Screening Vitis Genotypes for Responses to Botrytis cinerea and Evaluation of Antioxidant Enzymes, Reactive Oxygen Species and Jasmonic Acid in Resistant and Susceptible Hosts

Botrytis cinerea is a necrotrophic fungal phytopathogen with devastating effects on many Vitis genotypes. Here, a screening of 81 Vitis genotypes for leaf resistance to B. cinerea revealed two highly resistant (HR), twelve resistant (R), twenty-five susceptible (S) and forty-two highly susceptible (HS) genotypes. We focused on the HR genotype, ‘Zi Qiu’ (Vitis davidii), and the HS genotype ‘Riesling’ (V. vinifera), to elucidate mechanisms of host resistance and susceptibility against B. cinerea, using detached leaf assays. These involved a comparison of fungal growth, reactive oxygen species (ROS) responses, jasmonic acid (JA) levels, and changes in the anti-oxidative system between the two genotypes after inoculation with B. cinerea. Our results indicated that the high-level resistance of ‘Zi Qiu’ can be attributed to insignificant fungal development, low ROS production, timely elevation of anti-oxidative functions, and high JA levels. Moreover, severe fungal infection of ‘Riesling’ and sustained ROS production coincided with relatively unchanged anti-oxidative activity, as well as low JA levels. This study provides insights into B. cinerea infection in grape, which can be valuable for breeders by providing information for selecting suitable germplasm with enhanced disease resistance.


Introduction
Botrytis cinerea is a necrotrophic fungal phytopathogen that causes devastating gray mold disease in more than 200 dicotyledonous plant hosts, as well as some monocotyledonous species. This polyphagous pathogen is the second most prevalent phytopathogen responsible for pre-and postharvest decay and fruit quality deterioration in greenhouses, open fields, and during storage, including cold storage (0-10 • C) [1]. Grey mold is a major challenge to grape cultivation worldwide where periods of relative humidity (>90%) and cold temperatures (14-      In total, 81 Vitis genotypes were evaluated to investigate the resistance level against B. cinerea. All genotypes were classified according to their disease severity index (SI) at 96 h post-inoculation (hpi). Among the 81 genotypes, 42 (Table 1) were HS according to a disease SI of 5.51-7.0. Mycelium and sporulation was observed on these genotypes. A total of 25 genotypes (Table 1) were S, with mycelium production at 96 hpi, with no/less sporulation (SI of 3.51-5.50). A total of 12 genotypes (Table 1) were resistant, with much less mycelium production and no sporulation was observed in genotypes with SI values of 1.51-3.50. The 'Zi Qiu' and 'Ju Mei Gui' genotypes (Table 1) were HR with no mycelium or sporulation and with SI values of 0-1.50.
Two representative genotypes each from the HS, S, R and HR categories were chosen for macroscopic and microscopic evaluation of fungal growth at 96 hpi (  (Table S1).  (Table S1).

B. cinerea Growth on the HR Genotype 'Zi Qiu' and the HS Genotype 'Riesling'
To understand the development of B. cinerea on grape leaves from plants showing different resistance levels, we analyzed the two grape genotypes, 'Zi Qiu' (HR, V. davidii) and 'Riesling' (HS, V. vinifera), using scanning electron microscopy (SEM). At 4 hpi, their phenotypes were approximately the same ( Figure 3A,J). At 24 hpi, spore germination was clearly delayed, and fungal

B. cinerea Growth on the HR Genotype 'Zi Qiu' and the HS Genotype 'Riesling'
To understand the development of B. cinerea on grape leaves from plants showing different resistance levels, we analyzed the two grape genotypes, 'Zi Qiu' (HR, V. davidii) and 'Riesling' (HS, V. vinifera), using scanning electron microscopy (SEM). At 4 hpi, their phenotypes were approximately the same ( Figure 3A,J). At 24 hpi, spore germination was clearly delayed, and fungal growth was mostly blocked on 'Zi Qiu' leaves ( Figure 3A-I). The infection rate on 'Riesling' leaves was more significant and destructive than that of 'Zi Qiu' ( Figure 3J-R), which B. cinerea failed to infect. The presence of appressoria was first noted at 18 hpi ( Figure 3D) on 'Zi Qiu', after which time the progression of infection increased slowly until 48 hpi ( Figure 3D,G). At 8 hpi, appressoria were present on 'Riesling' ( Figure 3K), while penetrations became apparent at 12 hpi ( Figure 3L). The infection rate increased at 18 hpi, where infection pegs were clearly seen ( Figure 3M) and again at 24 hpi, ( Figure 3N), and infection hypha and necrotic spots appeared after 36 hpi ( Figure 3O). From 24 hpi, fungal growth was blocked on 'Zi Qiu', and the infection was abolished ( Figure 3E). Hollow conidia, as well as some appresoria ( Figure 3F), were present at 18 hpi until 96 hpi ( Figure 3H,I). In contrast, fungal germination and infection was noted at 24 hpi on susceptible 'Riesling' leaves, which progressively increased until 96 hpi ( Figure 3R). Some hyphae were branched ( Figure 3Q) with apparent lesions forming. From 48 hpi onward, the fungus spread and showed sporulation on 'Riesling' ( Figure 3P).

Activity of Peroxidase and Superoxide Dismutase in the HR Genotype 'Zi Qiu' and the HS Genotype 'Riesling' Infected by B. cinerea
We measured the activities of superoxide dismutase (SOD) and peroxidase (POD) in the infected and control leaves. Stress conditions disrupt ROS production leading to plant cell death but plants exhibit an array of anti-oxidant enzymes to scavenge harmful ROS and protect cells from oxidative damage [19]. SOD activities in the 'Zi Qiu' (V. davidii) and 'Riesling' (V. vinifera) control samples were approximately the same, except for a slightly elevated level at 4 hpi ( Figure 4A) in 'Riesling'. The activity in the inoculated 'Zi Qiu' was approximately twice that of the control throughout the experiment ( Figure 4A). The SOD activity in 'Riesling' was similar to that in 'Zi Qiu' at 4 hpi ( Figure  4A), but then increased from 4 hpi to 18 hpi, where it peaked before decreasing again until 96 hpi.  We measured the activities of superoxide dismutase (SOD) and peroxidase (POD) in the infected and control leaves. Stress conditions disrupt ROS production leading to plant cell death but plants exhibit an array of anti-oxidant enzymes to scavenge harmful ROS and protect cells from oxidative damage [19]. SOD activities in the 'Zi Qiu' (V. davidii) and 'Riesling' (V. vinifera) control samples were approximately the same, except for a slightly elevated level at 4 hpi ( Figure 4A) in 'Riesling'. The activity in the inoculated 'Zi Qiu' was approximately twice that of the control throughout the experiment ( Figure 4A). The SOD activity in 'Riesling' was similar to that in 'Zi Qiu' at 4 hpi ( Figure 4A), but then increased from 4 hpi to 18 hpi, where it peaked before decreasing again until 96 hpi. exhibit an array of anti-oxidant enzymes to scavenge harmful ROS and protect cells from oxidative damage [19]. SOD activities in the 'Zi Qiu' (V. davidii) and 'Riesling' (V. vinifera) control samples were approximately the same, except for a slightly elevated level at 4 hpi ( Figure 4A) in 'Riesling'. The activity in the inoculated 'Zi Qiu' was approximately twice that of the control throughout the experiment ( Figure 4A). The SOD activity in 'Riesling' was similar to that in 'Zi Qiu' at 4 hpi ( Figure  4A), but then increased from 4 hpi to 18 hpi, where it peaked before decreasing again until 96 hpi.    Figure 4B). However, in inoculated 'Zi Qiu' leaves, POD activity increased from 4 hpi to a maximum of 48 hpi, and then decreased until 96 hpi ( Figure 4B).  Three independent experiments were used for the means and standard errors. Small alphabetes indicate significant differences according to LSD test (p < 0.05) between "Zi qiu" and "Riesling". Three independent experiments were used for the means and standard errors. Small alphabetes indicate significant differences according to LSD test (p < 0.05) between "Zi qiu" and "Riesling". . Jasmonic acid (JA) levels in highly resistant 'Zi Qiu' and highly susceptible 'Riesling' leaves at 0, 12, 24, 48, 72, and 96 hpi with Botrytis cinerea and using sterile water as the control. Three independent experiments were used for the means and standard errors. Small alphabetes indicate significant differences according to LSD test (p < 0.05) between "Zi qiu" and "Riesling". Table 2 shows the Pearson's correlation coefficient values of the antioxidant enzyme activities, ROS levels and JA levels in 'Zi Qiu 'and 'Riesling' leaves. Significant positive correlations for SOD and POD activities with H2O2 and O2 − levels were observed. SOD activity was significantly negatively correlated with JA levels, but positively correlated with POD activity.

JA
O2 − H2O2 Figure 6. Jasmonic acid (JA) levels in highly resistant 'Zi Qiu' and highly susceptible 'Riesling' leaves at 0, 12, 24, 48, 72, and 96 hpi with Botrytis cinerea and using sterile water as the control. Three independent experiments were used for the means and standard errors. Small alphabetes indicate significant differences according to LSD test (p < 0.05) between "Zi qiu" and "Riesling".

Discussion
Grape genotypes vary in terms of their infection resistance, degree of fungal colonization, and disease severity to B. cinerea [20]. Of the 81 different Vitis genotypes evaluated here, two were categorized as HR, twelve as resistant, twenty-five as S, and forty-two as HS (Table 1). Resistant genotypes towards B. cinerea have been found in Vitis species for example, V. davidii, V. vinifera and in the progeny of crosses between V. vinifera and species like V. labrusca (Table 1). Numerous wild Chinese Vitis species show multi-fungal disease resistance [21], and they have been described as important resources for future disease resistance breeding programs [22,23].
Discrete colonization of B. cinerea on grape leaves was studied by SEM at different time points. In 'Riesling', the pathogen initially showed limited infection, as indicated by necrosis prior to 24 hpi, but then spread substantially, and showed signs of sporulation. Prior to 24 hpi, fungal growth in 'Zi Qiu' leaves was significantly delayed, as evidenced by the lower germination and infection rates. Most B. cinerea appressoria on the 'Zi Qiu' leaves did not develop into infection pegs, in contrast to those on 'Riesling' leaves, and consistent with previous reports [11]. It was also reported that sporulation densities on V. davidii var. Langao-5 and V. pseudoreticulata var. Baihe-35-1 was significantly lower than those on the HS cultivar V. vinifera cv. Pinot noir [24]. Here, we saw that at the initial stages B. cinerea colonization halted in 'Zi Qiu' leaves.
We investigated the possible basis of differences in growth of B. cinerea in the HR genotype 'Zi Qiu' compared with HS 'Riesling'. ROS is commonly produced in response to pathogen attack [25,26] and overall, higher levels of ROS accumulated in 'Riesling'. This is not consistent with a study suggesting that in host-pathogen interactions where the pathogen is a necrotroph, pathogen-induced cell death and ROS accumulation promote pathogen growth and disease development. Thus, ROS facilitate colonization on the leaves by the necrotrophic fungus B. cinerea [27]. In contrast, low ROS levels were observed here post-inoculation in 'Zi Qiu', suggesting that the anti-oxidative system maintains redox equilibrium [26,28] and protects cells from ROS damage [29].
Oxidative stress disturbs the redox equilibrium in infected tissues, thereby promoting disease development [30]. In the current study, ROS accumulation after inoculation was detected in leaves from both genotypes, but with higher levels in 'Riesling'. We conclude that 'Riesling' suffered significantly from the continued presence of ROS, and that 'Zi Qiu' did not experience substantial oxidative stress due to a high and timely anti-oxidative capacity. H 2 O 2 higher or lower levels increase either the susceptibility or resistance respectively to B. cinerea, while, O 2 − serves as a first substrate for H 2 O 2 formation [13,30,31]. Some reports have suggested that O 2 − plays a role in supporting B. cinerea invasion [32,33]. H 2 O 2 production is induced in plant cells, accompanied by O 2 − generation, which can promote programmed cell death and disease lesion development, thereby increasing B. cinerea infection [27]. We propose that the high and low levels of ROS in 'Riesling' and 'Zi Qiu' contribute to their susceptibility and resistance to B. cinerea infection, respectively [34]. We evaluated ROS accumulation and antioxidant enzyme activities during the interactions with B. cinerea [33]. Low ROS production and a timely increase in the activity of anti-oxidative enzymes were associated with the strong pathogen resistance of 'Pingli-5' and the HS cultivar 'Red Globe', which suffers from severe infection and sustained ROS production correlated with comparatively unchanged anti-oxidative activities [11]. These results support the significance of the ROS response in a timely detection of and defense against B. cinerea. We saw that the post-inoculated 'Riesling' leaves showed a slight variation in POD activity with lesion development. However, they showed increased SOD activity, which corresponded well with H 2 O 2 production and a reduction in O 2 − . The POD activity in 'Zi Qiu' increased during the experiment, and no significant change was observed in SOD activity. Low levels of ROS accumulation are necessary for the anti-oxidative system to sustain redox equilibrium [26], and we also observed that the infected 'Riesling' showed evidence of an insufficient anti-oxidative system, resulting in consistently elevated ROS levels. In contrast, 'Zi Qiu' showed relatively rapid changes in anti-oxidative capacity, especially POD activity, following inoculation, and thus likely experienced less ROS-induced stress. Given that substantial levels of ROS were induced in 'Riesling' but not in 'Zi Qiu', we propose that the coordination between ROS production and scavenging mechanisms associated with the anti-oxidative system during biotic stress [35] may be a key factor in the ability of genotype 'Zi Qiu' to shield itself against B. cinerea. JA levels were measured in both the HR and HS genotypes, and higher levels found in 'Zi Qiu'. We noted high levels of JA in the 'Zi Qiu' control, which were approximately equal to the JA levels seen in inoculated 'Riesling' (Figure 5), indicating that a continuously high presence of JA in 'Zi Qiu' may contribute to controlling B. cinerea, and possibly other pathogens. These results are consistent with another study [36], which stated that high JA levels block B. cinerea infection and strengthen grape resistance to B. cinerea. Moreover, JA is known to be a major hormone involved in plant defense responses [37] and is a crucial component in the plant defense responses against insects and microbial pathogens [38]. JA accumulation occurs relatively quickly in plant tissues and cells after exposure to fungal elicitors [39,40]. JA is involved in plant response to injury and biotic stresses, such as occurs during insect and pathogen attacks [7,41], and is associated with resistance to biotrophic and necrotrophic pathogens [16,42].

Plant and Fungal Materials
Plant materials were obtained from the Grape Repository (34 • [43,44], and detailed information for each of the species is available in Table S1. The material was used to evaluate gray mold disease development in 2016 and 2017. B. cinerea spores were isolated from the seedless cultivar 'Flame' (V. vinifera) in a greenhouse located on the North campus of the Northwest A&F University, Shaanxi, China. Spores were cultured on a potato dextrose agar medium at 22 • C. After 20 days, the conidia were removed, and 1.5 × 10 6 spores mL −1 were prepared in sterilized water, since this has previously been identified as the optimum inoculum [11]. The conidial suspension was confirmed to have a conidia/spore germination percentage of 95% or more before all experiments.

Detached Leaf Evaluation
Leaves of the same size and age (from the shoot at nodes 3 and 4) were arbitrarily selected from the grape plants. The detached leaves were washed with distilled water. For laboratory assessment, 24 leaves from each of three replicates of each genotype were evaluated. The leaves were quickly transferred to trays with 0.8% agar and sprayed evenly with the conidial suspension. Control leaves were sprayed with distilled water. The trays were placed in an incubator with a relative humidity of 90-100%, for the first 24 h in the dark and then 12/12 h light/dark at 22 • C.

Disease Severity Rating
Disease severity was evaluated and rated as previously described [45,46] with slight modifications. The disease symptoms observed on the leaves were ranked from 1 to 7 (Rank 1 = 0.1-5.0%, 2 = 5.1-15.0%, 3 = 15.1-30.0%, 4 = 30.1-45.0%, 5 = 45.1-65%, 6 = 65.1-85.0% and 7 = 85.1-100.0%) on the basis of the estimated percentage of lesion over the entire leaf surface. The ranking was then converted into a severity index (SI) according to the formula: SI = ∑(Rank × number of infected leaves in that rank) Total number of leaves × highest rank × 100 The resistance level was rated into four classes on the basis of the SI values. Disease resistance levels of the different genotypes were categorized as highly resistant (SI: 0-1.50), resistant (SI: 1.51-3.50), susceptible (SI: 3.51-5.50), and highly susceptible (SI: 5.51-7.0). Susceptibility data for the disease were collected in 2016 and 2017. The average SI values of the two-year data were used to evaluate the resistance level.

Light Microscopy
Two representative genotypes from each category were used to characterize the colonization of the grape leaves by B. cinerea using light microscopy. The following genotypes were used for each category: HR, 'Zi Qiu' and 'Ju Mei Gui'; R, 'Kang San' and 'Rizamat'; S, 'Flame' and 'Canner'; HS, 'Riesling' and 'Pinot noir'. The leaves were cut into 2-3 cm 2 segments and fixed and decolorized in 100% ethanol and in saturated chloral hydrate. The samples were stored in 50% glycerol and stained with anilineblue solution at the time of observation with an Olympus BX-51 microscope (Olympus, Tokyo, Japan) [47].

JA Quantification in HR and HS Genotypes
JA levels were quantified in inoculated and control 'Zi Qiu' and 'Riesling' leaves that were collected at different time points (0, 12, 24, 48, 72, and 96 hpi) and immediately frozen in liquid nitrogen. The samples were ground in liquid nitrogen with help of pestle and mortar and then stirred in 80% methanol at 4 • C overnight. The extract was methylated as previously described [52] and JA was quantified with a competitive enzyme-linked immunosorbent assay (ELISA) assay [53].

Statistical Analysis
Experiments were performed using three biological replicates in a randomized design. Means and standard errors were computed from independent replicates using SPSS 13.0 (SPSS Inc., Chicago, IL, USA). Least significant difference (LSD) 0.05 was employed to compute significant differences, and correlation data of the resistance evaluation from 2016 and 2017 were analyzed. All images were processed with Adobe Photoshop (Adobe Systems Incorporated, San Jose, CA, USA). All graphs were prepared using the Origin Pro 2016 32-bit software (Shenzhen, Guangdong, China) and correlation analysis was performed using the Pearson coefficient.

Conclusions
In this study, we investigated the resistance levels of different Vitis genotypes to B. cinerea. Most genotypes were susceptible, but a detached leaf assay revealed high resistance in clone 'Zi Qiu' (V. davidii) of Vitis germplasm. The results were further investigated by comparing fungal growth, ROS responses, JA levels, and changes in the antioxidant system, between the HS V. vinifera 'Riesling' genotype and the HR V. davidii 'Zi Qiu' genotype after B. cinerea inoculation. We observed that low ROS production, rapid elevation in antioxidant activities and high JA levels were associated with a high level of fungal resistance in 'Zi Qiu'. In contrast, the HS genotype 'Riesling' suffered from severe B. cinerea infection and sustained ROS production, together with relatively unchanged anti-oxidative activities and low JA levels. This study provides insights into B. cinerea infection of grape leaves, as well as information that may be valuable to breeders in selecting germplasm for increased disease resistance.