Two rhizospheric compatible bioagents P. fluorescens (OKC; GenBank accession JN128891) and T. asperellum (T42; GenBank accession JN128894) were used in the study (Patel et al., 2016a). A susceptible pea (Pisum sativum) cultivar AP3 was used as host of the obligate fungal pathogen E. pisi. Seeds of cv. AP3 were bioprimed with Trichoderma and Pseudomonas cells singly as well as in combination following the method reported by Yadav et al. (2013). Pseudomonas culture was inoculated in King’s B broth and incubated in incubator shaker at 27 ± 2°C for 24 h and then bacterial cells were harvested by centrifuging at 10000 rpm at 4°C. After harvesting, the cells were washed with distilled water (DW) three times and sterilized distilled water was added till the OD (optical density) 0.391 standardized for cell numbers 10⁸ ml^-1. A 5 mm mycelia disk of Trichoderma was inoculated on PDA in Petri plates after 6 days when the plates were filled with green spores of the fungus. The spores were harvested by adding an adequate amount of sterilized DW and filtered through sterilized cotton to remove mycelia debris. DW was added further till OD is 1.141 standardized for spore numbers 10⁸ ml^-1 of the suspension. Equal amount of both the spore suspensions were mixed for the co-inoculated treatment and CMC (carboxymethyl cellulose) was added @ 0.1% as sticker. Surface sterilized (0.1% HgCl₂) pea seeds were added to the culture suspensions and incubated at room temperature for 4–6 h for soaking. Soaked seeds were sown in sterilized soil amended with 20% vermicompost. All the pots having treated and untreated seeds were kept in growth chamber at 21 ± 2°C and 16/8 h light/dark ratio. The powdery mildew pathogen E. pisi, which is an obligate fungal parasite was collected from agricultural farm of Banaras Hindu University and inoculated on the treated plants by dusting with the help of a fine brush. Pathogen inoculation was done after 21 days of plant growth as described by Singh et al. (2002).

Pea leaves were collected after 24 and 48 h of pathogen inoculation for harvesting of RNA, studying the germ tube behavior on E. pisi conidia as well as for enzyme assays. Plant stem samples were also collected for histochemical staining for lignifications after 15 days of pathogen inoculation. Leaf samples were collected from both microbial treated and non-treated plants with or without pathogen challenge.

Leaf sample (0.5 g) from each treatment was taken and homogenized in 4 ml of borate buffer (pH 8.7; 4°C) and centrifuged at 13000 rpm for 15 min at 4°C. The supernatant was used as enzyme source. The reaction mixture contained 0.2 ml of enzyme extract, 1.3 ml distilled water and 1.0 ml of 0.1 M phenylalanine. The reaction mixture was incubated at 32°C for 30–60 min after the incubation and 0.5 ml of 1 M TCA (trichloro acetic acid) was added immediately to stop the reaction. OD optical density was measured at 290 nm by taking the control containing all the solutions except enzyme extract. Phenylalanine ammonia lyase (PAL) (EC 4.1.3.5) activity was also measured at 290 nm following formation of trans-cinnamic acid (Havir, 1987) and was expressed in terms of μmol l^-1 TCA per g fresh weight (FW).

Leaf sample (0.1 g) from each of the treatments was grinded with help of mortar and pestles in ice bath in 2.0 ml of 0.1% (w/v) of TCA. The crushed material was centrifuged at 12,000 × g for 10 min and 0.5 ml of the supernatant was used for consequent steps. 10 mM potassium phosphate buffer (pH 7.0) and 1 ml of 1 M potassium iodide solution was added to supernatant and incubated at room temperature for 5 min. The oxidation product formed was measured spectrophotometrically at 390 nm (Velikova, 2000). The amount of H₂O₂ formed was determined by correlating with the standard curve made with known concentrations of H₂O₂ and expressed in nmol H₂O₂ g^-1 FW.

Hand sectioned transverse sections (TSs) of pea stem from the first node were mounted in 1% phloroglucinol prepared in 95% ethanol, then transferred on glass slide and covered with cover-slip. Concentrated HCL was added near the edge of cover-slip so that HCL reach to the section (Guo et al., 2001). Pink color development showed lignin deposition. Observation was done under a compound light microscope (Nikon, Japan). For the germ tube behavior study leaf samples were destained in ethanol: acetic acid (3:1) solution. After leaf clearing leaves were stained with cotton blue and observed in a light compound microscope after mounting in glycerol. Germ tube lengths were calculated from 50 conidia per leaf from three different leaves of each treatment. For observation of HR, cleared leaves from different microbial treatments was observed under a light microscope after 72 h of pathogen challenge.

Quantitative estimation of lignin was done using the acid-solubilization method described by Dence (1992). Oven dry leaf and stem samples were crushed in mortar and pestle. 0.1 g of dry powdered sample was dissolved in 72% H₂SO₄. The acid added samples were incubated for 1 h at 30°C. 28 ml of sterilized distilled water was added after incubation. Further, autoclaving was done for 1 h at 121°C and 15 psi. Autoclaved samples were filtered in hot (80°C) condition by using Whatman No.1 filter paper. Filtrate was taken for spectrophotometric analysis at 205 nm. Results were described as lignin content in μg/g of dry weight.

Total RNA was harvested from leaf samples after 24 h of pathogen inoculation by using the method described by Patel et al. (2016b). Amount of RNA was analyzed by NanoDrop 2000 (Thermo). Approximately, 3 μg total RNA was digested using RNase-free DNase I at 37°C to remove remaining genomic DNA. Purity of RNA was tested by agarose gel electrophoresis (1.2% agarose, at 75 V). cDNA was prepared by following the standard protocol (Sambrook and Russell, 2007) with the help of oligo (dT) primers, RNase inhibitor and reverse transcriptase enzyme. The prepared cDNAs of different treatments were quantified by using NanoDrop 2000 and diluted up to the concentration of 50 ng/μl.

Gene specific primers were designed by using gene sequences retrieved from the NCBI database and primer-3 software according to specifications described by Thornton and Basu (2011). The accession numbers of the selected genes are provided in Supplementary Table S1. The primers designed for targeted genes are listed in Table 1. qRT-PCR was carried out in iQ5 Real-Time PCR Detection System (Bio-Rad Laboratories, Munchen, Germany) by using Eva Green SYBR^® Green Supermix Kit (Biorad) and gene specific primers at the concentration of 0.1 μM. Transcript level of mRNA was normalized and determined with the level of ubiquitin. PCR conditions were adjusted according to the modified program described by Marone et al. (2001). Three technical replicates were used with the following conditions: denaturation at 95°C for 2 min, 40 repeats at 95°C for 20 s, 60°C for 30 s, and 72°C for 25 s. The data obtained by real time PCR of different treatments were normalized with CT value of ubiquitin using the 2^-ΔΔCT method (Schmittgen and Livak, 2008). The fold changes were obtained relative to control (C) treatment for each gene.

Statistical analysis was done by using SPSS version 16. Experiments were repeated twice using a completely randomized design. The data are expressed as the mean of three independent replications ± standard deviations. The treatment mean values were compared by Tukey’s test at P ≤ 0.05 significance level.

Transcript accumulation pattern of NADPH oxidase (NOX) showed that its accumulation was highest (2.5 fold) in plants co-inoculated with the soil microbes and challenged with the pathogen E. pisi followed by pathogen challenged plants treated with only Trichoderma (2.25 fold) (Figure 1A). However, NOX accumulation was generally high in pathogen challenged plants treated with the soil microbes either singly or co-inoculated compared to non-challenged plants. H₂O₂ accumulation pattern also mimics the same pattern and it was observed that in pathogen challenged plants H₂O₂ accumulation was higher compared to pathogen non-challenged plants and highest accumulation being in the co-inoculated plants (Figure 1B). Further, accumulated levels of H₂O₂ were stable in pathogen-challenged and microbe inoculated plants at 48 h whereas in only pathogen-challenged plants without microbial treatments H₂O₂ levels declined significantly. HR development also correlated the observations of NOX and H₂O₂ and highest HR was observed in the co-inoculated plants (Figure 1C). Browning of cells was observed below the E. pisi colonies highest being in co-inoculated plants followed by Trichoderma inoculated plants and Pseudomonas inoculated plants. Least HR was observed in only pathogen challenged plants without microbial treatments. Germ tube development after 48 h (Figure 1E) and disease development after 10 days (Figure 1D) of pathogen inoculation were also least in co-inoculated plants and the disease development pattern was also in accordance to the pattern of NOX activities, H₂O₂ generation, and HR development.

H₂O₂ generation is often linked with activation of the MAPK pathway. However, in the current study we did not see transcript accumulation of MAPKs and their phosphate donor STK (Figure 2). Rather, the transcript levels gone down several fold in most of the pathogen challenged plants. However, slightly increased level of MAPK3 and 6 transcripts were observed in only Trichoderma treated plants compared to several fold reduction in only Pseudomonas and co-inoculated plants. The results clearly demonstrated a negative correlation between H₂O₂ generation and MAPK activities.

Phenylalanine ammonia lyase activity was increased in all microbial treatments but its activity was more in the pathogen challenged plants (Figure 3). Highest activity of PAL was observed in the co-inoculated plants challenged with the pathogen. Interestingly the co-inoculated plants without pathogen challenge also showed very high PAL activities and its activity is comparable with only Trichoderma treated plants and challenged with the pathogen. There was significant increase in PAL activity in only Pseudomonas treated plants also but its activity was low compared to only Trichoderma treated or co-inoculated plants. The pattern of PAL activities was also in accordance with the H₂O₂ accumulation pattern and therefore a positive correlation could be established between H₂O₂ generation and activation of PAL.

Transcript accumulation patterns of key enzymes in the monolignol biosynthetic pathway revealed that the monolignol biosynthesis was enhanced (Figure 4). The transcript accumulation patterns in all the enzyme synthesizing genes were nearly similar and their highest accumulation was observed in the co-inoculated treatments challenged with the pathogen compared to the single microbial treatments (Figure 5). Transcript accumulation of the first enzyme producing gene of the phenylpropanoid pathway PAL that converts phenylalanine to cinnamic acid was observed higher in the microbe treated plants compared to untreated control and its expression increased further after pathogen challenge. Nearly two fold increase in PAL transcript accumulation was observed in co-inoculated plants challenged with the pathogen compared to the control plants (Figure 5A). Enzymatic estimation of PAL also showed a similar trend with the transcript accumulation pattern (Figure 3). When the expression of the next enzyme producing gene C4H (cinnamate 4-hydroxylase) was observed, that converts cinnamic acid to 4-coumeric acid, it also revealed that its transcript accumulation was high in all pathogen challenged microbial treatments although highest being in the pathogen challenged control plants (Figure 5B). Nearly two fold increases in its transcript accumulation was observed in all microbial treated plants challenged with the pathogen compared to control. Strong transcript accumulation of HCT (hydroxycinnamoyl CoA:shikimate hydroxycinnamoyl transferase) (Figure 5C), F5H (ferulate 5-hydroxylase) (Figure 5D), COMT (caffeic acid O-methyltransferase) (Figure 5E) and CCoAOMT (caffeoyl CoA O-methyltransferase) (Figure 5F) further indicated that precursors of all monolignols were produced in greater amounts (Figure 4). Interestingly, transcript accumulation of the ABC transporter associated with lignifications was down regulated in all pathogen challenged treatments (Figure 6). However, higher transcript accumulation of laccase (Figure 5G) and peroxidase (Figure 5H) confirm conversion of monolignol precursors to monolignols and cross-linking of monolignols into complex lignins.

Histochemical staining of stem sections revealed lignifications took place in higher amounts in the intercellular spaces in plants treated with the soil microbes particularly in greater amounts in the pathogen challenged plants (Figure 7A). Quantitative estimation of lignifications also revealed a similar pattern in both stem and leaf (Figure 7B). High transcript accumulation pattern of the key enzymes in the phenylpropanoid pathway (Figure 5) positively correlate activation of the phenylpropanoid pathway, synthesis of monolignols in higher amounts and cross-linking of the monolignols into complex lignin polymers in pea treated with the soil microbes particularly in the co-inoculated plants when challenged with the pathogen.