Piriformospora indica recruits host-derived putrescine for growth promotion in plants

Abstract

Growth promotion induced by the endosymbiont Piriformospora indica has been observed in various plants; however, except growth phytohormones, specific functional metabolites involved in P. indica-mediated growth promotion are unknown. Here, we used a gas chromatography-mass spectrometry-based untargeted metabolite analysis to identify tomato (Solanum lycopersicum) metabolites whose levels were altered during P. indica-mediated growth promotion. Metabolomic multivariate analysis revealed several primary metabolites with altered levels, with putrescine (Put) induced most significantly in roots during the interaction. Further, our results indicated that P. indica modulates the arginine decarboxylase (ADC)-mediated Put biosynthesis pathway via induction of SlADC1 in tomato. Piriformospora indica did not promote growth in Sladc1-(virus-induced gene silencing of SlADC1) lines of tomato and showed less colonization. Furthermore, using LC–MS/MS we showed that Put promoted growth by elevation of auxin (indole-3-acetic acid) and gibberellin (GA₄ and GA₇) levels in tomato. In Arabidopsis (Arabidopsis thaliana) adc knockout mutants, P. indica colonization also decreased and showed no plant growth promotion, and this response was rescued upon exogenous application of Put. Put is also important for hyphal growth of P. indica, indicating that it is co-adapted by both host and microbe. Taken together, we conclude that Put is an essential metabolite and its biosynthesis in plants is crucial for P. indica-mediated plant growth promotion and fungal growth.

Piriformospora indica-induced elevated putrescine biosynthesis in tomato roots stimulates growth phytohormones auxin and gibberellin, resulting in growth promotion.

Introduction

Piriformospora indica (syn. Serendipita indica, Basidiomycota) is a root endophytic fungus with a broad host range including monocots and dicots (Varma et al., 1999; Qiang et al., 2012; Johnson et al., 2018). Piriformospora indica colonizes the root epidermis and cortex of many host plants including Arabidopsis, maize (Zea mays), tobacco (Nicotiana tabacum), Barley (Hordeum vulgare), rice (Oryza sativa), and Poplar (Varma et al., 1999; Waller et al., 2005; Vadassery et al., 2008; Jogawat et al., 2016). Increased nutrient uptake in the host plant is a major cause for P. indica-induced plant growth promotion (Pedersen et al., 2013; Rani et al., 2016; Bakshi et al., 2017; Prasad et al., 2019). Piriformospora indica also imparts biotic and abiotic stress tolerance by activating induced systemic resistance (ISR) in shoots (Waller et al., 2005; Baltruschat et al., 2008; Stein et al., 2008; Sun et al., 2010; Jogawat et al., 2016). Piriformospora indica manipulates multiple plant hormone pathways during colonization, such as jasmonates during early stages of interaction (Lahrmann et al., 2015) and auxin and cytokinin during plant growth promotion in diverse plants (Xu et al., 2018; Vadassery et al., 2008). In many plants like barley, where P. indica causes cell death‐associated colonization, the endophyte recruits gibberellin (GA) signaling to degrade DELLA transcription factors and establish cell apoptosis susceptibility (Schäfer et al., 2009; Jacobs et al., 2011). Plants also regulate and control the P. indica colonization through activation of basal defense pathway via CYCLIC NUCLEOTIDE GATED CHANNEL 19 (CNGC19) (Jacobs et al., 2011; Jogawat et al., 2020). Elevated levels of plant-secondary metabolite indole glucosinolate also restrict the propagation of P. indica and balance its growth on plant roots (Lahrmann et al., 2015).

Global transcriptome and metabolome analyses have revealed the beneficial effects of P. indica on host plants such as Arabidopsis (Vahabi et al., 2015; Strehmel et al., 2016), Barley (Molitor et al., 2011; Zuccaro et al., 2011; Ghabooli et al., 2013), and Chinese cabbage (Hua et al., 2017). Piriformospora indica-mediated reprogramming of host plant’s transcriptome, proteome, and metabolome under salt, water, and drought stress has also been explored (Waller et al., 2005; Molitor et al., 2011; Alikhani et al., 2013; Ghabooli et al., 2013). A nontargeted metabolite profiling study of Arabidopsis seedlings co-cultivated with P. indica under hydroponic conditions (14  d post-inoculation [dpi]), revealed altered primary metabolites (carbohydrates, organic acids, and aromatic amino acids) and secondary metabolites (aliphatic and indolic glucosinolates, oligolignols, coumarins, and flavonoids). Primary metabolites consisting of amino acids (e.g. Asn, Thr, Leu, 3-Cyano-Ala, beta-Ala, Val, Ala, Gln, ornithine, Pro, pyro-Glu, and γ-amino butyrate [GABA]), organic acids (e.g. citrate, 2-oxoglutarate, fumarate, malate, oxalate, glycerate, and fumarate), carbohydrates (e.g. 1-O-methylglucopyranoside, maltose, raffinose, trehalose, xylose, and ribose), polyols (erythritol and myo-inositol), phosphates (e.g. glycerol-3-phosphate, phosphate, and glycerophosphoglycerol), and sulfates (e.g. sulfate, thiamine, and thiamine-hex) showed increased levels in the inoculated roots. The metabolic profiling of roots also revealed significant alteration of phytohormone levels (jasmonates) and flavonoids (glycosylated kaempferol and quercetin), these variations were hypothesized to function as an effective signal for P. indica (Strehmel et al., 2016). In the Chinese cabbage, P. indica alters the levels of GABA, oxylipin-family compounds, poly-saturated fatty acids, and auxin and its intermediates (Hua et al., 2017). No study thus far has assigned functional role(s) for a specific metabolite in P. indica-mediated growth promotion across plants.

Though tomato (Solanum lycopersicum L.), with a worldwide annual production of ∼177 million tons, is one of the most important vegetables grown (Saeed et al., 2019), the beneficial effects of its interaction with P. indica, with regard to productivity, have been less explored. Piriformospora indica reduces the disease symptoms caused by the fungal pathogen, Verticillium dahliae and represses the amount of Pepino mosaic virus in tomato. It also increases tomato fruit biomass in hydroponic cultures and dry matter content (up to 20%) (Fakhro et al., 2010; Sarma et al., 2011). In addition, P. indica enhances growth, Na⁺/K⁺ homeostasis, antioxidant enzymes, and yield of tomato plants under normal and salt stress conditions (Abdelaziz et al., 2019). Furthermore, the metabolites involved in tomato-P. indica interaction is poorly understood despite the economic importance of this Solanaceous plant and the growth-enhancing role of P. indica. In this study, we investigated the alterations in the metabolome of tomato, as a consequence of P. indica colonization, to identify specific metabolites involved in P. indica-mediated growth benefits.

Polyamines (PAs) are low molecular weight carbon and nitrogen-rich aliphatic compounds, containing two or more amino groups that are essential for cell proliferation (Chen et al., 2019). In plants, PAs are mainly present in their free form as putrescine (Put), spermidine (Spd), and spermine (Spm), soluble conjugated (bound to small molecules including phenolics), and insoluble bound forms (bound to DNA, RNA, and proteins) (Chen et al., 2019). Spd and Spm are synthesized from Put by sequential additions of “amino propyl” groups derived from decarboxylated S-adenosyl-Met (Vuosku et al., 2012). Put is synthesized from the amino acid ornithine and arginine by arginine decarboxylase (ADC)- and ornithine decarboxylase (ODC)-mediated pathways (Kusano et al., 2008; Liu et al., 2015). PAs, including Put and Spd, are involved in plant growth and development (Kusano et al., 2008; Takahasi and Kakehi 2010; Liu et al., 2015), interactions of plants with growth-promoting rhizobacteria (Valette et al., 2019), and in-plant tolerance for abiotic stresses (Kumria and Rajam, 2002; Cuevas et al., 2008; Alcázar et al., 2010). In this work, we aimed to dissect the metabolomic alterations and characterize the functional role of the highly induced metabolite, Put, in tomato upon root colonization by P. indica.

Results

Endophytic fungus P. indica stimulates root and shoot growth of tomato

To study the time course of P. indica growth and colonization pattern in tomato, we conducted a growth promotion assay at different dpi. The roots were first observed 10 dpi with trypan blue staining to confirm colonization of P. indica-inoculated tomato roots. Though no growth promotion was observed in the plants, chlamydospores were visible in the root cortex at 10 dpi (Supplemental Figure S1). Evidence for P. indica-induced growth promotion was first observed after 30 dpi (Supplemental Figure S2A) and at 40 dpi (Figure  1A). For tracking of colonization, a green fluorescent protein (GFP)-tagged P. indica strain was utilized (Hilbert et al., 2012; Jogawat et al., 2020), and we observed both chlamydospores and a thick mat of fungal hyphae in roots at 40 dpi, confirming endophytic colonization (Figure  1B). Maximum growth promotion was observed at 40 dpi among the considered time points as compared to the control plants (Figure  1C). At this stage, root fresh weight (Figure  1, D and E), shoot, and root length (Supplemental Figure S2, B and C) were also significantly increased (P < 0.01) in P. indica-treated plants. We quantified the fungal DNA content in the roots and observed an increase at 30 and 40 dpi, respectively, over the control (Figure  1F). We observed maximum fungal colonization and growth promotion at 40 dpi among the considered time points (10, 30, and 40 dpi) and, therefore, selected this time point for subsequent investigations.

Figure 1.

Effect of P. indica (Pi) inoculation on tomato (S. lycopersicum) phenotype. A, Representative tomato shoot growth in non-inoculated (left) and P. indica-inoculated (right) pots. The experiment was independently replicated three times. B, GFP-labeled P. indica colonization pattern after 40 dpi in tomato root; (left) fluorescence, (middle) bright field, (right) merged image. One mycelia containing zone is zoomed in the fluorescence panel. Red and black arrows in the merged panel indicate P. indica chlamydospores and mycelia, respectively. C, Mean of shoot fresh weight ± se (n = 10) at different time points after P. indica inoculation. D, Quantification of root growth at 30 and 40 dpi with P. indica (n = 10). E, Visualization of root growth 40 d after inoculation with P. indica. The figure is the best representative of three independent experimental replications. F, Quantification of P. indica colonization (n = 4) in roots at 30 and 40 dpi. 0 dpi denotes un-inoculated control. Relative fungal colonization was measured by subtracting the C_T values of P. indica Tef1 from C_T values of tomato UBI3 gene and presented as percentage of maximum. Significance analysis was done by unpaired t test; ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001.

Colonization by P. indica alters tomato leaf and root metabolites

To unravel the global changes occurring in the metabolome of P. indica-colonized tomato plants (40 dpi), we carried out global untargeted gas chromatography–mass spectrometry (GC–MS) analysis of its shoots and roots separately. In addition, we also profiled the P. indica hyphal (mycelia) metabolome, to distinguish tomato-specific metabolites. Metabolite profiling revealed a total of 425 mass signals (124 for leaf, 163 for root, and 138 for P. indica mycelia), of which 55, 70, and 102 leaf-, root-, and mycelia-specific signals, respectively, were identified and annotated (Figure  2A; Supplemental Table S1). A comparative analysis of annotated metabolites revealed 76 mycelia-specific, 22 root-specific, and 9-leaf-specific metabolites (Figure  2B). Root and leaf shared 45 metabolites, root and mycelia shared 3 metabolites, and leaf and mycelia shared only 1 metabolite, while 22 metabolites were shared by root, leaf, and mycelia (Figure  2B; Supplemental Table S2). The annotated metabolites in tomato covered a broad range of primary metabolites including sugars and amino acids, while only three secondary metabolites (caffeic acid, chlorogenic acid, and benzoic acid) were identified (Supplemental Table S2). Normalized peaks of annotated leaf and root metabolites were used for calculating the “Pearson correlation coefficient” (PCC), which revealed a strong positive correlation between control and treated data sets of leaf (PCC = 0.89) and root (PCC = 0.87) metabolites (Supplemental Figure S3, A and B). These correlations indicate that upon colonization of the tomato roots by P. indica, both leaf and root metabolites are altered depending upon the basal level of the metabolites in the control plants. To study the variations between the root and leaf metabolome, metabolite fold-change was estimated. Metabolite fold changes (FCs) (ratio of fold up- or downregulation of each metabolite’s normalized peak area in P. indica-colonized plant compared to control plant) were calculated from GC–MS dataset for root- and leaf-specific metabolites (Supplemental Table S3). To obtain a global representation of metabolite FCs shared by leaf and root, we created a Pearson’s correlation-based clustered heat map (Figure  2C). Comparisons between leaf and root metabolite FCs revealed negatively correlated clusters. On further analysis of the correlations between leaf and root based on metabolite FCs, a correlation network was constructed showing negative correlations between roots and leaves (Figure  2D), indicative of the influence of colonization of tomato roots by P. indica leading to reversible alterations in root and leaf metabolome.

Figure 2.

Differential presence of annotated metabolites in tomato leaf, root and P. indica mycelia. A, Mean of total number of GC-MS mass signals (n = 3, each replicate is the pool of three individual plants) compared to the number of identified and annotated metabolites detected in leaf and root (control and P. indica-treated) of tomato and P. indica mycelia. B, Venn diagram to show comparative metabolite profiles and number of specific and common metabolites detected in leaf, root and P. indica mycelia. C, Heat map obtained using Pearson’s correlation-based clustering (algorithm: complete) of Log₂ FC of 45 common metabolites in tomato root and leaf (n = 3, denoted as FC1, FC2, and FC3). Scale shows FC values. Heat map was generated using MetaboAnalyst version 4.0. D, Pearson correlation network between root and leaf (n = 3) on the basis of Log₂ FC values of the 45 common metabolites identified in leaves and roots. Nodes represent leaves and roots, edges represent correlations (blue: negative correlation; red positive correlation), and thickness of the edges represents correlation strength. Pearson correlations were calculated using MetaboAnalyst; correlation algorithm and the network was generated using Cytoscape version 3.2.0 aided with MetScape.

Putrescine is the predominant metabolite altered in tomato roots upon P. indica colonization

After obtaining an overall picture of metabolite FC pattern of roots and shoots, we focused on identifying the major metabolites altered in the dataset. To achieve this, we made use of different univariate and multivariate computational work flows. First, we subjected the leaf metabolome data to volcano plot analysis. In leaf, volcano plots showed upregulation of five metabolites that is, 12-hydroxyoctadecanoic acid, glucaric acid, β-d-lactose, arabinose, and fructose and downregulation of four metabolites that is, L-alanine, chlorogenic acid, azelaic acid, and 2, 3-butanediol (Supplemental Figure S4). Metabolites having an alteration cutoff value of P < 0.05 and FC > 1.5 were considered as “significantly altered” in the volcano plot. These results suggested that colonization of root by P. indica altered both primary and secondary metabolome of the leaf, and therefore, suggested a systemic metabolic response.

As P. indica is a root infecting endophyte, we decided to specifically look at root-specific metabolites. First, we performed a partial least-squares discriminant analysis (PLS-DA), using the root metabolite data, to observe the divergence of two metabolite datasets (control and P. indica colonized). Based on the results obtained, control and P. indica-colonized root metabolites were classified into two different groups indicating a clear divergence in their metabolite levels (Figure  3A). A variable of importance (VIP)-score plot was generated from the PLS-DA model, which showed that the metabolite, Put has the highest VIP score (>1.38), indicating it as the most important variable among altered metabolites for the divergence in between the metabolome of control and P. indica colonized tomato roots (Figure  3B). Volcano plot revealed upregulation of eight metabolites that is, putrescine (1′), gluconic acid (2′), glucaric acid (3′), L-alanine (4′), propanoic acid (5′), L-glutamic acid (6′), lactic acid (7′), and acetoin (8′) and also significant downregulation of eight metabolites that is, benzoic acid (9′), myo-inositol (10′), azelaic acid (11′), phenylpyruvic acid (12′), fumaric acid (13′), L-valine (14′), 9,12-octadecadienoic acid (15′), and shikimic acid (16′) (Figure  3C; Supplemental Figure S5, A and B). In roots, Put showed maximum increase while benzoic acid showed a maximum decrease (Figure  3, C; Supplemental Figure S5, A and B). As GC–MS analysis did not detect Put in P. indica mycelia, therefore, for confirmation, we performed LC–MS/MS analysis, where Put was detected at very low levels (Figure  3D), indicating that its presence and increase was primarily root-specific. Further confirmation of Put being the most altered metabolite in roots was obtained after the root metabolite data were analyzed using computational metabolite marker selection approach in conjunction with orthogonal PLS to latent structures DA (OPLS-DA). The analysis generated an S-plot which showed Put to have the highest reliability and magnitude (p [1]6.1; p (corr) [1] 0.997) to be classified as a metabolite marker (Supplemental Figure S5C). Therefore, based on these results, Put is the most significant of all metabolites altered during the interaction between P. indica and tomato root.

Figure 3.

Analysis of differential root metabolites. A, PLS-DA score plot of control and P. indica-treated root samples on the basis of normalized peak areas of 77 metabolites. B, VIP score plot shows upper fifteen variables (metabolites) of importance in the root. Red arrow indicates Put with the highest VIP score identifying it as the most important variable (metabolite) during root and P. indica interaction. C: Control and T: P. indica treated. C, Volcano plot shows upregulation (red) of eight metabolites and downregulation (blue) of eight metabolites in root. Arrows indicate maximum up- and downregulated metabolites Put (1′) and benzoic acid (9′), respectively. Numbering of the denoted metabolites is described in figure S5. D, Comparative LC–MS/MS XIC showing Put (Q1–Q3: m/z 297–176) peaks in root and mycelia. E, Metabolite–metabolite interaction network constructed with the FC values of significantly altered root metabolites, which predicted 404 plausible nodes (metabolites) and 830 plausible edges (interactions). Here, plausible interactions of up- (blue) and downregulated (green) metabolites are highlighted. F, A zoomed-in view of the interaction network between up- and downregulated metabolites. G, Correlation network of interacting up- and downregulated metabolites with Put. H, Absolute amount (mean ± se, n = 3) of Put in control and P. indica-colonized tomato roots 40 dpi. Significance analysis was done by unpaired t test; ^*P < 0.05, ^**P < 0.01, ^***P < 0.005.

Next, we performed pathway analysis followed by metabolite interaction analysis to confirm whether the altered metabolic pathways included the Put-related pathway. Pathway analysis with significantly altered (P < 0.05) annotated root metabolite data set showed a plausible involvement of 46 metabolic processes (Supplemental Figure S6 and Supplemental Table S4). We considered pathways having P < 0.05 and impact >0.1 as truly impacted pathways. Among the top three pathways alanine, aspartate, and glutamate metabolism showed the highest impact (P = 2.48e⁻⁰⁷, impact = 0.327), followed by tyrosine metabolism (P = 5.92e⁻⁰⁶, impact = 0.108) and arginine and proline metabolism (P = 2.17e⁻⁰⁵, impact = 0.144). From these results, we observed that along with the other amino acids, arginine, the precursor of Put metabolism, was also significantly (P < 0.01) impacted upon colonization of roots by P. indica. Further, metabolite–metabolite interaction network, which predicted 404 nodes (metabolites) and 830 edges (interactions) (Figure  3E; Supplemental Table S5), showed functional relationships between the significantly altered metabolites. Moreover, the observation that upregulated Put, L-glutamic acid, and L-alanine have direct interaction with each other and also with downregulated phenylpyruvic acid and L-proline, indicated coordination among their related functional metabolism activities. However, no interaction was observed with the other two upregulated metabolites (acetoin and gluconic acid) and downregulated, octadecadienoic acid (Figure  3F). These results further emphasize the importance of the highly upregulated Put and hint at its significant involvement in metabolic interaction (functional relation) with other upregulated molecules during tomato and P. indica interaction. For confirmation of these interactions, we performed a Pearson’s correlation based network analysis between the interacting metabolites, and observed that Put was positively correlated with regard to observed changes in the levels of acetoin, gluconic acid, L-alanine, and L-glutamic acid and in contrast, negatively correlated with changes in proline, linoleic acid, and phenylpyruvic acid (Figure  3G). Interestingly, gluconic acid showed strong positive correlation with Put (PCC = 0.70583), but without significance (P = 0.117). Finally, an absolute and independent quantification of Put, using LC–MS/MS method, showed a highly significant elevation of Put content in tomato roots upon P. indica colonization (Figure  3H).

Piriformospora indica-induced Put biosynthetic gene in tomato

As Put level significantly increased in tomato roots upon P. indica colonization, we examined the expression levels of Put biosynthetic genes in tomato. As Put is synthesized from the amino acids arginine and ornithine by ADC- and ODC-mediated pathways and genes involved in the process being SlADC1, SlADC2, SlODC1, SlODC2, and SlODC3 (see Liu et al., 2018), we quantified the expression of all five genes in tomato roots colonized by P. indica. We observed a significant increase in transcript levels of only SlADC1, and no transcript induction was observed for any of the SlODC genes (Figure  4A). These results further confirmed P. indica-induced Put biosynthesis through ADC-mediated pathway in tomato.

Figure 4.

Effect of exogenous Put on the growth of P. indica and tomato and alteration in the expression of genes involved in the Put biosynthetic pathway in tomato upon P. indica (Pi) colonization. A, Expression of ADC (SlADC) and ODC (SlODC) transcript levels (mean ± se) 40 dpi with P. indica (n = 4). Note: Owing to small values, the error bars of the Controls for SlADC1 and SlADC2 have merged with the mean bars. B, Quantification of radial growth of P. indica upon 10 µM Put treatment (n = 10). Inset shows the visualization of the radial growth. C, Visualization of root growth in tomato seedlings upon Put treatment (10 µM). Scale bar: ∼4.5 cm. The figure is the representative of 10 replicates. Quantification of (D) shoot fresh weight and (E) root fresh weight of tomato and visualization of root growth upon treatment with Put (10 µM) (n = 10). Scale bar: ∼4 cm. F, Fresh weight of tomato seedlings upon treatment with 10 µM Gluconic Acid (GlucA) and 10 µM Put. For all box plots, box limit: percentile 25–75; whisker limit: percentile 5–95; centerline: mean. All significance analysis was done by unpaired t test; ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001.

Put-induced growth of both P. indica and tomato

PAs were previously reported to have a role in fungal cell differentiation and development (Stevens and Winther, 1979; Ruiz-Herrera, 1994). Therefore, we verified whether Put had a role in P. indica growth per se and its upregulation, in the host, provided P. indica a growth advantage. Five different Put concentrations (5–100 µM) were tested to assess its effect on the radial growth of P. indica. It was observed that 10 µM Put significantly induced P. indica radial growth, while Put concentrations higher than 10 µM did not induce further radial growth (Figure  4B; Supplemental Figure S7A). For further confirmation, we included 10 µM Put in P. indica broth cultures and results obtained indicated significant induction in both the fresh and dry biomass (Supplemental Figure S7, B and C) of P. indica in 14-d-old cultures. These results suggested that Put enhanced the growth of P. indica. In an attempt to identify the specific role of Put in tomato growth, 5-d-old tomato seedlings, were ectopically treated with Put. We observed a significant induction in multiple growth parameters of tomato (Figure  4C) after 21 d of treatment, including shoot fresh weight (Figure  4D), root fresh weight (Figure  4E), and shoot length (Supplemental Figure S8) and these results indicated that Put is indeed involved in growth promotion of tomato. In the volcano plot (Figure  3C), apart from Put, gluconic acid was the second most significantly upregulated metabolite upon P. indica colonization. Therefore, we further examined the comparative effect of Put and gluconic acid in 5-d-old tomato seedlings. In contrast, after 21 d of treatment, 10-µM gluconic acid did not show any growth promotion (Figure  4F).

Put-induced growth promotion in tomato by elevating auxin and GA biosynthesis

As a response to P. indica colonization, auxin and cytokinin were observed to be the major growth hormones involved in plant growth promotion in diverse plants (Sirrenberg et al., 2007; Vadassery et al., 2008; Meents et al., 2019). Furthermore, P. indica also induced GA biosynthesis during colonization (Cosme et al., 2016). Hence, we addressed the possible involvement of Put in altering the levels of this growth hormone in tomato. We measured indole-3-acetic acid (IAA), five GAs (GA₁, GA₃, GA₄, GA₇, and GA₈) and nine cytokinins (trans-zeatin [tZ], trans-zeatin riboside [tZR], dihydrozeatin riboside [DHZR], trans-zeatin riboside-O-glucoside [tZROG], trans-zeatin-7-glucoside [tZ7G], dihydrozeatin [DHZ], dihydrozeatin riboside-O-glucoside [DHZROG], dihydrozeatin-O-glucoside [DHZOG], and isopentenyladenine [iP]) in the Put-treated tomato seedlings using LC–MS/MS-based quantification. IAA level increased significantly in the Put-treated seedlings (Figure  5A). Similarly, GA₄ and GA₇ showed a significant increase (Figure  5B). No increase was observed in the cytokinin levels; instead, DHZR, tZ7G, DHZ, and iP levels decreased significantly (Figure  5C). These results signify the involvement of IAA and GAs in Put-induced growth promotion.

Figure 5.

Effect of exogenous application of Put on growth-inducing phytohormone levels in tomato. A, IAA (mean ± se (n = 4)). B, GAs (GA₁, GA₃, GA₄, GA₇, and GA₈) (mean ± se (n = 4)). C, Cytokinins (mean ± se (n = 4). tZ, trans-zeatin; tZR, trans-zeatin riboside; DHZR, dihydrozeatin riboside; tZROG, trans-zeatin riboside-O glucoside; tZ7G, trans-zeatin-7-glucoside; DHZ, dihydrozeatin; DHZROG, dihydrozeatin riboside-O-glucoside; DHZOG, dihydrozeatin-O-glucoside; iP, isopentenyladenine. Significance analysis was done by unpaired t test. ^*P < 0.05, ^**P < 0.01.

Put biosynthesis is crucial for P. indica-mediated growth promotion in tomato and Arabidopsis

Next, we evaluated the functional role of Put biosynthesis pathway in P. indica-mediated growth induction. First, a tomato knockdown line was generated in which SlADC1 was silenced by using a virus-induced gene silencing (VIGS) approach (Supplemental Figures S9 and S10). The silencing of the gene in the Sladc1-VIGS line was confirmed by estimating SlADC1transcript level compared to the control that is, empty vector (EV) transformed plants. It was observed that 7, 14, 30, and 40 d after agroinfiltration with the VIGS construct, SlADC1 was silenced upto 97.41%, 85.93%, 69.26%, and 50.98%, respectively (Figure  6A). For confirming SlADC1-specific gene silencing, we additionally estimated transcript levels of SlADC2 (Figure  6B), SlODC1 (Supplemental Figure S11A), SlODC2 (Supplemental Figure S11B), SlODC3 (Supplemental Figure S11C) in Sladc1-VIGS line and no difference in transcript levels of these genes were observed when compared with EV plants, and thus confirming SlADC1-specific gene silencing. At this stage, Put content was measured and found to be significantly reduced (Figure  6C), which further confirmed the SlADC1 silencing. WT, EV, and Sladc1-VIGS plants under control conditions showed no growth differences, however, upon co-cultivation with P. indica for 40 dpi, both the shoots and roots of Sladc1-VIGS plants showed no growth promotion when compared to WT +  P. indica and EV +  P. indica plants (Figure  6, D–F). We tested the hypothesis that loss of growth promotion in Sladc1-VIGS plants is due to decreased P. indica colonization, as Put enhanced the growth of P. indica in axenic cultures. We quantified P. indica colonization in all plants. At 30 dpi, P. indica colonization was significantly reduced in Sladc1-VIGS plants when compared to inoculated EV and WT (Figure  6G). We also made a comparative microscopic visualization of P. indica colonization in EV and Sladc1-VIGS root at 30 dpi and observed lower spores in Sladc1-VIGS plants compared to EV (Figure  6H). However, we did not observe significant difference in P. indica colonization levels in Sladc1-VIGS plants at 40 dpi (Figure  6G), which might also reflect reduced SlADC1 suppression (Figure  6A) at this time point. This observation suggested that loss of growth promotion in Sladc1-VIGS is due to lower Put biosynthesis and reduced P. indica colonization. In order to reiterate the importance of ADC-mediated Put biosynthesis, we treated the tomato seedlings with DL-α-(Difluoromethyl) arginine (DFMA), an irreversible inhibitor of ADC enzyme which has been previously reported to efficiently reduce cellular Put level in tomato (Fernández-Crespo et al., 2015). We transferred 5-d-old tomato seedlings in half MS media with and without 200 nM DFMA, and P. indica was inoculated for 21 d. Piriformospora indica showed reduced growth promotion in the DFMA treated tomato seedlings compared to control. Both the shoot and root fresh weights were reduced in the DFMA-treated seedlings (Supplemental Figure S12, A and B). On the other hand, 2  weeks of DFMA treatment did not show any inhibitory activity on P. indica growth (Supplemental Figure S12, C and D). These results reconfirm the importance of ADC1 in P. indica-mediated growth promotion in tomato.

Figure 6.

Piriformospora indica (Pi)-mediated growth in tomato Sladc1-VIGS plants. Mean + se of (A) ARGININE DECARBOXYLASE 1 (SlADC1) expression at different time points of P. indica colonization (n = 4) and silencing efficiencies at 7, 14, 30, and 40 dpi are 97.41%, 85.93%, 69.26%, and 50.98%, respectively. B, Mean + se of ARGININE DECARBOXYLASE 2 (SlADC2) transcript levels in EV and Sladc1-VIGS tomato after 7 and 14 dpi (n = 4). C, Put content in EV and Sladc1-VIGS plants (mean ± se; n =  4). Significance analysis of A–C was done by unpaired t test; ^*P < 0.05, ^**P < 0.01, ^***P < 0.001. D, Visualization of phenotypic difference among P. indica colonized (40 dpi) WT, EV, and Sladc1-VIGS plants. Mean + se of (E) shoot fresh weight (n = 15) and (F) root fresh weight (n = 15) in WT, EV, and Sladc1-VIGS plants. Significance analysis was done by ANOVA followed by Tukey’s test. Different letters denote significant differences (P < 0.05). Piriformospora indica colonization (G) Mean of relative fungal DNA amount (% of maximum)  ± se (n = 4–7) in WT, EV, and Sladc1-VIGS tomato at 30 and 40 dpi P. indica colonization and presented as percentage of maximum. H, Piriformospora indica spores visualized at 30 dpi using trypan blue staining, scale: 100 µM.

To study the role of Put biosynthesis in other plants, we checked its effect on Arabidopsis. Five-day-old Arabidopsis seedlings were grown on half MS media containing 10 µM Put for 9 d. We observed a significant induction in fresh weight (P < 0.01) (Figure  7, A and B). Previously, it was reported that overproduction of Put and other PAs reduced the growth in Arabidopsis (Alcázar et al., 2005), but our results indicate that exogenous application of low concentration (10 µM) of Put stimulates growth in Arabidopsis. For functional characterization of the adc mutants in Arabidopsis, we utilized previously characterized lines (Cuevas et al., 2008), adc1-2 (SALK_085350C), adc 1-2 (CS9658), adc1-3 (CS9657), adc2-3 (CS9659), and adc2-4 (CS9660) (Supplemental Figure S13, A and B) and treated them with P. indica. It was previously demonstrated that adc1-2, adc1-3, adc2-3, and adc 2-4 accumulated significantly less amount of free Put (Cuevas et al., 2008). In our experiment, we found that after 14 d of P. indica co-cultivation, fresh weight was significantly induced only in wild-type (WT), while adc mutants did not show any growth promotion (Figure  7, C and D). These results imply that P. indica fails to induce plant growth when Put biosynthesis is impaired. In a parallel experiment, we performed a complementation assay, where adc mutants (adc 1-2, adc 1-3, adc 2-3, and adc 2-4) were grown in 10-µM Put containing media and in another set of adc mutants were inoculated with P. indica in a media containing 10-µM Put. After 14 d, fresh weights in both the Put treated and P. indica with Put treated adc mutant seedlings (Figure  7, C and D) were increased and we did not observe any significant differences between the increased fresh weights in these two sets. These results indicate that exogenously provided Put complemented the mutant phenotypes, but P. indica does not have additional effect on Put supplemented mutants. These results confirm that Put is required for P. indica-mediated growth induction in Arabidopsis as well. We also checked the colonization of P. indica in WT and adc mutant lines (adc1-2 and adc1-3) at 14 dpi. Total fungal DNA quantification as well as microscopic visualization showed significantly lower P. indica colonization in adc mutant lines, than on WT (Figure  7, E; Supplemental S13, C), which indicate that host’s Put biosynthesis is necessary for fungal endophyte growth.

Figure 7.

Effect of Put on the growth of Arabidopsis. A, Representative figure of 14-d-old Arabidopsis seedlings treated with 10 µM Put (Put; Scale bar: 0.4 cm); B, Quantification of fresh weight of 14-d-old Arabidopsis seedlings upon 10 µM Put treatment (n = 20). Box limit: percentile 25–75; whisker limit: percentile 5–95; centerline: mean. Significance analysis was done by unpaired t test (^****P < 0.0001). C, Growth promotion assay in Put biosynthetic mutants along with gain-of-function assay by Put supplementation. Mean ± sd (n = 20). Significance analysis was done by ANOVA (^*P < 0.05) followed by post-hoc test. Pi indicates P. indica. D, Representative figure of n = 20 biological replicates of Put -induced P. indica-mediated growth promotion assay in two adc mutant lines of Arabidopsis (14-d-old). Scale bar: 0.4 cm. The seedlings were transferred from plates to a black surface for the purpose of photography. Pi indicates P. indica. E, Mean ± se of P. indica relative DNA content (n = 4) in roots at 14 dpi. Relative fungal colonization was measured by subtracting the C_T values of P. indica Tef1 from C_T values of Arabidopsis Actin2 gene and presented as percentage of maximum. Significance analysis was done by unpaired t test; ^****P < 0.0001.

Discussion

Mutualistic interactions of plants with P. indica can enhance growth through mechanisms such as nutrient uptake (phosphate and nitrate uptake, sugar transport), phytohormone production (auxin and cytokinin) and indirectly through ISR (Sherameti et al., 2005; Sirrenberg et al., 2007; Vadassery et al., 2009; Pedersen et al., 2013; Rani et al., 2016; Vahabi et al., 2018). Plant roots release many metabolites into the rhizosphere to kick-start symbiotic interactions, including flavonoids that act as chemo-attractants for rhizobial bacteria (Liu and Murray 2016; Oldroyd 2013) and strigolactones for mycorrhizal fungi (Besserer et al., 2006). In addition, roots also accumulate metabolites and transport it to shoots, as in the case of blumenol C-glucoside upon mycorrhizal colonization in Nicotiana attenuata (Wang et al., 2018). Host plant metabolites that are responsible for P. indica-mediated plant growth promotion are not known. In Chinese cabbage roots, P. indica stimulated the synthesis of metabolites involved in the tryptophan and phenylalanine metabolism as well as that of GABA. Tryptophan and indole metabolism were speculated to be used for de novo biosynthesis of auxin in P. indica-colonized roots, facilitating growth promotion in Chinese cabbage (Hua et al., 2017). Here, we report the identification and functional characterization of a specific metabolite, Putrescine, to be involved in the growth promotion response of tomato upon P. indica colonization.

Metabolite analysis shows that pathways targeted by P. indica in host roots belong to primary metabolism, including amino acids and PAs. In tomato roots polyamine, Put was the highly induced metabolite upon establishment of symbiotic interaction by P. indica. Multivariate statistical analysis and metabolite marker finding approach also identified Put as the most highly induced metabolite, among the eight upregulated root metabolites. Pathway analysis with significantly altered root metabolites also revealed amino acids arginine and glutathione metabolism to be altered in tomato root. Arginine and glutathione metabolisms regulate Put and glutamic acid levels in the cell, respectively; arginine is the precursor in Put biosynthesis, while glutamic acid is the precursor as well as degradation product of glutathione (Liu et al., 2010; Mo et al., 2015). During antioxidant-based defense, exogenous PAs induce glutathione level to reduce the overproduction of reactive oxygen species (Nahar et al., 2015a, 2015b). Significant impact of arginine metabolism indicates the primary involvement of the Put biosynthetic pathway during P. indica interaction with tomato. In our study, transcript level of the Put biosynthesis gene, SlADC1, was induced in P. indica-colonized root. The knockdown of adc genes in Arabidopsis and tomato also resulted in loss of growth promotion response. A balance of P. indica colonization is crucial for growth promotion response. The loss-of-function of calcium channel, CNGC19 results in increased colonization due to loss of controlled P. indica growth, resulting in a growth inhibition phenotype in Arabidopsis (Jogawat et al., 2020). The increased P. indica colonization in Arabidopsis indole glucosinolate cyp79b2/b3 mutant led in turn to plant death and not growth promotion suggesting compromised mutualism (Lahrmann et al., 2015). Our study showed reduced in vivo P. indica colonization in the tomato roots upon silencing of SlADC1 gene (30 dpi), which proves that in vivo colonization of P. indica depended on the Put biosynthesis. In stable knockout lines of Arabidopsis ADC genes, the colonization was also highly reduced. Piriformospora indica thus induces Put biosynthesis by the ADC-mediated pathway which is functionally important for both colonization and growth promotion. In tomato, ADC and ODC have differential tissue expression with the expression of ADC being predominantly in the roots (Kwak and Lee, 2001; Acosta et al., 2005). Piriformospora indica also produces Put in its mycelia; however, these low levels do not complement the adc mutants suggesting the critical role for plant-generated Put for the observed growth inducements in P. indica-treated tomato and Arabidopsis. However, our study does not rule out the involvement of additional functional metabolites at different stages of P. indica colonization.

PAs are known to be involved in plant embryogenesis and growth (Kusano et al., 2008; Takahasi and Kakehi 2010; Liu et al., 2015). Thermospermin plays a role in stem elongation of Arabidopsis (Knott et al., 2007). A double mutant of Put biosynthesis gene ADC (adc1⁻/adc2⁻) and Spd biosynthesis gene Spd synthase (spds1⁻/spds2) showed lethal defect in embryo development in Arabidopsis (Imai et al., 2004; Urano et al., 2004). PAs, including Put, are induced upon abiotic and biotic stress (Minocha et al., 2014; Liu et al., 2015; Seifi and Shelp 2019). Upon salt stress, Put levels were elevated in O. sativa and N. tabacum, and in addition, drought stress also induced Put in Arabidopsis and O. sativa (Roy and Wu, 2001; Kumria and Rajam, 2002; Capell et al., 2004; Alcázar et al., 2010). Put biosynthetic mutants (adc1 and adc2) of Arabidopsis, with decreased levels of Put, resulted in altered responses against freezing (Cuevas et al., 2008). The role of Put is also known from other symbiotic interactions. In rice, a common metabolomic signature, upon interaction with plant growth-promoting rhizobacteria, is the increased accumulation of hydroxycinnamic acid amides, identified as N‐p‐coumaroylputrescine and N‐feruloylputrescine (Valette et al., 2019). An accumulation of coumaroylputrescine and N‐feruloylputrescine is also reported during early developmental stages of barley mycorrhization (Peipp et al., 1997). Put when supplied exogenously induced the growth of tomato and P. indica, indicating that the PA is co-adapted by both host and microbes. In plants, it is known that low amounts of these exogenous PAs can act as growth stimulants (Martin-Tanguy 2001). However, PA homeostasis is tightly regulated because higher levels of PAs are toxic to cells and cause cell death (Kusano et al., 2008). For example, overexpression of polyamine biosynthesis gene ADC2 reduced growth in Arabidopsis (Alcázar et al., 2005). Elevation of in vitro growth of P. indica upon exogenous Put treatment suggests its role in hyphal growth. Interestingly, higher concentration of Put leads to the saturation of growth, which indicates optimal concentration of Put for growth promotion. Put and Spd are involved in regulating hyphal growth of arbuscular mycorrhizal fungi, Glomus mosseae due to endogenous concentration of these compounds in spores being a growth limiting factor (Ghachtouli et al., 1996). Therefore, here, P. indica induces Put in tomato roots for better colonization and growth.

PAs interact with hormones to regulate the growth and development of plants (Liu et al., 2013; Li et al., 2018). Auxin and cytokinin are major growth hormones involved in plant growth promotion response of P. indica in diverse plants (Xu et al., 2018). Piriformospora indica induces a rapid increase in auxin levels during early recognition- a phase which is crucial for reprogramming root development (Meents et al., 2019). Auxin level increased upon P. indica infestation in Arabidopsis roots (Vadassery et al., 2008) and P. indica also produces IAA in liquid culture (Sirrenberg et al., 2007). Several reports demonstrate that P. indica interferes with auxin production and signaling in the hosts and contribute to root growth (Sirrenberg et al., 2007; Vadassery et al., 2008; Lee et al., 2011; Dong et al., 2013; Ye et al., 2014; Kao et al., 2016; Hua et al., 2017). A significant increase in auxin (IAA) level in tomato seedlings upon Put treatment confirmed Put -mediated auxin alteration. Though the exact mechanism of Put-mediated auxin elevation is not yet known, a previous report showed overexpression of SPMSYN (35S: AtSPMS-9), a Spm biosynthetic gene, differentially altered expression of auxin biosynthesis gene YUCCA (Gonzalez et al., 2011). Interestingly, auxin-responsive elements are located in the promoter region of SlADC1 (Liu et al., 2018) and could, therefore, play a role in its transcriptional regulation. Interestingly, Put did not mediate a cytokinin increase and points to specificity in induction of growth hormone pathways in tomato roots. PAs are observed to be involved in GA-induced development in grape berries and peas (Shiozaki et al., 1998; Smith et al., 1985). PAs are known to be required for GA-induced growth in peas by supporting cell proliferation (Smith et al., 1985). Piriformospora indica colonization is known to upregulate several GA biosynthesis genes. GA‐deficient ga1‐6 mutant reduces P. indica colonization, whereas the quintuple‐DELLA mutant can increase colonization (Schäfer et al., 2009; Jacobs et al., 2011). Piriformospora indica promotes early flowering in Arabidopsis by increasing the GA content (Pan et al., 2017). In the tripartite interaction between rice—P. indica and rice water weevil, GA elevation in rice roots by P. indica helps to tolerate root herbivory. Piriformospora indica‐elicited GA biosynthesis suppressed the herbivore‐induced JA in roots and recovered plant growth (Cosme et al., 2016). In our study, increased level of GA₄ and GA₇ in Put -treated tomato seedlings also contributes to growth promotion along with auxin. The exact mechanism involved in Put mediated direct or indirect regulation of growth hormones is unknown and could involve hormonal crosstalk. Based on the leads obtained from the current study, future studies will focus on how Put, a primary metabolite involved in diverse plant processes, has been co-adapted by a symbiotic microbe for enhancing bi-directional growth.

Conclusion

Piriformospora indica colonization realigns the tomato metabolism including polyamine biosynthetic pathway in root, where biosynthesis of a major polyamine, Put, is significantly induced through ADC-mediated pathway. Put plays a vital role during interaction between P. indica and tomato, as well as Arabidopsis, and is inevitable for their growth promotion by P. indica. Our work sheds the light on a mechanism where P. indica employs ADC-mediated Put biosynthesis in tomato to promote its own growth into the host plant’s roots, and simultaneously, host plant allows the induction of Put biosynthesis as it helps in plant’s growth promotion via elevation of auxin and GA levels (Figure  8).

Figure 8.

Schematic representation of P. indica-induced Put biosynthesis in plants that promote the growth of both plants and P. indica. Piriformospora indica realigns cellular metabolism and induces ADC-mediated Put biosynthesis in the host. The other Put biosynthesis pathway, mediated by ODC, is not induced by P. indica (shown in dashed arrow). The increased biosynthesis of Put induces IAA and GAs which in turn promotes growth in plants. Induced Put levels in the plants also help P. indica growth.

Materials and methods

Plant growth and P. indica co-cultivation

Piriformospora indica (Verma et al., 1998) culture was grown and maintained on Kaefer medium (Varma et al., 1999) at 28 ± 2°C at 110 rpm in an orbital shaker. Tomato (S. lycopersicum, cv Pusa Ruby) seeds were pre-soaked in water overnight, kept on a moist tissue paper for 5 d in dark. After germination, tomato seedlings were transferred to pots containing soilrite (horticulture grade expanded perlite, irish peat moss, and exfoliated vermiculite in equal ratio i.e., 1:1:1, w:w:w) mixed with 1% P. indica (w/w) mycelia (1 g P. indica mycelia was mixed with 1 kg soilrite) for co-cultivation, where each pot contained single seedling. The plants were grown at 26°C (day/night: 16/8 h), relative humidity 60%, and light intensity 300 µmol/m²/s for different time points (10, 20, 30, and 40 dpi). Control plants were cultivated without P. indica inoculation. Fungal colonization was detected using trypan blue staining of root segments (Supplemental Figure S1). For experiments with Arabidopsis, seeds of WT (ecotype Columbia) and adc 1-2 (SALK_085350C), adc 1-2 (CS9658), adc 1-3 (CS9657), and adc 2-3 (CS9659), adc 2-4 (CS9660) mutant lines with T-DNA insertion in the exons (Alonso et al., 2003) from TAIR were used. Seeds were surface-sterilized, stratified, and placed on half-strength MS plates supplemented with 1% (w/w) sucrose and 0.8% agar (w/w) and allowed to germinate for 7 d. The seedlings were grown at 22°C, 10 h light/14 h dark photoperiod, and a light intensity of 150 µmol m⁻² s⁻¹ in a growth chamber (Percival). These seedlings were subsequently transferred to 1 × plant nutrient medium for co-cultivation with P. indica (Hilbert et al., 2012; Johnson et al., 2013). Samples were harvested at 14 dpi stage.

Visualization and study of P. indica colonization

For observing P. indica colonization of tomato, the roots were harvested, washed, and softened using 10% KOH, acidified in 1 M HCl and then stained with 0.02% Trypan blue for 1 h at 65°C. The samples were then de-stained in 50% lactophenol for 2 h (Dickson and Smith, 1998). The roots were observed under a light microscope (Nikon 80i). Simultaneously, tomato roots colonized with GFP-tagged P. indica (Hilbert et al., 2012; Jogawat et al., 2020) were also used for visualization of colonization under a confocal microscope (Leica TCS M5). Confocal images were acquired using Leica microsystems LAS-AF confocal microscope with HCX PL APO CS 20X objective. Excitation wavelength for GFP imaging was 489 nm and emitted fluorescence was acquired at 505–558 nm wavelength. Percentage gain for white light laser was set at 70% at 1024 × 1024 resolution. GFP-tagged P. indica was received from Prof. Ralf Oelmüller (Friedrich Schiller University, Jena, Germany). At 40 dpi, the roots were harvested and observed under a confocal microscope (Leica TCS M5) at an emission wavelength of 505–530 nm, 470 nm excitation and digital sectioning of 4–5 µm of root thickness (Jogawat et al., 2020). The relative amount of fungal DNA was quantified in 30 and 40 dpi roots using real-time-qPCR utilizing plant UBI3 and P. indica Tef1, keeping 0 dpi as untreated control (Bütehorn et al., 2000). Relative changes in fungal DNA content were calculated using C_T of PiTef1 which were normalized by C_T of UBI3 (tomato) and Actin2 (Arabidopsis) using the ΔΔC_T method (Rao et al., 2013) and P. indica DNA content in control roots (0 dpi) was defined as 1.0 which indicates a level of no change (Vadassery et al., 2008; Jogawat et al., 2020). The values are presented as percentage of maximum. The primer pairs used in the gene expression studies are provided in Supplemental Table S1.

Untargeted metabolite profiling through GC–MS

Leaves (all the leaflets) and roots of P. indica-colonized tomato (40 dpi) and P. indica mycelia were harvested and lyophilized. Three biological replicates (plant samples) were taken for both sets of control and colonized plants. Each replicate represented a pool of three individually harvested plants. Lyophilized samples were extracted and derivatized for GC–MS analysis according to Kundu et al. (2018). The derivatized samples were analyzed through GC–MS on a Shimadzu GC–MS-QP2010TM coupled with an auto sampler-auto injector (AOC-20si). Chromatography was performed using an Rtx-5 capillary column (Restek Corporation, Bellefonte, PA, USA) and helium as carrier gas. Peak integration and mass spectra analysis were done through GC–MS solution software (Shimadzu, Kyoto, Japan). Derivatized metabolites were identified through aligning and matching the mass spectra with NIST14s spectral library. Normalization of each peak area was done by comparison with internal standard (ribitol) peak area used in each sample. For some metabolites with multiple peaks, summation of the normalized peak area was considered after confirming the mass spectra as per the published protocol (Lisec et al., 2006). Venn diagram was generated in Venny version 2.1 (https://bioinfogp.cnb.csic.es/tools/venny/). Logarithmic values of normalized peak area and metabolite FC were used for all multivariate statistical analyses considering FDR. Volcano plots were prepared in Origin version 6.0 (https://www.originlab.com/) by using Log₁₀ values of the FC, where FC > 1.5 was taken as cutoff value and P < 0.05 was taken as cutoff for significance. Pearson’s correlation-based clustered heat-map, PLS-DA, OPLS-DA, network analysis, and pathway impact analysis were done using MetaboAnalyst version 4.0 (http://www.metaboanalyst.ca/). Metabolite network analysis was done using significantly altered metabolites and pathways with P < 0.05 significance were selected. Correlation network was constructed by Cytoscape version 3.2.0 aided by MetScape by uploading correlation values calculated in correlation algorithm of MetaboAnalyst version 4.0.

Put quantification through LC–MS/MS

Extraction and derivatization of Put (through benzoylation) was carried out according to Lou et al. (2016) with slight modification. In brief, around 200–250 mg of fresh plant sample was ground in liquid nitrogen and extracted with 1 mL of 10% perchloric acid. The extract was centrifuged at 12,000 g for 10 min at 4°C. Supernatant was collected and 500 µL of supernatant was derivatized with 500 µL of 2N sodium hydroxide and 20 µL of benzoyl chloride by incubating the mixture at 48°C for 20 min. An aliquot of 1 mL of sodium chloride was added to the sample and extracted with 1 mL of diethylether. Centrifugation was done at 5,000 g for 10 min and the upper phase of the sample was collected and evaporated to dryness. The dried derivatized sample was again dissolved in 500 µL of 50% acetonitrile. This sample was diluted (1:1000, v/v) with 50% acetonitrile before analyzing it in LC–MS/MS. Derivatized Put was analyzed in Exion LC (SCIEX) coupled with triple-quadruple-trap MS/MS equipped with a Turbospray ion source (SCIEX 6500+). Chromatography was performed on a Zorbax Eclipse XDB‐C₁₈ column (50 × 4.6 mm, 1.8 μm, Agilent Technologies, Palo Alto, CA, USA) by using 1% formic acid (solvent A) and acetonitrile (solvent B) as mobile phase. A linear gradient (0–1 min, 5% B; 1–7 min, 5–95% B; 7–7.6 min, 95–5% B; 7.6–9 min, 5% B) was applied for separation of derivatized Put. For detection, the mass spectrometer was operated in Multiple-reaction monitoring (MRM) mode to monitor analyte parent ion → product ion (297.0→105). Settings were as follows: ion spray voltage, 5,500 eV; turbo gas temperature, 650°C; nebulizing gas, 70 p.s.i.; curtain gas, 45 p.s.i.; heating gas, 60 p.s.i.; DP, 60; EP,10; CE, 20; CXP,10. Quantification was done by using external calibration curve prepared with authentic Put standard (Sigma, St Louis, MO, USA).

Estimation of growth phytohormones

An amount of 250 mg of tomato seedlings was ground in liquid nitrogen and extracted with 1 mL of cold extraction buffer (MeOH:H₂O:HCOOH, 15:4:0.1) containing 25 ng of trans-[²H₅] zeatin, trans-[²H₅] zeatin riboside, [²H₆] N⁶-iP, [²H₅]-IAA, and [²H₂]-GA₁ as internal standards. Homogenized sample was centrifuged at 10,000 g for 10 min at 4°C. Supernatant was loaded onto a Strata-X (Phenomenex) C18 solid-phase extraction (SPE) column pre-conditioned with 1 mL of methanol and 1 mL of 0.1% formic acid in water. After loading, the SPE column was washed twice with 0.1% formic acid and 5% methanol. Finally, the elution was done with 1 mL 0.1% formic acid in acetonitrile and dried in speed-vac. Dried sample was re-dissolved in 100 µL 5% methanol and analyzed by liquid chromatography coupled with a SCIEX 6500⁺ triple-quadruple-trap MS/MS. LC–MS/MS was performed according to (Schäfer et al., 2013) with slight modifications. In brief, separation of phytohormone was done using a Zorbax Eclipse XDB C18 column (50 × 4.6 mm, 1.8 μm, Agilent Technologies). The mobile phase comprised solvent A (water, 0.05% formic acid) and solvent B (acetonitrile) with the following elution profile: 0–0.5  min, 95% A; 0.5–5 min, 5–31.5% B in A; 5.01–6.5 min 100% B and 6.51–9 min 95% A, with a flow rate of 1.1 mL min⁻¹. The column temperature was maintained at 25°C. For detection of IAA and cytokinins, the mass spectrometer was operated in positive ionization mode (MRM modus) to monitor analyte parent ion → product ion (176.0 → 130.0 for IAA; 220.2 → 136.3 for trans-zeatin; 352.2 → 220.3 for trans-zeatine riboside, 354.2 → 222.1 for DHZR, 514.1 → 382.1 for tZROG, 382.1 → 220.19 for tZ7G, 222 → 136 for dihydrozeatin, 516.2 → 222 for DHZROG, 384.2 → 222 DHZOG, 225.2 → 136.3 for trans-[²H₅]zeatin; trans-[²H₅]zeatin riboside, 204.1 → 136 for iP, 210.1 → 136 for [²H₆] N6-iP, 181.0 → 134.0 for [²H₅]-IAA). Settings were as follows: ion spray voltage, 5,500 eV; turbo gas temperature, 650°C; nebulizing gas, 70 psi; curtain gas, 45 psi; heating gas, 60 psi Analyst 1.5 software (Applied Biosystems, Foster City, CA, USA) was used for data acquisition and processing. For detection of GAs, MS analysis triple Quad 6500+ was operated in negative ionization mode with Ion Spray Voltage of −4,500 eV, CUR gas 45 psi, CAD—medium, Temperature— 650°C, GS1 and GS2 60 psi. MRM was used to monitor analyte parent ion → product ion (331.1 → 213.1 for GA₄, 345.1 → 143.1 for GA₃, 347.1 → 273.1 for GA₁, 329.1 → 223.1 for GA₇, 363.1 → 275.1 for GA₈ and 349.0 → 276.0 for [²H₂]GA₁) with detection window of 60 s.

Preparation of the VIGS construct for silencing SlADC1

To knockdown the expression of ADC1 in S. lycopersicum, the TRV-VIGS technique was used as previously reported (Lee et al., 2017). The sequence of SlADC1 gene (2,124 bp) was obtained from the genome version: S. lycopersicum ITAG V3.2. For the SlADC1-VIGS silencing construct, a short sequence corresponding to 101–476 bp in the CDS sense strand was selected by SGN-VIGS TOOL (vigs.solgenomics.net/). The 375-bp short sequence was amplified and cloned into the TRV2 vector. Details of the primer pair used are provided in Supplemental Table S6. VIGS in tomato was carried out as described (Senthil-Kumar and Mysore, 2014) with slight modifications using the method of needleless syringe inoculation (Ryu et al., 2004). The silencing ability of the TRV2 vector was confirmed by the knockdown of Phytoene Desaturase gene responsible for chlorophyll biosynthesis in tomato. The positive plasmids for TRV2: SlADC was confirmed by sequencing and transformed into the Agrobacterium tumefaciens strain GV3101. Briefly, pTRV1, pTRV2, and pTRV2-SlADC1 constructs were grown till OD₆₀₀ = 1.0 separately and mixed in a 1:1 ratio before infiltration. Agro-inoculation (infiltration) was done in a 14-d-old tomato seedling using 1-mL needleless syringe on the abaxial surface of the leaf of the two independent sets of plants (TRV:00, and TRV-SlADC1) and were maintained in a plant growth room at 26°C (day/night: 16/8 h). For P. indica-tomato experiments, 14-d-old plants, post germination, were co-cultivated with P. indica by applying 1 mL of spore suspension at a concentration of 5 × 10⁵ spores/mL, at the crown region of the plant in soilrite, simultaneously during agro-inoculation and were allowed to grow for 40 d. Subsequently, at definite time points (7, 14, 30, and 40 d, post-infiltration and fungal co-cultivation), root samples were harvested for RNA isolation followed by cDNA synthesis. Confirmation of SlADC1 silencing was done in the root sample of tomato and Arabidopsis by reverse transcription-quantitative PCR (RT-qPCR) with gene-specific primers (Supplemental Table S6). Significant downregulation of the gene, SlADC1 was also found in the root system. The protocol is shown in detail in Supplemental Figure S9.

RNA extraction and gene expressions analysis

Harvested plant samples were homogenized using liquid N₂ and total RNA was extracted using TRIzol Reagent (Invitrogen, Waltham, MA, USA). Extracted RNA was treated with DNase (TURBO DNase, Ambion, Austin, TX, USA) to remove DNA contamination, and quantified using Nano Drop (Thermo Scientific, Waltham, MA, USA). cDNA was prepared by using High capacity cDNA reverse transcriptase kit (Applied Biosystems). Gene sequences were obtained from Sol Genomics Network (https://solgenomics.net/), and gene-specific primers were designed using NCBI primer designing tool (https://www.ncbi.nlm.nih.gov/tools/primer-blast/). PowerUp SYBR green Master Mix (Applied Biosystems) was used for RT-qPCR analyses. RT-qPCR was performed on a Bio-Rad CFX connect Real-Time PCR machine (Bio-Rad, Hercules, CA, USA). Relative expression of the genes in the treated samples was calculated as FC relative to untreated samples. Expression of the housekeeping gene, ubiquitin (UBI3), was used to normalize gene expression data. The primer pairs used in the gene expression studies are provided in Supplemental Table S1.

Treatment of P. indica, tomato, and Arabidopsis with Put and DFMA

For examining the effect of Put on P. indica growth in axenic cultures, five different concentrations (5, 10, 20, 50, and 100 µM) of Put (Sigma) were directly supplemented to fungal defined minimal medium (Jogawat et al., 2016) and the fungus was allowed to grow at 28 ± 2°C in an incubator. After 2 weeks, fungal radial growth and fresh weight were measured. Five-day-old seedlings of tomato and A. thaliana, grown under sterile conditions, were transferred to half MS plates containing 10 µM of Put. Tomato seedlings were allowed to grow for 21 d and Arabidopsis seedlings were allowed to grow for 9 d. In a separate experiment, 5-d-old tomato seedlings were transferred to half MS containing 200 nM of the ADC inhibitor, DFMA (Santa Cruz Biotechnology, Dallas, TX, USA) and allowed to grow for 21 d.

Statistical analysis

Significance analysis (t test and one-way analysis of variance [ANOVA]) was done using Sigma Plot version 13 (www.sigmaplot.com). Plots of the figures were generated using Origin version 6.0 (www.originlab.com). PCC was calculated using Social Science Statistics (https://www.socscistatistics.com/tests/pearson/) online tool and MetaboAnalyst version 4.0 (https://www.metaboanalyst.ca/).

Accession numbers

Sequence data from this article can be found in the GenBank/EMBL/Solgenomics data libraries under accession numbers: SlADC1 (Solyc01g110440), SlADC2 (Solyc10g054440), SlODC1 (Solyc04g082030), SlODC2 (Solyc03g098300), SlODC3 (Solyc03g098310); ATADC1 (AT2G16500), ATADC2 (AT4G34710).

Supplemental data

The following materials are available in the online version of this article.

Supplemental Figure S1. Trypan blue staining shows P. indica colonization in tomato root after 10 d of co-cultivation.

Supplemental Figure S2. Different phenotypic changes and growth induction of tomato upon P. indica colonization.

Supplemental Figure S3. Pearson’s correlation of metabolite alteration in leaf and root.

Supplemental Figure S4. Alteration of leaf metabolites in tomato upon P. indica treatment.

Supplemental Figure S5. Tomato root metabolites alteration upon P. indica colonization (40 dpi).

Supplemental Figure S6. Scatter plot of pathway impact in root shows specific metabolism in root affected by P. indica colonization.

Supplemental Figure S7. Effect of Put on P. indica growth.

Supplemental Figure S8. Effect of 10 µM Put on growth of tomato.

Supplemental Figure S9. Schematic representation of the VIGS protocol in tomato.

Supplemental Figure S10. VIGS confirmation in tomato.

Supplemental Figure S11. Ornithine decarboxylase transcript expression level in Sladc1-VIGS tomato root.

Supplemental Figure S12. Tomato and P. indica growth assay upon DFMA treatment.

Supplemental Figure S13. Confirmation of T-DNA insertion in Arabidopsis adc mutant lines and difference of P. indica colonization in WT and adc mutant lines.

Supplemental Table S1. Annotated metabolites in tomato and P. indica mycelia with derivatization level and obtained molecular mass.

Supplemental Table S2. Identity of metabolites distributed in Venn's Diagram.

Supplemental Table S3. FC of metabolites found both in control and treated leaf and root samples.

Supplemental Table S4. Pathway impact analysis table for tomato root treated with P. indica.

Supplemental Table S5. Metabolite-metabolite interaction analysis output.

Supplemental Table S6. Primer list.

 

J.V. and A.K. designed the study and planned the experiments. A.K., A.J., S.M., and P.K. performed the experiments. A.K., A.J., P.K., S.M., and J.V. analyzed the data and wrote the manuscript. All authors have read and approved the final version of the manuscript.

The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (https://academic.oup.com/plphys/pages/general-instructions) is Jyothilakshmi Vadassery-jyothi.v@nipgr.ac.in.