Tomato plants of Pusa Ruby variety (PR) (Sutton Pvt. Ltd., India) was grown on commercial soilrite (a mixture of horticulture grade perlite, Irish peat moss and exfoliated vermiculite in 1:1:1 ratio) and maintained in plant growth chambers with 16 h light and 8 h dark cycles, at a temperature of 26 ± 1°C under a light intensity of 50 μM m^–2 s^–1. Pure culture of the pathogen Rhizoctonia solani Kuhn. (AG1-1A isolate of the fungus was obtained from Rice Research Station, Chinsurah, West Bengal, India, Basu et al., 2016) was used for this study. The fungus was maintained at 28°C in potato dextrose agar plates. For inoculation of plants, both sclerotia and mycelial discs of 3 mm diameter from the growing edge of 10-day-old culture were used according to our previous report (Basu et al., 2016).

Preparation of hormone solutions was done as per our earlier published article (Chowdhury et al., 2017a). Tomato plants were primed with either of the two phytohormones, MeJA (Sigma-Aldrich) or SA (Himedia, India). For priming of plants, 0.08 mM concentration of the phytohormones was used as it did not affect the fungal growth when the media was supplemented with the phytohormones at this concentration (Supplementary Figures 1–3). One hundred millimolars of phytohormone stocks were prepared by dissolving in ethanol. This stock was used to make the final concentrations in the experiment in distilled water. Hormone solutions were sprayed according to Chowdhury et al., 2017a on 4-week-old plants and covered for 24 h prior to inoculation with R. solani. Control plants of similar age were sprayed with the diluted solvent only, without the phytohormones.

This study was carried out using detached leaf assay according to our earlier published protocol (Basu et al., 2016). For behavior of pathogen in the vicinity of leaves, fully expanded leaves from 4-week-old plants were placed on water agar media in petri plates, keeping the dorsal sides up. One 3 mm mycelial disc was placed 1 cm away from the edge of the leaf and allowed to grow toward the leaf. Three microscopic fields were observed for each leaf at 3 and 5 dpi. Photographs were taken with the help of a stereomicroscope (Radical Ltd., India). Results were obtained from three independent experiments three replicates each.

This was done according to our protocol (Basu et al., 2016). Sclerotia of similar sizes were placed singly at the center of tomato leaves of control, MeJA, and SA-primed leaves. Leaves were incubated at 26°C (± 1°C) in petri plates with humidity maintained by moist tissue papers. Germination of these sclerotia and progression of the emerging hyphae was observed at 4-h intervals (4, 8, 12 and 24 hpi) by stereomicroscope (Radical Ltd., India) and compound microscope (Leica DMLS). For each priming regime, three independent experiments with three replicates each were used for data analysis.

Leaves were prepared for scanning electron microscopy (SEM) according to Basu et al., 2016. First, the leaves were fixed using 3% glutaraldehyde in0.1 M sodium cacodylate buffer (pH 7.2) overnight at 4°C. The next day, the excess fixative was removed by washing the leaves thoroughly with0.1 M sodium cacodylate. Then, the leaves were dehydrated with a serial dilution in an ethanol series of 25, 50, 70, 85, 95, and 100% v/v. The plant samples were processed through a critical point drying technique and then mounted on metal stubs coated with gold. Samples were observed under a scanning electron microscope (Carl Zeiss EVO 18, Germany).

For confocal microscopy, tomato plants were inoculated with mycelial discs of R. solani transformed with gfp gene. R. solani had been previously transformed in our laboratory by Basu et al., 2016. Infected leaf and stem samples were processed for removal of chlorophyll through acetoethanol (1:3, v/v). All the samples were subjected to CLSM analysis using the model IX81, Olympus Singapore Pte Ltd., equipped with FLUOVIEW FV1000 software. Representative fluorescence fields were chosen from at least three independent plants. For visualization of green fluorescence protein (GFP) fluorescence, the excitation wavelength was 473 nm and the emission window was set at 485–585 nm.

Disease index was calculated according to Basu et al., 2016. The characteristic brown colored lesions that developed on tomato leaves during R. solani infection was considered for disease index study on both detached leaves and whole seedlings. Nine plants were observed, and the experiment was repeated four times. Six-week-old seedlings were inoculated with single sclerotium at the center of each leaf to calculate disease index of whole seedling. For detached leaf assay, each leaf was inoculated with a sclerotium at the center. After inoculation, the whole seedlings and detached leaves were incubated at 26 ± 1°C in humid condition. The area of the brown necrotic lesions developed on seedlings and detached leaves were calculated and represented as the percentage of necrotic area against the whole leaf area. Disease index necrotic areas of the infected leaves at 2 or 3 dpi was graded into five classes (0 = no infection, 1 = 1–25%, 2 = 26–50%, 3 = 51–75%, and 4 = 76–100% infected leaf area). Disease index (DI) was calculated according to Basu et al., 2016.

Tomato leaves of control and experimental sets were collected at different time points post-inoculation (12, 24, 48, and 72 hpi). After removal of chlorophyll with acetoethanol solution (1:3), leaves were stained with lactophenol-trypan blue solution (0.5% trypan blue, lactic acid:phenol:H₂O 1:1:1 v/v/v) and kept overnight. The excess stain was later washed off with distilled water and the leaves were mounted with 50% glycerol on glass slides for viewing under compound microscope (Olympus BX-51). Number of infection cushions per microscopic field on the leaf surfaces was counted. Each leaf was scored for three microscopic fields. Three leaves taken from three different experimental sets each of control and primed plants were evaluated for each time point. Evaluation was done according to Basu et al., 2016.

Cellular deposition of phenolics, peroxide, and callose was detected by toluidine blue and DAB stain according to our protocol (Ray et al., 2015), and aniline blue (Chowdhury et al., 2017a) respectively. Six independent experiments with two leaves each were used for polyphenol and peroxide accumulation study. For both assays, chlorophyll from leaves was removed with 1:3 aceto-ethanol solution. For detection of polyphenol, leaves were immersed in 0.05% toluidine blue stain (dissolved in 0.1 M potassium phosphate buffer, pH 5.5) for 6 hours. Accumulation of phenols was indicated by blue patches formed on the leaf surface.

Peroxide was detected by soaking the leaves in 1 mg/ml 3’3-Diaminobenzidine (DAB) solution (pH 3.8) for 8 h. The dark brown patches on the leaf surface indicated the accumulation of peroxide. Data were presented as the percentage of colored area with respect to total leaf area.

Pathogen-mediated callose deposition on the leaves was observed under a fluorescence microscope (Leica DMLS, excitation maximum 330–385nm, dichroic mirror DM 400, barrier filter > 420 nm). Deposited callose in the leaves was observed for the bluish-green fluorescence developed after aniline blue staining. Four leaves were observed with three microscopic fields from each leaf according to Chowdhury et al., 2017a and representative fields were considered.

These assays were done according to our standardized protocols (Ray et al., 2015; Chowdhury et al., 2017a,b) with three independent experiments using at least three individual plants in each experiment. For quantification of total phenol, leaves were homogenized in 80% methanol. Tissue extracts were then mixed with water and folin-ciocalteau (Sigma Aldrich) reagent followed by 5 min of dark incubation. The reaction was started after the addition of 20% sodium-tartarate followed by 30 min of incubation at room temperature. Absorbance was recorded at 720 nm, and the amount of total phenol was expressed per gram of fresh tissue against the standard curve of gallic acid.

Flavonoid was quantified according to our protocol (Chowdhury et al., 2017a,b), by extracting plant tissue in 80% ethanol. A reaction mix was made containing 500 μl tissue extract along with 10% aluminum chloride, 1 mM potassium acetate, and distilled water to a final volume of 5 ml. Data was collected as the absorbance at 415 nm and represented per gram of fresh tissue according to our Chowdhury et al. (2017a,b).

Proline and malondialdehyde contents were assayed in accordance with Chowdhury et al. (2017b).

Total chlorophyll content was determined by extraction of fresh leaves in 80% acetone according to Chowdhury et al. (2017b).

This was done according to our published protocol (Chowdhury et al., 2017a,b). Total RNA was isolated from leaf samples by Trizol (Invitrogen) method, followed by quantification with a nanodrop spectrophotometer (Eppendorf). Two micrograms of RNA were used to synthesize cDNA using random primers. Real time PCR (RT-qPCR) was carried out in 96-well blocks in 20μl reaction volume (Applied Biosystems 7500 Real-Time PCR System, USA) using Power SYBR Green master mix (Applied Biosystems, USA). Transcript sequences were amplified by gene-specific primers. Relative fold changes in gene expression were determined by a standard method, particularly by calculating ΔΔCt value. The specificity of the PCR reactions was adjusted based on a melting curve analysis of the amplified products utilizing the standard method installed in the system. The pathogen gene 18s rDNA was used as an internal control for the pathogen (Sayler and Yang, 2007). Tomato actin gene was used as internal control according to our protocol (Ray et al., 2015). The list of primers both for the pathogen and the tomato plants used in this study is given below.

List of primers used for quantitative real-time PCR:

The effect of MeJA and SA on the growth of R. solani was determined by allowing the fungus to grow on potato dextrose agar (PDA) plates supplemented with varying concentrations of MeJA or SA. The concentrations used were 0.05, 0.08, 0.1, 0.5, and 1 mM for each of the phytohormones. A mycelial disc of 3 mm diameter from the growing culture of R. solani was placed in the center of each phytohormone supplemented PDA plate. These plates were incubated in 28°C for the next three days and colony diameter was recorded. Data were collected from three independent replicates.

Salicylic acid (SA) and JA were extracted from tomato leaves according to de Sa et al. (2013). For extraction of SA and JA, 500 mg tissue was ground in liquid nitrogen. Ground tissues were homogenized in 85% MeOH (v/v). Samples were then centrifuged at 10,000 rpm for 10 min at 4°C. The supernatants were mixed with0.2 mol/L NaOH. The mixtures were filtered using a Whatmann grade 1 filter paper. Filtrates were mixed with 5% Trichloroacetic acid (TCA) and were partitioned using ethyl acetate and cyclohexane (1:1 v/v). The organic phases were collected, and 0.1 mol/L sodium acetate was added and then dried. Dried samples were suspended in 10% methanol for analysis with HPLC. Three sets of extracted samples for each time point and SA and JA standards were analyzed with HPLC using standard calculations according to Trapp et al. (2014). HPLC system (Agilent Technologies, USA) with 4.6 × 100 mm, 3.5 μm column was used. Formic acid (0.05%, v/v) and MeOH with0.05% (v/v) of formic acid were used as mobile phases A and B, respectively. The elution gradient used was 0-10 min, 42-55% B in A; 0-13 min, 55-100% B; 13-15 min 100% B; 15-15.1 min 100-42% B in A; and 15.1-20 min 42% B in A at a constant flow rate of 1.1ml/min. All the chemicals were purchased from Sigma Aldrich and were of HPLC grade as required.

All the data were analyzed using software according to our previous publications (Chowdhury et al., 2017a,b). Each experiment was done in a completely randomized design (CRD) with three independent experiments with three to four replicates each and values represented as the standard error mean ± SEM. The data were subjected to one-way analysis of variance (ANOVA) with different letters indicating significant differences between treatments at p < 0.05, according to Duncan’s multiple range test (DMRT), using a software package, SPSS version 16, 2007.

We wanted to observe the behavior of R. solani and see if the fungus showed any preference toward unprimed versus primed host. The fungal hyphae were allowed to grow toward the host leaves from a distance (Figures 1A–F). We found that there was a marked difference in the behavior of R. solani in the vicinity of primed versus unprimed leaves which became more visibly pronounced at 3 and 5 dpi when the hyphae reached closer to the leaves (Figures 1A–F). In the case of control leaves, numerous hyphae approached the leaf, some of which almost touched the leaf edges by 3 dpi. In contrast, only a few hyphae reached the vicinity of MeJA and SA pretreated leaves (Figure 1G). Almost a 3 times greater number of hyphae grew in bulk toward the control leaves, straight from the germinating sclerotia approaching the leaf edge in a perpendicular direction by 5 dpi (Figure 1H). The SA-primed leaves were least preferred by the pathogen at 3 and 5 dpi (Figures 1G,H). In the case of the SA-treated leaves, the hyphae ran parallel to the leaf margin at a distance, avoiding direct contact even at 5dpi (Figure 1F). The infection progressed quickly in control leaves such that the leaves started to necrose on the 5th day of the experiment (Figure 1B). In congruence with the behavior of the pathogen in the vicinity, the phytohormone treated leaves remained green and fresh even at the end of the experiment at 5 dpi (Figures 1D,F).

For R. solani, the compacted mass of hyphae forming the sclerotia serve as primary inocula. In order to find out if host priming affected the behavior of the sclerotia, the steps of the sclerotial germination was studied on primed and unprimed control leaves. The lengths of hyphae from germinating sclerotia were measured at fixed time intervals in three experimental sets of control, MeJA, and SA-primed host leaves (Figure 2). In the control leaves, the sclerotia started to germinate as early as 4 h post-inoculation. The hyphae continued to spread out at 8 and 12 hpi. At 24 h, the sclerotia became obscured in the surrounding dense cottony mass of hyphae (Figure 2A). In contrast, the sclerotia did not germinate at 4 hpi and barely started to germinate at 8 hpi on the primed leaves. The emerging hyphae were visibly spreading onto the leaf surface only at 12 and 24 hpi, with much less growth compared to the control. The most surprising observation was that sclerotial germination was least in the SA pre-treated leaves compared to the MeJA pre-treated. There was nearly double the growth of hyphae from sclerotia placed on MeJA-primed leaves compared to SA-primed leaves. This shows that the MeJA pre-treatment was less of a deterrent compared to SA pre-treatment (Figure 2B).

In the previous experiment, since the R. solani sclerotia showed least growth on SA pre-treated leaves, we wanted to observe in greater detail the behavior of sclerotia and the emerging hyphae on SA-primed leaves using SEM and confocal microscopy (Figure 3). At 24 hpi on the control sets, numerous infection hyphae had emerged from the sclerotia (Figure 3A). The lateral branches stayed in intimate contact with the host leaf surface and formed hyphal aggregation at places shown by arrows in Figure 3A, giving rise to infection cushions. In contrast, the sclerotia on SA pre-treated leaves showed only a few emerging hyphal strands with complete absence of any aggregation (Figure 3B). At 48 hpi, the infection hyphae already formed infection cushions on the control leaves (Figures 3C,E), while on SA-primed leaves, the hyphae grew in straight lines avoiding intimate contact with the leaf surface (Figure 3D) with rare hyphal aggregates (Figure 3F). Fluorescence microscopy of control and SA-primed leaves infected with gfp transformed R. solani at 24 hpi showed typical right angles branching of hyphae with septa near the branch points (Figures 3G–J). The hyphae formed dense mats with profuse branching on the control leaves while showing scant growth on the SA-primed leaves (Figures 3G–J).

During the establishment phase of the fungus on the control leaf surface, the hyphae interacted with the trichomes of the control leaves from 12 hpi (Figure 4A) and formed a hyphal network closely pressed to the epidermal terrain by 24 hpi (Figure 4B). By 48 hpi, the pathogen formed infection cushions that coalesced to form larger masses on the control leaves (Figures 4C,J). On the MeJA-primed leaves, the hyphae showed superficial interaction with the trichomes and avoided close contact with the leaf surface, only forming hyphal aggregated at places at 48 hpi (Figures 4D–F,J). The hyphae on the SA-primed leaves showed no interaction with trichomes at 24 hpi and showed least hyphal growth even at 48hpi (Figures 4G–J).

The above behavior of the pathogen on the leaf surface was also reflected on the stem and root surfaces. The dense, profuse fungal growth on the surface of the stem in the control plants showed close interaction with the stem trichomes (Figure 5A). On MeJA pre-treated leaves, the hyphal growth was reduced to a loose dispersed network with superficial interaction with stem trichomes (Figure 5B). In the case of SA pre-treated plants, an interesting behavior was observed. Here, the mycelium strands grew parallel to each other without producing many lateral branches (Figure 5C). It appeared like the pathogen in this particular phytohormone treatment and tried to grow forward without branching. Similar behavior was also noticed in the case of roots (Figures 5D–F). Cross-sections through stems of control plants showed intracellular growth of hyphae (Figure 5G, solid arrow) and thick bunches of hyphae growing intercellularly outlining the cortical cells (dashed arrow). The primed stems showed thin hyphae with intracellular growth at places in the cortical tissues (Figures 5H,I).

Confocal laser scanning microscopy (CLSM) of leaves and stems of plants inoculated with transformed R. solani showed significant establishment of the fungus on the control leaves and stems, less infection in the MeJA-primed plants, and least in the case of SA-primed plants (Figures 6A–L). As seen with SEM, confocal microscopy of cross-section of the control stem revealed significant interaction of fungal hyphae with the trichomes and invasion of underlying tissues (Figures 6M,N).

The avoidant behavior of the pathogen on the primed host surface was reflected in the reduction of overall disease severity, especially in the SA-primed plants. Disease Index was measured by detached leaf assay (Figure 7) and by whole seedling assay (Figure 8). Phytohormone treatment significantly delayed the onset of symptoms which was evident even from the first-day post-inoculation. Necrotic symptoms started to be visible at 2 dpi in the case of control plants (Figure 7A), but these lesions quickly increased in size so that by 3 dpi, the control leaves were almost entirely necrosed (Figure 7A). In comparison with the control, the MeJA pretreated leaves showed almost no necrotic symptoms up to 2 dpi and only on the 3rd day after inoculation, did the MeJA pretreated leaves show small necrotic spots. The SA pre-treated leaves showed yellowing at 3 dpi with very few necrotic lesions (Figure 7B). The SA-primed plants also had the least disease index (Figure 7C).

The whole seedling assay under these experimental conditions showed similar results. The MeJA pre-treated plants showed less severe symptoms than the control and more severe than SA-treated ones at the corresponding time points (Figure 8A). The severity of the disease was in the order of Control > MeJA > SA from maximum to minimum (Figures 8A,B). Overall, at 2 dpi, there was about 50% less disease index in the SA-primed leaves compared to MeJA-primed leaves and almost 40% less at 3 dpi (Figure 8B).

To see if the initiation of the infection cushions and appressoria was altered by the priming status of the host, the initial stages of establishment of the pathogen were observed over three days. Development of lobate appressoria started as early as 12 hpi in the case of control leaves (Figure 9A). It then advanced rapidly over time and ended up in the formation of dense coalesced masses of hyphae by 72 hpi (Figure 9D). On the surface of MeJA and SA pretreated leaves, however, the pathogen could not initiate the formation of infection structures up to 48 hpi, and infection cushions started to develop as late as 72 hpi (Figures 9E–L). The frequency of infection cushion was quantified as the number of infection cushions developed per unit area on the leaf surfaces. The development of infection cushion on the control was three times greater than that of the two phytohormone pretreated leaves at 72 hpi (Figure 9M). When compared between the two primed leaves, SA pretreatment supported approximately 25% less infection cushions compared to the MeJA pretreated leaves.

To confirm the effect of pre-treatment on the formation of infection cushion, we observed the expression of the Appressorial Penetration-associated gene (GAS1) in the pathogen. This gene was expressed significantly in the control plants from 4 hpi onward, whereas it was significantly less in MeJA and least in the SA-primed background (Figure 9N). This low expression of the GAS1 gene directly correlated with a much lower number of infection cushions in MeJA and SA-primed plants than control at all time points (Figure 9M).

Callose was found to be deposited in several parts of primed tomato leaves after R. solani infection. Stomata, veins, and trichomes were most frequently found to be the sites of greater deposition (Figure 10). We observed interesting differences in callose deposition in response to pathogen under different priming status of the host. The stomatal fluorescence was more intense in the case of SA pretreated leaves compared to control leaves and MeJA-primed leaves (Figures 10A–I). In MeJA-primed leaves, callose deposition was prominent only at 3 dpi (Figure 10F) where both of the guard cells of each stoma fluoresced (Figure 10J). At 1 dpi, in control and SA pretreated leaves, most guard cells fluoresced only at the two poles of the stoma, which increased considerably with increasing time (Figure 10J). Following this stage, only one of the pair of guard cells in each stoma fluoresced in both control and SA plants at 2 dpi. At 3 dpi, in SA-primed leaves, the callose deposition spread to the other guard cell, making a completely fluorescing stoma. In the control sets, by 3 dpi, the leaves showed necrosis and no stomatal fluorescence was detectable (Figure 10J).

More deposition of callose was observed on the veins and trichomes in the case of SA-primed leaves compared to the MeJA and Control leaves. In the MeJA pre-treated leaves, the fluorescence was intermediate (Figure 10K).

Secondary metabolites are an important component of inducible defense, and phytohormone priming is known to induce their production. These parameters were compared in the MeJA and SA-primed backgrounds after infection to understand why SA imparted better tolerance to the necrotroph, contrary to expectation. Differential accumulation of phenol was recorded in response to R. solani infection on the leaves of control and phytohormone treated leaves via toluidine blue assay. SA pretreatment was found to be highly inducive to the increase of the total phenol content in leaves in the early phase of infection at 2 dpi, showing significantly more phenol accumulation than the control and MeJA pre-treated leaves (Figure 11). Interestingly, in the case of MeJA pretreatment, the accumulation was less than that of SA at 2 dpi but increased in the later stages of infection at 3dpi (Figure 11).

Secondary metabolites, especially polyphenolic compounds, are an integral part of the plant basal defense. The total phenol content in tomato plants during R. solani pathogenesis assayed through Folin Ciocalteau method remained almost unaltered up to 24 hpi in control and MeJA-primed plants but was elevated in SA-primed plants (Figure 12A). As with toluidine blue assays, the order of phenol deposition was SA > MeJA > control at 24 and 48 hpi. However at 72 hpi, MeJA pretreatment was found to be highly inducive to total phenol content, and the pattern is altered to MeJA > SA > Control. In the case of flavonoid content, the difference was in the order SA > MeJA > Con for almost all time points, although the difference was more pronounced in 48 hpi (Figure 12B).

The above result of enhanced polyphenol content was corroborated by altered regulation of two key flavonoid synthesizing gene, namely, Chalcone biosynthetic genes CHS1 and CHS2. CHS1 and CHS2 did not significantly change at early time points of pathogenesis and were only upregulated at 24 hpi. At 24 hpi, the expression of these genes was significantly more in SA pre-treated plants than MeJA (Figures 12C,D). The differences were nearly two-fold between SA and MeJA treatments in the case of CHS2 (Figure 12D).

Chlorophyll content was found to be significantly affected after infection. Chlorophyll was degraded vigorously with the progresses of pathogenesis in control plants. However, both MeJA and SA priming successfully limited the extent of degradation of Chl a throughout the course of pathogenesis (Figure 12E). Interestingly Chl a was least degraded in SA-treated plants in the early time points of infection at 24 and 48 hpi, which evened out at 72 hpi as infection set in for all three sets. Chl b was also degraded gradually after infection. However, it did not differ significantly between sets (Figure 12F).

Reactive oxygen species (ROS) is one of the key contributing factors in determining the successful establishment of the pathogen. The amount of ROS was determined by DAB staining, which was detected as brown patches on the leaves. Control leaves had significantly higher amount of H₂O₂ at all time points compared to primed leaves as expected (Figure 13). Surprisingly, MeJA pre-treated leaves showed almost double the amount of H₂O₂ than that of the SA pre-treated leaves at different time points (Figure 13I), indicating that SA pretreatment created less stress and offered better protection against the pathogen than MeJA. The correlation of H₂O₂ content with degradation of chlorophyll and the pathogen as the source of the excess H₂O₂ have been discussed in later sections.

Significant reduction of oxidative burst by phytohormone priming was also validated by suppression of lipid peroxidation (MDA) (Figure 13J) and proline content in the primed sets (Figure 13K). The results were similar to H₂O₂ accumulation. Evidently, salicylate pretreatment can better limit the ROS production stimulated by R. solani infection than Jasmonate pre-treatment which also explains the overall least disease index in SA-primed plants.

To date, there is no in-depth study on the MeJA and SA-mediated defense gene regulation during post-necrotrophic infection in plants primed with SA compared with MeJA priming. To investigate the basis of differential defense responses to R. solani when primed with SA and MeJA, we selected a total of 10 marker genes from salicylate and jasmonate signaling cascades. Their expression profiles were monitored over a time course during pathogenesis in these primed backgrounds.

We included the F-box protein CORONATINE INSENSITIVE 1 gene (COI1), the repressor protein JASMONATE ZIM DOMAIN gene (JAZ), along with Allene Oxide synthase gene (AOS), a leucine-rich repeat systemin receptor kinase (LRR-RK) gene (SR 160), and a protease inhibitor gene (PIN II) as important components of MeJA signal transduction pathway.

There was considerable upregulation of the COI1 gene within 24 h of spraying of phytohormones, i.e., at 0 hpi, with SA priming resulting in nearly double COI1 transcript levels compared to MeJA priming (Figure 14A). After SA pre-treatment, the infection resulted in two peaks of COI1 transcripts at 8 and 24 hpi while in MeJA-primed plants, COI1 showed a steady rise after an initial dip at 4 hpi, which is 28 h post priming. The COI1 transcripts remained high at 48hpi, then went down at 72 hpi near to the control level (Supplementary Figure 1A).

Similar biphasic upregulation was observed in the case of JAZ in SA-primed plants (Figure 14B). JAZ was significantly upregulated at 24 hpi in the MeJA-primed plants coinciding with COI1, likely as a means to regulate JA signaling at this time. Interestingly, in SA-primed plants, although COI1 was significantly upregulated at 24 hpi, the repressor, JAZ, was not (Figures 14A,B). Moreover, this low level of expression of JAZ was maintained in the next two time points of 48 and 72 hpi in the SA-primed plants (Supplementary Figure 1B). Hence, contrary to expectation, this shows that R. solani infection after priming with exogenous SA induced the JA marker genes even more than priming with MeJA itself in the early time points post-infection, i.e., in the first eight hpi. MeJA priming, on the other hand, induced the JA marker genes more than SA priming, but only at 12 hpi onward. This is contrary to expectation because R. solani is considered as a necrotroph and not a hemibiotroph. This will be discussed in greater detail in the next section. Overall, these results indicate that along with the conventional COI1 dependent activation of jasmonate signaling against the invading pathogen, SA priming played an even greater role in the early time points post-infection than MeJA priming. This explains why the disease index was least for SA-treated plants compared to the control and MeJA pre-treatment (Figures 7, 8).

Jasmonic acid (JA) biosynthesis gene AOS was upregulated at 24 hpi in all experimental sets. However, AOS transcript levels were highest in the case of jasmonate priming at that time point, in the order MeJA > SA > Con sets (Figure 14C). At 48 hpi, AOS transcript level was doubled in MeJA-primed plants which declined sharply in the next time point (72 hpi). Contrastingly AOS induction decreased in both control and SA-primed plants in the next two time points (Supplementary Figure 1C).

One interesting fact was observed in the case of expression of the early gene Systemin Receptor 160 (SR160) (Figure 14D). In the control plants, SR160 was upregulated within 4 hpi and then, the transcript level gradually fell. Contrastingly, exogenous application of MeJA and SA was able to delay the upregulation of the Systemin receptor gene till 24 hpi. At the next time point of 48 hpi, SR160 transcript level peaked in MeJA-primed background (Supplementary Figure 1D). One explanation would be that the phytohormone pre-treatment delayed the disease progression and hence the release of systemin from infected tissues was automatically delayed. Most importantly, SA pre-treatment showed significantly less SR160 transcript levels even at 24 hpi compared to MeJA-primed plants, corroborating the fact that SA priming contributed to greater tolerance to the pathogen than MeJA priming.

A protease inhibitor gene (PIN II) was significantly upregulated in the MeJA-primed plants from 0 hpi, i.e., from 24 h after priming (Figure 14E). The transcript level remained significantly elevated up until 8 hpi and then was reduced from 12 hpi onward. In the control and SA-primed plants, PIN II was upregulated during 4-12 hpi although not nearly as high as the MeJA-primed plants. In the later time points, the expression of this gene remained at levels nearer to that of the control in all sets (Supplementary Figure 1E).

In general SA-mediated defense signaling is thought to be directed against hemibiotrophic and biotrophic pathogens. To compare defense response against R. solani after priming with SA and MeJA, we included five SA pathway marker genes, including two major SA biosynthetic genes, viz., Isochorismate synthase (ICS), Phenylalanine ammonia lyase 5 (PAL5), along with important components of SA signal transduction pathway such as BA/SA carboxyl methyltransferase 1 gene (BSMT1), Pathogenesis-related protein-1a gene (PR1a), and the Phytoalexin-deficient 4 gene (PAD4).

It was observed in the case of SA pre-treated plants that ICS (Figure 15A) and PAL5 (Figure 15B) were both activated during the early time points of pathogenesis, especially at 8 hpi. Later on, at 24 hpi, the SA production was shifted mainly through the isochorismate pathway in the SA-primed plants showing a high ICS transcript level, which gradually went down in the 48 and 72 hpi (Supplementary Figure 2A). Interestingly, in MeJA pre-treated background, the plants relied solely on the PAL5 pathway for SA biosynthesis as evidenced by its induction throughout, especially high induction at 24 hpi (Figure 15B), while avoiding the ICS pathway as there is little or no expression of this gene in MeJA-primed plants (Figure 15A). PAL5 subsequently went down in MeJA-primed plants. In the SA-primed plants, PAL5 rose till 48 hpi before going down at 72 hpi (Supplementary Figure 2B).

The expression of the SA regulatory gene BSMT (required to mobilize SA out of the site) in the control plants showed a peak at 8 hpi after that it declined gradually till 24 hpi (Figure 15C). For SA-primed plants, BSMT remained low in its expression until at 72hpi in which it reached control levels. In MeJA-primed plants BSMT showed a sharp rise at 24 hpi likely to mobilize SA out of the site as JA defense sets in at this time point. In the next two days BSMT expression declined gradually in the MeJA-primed plants (Supplementary Figure 2C).

One of the key marker genes for SAR is pathogenesis-related proteins (PR) which is not only diverse in their genetic constituents and functions, but also in its response to different components as well as different pathogens. In this study, the acidic PR protein gene, PR1a has been found to be induced gradually in the primed plants while there was less expression in the control plants. The PR1a expression showed a steady gradual increase in response to infection in SA-primed plants and became significantly higher than MeJA-primed plants at 24 hpi (Figure 15D) as expected. Over the next two days, PR1a expression level gradually went down in the SA-primed plants, whereas it was significantly low for control and MeJA-primed plants (Supplementary Figure 2D). In the case of PAD 4, there was many-fold increase in the SA-primed plants at 24 hpi compared to both MeJA and control (Figure 15E). This elevated PAD4 level was maintained at 48 hpi also and then started declining in the SA-primed plants (Supplementary Figure 2E).

Time course quantification of phytohormones in plants after R. solani infection showed a spike in SA content to about 15,000ng/gm fresh weight of leaves at 24 hpi in the SA-primed plants (Supplementary Figure 3A). In the next phases of up to 72 hpi, a decline in the SA content was seen in these plants. The JA content gradually increased and reached a peak amount of about 5,200ng/gm at 48 hpi in these SA-primed plants (Supplementary Figure 3A). On the other hand, although a similar trend was seen in the case of the JA-primed plants, the peak of SA reached only 5,300 ng/gm at 24 hpi which is only 1/3 of the amount seen in the SA-primed plants (Supplementary Figure 3B). There was considerable increase in JA content right from 0 hpi (i.e., 24 h after priming), reaching a peak at 48 hpi with a JA content of 8,500ng/gm in the JA-primed plants. So, in both SA and JA-primed plants, there was a peak in SA in early phase of infection (24 hpi) and a later peak in JA at 48 hpi. However, the SA peak in the SA-primed background was three times more than that of JA-primed plants (Supplementary Figures 3A,B). This indicates the importance of increased amounts of SA in the early phases of R. solani infection, even more so than the amount of JA in the later phases, which was likely a decisive factor in the outcome of the disease. In the infected control plants, the amounts of both SA and JA were significantly less in all-time points, although the profile was similar. The mock plants (no priming, no infection) did not exhibit significant differences throughout the time-course in terms of SA and JA content, which was less than even the control infected plants.

Quantification of SA and JA was also done on plants that were primed but not infected to see the effect of priming alone (Supplementary Figure 4). Priming with SA (without infection) contributed to elevated level of SA content in leaves and it reached approximately 5,000 ng/gm of fresh weight 24 hours post-priming. Subsequently, the SA content gradually went down (Supplementary Figure 4A). Priming with JA (without infection) also contributed to increase in SA level, although it was significantly less than that of SA priming. In the case of priming with JA, an initial sharp rise in JA was observed amounting to approximately 6,800 ng/gm, but at later stages, there was rapid decline (Supplementary Figure 4B). The SA priming also resulted in the increase in the JA content, although it was not nearly as high as JA priming did.

In order to check if the phytohormones methyl jasmonate or salicylate had any effect on the growth of R. solani, the pathogen was grown on PDA that was supplemented with either MeJA or SA at five different concentrations, viz., 0.05, 0.08, 0.1, 0.5, and 1 mM. Mycelial growth was recorded for 3 consecutive days (Supplementary Figures 5, 6). Results show that MeJA had no effect on the growth of mycelia up to the concentration of0.1mM (Supplementary Figure 7A), which was above the concentration of0.08 mM used in the priming experiments. For SA, there was no effect on the mycelial growth at a concentration as high as0.5 mM, which was above the concentration of0.08 mM used for priming of tomato plants (Supplementary Figure 7B).