ABSTRACT
Below-ground microbes can induce systemic resistance (ISR) against foliar pests and pathogens on diverse plant hosts. The prevalence of ISR among plant-microbe-pest systems raises the question of host specificity in microbial induction of ISR. To test whether ISR is limited by plant host range, we tested the ISR-inducing ectomycorrhizal (ECM) fungus Laccaria bicolor on the non-mycorrhizal plant Arabidopsis. We found that root inoculation with L. bicolor triggered ISR against the insect herbivore Trichoplusia ni and induced systemic susceptibility (ISS) against the bacterial pathogen Pseudomonas syringae pv. tomato DC3000 (Pto). We found that L. bicolor-triggered ISR against T. ni was dependent on jasmonic acid (JA) signaling and salicylic acid (SA) biosynthesis and signaling. We found that heat killed L. bicolor and chitin are sufficient to trigger ISR against T. ni and ISS against Pto and that the chitin receptor CERK1 is necessary for L. bicolor-mediated effects on systemic immunity. Collectively our findings suggest that some ISR responses might not require intimate co-evolution of host and microbe, but rather might be the result of root perception of conserved microbial signals.
INTRODUCTION
Plants associate with complex communities of microorganisms. Interplay of host and microbial genotype determine whether the outcome of specific plant-microbe interactions is beneficial or detrimental to plant health1–3. Plants possess receptors to sense potential pathogens including transmembrane pattern-recognition receptors that recognize conserved microbe-associated molecular patterns (MAMPs), and intracellular receptors that directly or indirectly recognize effectors to help limit pathogen growth4,5. MAMP perception by plant receptors triggers pattern-triggered immunity (PTI) responses including a reactive oxygen species (ROS) burst, Mitogen-Activated Protein Kinase (MAPK) signaling, ion influx and callose deposition6–9. In addition to local immune mechanisms, root-associated microbes can induce systemic resistance (ISR) against a diverse spectrum of above-ground threats2,10. While the mechanisms by which plants perceive MAMPs and effectors are well understood, there is a more limited understanding of the mechanisms by which plant perceive the microbes that trigger ISR.
In response to perception of microbes or pathogens, plants activate systemic defense mechanisms including commensal-triggered ISR and pathogen-triggered systemic acquired resistance (SAR)11,12. While both SAR and ISR confer resistance to pests and pathogens, they promote resistance through distinct mechanisms2. SAR occurs upon local perception of a pathogen, MAMP, or effector and results in systemic induction of SA-dependent gene expression and accumulation of secondary metabolites13,14. Unlike SAR, ISR is associated with small or no changes in systemic gene expression and phytohormone levels15,16. Additionally, while SAR is dependent on the SA signaling protein NPR113, the pathways required for ISR can vary between systems17,18.
ISR triggered by a single microbial strain can be effective across diverse hosts and against diverse pathogens. For instance, Pseudomonas simiae WCS417 was isolated from barley fields19 and can trigger ISR on carnation, tomato, wheat and Arabidopsis20–22. Similarly, P. defensor WCS374 and P. capeferrum WCS358, which were isolated from potato fields, trigger ISR in radish and Arabidopsis respectively21,23. That ISR by a single bacterial strain is effective against phylogenetically diverse plants suggests that ISR is likely dependent on broadly conserved plant perception and signaling mechanisms.
In addition to interactions with rhizobacteria, most vascular plants (>90%) form beneficial symbioses with arbuscular mycorrhizal (AM) fungi or ectomycorrhizal (ECM) fungi24. Both AM and ECM fungi can induce ISR on their host plants against foliar pests and pathogens25–29. AM symbiosis requires the Common Symbiotic Pathway (CSP) for initiation of successful symbiosis30; recent evidence suggests that ECM symbiosis may also depend on the CSP31. It is unknown whether mycorrhizal-triggered ISR is dependent on the CSP and is dependent on the same or distinct mechanisms from rhizobacterial-triggered ISR.
Here, we tested the hypothesis that ISR occurs independently of mutualistic symbiosis or host specificity. To address our hypothesis, we tested whether Laccaria bicolor, an ECM fungus that colonizes a wide range of tree host species32 could trigger ISR on Arabidopsis thaliana, a non-mycorrhizal plant24. Arabidopsis lacks the CSP and so allows for direct testing of whether ISR by mycorrhizal fungi is dependent on the CSP. We rationalized that if ISR is the result of a mutualistic symbiosis, then we should not observe ISR by L. bicolor on Arabidopsis. In contrast, if ISR is the result of plant perception of a general microbial factor, we might observe ISR by L. bicolor on Arabidopsis.
Consistent with the hypothesis that ISR against herbivores can occur independently of highly co-evolved mutualism, we found that L. bicolor can trigger ISR on Arabidopsis against caterpillars of the generalist herbivore Trichoplusia ni and induce systemic susceptibility (ISS) against P. syringae pv. tomato DC3000 (Pto). We found that L. bicolor-triggered ISR against T. ni and ISS against Pto requires CERK1-dependent chitin perception. We show that chitin and heat-killed fungi are sufficient to modulate systemic immunity and this response is distinct from root perception of bacterial MAMPs. Collectively this work shows that chitin perception by roots is sufficient to modulate systemic plant immunity and that ISR against herbivores might be a general response to perception of chitin, independent of host adaptation.
RESULTS
L. bicolor triggers ISR on Arabidopsis against T. ni
To test the hypothesis that ISR occurs independently of mutualist symbiosis, we treated the non-mycorrhizal plant Arabidopsis thaliana with the ECM fungus Laccaria bicolor and measured ISR against the generalist herbivore Trichoplusia ni17,33. The roots of 9-day-old Arabidopsis seedlings were treated with L. bicolor (Methods), and the rosettes were challenged with T. ni larvae 3 weeks later. After 1 week of feeding, we found that T. ni larvae that fed on L. bicolor-treated Arabidopsis had a 27% reduction in larval weight compared to buffer-treated plants (Figure 1a, p < 0.006). The observed L. bicolor-induced ISR is similar in magnitude to previously described bacterial treatments that induce ISR against T. ni on Arabidopsis17,34,35.
These data suggest that ISR by the ECM fungus L. bicolor does not require host colonization. As ISR can occur by distinct mechanisms against different herbivores and pathogens, we sought to compare and contrast L. bicolor induced ISR to other well-studied mechanisms of ISR on Arabidopsis.
ISR by L. bicolor requires SA and JA signalling
ISR against herbivores triggered by beneficial Pseudomonas spp. is dependent on JA signalling17,34,35. To determine if L. bicolor-induced ISR against T. ni is also dependent on JA signalling, we tested ISR by L. bicolor on the JA-Ile receptor mutant, coi1-16. As previously shown, we found that the coi1-16 mutant has enhanced susceptibility to T. ni17,36. We found that the coi1-16 mutant did not show reduced caterpillar weight gain after L. bicolor treatment (Figure 1a). Therefore, like ISR by Pseudomonas spp. against herbivores, ISR by L. bicolor in Arabidopsis is dependent on JA-Ile perception via COI1.
ISR against herbivores triggered by Pseudomonas spp. is dependent on SA biosynthesis via SID2 and is independent of SA signalling via NPR117. We found that a sid2-2 mutant (deficient in SA biosynthesis) did not show ISR in response to L. bicolor (Figure 1a). In contrast, L. bicolor treatment of the SA signalling mutant npr1-1 resulted in a 30% reduction in T. ni weight gain (Figure 1a, p < 0.11). Consistent with previous studies, we found that the sid2-2 and npr1-1 mutants show enhanced resistance to T. ni17,33. Though not statistically significant, the ISR induced on the npr1-1 mutant was similar in magnitude to wild-type plants (Figure 1a). T. ni larvae feeding on an npr1-1 mutant are extremely small leading to challenges in quantifying further reduction in weight17,33. Given the consistent and similar magnitude in resistance by L. bicolor treatment with npr1-1 mutant, it seems likely that L. bicolor triggered ISR is independent of or partially dependent on NPR1. Overall, these data show that similar to ISR against T. ni by P. simiae WCS417, ISR by L. bicolor requires both JA signalling and SA biosynthesis17.
We found that ISR by L. bicolor against T. ni is dependent on SID2 but partially dependent or independent of the positive regulator, NPR1. This raises the possibility that negative regulators of SA signalling are required for ISR by L. bicolor. Ding et al. (2018) reported that NPR3 and NPR4 act as SA receptors that repress SA defence gene expression in an NPR1 independent manner. To determine whether L. bicolor mediated ISR is dependent on SA transcriptional co-repressors, the npr3-2 npr4-2 (npr3/4) double mutant, was tested for ISR. We found that in contrast to wild type plants, L. bicolor treatment of an npr3/4 mutant resulted in a significant increase in T. ni larval weight gain (Figure 1b, p < 0.003). The variance in the weight gain of T. ni larvae feeding on control plants relative to the data in Figure 1a for this and subsequent experiments can be attributed to spatial and temporal factors17,36; replicates of the same experiment were done in the same location and over a relatively short period of time while distinct experiments were done over longer time scales and in multiple locations leading to variance in the absolute weight gain of the controls17,36 (Supplementary Figure 1). These data indicate that L. bicolor-triggered protection of Arabidopsis from caterpillar feeding is dependent on the negative regulators of SA signalling, NPR3/4.
L. bicolor induces systemic defence-gene expression
Previous reports of ISR by beneficial Pseudomonas against insect herbivores are associated with a low-level priming of expression (1.5-to 3.5-fold) of JA-Isoleucine (JA-Ile) dependent genes17,38. In contrast, SAR is associated with strong priming (4-to 10-fold) of SA-dependent gene expression in distal leaves39,40. To determine if ISR by L. bicolor is associated with priming of systemic defence-gene expression, we performed RNAseq to analyse gene expression in Arabidopsis leaves after root treatment with L. bicolor (Figure 2a and Supplementary Table 1). In response to Arabidopsis root treatment with L. bicolor, we found a significant increase in systemic expression of hormone-responsive genes including JA- and Ethylene-dependent genes (Figure 2a). We also found an increase in expression of genes involved in secondary metabolism and chitin responses (Figure 2a). We found that L. bicolor treatment of roots resulted in significant induction in shoots of JA-Ile marker genes including VSP1, VSP2, MYC and MAMP-responsive genes including MYB51 and WRKY70 (SA), which have been shown to be induced by root treatment with Pseudomonas spp.17,38 (Figure 2b). The initial RNAseq experiments were performed on gnotobiotic plants growing on solid media (Methods); to determine if the responses we saw were further enhanced after challenged with herbivores, we grew plants in soil and treated with L. bicolor. In soil-grown plants, we found no significant systemic changes in gene expression in response to L. bicolor treatment by RNAseq (Methods) or qPCR (Figure 2c). While challenging leaves with T. ni resulted in significant induction of JA-Ile-dependent defence-gene expression relative to unchallenged plants, we did not find any further enhancement of gene expression up L. bicolor treatment (Figure 2c). The lack of significant changes in soil grown plants during ISR is consistent with previous reports of no or slight priming of systemic gene expression17,22. The slight increase in systemic gene expression after L. bicolor treatment is consistent with an ISR-like rather than with a SAR-like mechanism (Figure 2).
SAR is associated with elevated SA levels in distal tissues41, while ISR is not associated with significant changes in systemic hormone accumulation2. Phyto-hormone analyses was performed using leaves from the same soil grown plants used for the qRT-PCR experiment above. We found that similar to the qRT-PCR results, L. bicolor treatment of Arabidopsis roots resulted in no significant increase in accumulation of the phytohormones SA, ABA and JA-Ile (Supplementary Figure 2). As expected, plants challenged with T. ni had a significant increase in JA levels; however, these effects were not further enhanced after L. bicolor treatment. Collectively, these data indicate that L. bicolor triggers systemic resistance against herbivory with moderate priming of defence gene expression and no significant accumulation of phytohormones in Arabidopsis leaves.
Low-molecular weight secondary metabolites contribute to plant defence and to systemic resistance against invading pathogens and other threats29,35,42. Our RNAseq data showed that L. bicolor treatment resulted in significant upregulation of the camalexin biosynthesis genes PAD3 and CYP71A13 (Figure 2 and Supplementary Table 1). To determine whether secondary metabolites accumulate in systemic leaves after L. bicolor treatment, targeted metabolite analysis was performed with Arabidopsis plants with and without L. bicolor treatment and T. ni feeding. We found that similar to Arabidopsis treatment with beneficial Pseudomonas spp.35, camalexin levels in leaves were significantly higher in L. bicolor-treated plants than in buffer-treated plants (Figure 3a, p < 0.05). Camalexin levels were increased by T. ni challenge, and L. bicolor-pretreated plants accumulated higher levels of camalexin after T. ni challenge than the buffer-treated controls (Figure 3a). In contrast, we did not observe significant accumulation of indolic glucosinolates like glucobrassicin and their by-product raphanusamic acid, which are involved in pathogen- and insect-triggered defense43,44 (Supplementary Figure 3). Camalexin is derived from the tryptophan pathway, which is also the precursor for many metabolites with insecticidal properties42. The mutant cyp79b2/b3, which is impaired in the biosynthesis of indole-3-acetaldoxime (IAOx) and camalexin from tryptophan, was tested for L. bicolor induced resistance against T. ni larvae. Pre-treatment of the cyp79b2/b3 mutant with L. bicolor did not result in significant decrease in T. ni larval weight (Figure 3b). This indicates that the IAOx pathway is required for L. bicolor triggered systemic protection against herbivore damage.
CERK1 is necessary for L. bicolor induced ISR
During colonization of plant roots, beneficial Pseudomonas spp. both suppress45,46 and induce23 distinct subsets of defence-related genes indicating that beneficial microbes can induce and suppress PTI. To test whether L. bicolor induces PTI, Arabidopsis seedlings were treated with live and heat-killed L. bicolor and MAP Kinase phosphorylation was used as a readout for PTI. We found that both live and dead L. bicolor triggered phosphorylation of MPK3, MPK4 and MPK6 in Arabidopsis seedlings (Figure 4a). This observation indicates that both live and dead L. bicolor trigger PTI in the local tissue of Arabidopsis.
That L. bicolor can induce PTI led us to ask whether L. bicolor triggered ISR against T. ni might be the result of fungal MAMP perception by plant roots. To test this hypothesis, Arabidopsis seedlings were inoculated with heat-killed L. bicolor and challenged with T. ni. We found that treating Arabidopsis roots with heat-killed L. bicolor caused a 33% reduction in T. ni larval weight gain (Figure 4b, p < 0.0001). This indicates that live L. bicolor is not necessary to trigger ISR and that MAMP perception may underlie L. bicolor-mediated ISR against T. ni.
Chitin is a structural component of the fungal cell wall and can induce PTI in plants via the CERK1 receptor7,47. We tested whether chitin is sufficient to trigger ISR against T. ni when applied to roots. We observed that chitin treatment of Arabidopsis roots resulted in a 38% reduction in T. ni larval weight gain relative to buffer-treated plants (Figure 4b, p < 0.0001). This indicates that chitin is sufficient to trigger ISR against T. ni. To test whether plant perception of chitin is necessary for ISR by L. bicolor, ISR experiments were performed with the chitin receptor mutant, cerk1-2. We found that the treatment of cerk1-2 mutant roots with L. bicolor did not inhibit T. ni larval weight gain (Figure 4c). Thus, ISR by L. bicolor is dependent on CERK1, which indicates that ISR by L. bicolor is due to chitin perception by Arabidopsis roots.
Like Pseudomonas-triggered ISR against herbivores, L. bicolor-mediated ISR against ni depends on JA signalling and of SA biosynthesis (Figure 1)17. Many microbes that trigger ISR against herbivores can affect defence responses against other pests and pathogens. For instance, P. simiae WCS417 triggers ISR against bacterial pathogens while others trigger induced systemic susceptibility (ISS)17. To determine if L. bicolor can protect Arabidopsis from other threats, Arabidopsis plants were treated with L. bicolor and infected with Pto. When Col-0 leaf disks were harvested and CFUs were counted, we found that L. bicolor (both live and dead) and chitin treatment induced systemic susceptibility (ISS) to Pto (Figure 5a; p < 0.001). As with L. bicolor-induced ISR against T. ni, we found that L. bicolor triggered ISS against Pto is dependent on CERK1 (Figure 5b). These data indicate that in some instances, ISR against T. ni and ISS against Pto might be due to induction of PTI in plant roots.
If chitin is sufficient to induce resistance against T. ni and susceptibility against Pto, treatment of Col-0 roots with Paxillus involutus, another ECM fungus or a pathogen like Ustilago maydis should also result in ISS against Pto. However, U. maydis can cause necrosis and stunting of Arabidopsis48. To avoid fungal pathogenicity confounding the data interpretation, Arabidopsis roots were treated with heat-killed U. maydis D132. We found that Arabidopsis plants treated with live mycelium of the ECM fungus P. involutus and heat-killed maydis D132 also triggered ISS against Pto [Figure 5c; p < 0.01 (P. involutus) & p < 0.001 (U. maydis D132)]. To test if additional MAMPs also induce ISS, Arabidopsis roots were treated with flg22, elf18 and leaves were infiltrated with Pto. In contrast to chitin, flg22 and elf18 treatments significantly reduced Pto CFUs in Arabidopsis leaves (Figure 5d, p < 0.0097 and p < 0.00016 respectively). These data show that perception of chitin in the roots modulates systemic immunity against both insect herbivores and biotrophic pathogens and that not all MAMPs have the same effect on systemic immunity.
DISCUSSION
The evolutionary history of some symbiotic plant-microbe interactions is well defined, while others remain more nebulous. Nitrogen-fixing rhizobia and AM and ECM fungi engage in highly specific signal exchange to initiate nutritional symbiosis with their plant hosts30. Whether similar co-evolution and symbiotic specificity are necessary for ISR triggered by mycorrhizal fungi and rhizobacteria is unknown. We used a non-mycorrhizal model consisting of Arabidopsis and the ECM fungus, L. bicolor to address the specificity of ISR. We found that L. bicolor can trigger ISR in non-mycorrhized host, indicating that ISR can be independent of mutualistic symbiosis (Figure 1a). This is consistent with ISR-induction on distinct hosts from that which a strain was originally isolated20–23.
We found that L. bicolor treatment of Arabidopsis roots results in a local PTI response (Figure 3a). Our data show that chitin, a fungal MAMP, is sufficient to trigger ISR against T. ni and ISS against Pto (Figure 4b). The bacterial MAMPs flg22 and lipopolysaccharide (LPS) have previously been shown to be sufficient to trigger ISR against bacterial pathogens49 and both flagellin and WCS417 trigger overlapping transcriptional responses in Arabidopsis roots23. Interestingly, in addition to a role for chitin perception and immunity signalling, AM symbiosis in rice is also dependent on CERK150. While not all MAMPs result in the same effects on systemic defences, collectively these observations suggest that in some cases, ISR, a potentially beneficial consequence of plant-microbe interactions, might be due to MAMP exposure rather than host colonization.
Thus far, ISR has been studied in mutualistic symbioses (AM and ECM fungi), or where the evolutionary history of the interaction is unknown. Our findings show that plants also respond to non-adapted microbes resulting in ISR against herbivory. This suggests that testing isolates known to induce ISR, even if on a phylogenetically distant plant host, may be a rational approach to identifying ISR-inducing strains. These findings also suggest that chitin from a variety of sources may be useful in promoting ISR in agricultural plants against herbivores.
MATERIALS AND METHODS
Plant growth conditions
Arabidopsis seeds used in this study include the wildtype accession Columbia-0 (Col-0) and the mutants coi1-1651, npr1-152, sid2-253, npr3-2/npr4-254, cyp79b2/b355 and cerk1-27. Seeds were sterilized with 70% ethanol for 1 min, 10% bleach for 2.5 min and washed thrice with sterile distilled water. Surface sterilized seeds were cold stratified at 4 °C for two days and planted on Jiffy-7 Horticulture peat pellets. Trays containing the pellets were covered with clear domes to maintain high humidity during seed germination. One-week-old seedlings were thinned to one seedling per pellet. The plants were watered twice a week and grown under 12 hours of light (80-100 µE) at 23 °C, 12 hours of dark at 20 °C, 70-80% relative humidity for the duration of the experiment.
Culturing Laccaria bicolor
Laccaria bicolor (monokaryotic strain CII-H-82, S238N) was cultured on solid modified melin norkrans (MMN) medium (Composition (g/L): glucose 10.0, (NH4)2C4H4O6 2.5, (NH4)2SO4 0.25, KH2PO4 0.5, MgSO4.7H2O 0.15, CaCl2.2H2O 0.05, NaCl 0.025, 0.1% thiamine-HCl 0.1 mL and 1% FeCl3.6H20 1mL; pH 5.2 - 5.4) at 23-26 °C in the dark56. Three-week-old colonies were dissected using a sterile scalpel and at-least 25 agar plugs (0.25 cm2) were inoculated per 200 mL of liquid MMN medium. Media with agar plugs were incubated at 23 °C in the dark with shaking at 100 rpm. Mycelium that grew from the fungal plugs after 3 weeks was homogenized using a sterile mechanical shearer for less than 10 seconds. The homogenized fungal solution was poured into sterile 50 mL FALCON polypropylene conical tubes (falcon) and screw caps were partially tightened. The falcon tubes with L. bicolor were returned to the shaker at 23 °C in the dark for two days.
Fungal inoculation and chitin treatment of plant roots
L. bicolor cultures in tubes were centrifuged at 4500 rpm for 5 min at 4 °C. The supernatant was discarded and the hyphae in the pellet was re-suspended in 10 mM magnesium sulphate (MgSO4) buffer. The optical density of the solution was measured at 600 nm using 10 mM MgSO4 as the blank. An inoculum with OD600 of 0.1 was used for Arabidopsis root treatment. 9-days old Arabidopsis seedling roots were inoculated with 2 ml of the L. bicolor solution or 10 mM MgSO4 for the control plants. Paxillus involutus culture was prepared as described above for L. bicolor in MMN medium and 2 mL of OD600 = 0.1 was inoculated on 9-days old Arabidopsis seedling roots.
Dead L. bicolor and Ustilago maydis D132 were prepared by heat treating a 20-fold concentration (OD600 of 2) of the L. bicolor and U. maydis inoculum in a water bath at 65-85 °C for 25 mins. The heat-killed cultures were inoculated on MMN and PDA agar media respectively and incubated at 23 °C in the dark to check for survival. After cooling the inoculum to room temperature, plant roots were treated with 2 mL of the heat-killed L. bicolor and U. maydis. A chitin stock (10 mg/ml) was prepared using chitin from shrimp shells (Sigma Aldrich), autoclaved for 30 mins and centrifuged to collect the supernatant45. Chitin treatment solution (500 µg/ml) was diluted from the stock using 10 mM MgSO4 and 2 ml was pipetted onto the soil surrounding each individual seedling.
Caterpillar feeding experiments
ISR experiments were done by challenging control and L. bicolor treated plants with the cabbage looper, Trichoplusia ni17,57. Fresh batches of T. ni eggs were obtained from Natural Resources Canada with instructions that the eggs should be collected over 24 hours and shipped immediately to allow for synchronous hatching. The eggs were incubated at 23 °C in the same light regime as the plants to synchronize them with the plant circadian rhythm. First larval instars emerged 2 days after incubation. A single larva was placed on a single plant using a fine paint brush and the entire plant was covered with a breathable nylon mesh net. The hatchlings could feed on the plant for one week, after which the larvae were collected and weighed to the nearest 0.1 mg. The initial weight is negligible, and so the final weight represents the weight gain over the 1 week of feeding58. Each treatment or plant genotype was tested in at least 3 independent experiments (performed on different days and different batches of plants). A minimum of 20 plants per treatment per treatment were used.
For phytohormone, metabolite and qRT-PCR analyses, each plant was challenged with two T. ni larvae. Untreated controls were netted without T. ni larva and incubated under same conditions. 8 plants were used for every treatment and one leaf from each plant per treatment was harvested after 24 hours, pooled together and immediately frozen in liquid nitrogen. Samples were stored at -80 °C until processing. Only leaves with visible feeding damages were collected for T. ni treatment samples. Samples were collected from at least 4 independent experiments.
Pseudomonas syringae pv. tomato DC3000 (Pto) infection assays
Pto59 infections were performed as described60. Pto inoculum was prepared by diluting an overnight culture to OD600 = 0.0002 using 10 mM MgSO4. 5-weeks old plants were watered and covered with humidity dome for a minimum of an hour. Three leaves per plant were marked using a sharpie and infiltrated with Pto inoculum using a blunt 1 ml syringe. Trays were covered with humidity dome and placed back under plant growth conditions. Leaf disks (0.9 cm diameter) were collected from 2 infected leaves per plant (six plants per treatment) at two days post infection. The leaf disks were homogenized, serially diluted and plated on Luria Bertani (LB) medium with rifampicin (50 µg/ml). Colony forming units (CFU) were counted 2 days after incubation at 29 °C and experiments were performed at least 3 independent times.
Arabidopsis treatment with L. bicolor on plates
For RNAseq experiments with plants grown on solid media, 20 surface-sterilized, stratified A. thaliana (Col-0) seeds were germinated on MS medium agar plates61. After 7 days, the roots were covered with cellophane strips containing pre-cultivated mycelium of L. bicolor. The controls were covered with cellophane strips. After two days of exposure, the leaves which were not in contact with the fungus were harvested and frozen at -80 °C. The leaves from 5 plates were pooled. The experiment was repeated five times independently for RNA extraction and sequencing.
RNA extraction and qRT-PCR analyses
RNA was extracted from samples collected as described above by using Qiagen RNAeasy extraction kit. The extracted RNA was quantified using Nanodrop. TURBO DNAse (Ambion) was used to remove DNA contamination from RNA samples. Single cDNA synthesis was performed using 1µg DNA-free RNA using Superscript III (Invitrogen) and Oligo dT primers in a 40 µL volume. Gene expression data was obtained by running quantitative PCR reactions with 10 µL reaction volume containing 0.5 µL cDNA, 1 µL of 5 µM primer mix containing forward and reverse primers (2.5 µM each) and 2x PowerUpTM SYBR Green master mix (Thermo Fischer). Expression values were normalized to the housekeeping gene EIF4A45 and samples from 5 biological replicates were tested. The other genes and the respective primers used in this analysis are listed in Supplementary 2.
RNAseq and transcriptomics analyses
Quality of RNA extracted from the samples was checked using a Bioanalyzer (Agilent 2100). The RNA integrity numbers (RIN) ranged from 6.8 – 8.2 (Supplementary Tables 3 and 4). Library construction and sequencing were conducted at Chronix Biomedical (Chronix Biomedical, Inc., Göttingen, Germany). RNA libraries were prepared using the TrueSeq RNA Library Prep Kit (Illumina). Single-end reads were sequenced with a length of 75 bp on an Illumina HighSeq 2000.
Sequencing yielded 16 to 23 million reads per sample (Supplementary Tables 3 and 4). Raw sequence data was processed with the FASTX or FASTp toolkit62. Using FASTQ Trimmer, all nucleotides with a Phred quality score below 20 were removed from the ends of the reads, and sequences smaller than 38 bp or sequences with a Phred score below 20 for 10% of the nucleotides were discarded by the FASTQ Filter; adapter sequences and primer sequences were trimmed with the FASTQ Clipper (http://hannonlab.cshl.edu/fastx_toolkit/). Read numbers per sample after processing remained between 16 and 22 million (Supplementary Tables 3 and 4). The raw data have been deposited in https://www.ebi.ac.uk/arrayexpress/experiments/E-MTAB-8544 (gnotobiotic experiments) and https://www.ebi.ac.uk/arrayexpress/experiments/E-MTAB-8523 (soil experiments).
The processed sequences were mapped against the Arabidopsis thaliana transcriptome TAIR10 (downloaded from www.arabidopsis.org63 using Bowtie 264. Bowtie mapping files were summarized to transcript count tables in R. To find transcripts with significantly increased or decreased abundance, the DEseq2 package65 implemented in R66 (R Core Team 2017) was used.
Phytohormone and metabolite measurements
Extraction of phytohormones was carried out from the samples described above (Caterpillar feeding experiments) with methyl-tert-butyl ether (MTBE)67. Reversed phase separation of constituents was performed as previously described using an ACQUITY UPLC® system (Waters Corp., Milford, MA, USA) equipped with an ACQUITY UPLC® HSS T3 column (100 mm x 1 mm, 1.8 µm; Waters Corp., Milford, MA, USA). Nanoelectrospray (nanoESI) analysis was carried out as described in67 and phytohormones were ionized in a negative mode and determined in a scheduled multiple reaction monitoring mode with an AB Sciex 4000 QTRAP® tandem mass spectrometer (AB Sciex, Framingham, MA, USA). Mass transitions are described in Supplementary Table 5.
Mitogen-activated protein kinase (MAPK) experiments
Two sterile seeds were germinated per well in a 24 well plate with liquid Murashige & Skoog (MS) medium and 0.5% sucrose. The media was replaced with fresh MS + sucrose medium on day 7 and with water on day 14. Elicitor and treatment solutions were prepared with sterile distilled water to the final concentration of L. bicolor (OD600 = 0.2) and 20x-dead L. bicolor. 15-day-old seedlings were treated with 500 µl of L. bicolor treatments. Seedlings were treated with for 15 minutes and immediately frozen in liquid nitrogen after 15 minutes of treatment. The samples were stored at -80 °C before being weighed. Samples were homogenized while frozen prior to protein extraction. MAP Kinase activation in the samples were analyzed according to68. Separated proteins were quantified by staining with Ponceau Red and de-stained using stripping buffer.
Statistical analyses
Statistics were computed in R66. For the caterpillar feeding experiments, linear mixed-effect models were applied to log-transformed T. ni larval weight gain from all experiment repetitions. For the Pto CFU/mL experiments, linear mixed-effect models were applied to the untransformed data from all experiment repetitions. Experiment repetition was applied as random effect (function ‘lmer’ from package ‘lme4’69). Normal distribution and homogeneity of variance was verified by visual inspection of the residual plots. Statistical significance was determined by performing two-way ANOVAs on the linear mixed-effects model fit to the data (function ‘Anova’ from package ‘car’,70. Homogeneous subsets were determined by a post-hoc test (Tukey’s HSD, function ‘glht’ from package ‘multcomp’71).
For the phytohormone, metabolome and qRT-PCR analyses, the data obtained from the treatments were normalized to mock treated samples. Linear models were applied to the data. Normal distribution and homogeneity of variance was verified by visual inspection of the residual plots. Statistical significance was determined by performing two-way ANOVAs on the linear models. Homogeneous subsets were determined by a post-hoc test, Fisher’s LSD for phytohormone and metabolite analyses, and Tukey’s HSD for qRT-PCR data. RNAseq analyses were performed using the ‘DESeq2’ package65.
ACKNOWLEGDEMENTS
This work was supported by funding in the frame of the International Research Training Group 2172: PRoTECT – Plant Responses To Eliminate Critical Threats by the Deutsche Forschungsgemeinschaft (DFG) hosted by the University of Goettingen (Germany), for the Service Unit for Metabolomics and Lipidomics (DFG, INST 186/822-1) to I. F. and the University of British Columbia (Canada). This work was also supported by a Natural Sciences and Engineering Research Council (NSERC) Discovery Grant (NSERC-RGPIN-2016-04121) awarded to C.H.H. Y.L was supported by an NSERC CREATE-PRoTECT award and a Chinese Graduate Scholarship Council Award. We thank K. Ziesing and Sabine Freitag for technical assistance and Volker Lipka and Sina Barghahn for support with MAPK assays and for critical reading of the manuscript.