level of significance (parenthesis). ‘-‘ Indicates no significant correlation among the variables.

Peroxidase activity

Grapevine peroxidases belonging to class III group and located in cell walls or vacuoles, catalyze the oxidation of a broad range of substrates, such as the cell wall protein extensin, plant growth regulators and in particular the auxin indole-3-acetic acid (IAA), and phenolics (benzoic acids, stilbenes, flavonols, cinnamyl alcohols and anthocyanins) reducing H₂O₂[27]. In plant-pathogen interactions, peroxidases play a role in cell wall rigidity, through a cross-linking activity on phenolic monomers that leads to the formation of suberin and the oxidative coupling of lignin subunits, and, through the hydroxylic cycle, contribute to the oxidative burst producing reactive oxygen species (ROS) that are involved in the hypersensitive response and more generally in cell death [28].

Peroxidase activity (POX) in both inoculated and not inoculated leaves was evaluated qualitatively by transmission light microscopy, and quantitatively by spectrophotometry, exploiting the chromogenic property of TMB that undergoes a change in colour upon oxidation by peroxidase in presence of hydroperoxide [29,30].

The spectrophotometric assay indicates that, compared to the susceptible Chardonnay and the other samples tested, the resistant individuals Bianca and 18120 are characterized by a higher basal POX (Figure ​(Figure55).

Open in a separate window

Figure 5

Basal POX of the different cv/hybrids. Bars represent standard deviation

Wilcoxon signed-rank test showed a significant increase in POX followed the inoculation with P. viticola in only 38% of the cases (Table ​(Table5).5). Just in 4 samples (18088, 18110, 18106 and 18120) the increase in POX occurred between 1–2 dai. While in the first three cases (18088, 18110, 18106) it ranged between 0.005 and 0.014 and no significant reduction in the host colonization was observed, in the resistant hybrid 18120 the increase was more consistent (0.154) and associated with a reduced colonization of the tissues. In the other samples, including Chardonnay and Bianca, the increase occurred between 3 and 6 dai, when the pathogen was already diffusing inside the leaf tissues and differentiating the first sporangiophores. A generalised increase of POX was also detected at 4–5 dai in both resistant and suceptible genotypes by Kortekamp and Zyprian [26]. The resistant individuals Bianca and 18120 showed an higher basal POX than susceptible plants, as in the case of the resistant V. riparia selections Gloire de Montpellier and Lake St. George and the interspecific hybrid Seyval [26]. In these selections, the infection caused an increased production of ROS, hypersensitive response and POX in cells flanking the infection point. Apart from 18120, the early increase in POX activity was not associated with the host phenotype.

Table 5

POX (ΔE) of the samples inoculated with P. viticola from 1 to 6 dai

cv/hybrid line	1	2	3	4	5	6
Chardonnay	0.003	0.006	0.010*	0.110*	0.025	0.006
Bianca	0.058	0.014	−0.009	0.057*	−0.044	0.056*
18001	−0.019	−0.019	0.023*	0.064*	−0.020	0.016
18017	−0.030*	0.045	0.008	0.017	0.012*	0.026*
18048	−0.053	0.006	0.006	0.056*	−0.018	−0.023
18088	0.005*	−0.017*	0.011*	0.035*	0.012*	−0.009*
18100	0.006	−0.017*	0.023*	0.016	0.009*	0.009*
18106	0.033	0.009*	−0.007	0.036	0.034*	0.009*
18110	0.014*	−0.021	−0.063*	0.119*	0.033*	−0.029*
18120	−0.027	0.154*	0.003	0.032	0.120	−0.033

Open in a separate window

*Indicates a significant difference between not inoculated and inoculated samples at P = 0.05 level of probability.

Microscopic observations showed the occurrence of POX in all the host genotypes on the guard cell walls and vacuoles of the stomata close to the areas infected by P. viticola both in the early phases of the infection process, at 1–3 dai, and in the later stages corresponding to the sporangiophore differentiation, at 4–6 dai (Figure ​(Figure3a,c3a,c and g). Traces of POX were also visible at the base of the sporangiophore, a structure in close contact with the stoma (Figure ​(Figure3h).3h). The susceptible cv Chardonnay did not show any blue staining inside the stomata before the beginning of sporulation (4 dai). POX, localized on both stomata and cell walls, was always associated with browning of the host cells (Figure ​(Figure3c)3c) and in the resistant individuals Bianca and 18120, with altered pathogen structures. However, it did not completely prevent the colonization of the host tissues by P. viticola, that was in most of the cases able to grow and sporulate (Figure ​(Figure3b,3b, f and h) even if over a reduced area. For example, 3 days after inoculation on cv Bianca POX was observed in the plasma membrane of the epidermal cells surrounding a stoma showing brown cells underneath (Figure ​(Figure3c).3c). In the area immediately under the stoma, aniline blue staining revealed an altered structure of P. viticola mycelium, surrounded by callose depositions and with no recognizable haustoria (Figure ​(Figure3d3d m₁). However, the pathogen differentiated a normal branching mycelium and haustoria with the callose neck clearly visible (Figure ​(Figure3d3d m₂) in the surrounding tissues. The synthesis of ROS by NADPH oxidases, balanced by detoxification systems able to maintain redox homeostasis, during different growth and reproduction phases such as hyphal growth, sporulation and spore differentiation and germination is a common feature in fungi [31]. Since the sporangia were viable, the presence of POX could be related to the differentiation of the zoospores, whereas in the case of sporangiophores, since POX is mainly located inside fungal structures in close contact with the stomata, it could be more likely a response to the oxidative stress of the host leading to the necessity of maintaining the redox balance.

Callose

Callose deposition localised in the stomata, the spongy cells and around the advancing hyphae (Figures ​(Figures1c,o;1c,o; 3d; 6h) was the first structural response detectable in the infected tissues of cv Bianca and its siblings from 1 dai onwards. In Chardonnay, callose was limited to some stomata and detected in later phases, starting from 5 dai. Callose contributes to the penetration resistance against some agents of powdery mildew [32] and it is involved in the restriction of bacterial colonization. Callose does not seem to play an important role in limiting the development of P. viticola in other tolerant or resistant Vitis cultivars, since it is not detected until 5 dai [25,33]. In Bianca and its hybrids, on the contrary, callose deposition at the penetration points occurrs rapidly and is usually followed by the encasement of the developing hyphae and the penetrating haustoria. Callose plugging of the stomata represents a physical barrier to both the penetration and evasion of the pathogen and could be related to the reduction of sporangial amount produced in some resistant genotypes. In the host cells callose can hamper the nutrient exchange mediated by the haustorium, while its deposition around the hyphae efficiently contributes to delay or block the pathogen colonization of the host [34].

Phenolic compounds

Phenolics are ubiquitous compounds in plants that include numerous and diverse compounds that have a role in a large number of functions in growth, development and defence from both abiotic and biotic stresses [35]. They are mainly involved in postinfectional responses to disease agents but they also act as preinfectional barriers [32,36]. During the interaction with P. viticola the main phenolic compounds detected in different grapevine varieties are stilbenes, flavonoids [37-40] and phenolic compounds oxidized by POXs [41], which are accumulated more rapidly and to a higher extent in resistant genotypes in comparison to V. vinifera. The phenolic compounds detected in the leaves of cv Chardonnay not inoculated with P. viticola were mainly located in the leaf veins, whereas in Bianca they were also found in a few stomata (data not shown). Upon nitrous-acid reaction, phenolic compounds such as catechols and chlorogenic acid develop a dark cherry-red colour. Positive reactions were observed on the leaf veins, epidermis and mesophyll of resistant and susceptible plants inoculated with P. viticola at 5 and 7 dai, but while in susceptible tissues phenolics did not interfere with the pathogen colonization (Figure ​(Figure6a),6a), in resistant samples they concentrated inside the vacuoles of the spongy mesophyll cells immediately underneath the penetration sites (Figure ​(Figure66c).

Open in a separate window

Figure 6

Localization of phenolic compounds, flavonoids and lignin in infected leaf tissues. Sensitive (a,b,e,f,i,j) and resistant (c,d,g,h,k,l) leaves have been inoculated with P. viticola. Phenols (a,c) and lignin (i,k) are visualized in bright field microscopy following nitrous reaction and phloroglucinol-HCl staining respectively, whereas flavonoids (e,g) and pathogen structures (b,d,f,h,j,l) in fluorescence microscopy following Wilson reaction and aniline blue staining respectively. (a) Phenolic compounds (red) on palisade cells of cv Chardonnay at 5 dai. (b) Wide diffusion of P. viticola mycelium inside the same tissue showed in (a). (c) Phenolic compounds (red-brown) inside the spongy mesophyll cells in close proximity to a failed penetration site. Hybrid Bianca at 5 dai. (d) Sporangia (s) and germinating zoospore (arrow) on the leaf epidermis showing fluorescent cells (asterisk), indicative of cell death. (e) Absence of a specific reaction to Wilson reagent in cv Chardonnay at 5 dai. (f) Regular diffusion of the mycelium in (e) sample. (g) Specific reaction to Wilson reagents (orange) inside the spongy mesophyll cells of cv Chardonnay at 5 dai. (h) Infected stoma covered by callose and poorly fluorescent, probably degenerating, P. viticola mycelium (m) with haustorial necks clearly visible in the same area showed in (g). (i) Presence of lignin (red-brown) inside the leaf veins of hybrid 18048 at 15 dai. (j) Wide diffusion of the pathogen inside sample (i). (k) Lignin on the host cell walls and inside cytoplasmic vesicles in the cells surrounding an infected stoma. Bianca at 15 dai. (l) Absence of mycelium, probably degenerated, inside (k) sample.

Neu's and Wilson's reagents, used for the detection of flavonoids, gave analogous results: no specific reaction to the presence of P. viticola was observed in susceptible plant tissues (Figure ​(Figure6e),6e), whereas the spongy mesophyll cells of the resistant hybrids showed a strong fluorescent signal and a shrinked structure (Figure ​(Figure6g).6g). Flavonoids were also detected in guard cells and on the walls of the surrounding epidermal cells.

Lignin, a complex polyphenolic compound, is a normal component of the host cell wall, particularly in the vascular system and can be generated by mechanical wounding or act as a non-degradable, mechanical barrier to fungal pathogens at a preinfectional level or in response to a pathogen attack [42,43]. Lignin was observed in both inoculated and not inoculated samples of susceptible and resistant cv/hybrids at 7 and, at higher levels, at 15 dai. It localized in the leaf veins, stomata and walls of the surrounding epidermal cells. No specific reaction associated with the pathogen presence in susceptible plants (Figure ​(Figure6i,j),6i,j), whereas in the resistant hybrids a strong reaction to phloroglucinol-HCl could be found inside some infected stomata (Figure ​(Figure6k)6k) where the pathogen growth was blocked at the substomatal vesicle level (Figure ​(Figure6l).6l). In the leaf tissues of the susceptible plants where phenolic compounds have been detected, the pathogen mycelium showed regular structures and a large diffusion (Figure ​(Figure6b,f,j).6b,f,j). In resistant hybrids, despite the high presence of sporangia and germinating zoospores (Figure ​(Figure6d),6d), the pathogen was not always able to colonize the tissues, since its development was either blocked at the substomatal vesicle level (Figure ​(Figure6l)6l) or it was limited to the areas close to the penetration site (Figure ​(Figure66h).

Conclusions

Durability i.e. the maintenance of an adequate level of resistance throughout the useful lifetime expected from a variety [22] is one of the most important objectives of plant breeding programmes for resistance. Useful lifetime is usually short for annual crops which can be easily replaced, but it could last several years in perennial cultivated plants, which are very costly to replace or to protect by other means against virulent pathogens. Resistant annual crops are often characterized by the presence of single resistance genes (qualitative resistance), which generally induce a strong selection pressure on the pathogen populations and, as a consequence, a rapid increase in the frequency of strains that are able to overcome the host defence reactions. Quantitative resistance due to multiple genes and resulting in reduced disease incidence is usually more useful for perennial crops, since it assures more durable results, even if adaptation to the host has also been observed in the case of polygenic traits [14]. As already pointed out, resistance to P. viticola is associated with the synthesis of physical barriers, such as callose [33] and lignin [37] appositions, and substances with antifungal properties such as pathogenesis-related proteins [44], peroxidases [41] and phenolic compounds [37,45]. The host response depends on many factors linked to both the environment and the pathogen. In particular, the pathogen virulence can modulate the host defence responses. Field populations of P. viticola that are presumably composed of different genotypes are normally used for the screening for resistance in grapevine [19,20,46], but this method prevents a comparative evaluation of the interaction between different host and pathogen genotypes. To the best of our knowledge, no prior attempt has been made to evaluate grapevine resistance as performed in the present study by combining the use of individual genotypes, isolated from single germinated oospores sampled from geographically distinct vineyards, and the quantitative assessment of the disease intensity. The results clearly show that the disease level expressed as AUDPC is determined not only by the isolate virulence and the host resistance level but also by their interaction. Therefore particularly virulent strains such as G are likely to cause conspicuous damages, at least to the leaf tissues, in hybrids characterized by a resistant behaviour against less aggressive strains. On the other hand, it must be pointed out that less severe epidemics would develop on the resistant hybrids in comparison with the susceptible varieties due to a constant reduction of the sporulation efficiency. The reduction of sporulation efficiency is of paramount importance in the field, since the amount of sporangia produced by P. viticola greatly affects the epidemic growth rate of the pathogen. In presence of favourable climatic conditions, in fact, even from a limited number of infection foci the pathogen can potentially spread in the host population thanks to an abundant production of sporangia, leading to severe infections on both leaves and clusters.

Histological investigations confirmed the differences in the interaction between P. viticola strain G and its hosts, particularly evident from 4 dai onwards. In the susceptible individuals, P. viticola colonization followed a regular pattern leading to an extensive growth in the host tissues and sporulation. On the contrary, the pathogen development in the resistant plants was characterized by alterations in the hyphal diameter in the early growth phases and a reduced colonization of the tissues. The resistant reactions therefore likely occurr early in the colonization process and seem to limit the pathogen growth instead of completely blocking it.

Colonization patterns in resistant and intemediate genotypes were associated with numerous changes in both the structural and chemical leaf characteristics, due to the activation of complex defense responses which eventually cause the necrotic lesions and the reduced colonization already described in Bianca [18]. A rapid callose deposition at the penetration point and around the developing hyphae seems to characterize the interaction between P. viticola and the resistant genotypes. The cytochemical analysis carried out in this study showed that peroxidatic activity and phenolic compounds synthesis are involved in disomogenous defense reactions of the resistant plants. In fact the pathogen colonization is stopped in the necrotic lesions but continues in the nearby tissues where no detectable host reaction occurs. Flavonoids and other phenolics were observed inside the infected stomata and surrounding epidermal cells together with shrinked and brown mesophyll cells in close contact to the developing hypha. No close association between pathogen and phenolics could be detected in the susceptible cv Chardonnay. Therefore the browning content of the cells close to the areas colonized by the pathogen in resistant plants could be the result of the oxidation of flavonoids to semiquinones and quinones, which in turn react with other phenols, amino acids or proteins forming brown products [47]. In healthy cells, the oxidases and their substrates are compartmentalized in different parts (cytoplasm and vacuoles respectively), and come in contact with each other when the cells are in some way damaged [48]. No other structural changes in the infected areas were observed until 7 dai when in the resistant genotypes lignin deposition was detected in the walls of the infected stomata and surrounding cells. In all the tested genotypes, lignin has been detected from 1 dai in the leaf veins, but its association with the pathogen occurred only in the resistant varieties in the final infection stages, as already observed by Dai and co-workers [37,49]. Lignin apposition probably represents the final stage of the defence reaction started with POX.

In conclusion, the response of Bianca and its siblings to P. viticola varied according to the inoculated strain, which in turn demonstrated different virulence levels. The most virulent strain, without any reduction in its overall fitness, induced the same disease intensity in Chardonnay and Bianca, suggesting that no fitness costs are associated with the resistance breakdown. However some hybrids showed an intermediate resistant phenotype and could be used in further field trials. Resistance is associated with a limitation of the pathogen growth, due to multiple reactions, ranging from callose deposition, POX activity and accumulation of phenolic compounds. From an epidemiological point of view, the reduction of sporangial production, a phenotypic trait common to to cv Bianca and all its siblings, is particularly interesting and can contribute to a more efficient downy mildew control.

Methods

Abbreviations

AUDPC: Area under disease progress curve; dai: Days after inoculation; DAPI: 4’,6-diamidino-2-phenylindole; IA: Infected Area; IEI: Infection efficiency index; IR: Intermediate-resistant phenotype; IS: Intermediate-susceptible phenotype; MGR: Maximal growth rate; PF: Pathogen fitness; POX: Peroxidase activity; PSA: Percentage sporulating area index; QTL: Quantitative Trait Loci; SPOR: The natural logarithm of the average number of sporangia per cm² of infected leaf area; T₁₀: Time expressed in hours required for disease severity to reach 10%; TMB: 3,3’,5,5’- Tetramethylbenzidine; V: Variable phenotype; WAK: Water Agar Kinetin medium.

Competing interests

The authors declare that they have no competing interests.

Authors’ contributions

ST carried out all the studies and drafted the manuscript, GV participated in the statistical analysis of the data and helped to draft the manuscript, DM helped to carry out microscopic examinations, AV conceived the study, participated in all steps of the analysis and drafted the manuscript. All authors read and approved the final manuscript.

References