Bacterial type III effector-induced plant C8 volatiles elicit antibacterial immunity in heterospecific neighbouring plants via airborne signalling.

Upon sensing attack by pathogens and insect herbivores, plants release complex mixtures of volatile compounds. Here, we show that the infection of lima bean (Phaseolus lunatus L.) plants with the non-host bacterial pathogen Pseudomonas syringae pv. tomato led to the production of microbe-induced plant volatiles (MIPVs). Surprisingly, the bacterial type III secretion system, which injects effector proteins directly into the plant cytosol to subvert host functions, was found to prime both intra- and inter-specific defense responses in neighbouring wild tobacco (Nicotiana benthamiana) plants. Screening of each of 16 effectors using the Pseudomonas fluorescens effector-to-host analyser revealed that an effector, HopP1, was responsible for immune activation in receiver tobacco plants. Further study demonstrated that 1-octen-3-ol, 3-octanone and 3-octanol are novel MIPVs emitted by the lima bean plant in a HopP1-dependent manner. Exposure to synthetic 1-octen-3-ol activated immunity in tobacco plants against a virulent pathogen Pseudomonas syringae pv. tabaci. Our results show for the first time that a bacterial type III effector can trigger the emission of C8 plant volatiles that mediate defense priming via plant-plant interactions. These results provide novel insights into the role of airborne chemicals in bacterial pathogen-induced inter-specific plant-plant interactions.

INTRODUCTION

Microbe‐induced plant volatiles (MIPVs) act as infochemicals that mediate communication between plants and other members of the phytobiome (Sharifi & Ryu, 2018a, 2018b). Given their vaporous characteristic, volatile organic compounds (VOCs) can diffuse freely through air, allowing conspecific and heterospecific communications (Weisskopf, Schulz, & Garbeva, 2021). For instance, multi‐trophic responses have been observed in plants exposed to airborne signalling compounds produced by a diverse range of plant species upon treatment with diverse chemicals (Farmer & Ryan, 1990; Godard, White, & Bohlmann, 2008), biological inducers (Arimura et al., 2000; Bruin, Dicke, & Sabelis, 1992; Engelberth, Alborn, Schmelz, & Tumlinson, 2004; Rhoades, 1983; Shulaev, Silverman, & Raskin, 1997) and physical inducers (Baldwin & Schultz, 1983; Dolch & Tscharntke, 2000; Glinwood, Ahmed, Qvarfordt, Ninkovic, & Pettersson, 2009; Karban, Baldwin, Baxter, Laue, & Felton, 2000; Karban, Huntzinger, & McCall, 2004). In lima bean (Phaseolus lunatus L.), treatment with the chemical inducer benzothiadiazole (BTH) triggered metabolic influx into non‐anal biosynthesis, which in turn elicited systemic acquired resistance (SAR) in undamaged conspecific neighbours via the expression of the pathogenesis‐related protein 2 (PR2) gene, and subsequently reduced symptom appearance when challenged by Pseudomonas syringae pv. syringae (Yi, Heil, Adame‐Alvarez, Ballhorn, & Ryu, 2009). In wheat (Triticum aestivum L.), infection with the leaf rust pathogen, Puccinia triticina Erikss, induced the emission of ocimene, which acts as an airborne inducer, triggering resistance in neighbouring plants, as indicated by the increased expression of two defense response marker genes encoding β‐1,3‐glucanase and PR1 (Castelyn, Appelgryn, Mafa, Pretorius, & Visser, 2015).

To resist invading pathogens, plants have evolved two innate immune systems. The first system is activated upon recognition of microbe‐associated molecular patterns (MAMPs) or pathogen‐associated molecular patterns (PAMPs) such as lipopolysaccharides, flagellin, elongation factor Tu (EF‐Tu), peptidoglycan (PGN) and lipopeptides (Böhm, Albert, Fan, Reinhard, & Nürnberger, 2014; Zipfel, 2014) by pattern recognition receptors (PRRs) (Boller & Felix, 2009). This process is called pattern‐triggered immunity (PTI). The activation of PTI leads to an array of signalling events including the activation of mitogen‐activated protein kinases (MAPKs) and calcium‐dependent protein kinases (CDPKs), rapid generation of reactive oxygen species (ROS) and reinforcement of the cell wall (Macho & Zipfel, 2015), thereby suppressing the growth of invading pathogens.

To overcome PTI, several Gram‐negative bacterial pathogens, such as Pseudomonas syringae pv. tomato (Pto), have evolved a type III secretion system (T3SS) (Büttner & He, 2009). This system resembles a molecular syringe and directly injects the bacterial effector proteins into the host cell cytosol, rendering the plant immune system dysfunctional and promoting bacterial infection (Büttner & He, 2009; Guo, Tian, Wamboldt, & Alfano, 2009; Y. J. Kim, Lin, & Martin, 2002; Lewis et al., 2015). For instance, AvrPtoB, a type III effector of Pto, promotes the degradation of FLS2, a flagellin receptor of Arabidopsis thaliana, by functioning as an E3 ligase (Göhre et al., 2008). In return, plants have evolved a second layer of immunity, known as effector‐triggered immunity (ETI), which senses the presence of effector proteins via the resistance (R) proteins (Büttner & He, 2009; Cunnac, Lindeberg, & Collmer, 2009; Laflamme et al., 2020). After recognition, R proteins leads to the accumulation of salicylic acid (SA) and the induction of PR genes as well as programmed cell death (PCD), that is, hypersensitive response (HR), in the infected host cells to suppress the migration of the pathogen to neighbouring cells (Hammond‐Kosack & Jones, 1997; Hatsugai et al., 2017). In addition, PTI and ETI are involved in non‐host resistance that provides broad‐spectrum immunity against a broad spectrum of phytopathogens, which are pathogenic to other plant species (Lee et al., 2017; Mysore & Ryu, 2004; Senthil‐Kumar & Mysore, 2013). A pathogen that is avirulent to a non‐host plant is referred to as a non‐adapted/non‐host pathogen (Mysore & Ryu, 2004) and can provide durable resistance to infected plants.

ETI not only triggers the local HR response but also activates systemic resistance, especially SAR, through local SA biosynthesis (Durrant & Dong, 2004; Fu & Dong, 2013). Whereas SAR is triggered by avirulent bacterial pathogens and its non‐host resistance in association with SA, induced systemic resistance (ISR), another form of systemic resistance, is triggered by the interaction of plants with beneficial microbes, particularly plant growth–promoting rhizobacteria (PGPR). In addition, ISR generally shows a strong association with jasmonic acid (JA) and ethylene (ET) signalling, which is primarily regulated by the JA‐/ET‐associated transcription factors including MYC2 (Choudhary, Prakash, & Johri, 2007; Pozo, Van Der Ent, Van Loon, & Pieterse, 2008). Collectively, SAR and ISR play a pivotal role in priming the defense response of plants to prevent secondary infection (Conrath, Beckers, Langenbach, & Jaskiewicz, 2015; Martinez‐Medina et al., 2016; Vlot et al., 2021). Such defense priming permits plants to save resources rather than continuously utilizing them to enhance pathogen resistance, which can negatively influence plant growth and yield (van Hulten, Pelser, Van Loon, Pieterse, & Ton, 2006).

The suppression of SA signalling and volatile production can be accomplished by injecting the effector proteins of virulent bacterial strains into the plant cell cytosol (M. Huang et al., 2012). The 2b effector protein of the Cucumber mosaic virus (CMV) suppressed JA signalling in the host plant and altered volatile emission to attract its insect vector (Tungadi et al., 2017), suggesting that pathogen effectors play a key role in producing a particular blend of MIPVs, which affects the immune response of the receiver plant. In a recent study, the PGPR Bacillus amyloliquefaciens strain GB03 triggered the emission of MIPVs, particularly β‐caryophyllene, in tomato (Solanum lycopersicum L.), which elicited the release of a substantial amount of SA in root systems, rendering the rhizosphere microbial community of the neighbouring plant similar to that of the inoculated plant (Kong, Song, Sim, & Ryu, 2021). However, it remains unclear whether pathogen‐triggered MIPV emission contributes to resistance in neighbouring plants and which molecular determinants in the pathogen affect the challenged plants.

In the present study, we investigated the effect of a pathogenic bacterium‐triggered MIPV on heterospecific vegetation to identify airborne immune responses in neighbouring plants. A bacterial effector protein HopP1 was identified as a factor responsible for the triggering of the emission of MIPVs from emitter plants and eliciting systemic resistance to pathogens in receiver plants. To the best of our knowledge, this is the first demonstration of type III effector–triggered emission of MIPVs, which conferred immune activation in heterospecific neighbouring plants.

MATERIALS AND METHODS

Plant material, growth conditions and treatments

Seeds of tobacco (Nicotiana benthamiana), lima bean (Phaseolus lunatus L.), pepper (Capsicum annuum L. cv. Bukwang) and cucumber (Cucumis sativus L.) were surface‐sterilized with 3% sodium hypochlorite and then washed four times with sterile distilled water. The sterilized seeds were plated on half‐strength Murashige and Skoog (1/2 MS) medium containing 0.6% agar and 1.5% sucrose (pH 5.8) and incubated at 23°C under a 16 hr light/8 hr dark cycle for 3–5 days until germination. The germinated seeds were then placed on soilless medium (Bunong), and seedlings were grown in an environmentally controlled growth room maintained at 25°C ± 2°C, 12 hr light/12 hr night cycle and approximately, 7,000 lx light intensity using fluorescent bulbs. Exposure experiments under in vivo conditions were conducted in closed transparent acrylic plastic boxes (20 cm × 60 cm × 20 cm). A 10 ml solution of 0.5 mM benzothiadiazole (BTH) was applied to 3‐week‐old lima bean, cucumber, tobacco and pepper plants as a direct treatment. To expose the receiver plant to volatile emitter plants, five plants of the same age were placed adjacent to BTH‐treated plants while avoiding direct physical contact between them. Plants treated only with sterile distilled water and plants placed adjacent to these water‐treated plants served as controls for the direct and indirect treatments, respectively. Airborne induction and priming of the defense response in receiver plants were evaluated by inoculating the plants with foliage pathogens, namely Pseudomonas syringae pv. syringae (causal organism of bacterial spot disease in lima bean), P. syringae pv. tabaci (Pta) (causal organism of wildfire disease in tobacco) and P. syringae pv. lachrymans (causal organism of angular leaf spot disease in cucumber). These pathogens were grown on solid King's medium, while Xanthomonas axonopodis pv. vesicatoria (Xav) (causal organism of bacterial spot disease in pepper) were grown on Luria‐Bertani (LB) agar medium at 30°C for 2 days. Pathogen population size was evaluated at 0 and 3 days post inoculation (dpi).

To evaluate whether a biological elicitor induces MIPVs that conspecifically confer defense priming in neighbouring plants, the avirulent (non‐host) pathogen Pto (known as a SAR inducer) and its hrp (T3SS) null mutant were spray inoculated onto 3‐week‐old lima bean plants with 8–10 fully expanded leaves. Thereafter, tobacco seedlings were placed in close proximity to the emitter plant, while avoiding direct physical contact between them (Figure 1). One week later, the tobacco plants were spray inoculated with Pta (H.‐C. Huang et al., 1988) to evaluate the SAR‐triggering ability of the volatiles.

Figure 1

Schematic representation of the experimental design. In an acrylic chamber (width, length and height = 55 × 95 × 65 cm3), a 3‐week old lima bean (emitter) plant was inoculated with Pseudomonas syringae pv. tomato DC3000 (Pto) to induce volatile production. Tobacco (receiver) plants placed in the same chamber were exposed to these volatiles for 1 week and then challenged with Pseudomonas syringae pv. tabaci (Pta). The activation of disease resistance in receiver plants by the airborne signalling molecules was evaluated by assessing the Pta population size [Colour figure can be viewed at wileyonlinelibrary.com]

To determine the effectors that were responsible for systemic resistance in receiver plants, individual type III effectors of Pto (AvrPto1, AvrE, HopA1, HopAB2, HopAF1, HopAm1‐1, HopC1, HopE, HopF2, HopI, HopM1, HopO1‐1, HopP1, HopQ1‐1, HopX and HopY) were expressed in cells of emitter plants using the Pseudomonas fluorescens (Pf) effector‐to‐host analyser (EtHAn), a non‐pathogenic strain with no effector (Thomas, Thireault, Kimbrel, & Chang, 2009) (Figure 2c). All 16 effector constructs in the Pf EtHAn strain and the Pf EtHAn strain with an empty vector were obtained from Dr. Jeff Chang, Oregon State University, Oregon, USA. To inoculate plants, bacteria were scraped off of the plates and resuspended in sterile distilled water. The concentration of each bacterial suspension was adjusted to 107 cfu/ml, based on the OD600 value. The bacterial suspensions were sprayed onto the plants, and the number of bacterial cells on leaf discs (diameter = 1 cm) sampled from five leaflets was counted at 0 and 3 days post inoculation (dpi). This experiment was performed in three independent biological replicates, with each replicate containing five plants.

Figure 2

Bacterial type III effector‐induced plant volatile emission in the emitter plant elicits immunity in the neighbouring receiver plants. (a) Pathogen population in receiver plants. Tobacco (receiver) plants were exposed to volatiles released by the Pto‐inoculated lima bean (emitter) plant for 7 days. The Pta population sizes in receiver plants that were pre‐exposed to Pto and non‐exposed control plants were measured at 0 and 3 dpi with Pta. The hrpL mutant was used as a negative control to study the function of the effector. The represent pictures regarding Pta resistance in Pto pre‐exposed tobacco leaf are provided in Figure S4 (b) Expression analysis of disease resistance genes, NbPR1a, NbPR1c, NbHIN1 and NbSAR8.2, in receiver and control plants by quantitative reverse‐transcription PCR (qRT‐PCR) at 0 and 6 hpi. (c) Schematic of the delivery system used to stably introduce individual type III effectors from Pto into plant cells using P. fluorescens (Pf) EtHAn, a non‐pathogenic strain with no native effector. IM: inner membrane; OM: outer membrane; PM: plasma membrane. Data represent mean ± standard error (SE; n = 4 plants per treatment). Different letters indicate significant differences between treatments (p < .05; least significant difference [LSD] test). The experiment was repeated three times [Colour figure can be viewed at wileyonlinelibrary.com]

To assess the direct effect of synthetic C8 compounds on plant immunity, tobacco plants were exposed to 1‐octen‐3‐ol, 3‐octanone and 3‐octanol (Sigma‐Aldrich) at different concentrations (100, 1,0.01, and 0.0001 mM) for 1 week. Tobacco grown for 3 weeks were placed in an acrylic chamber (15 cm × 30 cm), and 5 ml of 1‐octen‐3‐ol volatile was exposed to the tobacco plants. Both sides of the acrylic chamber were sealed with plastic wrap. Pathogen population was evaluated as described above.

Bacterial culture

Pseudomonas syringae pv. tomato DC3000 (Pto; non‐host pathogenic bacterium), Pto hrp (T3SS) null mutant and Pseudomonas syringae pv. tabaci (Pta; foliage pathogen) (H.‐C. Huang et al., 1988) were cultivated on Pseudomonas Agar F medium (Difco, Detroit, MI). Pto suspension was prepared in 10 mM MgCl2 at an optical density (OD600) of 0.01 for eliciting volatile emission in emitter plants, while Pta suspension (OD600 = 1; concentration = 108 colony forming units [cfu]/ml) was prepared for spray inoculating receiver plants.

Plant RNA extraction, cDNA synthesis and gene expression

Total RNA was isolated from tobacco leaves harvested at 0, 6 and 24 hr post inoculation (hpi) with Pta using the RNeasy® Plus Mini Kit (Qiagen, USA), according to the manufacturer's instructions. Quantitative reverse‐transcription PCR (qRT‐PCR) assays were performed as described previously (Song, Sim, Kim, & Ryu, 2016). The expression of candidate defense priming genes was analysed using the following primer pairs: NbPR1a‐F/R (5′‐AATATCCCACTCTTGCCG‐3′ and 5′‐CCTGGAGGATCATAGTTG‐3′, respectively), NbSABP‐F/R (5′‐ACCATCAGACCAAGATGT‐3′ and 5′‐TGGCTAAGAGTGGAAGGT‐3′, respectively), NbHIN1‐F/R (5′‐AATATCCCACTCTTGCCG‐3′ and 5′‐CCTGGAGGATCATAGTTG‐3′, respectively), and NbPR2‐F/R (5′‐ACCATCAGACCAAGATGT‐3′ and 5′‐TGGCTAAGAGTGGAAGGT‐3′, respectively). PCR was performed using the following thermocycler parameters: 95°C for 10 min, followed by 44 cycles at 95°C for 30 s, 60°C for 30 s and 72°C for 42 s. Transcript levels of each tobacco gene were calibrated and normalized using the housekeeping gene NbACT (GenBank accession no. U60489), and relative quantification of specific mRNA levels was performed using the comparative 2–Δ(ΔCt) method (Livak & Schmittgen, 2001).

Preparation and conditioning of silicone laboratory tubing

Polydimethylsiloxane (PDMS) tubing pieces were prepared as described previously (Song et al., 2016). Briefly, silicone tubing (1 mm i.d. × 1.8 mm o.d.; Carl Roth GmbH, Germany) was cut into 0.5 cm pieces, soaked in acetonitrile:methanol solution (4:1, v/v) overnight and baked at 210°C for 3 hr in glass columns under nitrogen (N2) gas flow. The baked PDMS tubing pieces were cooled under N2 gas flow and stored in aliquots after purging with argon (Ar) gas.

Analysis of volatile compounds

Volatile compounds of lima bean (emitter) plants were analysed by thermal desorption coupled with gas chromatography–mass spectrometry (TD‐GC–MS) on a GC–MS‐QP 2010 Ultra (Shimadzu Corporation, Japan) gas chromatograph–mass spectrometer, equipped with an Rtx‐5MS column (30 m long, 0.25 mm i.d., 0.25 μm film thickness; Restek, USA) (Kallenbach et al., 2014). An individual piece of PDMS tubing was placed in an 89 mm glass TD tube (Supelco, USA) and desorbed under a stream of N2 gas (flow rate = 60 ml/min) at 200°C for 8 min. All substances desorbed from the PDMS tubing were cryo‐focussed at −20°C onto a Tenax® adsorbent trap in front of the column. After desorption, the Tenax® trap was heated to 230°C within 10 s, and analytes were injected using a 1:20 split ratio at a constant linear velocity of 40 cm s−1 with helium (He) as the carrier gas. The TD–GC interface was held at 230°C. The GC column temperature was held at 40°C for 5 min and then ramped up to 185°C at a rate of 5°C/min and finally to 280°C at a rate of 30°C/min, where it was held for 0.83 min. The electron impact (EI) spectra were recorded at 70 eV in scan mode at a mass‐to‐charge ratio (m/z) ranging from 33 to 400. The transfer line was held at 240°C, and the ion source was held at 220°C. Data were processed using the GC–MS solution software (version 4.20; Shimadzu Corporation).

Virus‐induced gene silencing assay

To perform the VIGS assay, 2‐week‐old tobacco seedlings were infiltrated with the tobacco rattle virus (TRV)‐based VIGS vector (pTRV2) containing a plant defense‐ or hormone‐related gene (NbSABP1, NbMEK1, NbSKP1, NbMYB1 or NbTGA2.1) or empty vector control (pTRV2::00), as described previously (Anand et al., 2008; K. H. Kim et al., 2016; Senthil‐Kumar & Mysore, 2014). The TRV‐VIGS vectors were infiltrated into tobacco leaves using Agrobacterium tumefaciens strain GV2260, which was grown at 28°C in LB broth containing appropriate antibiotics. The inoculation buffer (10 mm MgCl2, 10 mm MES [pH 5.6] and 150 μM acetosyringone) containing Agrobacterium (transformed with both pTRV1 and pTRV2) was adjusted to a final OD600 of 1.0 and shaken at room temperature for 4–6 hr before infiltration. Then, Agrobacterium cultures containing pTRV1 and pTRV2 vectors were mixed in a 1:1 ratio and infiltrated into the leaves of 2‐week‐old tobacco plants with a needleless syringe. The infiltrated plants were monitored for approximately 14 days. This experiment was performed in three independent biological replicates, with each replicate containing nine plants.

Statistical analysis

The experimental data were subjected to the analysis of variance (ANOVA) using the JMP software version 5.0 (SAS Institute Inc., Cary, NC; www.sas.com). The normality and homogeneity of variance of the data were assessed and when two assumptions were not met, the data were transformed using Box‐Cox using a package MASS with R Studio software. Significant treatment effects were determined based on the magnitude of the F‐value (p = .05). When a significant F‐value was obtained, separation of means was accomplished by the protected Fisher's least significant difference (LSD) test at p = .05.

RESULTS

Experimental set‐up and combination selection for interspecific plant–plant interactions

To evaluate intra‐specific and inter‐specific airborne priming of disease resistance, we performed an in vivo disease assay with four plant species: lima bean, tobacco, cucumber and pepper (Figure S1). Among these plants, BTH‐induced lima bean displayed the highest intra‐specific disease priming effect, suppressing the pathogen population in leaf by 40% compared with the control at 3 dpi (Figure S1b). Thus, lima bean was chosen as the emitter plant, and BTH‐induced volatiles of lima bean plants were used to treat tobacco, cucumber and pepper plants (heterospecific interaction). Among receiver plants, the only significant pathogen suppression occurred with tobacco (Figure S1b). Thus, the heterospecific lima bean–tobacco plant combination was used as a model system to further evaluate defense priming in the receiver plant upon exposure to biologically induced volatiles from emitter plants.

Immune system activation by a pathogenic bacterium and its type III effectors through airborne communication

The immune activation effect of MIPVs was determined by examining the suppression of pathogenicity in neighbouring (receiver) tobacco plants pre‐exposed to biologically altered volatiles released by the Pto‐treated lima bean (emitter) plant. The results showed that the pathogen population in tobacco plants exposed to volatiles emitted from Pto‐treated lima bean plants was significantly reduced at 3 dpi with Pta (7.6 log CFU·per leaf disc) compared with water‐treated (control) lima bean plants (8.1 log CFU·per leaf disc) (Figure 2a). By contrast, exposure to volatiles emitted from Pto hrpL mutant‐treated lima bean plants showed no change in pathogenic cell numbers on tobacco leaves upon Pta inoculation, indicating that the induction of pathogen resistance in receiver plants by Pto‐induced plant volatiles released by emitter plants is strongly associated with the bacterial hrpL gene.

Next, we examined the expression of disease resistance genes, NbPR1a, NbPR1c, NbHIN1 and NbSAR8.2, in tobacco plants pre‐exposed to volatiles emitted from Pto‐ and Pto hrpL mutant‐treated lima bean plants. At 6 hpi with Pta, the relative expression levels of NbPR1a and NbPR1c were significantly lower in Pto hrpL‐null mutant pre‐treated tobacco by 59% and 64%, respectively, than in wild‐type Pto pre‐treated tobacco (Figure 2b); however, no difference in NbHIN1 and NbSAR8.2 expression was observed in tobacco plants pre‐exposed to Pto hrpL‐treated lima bean plants compared with the control (Figure 2b). These results indicate that the Pto hrpL gene, which encodes a regulatory component of the T3SS, is essential for triggering SA‐dependent defense priming in receiver plants. To further investigate the effects of individual type III effectors of Pto on airborne immune activation in tobacco, we employed a type III effector delivery system to translocate individual effectors from Pf EtHAn, a non‐pathogenic strain lacking the T3SS and effectors, into the cytosol of plant cells. Thomas and colleagues transformed Pf with the structural genes of the Pto T3SS and named the resulting strain as EtHAn (Thomas et al., 2009). Of the 16 type III effectors tested in this study, we identified three effectors (HopP1, HopL and HopE) that conferred increased resistance to a pathogen in a neighbouring plant (Figure 3). Finally, we selected the effector HopP1, as it showed the greatest reduction in the population size of Pta on tobacco leaves (Figure 3). Tobacco plants pre‐exposed to Pf EtHAn::HopP1‐treated lima bean plants showed the smallest Pta population size (5.9 log cfu/disc) compared with those pre‐exposed to Pf EtHAn‐treated and control lima bean plants (7.4 and 7.3 log cfu/disc, respectively), indicating that HopP1 is the main effector of Pto affecting airborne suppression in the Pta population size in neighbouring plants. By contrast, the HopY effector increased the population size of Pta on tobacco leaves (8.0 log cfu/disc), thus suppressing immune activation in neighbouring plants.

Figure 3

Bacterial type III effector‐mediated immune activation in receiver plants via emitter plant‐derived volatiles. Pathogen populations were measured in tobacco (receiver) plants at 3 dpi with Pta (OD600 = 1). Emitter plants were treated with P. fluorescens (Pf) EtHAn expressing various bacterial effectors to induce effector‐mediated volatile compound emission. Data represent mean ± SE (n = 4 plants per treatment). Different letters indicate significant differences between treatments (p = .05; LSD). E.V.: empty vector. The experiment was repeated three times [Colour figure can be viewed at wileyonlinelibrary.com]

Bacterial type III effector HopP1‐induced plant C8 volatiles

To identify VOCs released by type III effector‐treated emitter plants, we performed GC–MS analysis of volatile compounds collected at 24‐hr intervals for 3 days. The GC chromatogram profiles exhibited a continuous shift in the composition of volatile compounds released by Pto‐ and HopP1‐treated emitter plants (n = 6) (data not shown). Among all VOCs identified, 1‐octen‐3‐ol, 3‐octanone and 3‐octanol were released in significantly large amounts in response to Pto and HopP1 treatments compared with the control treatment. These three VOCs were selected as keystone compounds, and their levels were further analysed in the VOC profiles of wild‐type Pto‐, Pf EtHAn::HopP1‐ and Pf EtHAn‐treated emitter plants (Figure 4a). Our results showed that the syntheses of 1‐octen‐3‐ol, 3‐octanone and 3‐octanol were strongly increased in response to HopP1 (Figure 4b–d). By contrast, no significant increase in the amounts of these molecules was observed in emitter plants treated with T3SS mutants, Pto hrcC and Pto hrpL, compared with control plants (no bacteria).

Figure 4

Identification of volatile compounds produced by lima bean plants upon treatment with the bacterial effector HopP1. (a) Gas chromatography (GC) profiles of volatiles emitted from lima bean plants at 72 hpi after treatment with wild‐type Pto, Pf EtHAn HopP1, Pf EtHAn and control. (b–d) Peak areas measured for the three C8 volatiles, 1‐octen‐3‐ol (b), 3‐octanone (c), and 3‐octanol (d), at 72 hpi with Pto, Pto hrcC‐, Pto hrpL‐, Pf EtHAn HopP1, Pf EtHAn and non‐treated control. Data represent mean ± SE (n = 4 plants per treatment). Different letters indicate significant differences between treatments (p = .05; LSD test). The experiment was repeated three times [Colour figure can be viewed at wileyonlinelibrary.com]

Elicitation of ISR by 1‐octen‐3‐ol

Immune system activation in tobacco by airborne C8 volatiles was tested using standard VOCs at different concentrations. At 3 dpi, treatment with 1 mM 1‐octen‐3‐ol led to a significant reduction in Pta population size on tobacco leaves (7.8 log cfu/disc); however, treatment with 1‐octen‐3‐ol at concentrations <.01 mM resulted in no suppression of pathogen growth compared with the control (Figure 5b). On the other hand, 3‐octanone and 3‐octanol induced SAR in receiver plants at 100 mM concentration, indicating that 1‐octen‐3‐ol was the most effective airborne signalling molecule to trigger immune activation in receiver plants (Figure S2).

Figure 5

Effect of 1‐octen‐3‐ol, a keynote volatile, on the immunity of receiver plants. (a) Representative photograph showing Pta‐induced symptoms on tobacco (receiver) plants at 7 dpi. Prior to pathogen infiltration, plants were exposed to varying concentrations of 1‐octen‐3‐ol for 7 days. (b) Quantification of pathogen population at 0 and 3 dpi with Pta in receiver plants exposed to different concentrations of 1‐octen‐3‐ol. (c) Expression levels of NbPR1a, NbPR1c, NbHIN1 and NbSAR8.2 in tobacco plants exposed to 1‐octen‐3‐ol. Gene expression was analysed by quantitative real‐time polymerase chain reaction (qRT‐PCR) at 0, 6 and 24 hpi. (d) Quantification of pathogen population in SABP2‐ and CO1‐silenced tobacco plants treated with 1‐octen‐3‐ol or water (control). After treatment with 1‐octen‐3‐ol or water for 7 d, Pta (108 cfu/ml) was infiltrated into tobacco leaves, and pathogen population size was measured at 0 and 3 dpi. Data represent mean ± SE (n = 5 plants per treatment). Different letters indicate significant differences between treatments (p = .05; LSD test) in each silencing plant. The experiment was repeated three times [Colour figure can be viewed at wileyonlinelibrary.com]

To confirm that 1‐octen‐3‐ol induces SAR in receiver plants, we performed qRT‐PCR analysis and the VIGS assay for defense‐related genes. Expression analysis revealed that NbPR1a, NbPR1c, NbHIN1 and NbSAR8.2 were expressed to higher levels in receiver plants exposed to 1‐octen‐3‐ol for 6 and 24 hr compared with the control (Figure 5c). In addition, silencing of SA biosynthesis‐ and signalling‐related genes SABP2, NbSIPK, NbICS and NbNPR1 disrupted the protective effect of 1‐octen‐3‐ol on tobacco plants against Pta (Figure 5d and Figure S3), whereas silencing the JA signalling–related gene, COI1, enhanced the protective effect of 1‐octen‐3‐ol (Figure 5d) on receiver plants by suppressing pathogen population by 13.8%. Collectively, these results indicate that 1‐octen‐3‐ol treatment enhances plant immunity through the SA‐dependent signalling pathway.

DISCUSSION

Airborne plant–plant communication is highly advantageous for plant fitness under natural conditions (Heil, 2008). Although HIPVs have been studied for a long time, MIPVs have not been intensively evaluated (Sharifi, Lee, & Ryu, 2018). In this study, we demonstrated for the first time that bacterial type III effector‐mediated MIPV emission can elicit immune activation in the neighbouring plant through airborne communication. Interspecific airborne communication is not a new phenomenon. However, while previous studies primarily focused on the ecological impact of this mode of communication, for instance, demonstrating VOC‐mediated interaction between two different plant species in the field (Karban et al., 2014; Karban & Yang, 2020; Pearse, Hughes, Shiojiri, Ishizaki, & Karban, 2013), the agricultural value of VOC‐mediated inter‐specific communication has been largely overlooked. Here, we used legume (lima bean) and solanaceous (tobacco) plants as emitter and receiver plants, respectively, to explore to advantages of VOC‐mediated interspecific communication in agriculture. The emitter plant was biologically primed by Pto, a non‐host (avirulant) pathogen, to provoke MIPV emission, and its ability to induce SAR against a virulent Pta was evaluated in receiver plants (Figure 1). The results of the current study broaden our knowledge of the modulation of MIPVs by the secretion of bacterial type III effectors and their role in eliciting defense against pathogens in neighbouring plants.

T3SS and effector proteins act as the main components of the virulence machinery of Pto, a non‐host bacterial pathogen of lima bean (Xin & He, 2013). Here, we showed that Pto is capable of triggering the production of MIPVs, and HopP1 plays a key role in inducing specific signalling MIPV‐associated SAR in neighbouring plants. Similar to HopP1, the delivery of several other effectors such as HoPC1, HopI and HopE into the leaf cell cytosol of lima bean plants also elicited airborne immune activation in tobacco (Figure 3). Among bacterial effectors, structural proteins such as harpins, which induce HR and pathogenicity in host plants, were characterized as the first pathogen‐independent HR elicitors in plants (Kvitko, Ramos, Morello, Oh, & Collmer, 2007; Wei et al., 1992). Kvitko and colleagues identified HopP1 and HopAK1 as harpin genes in P. syringae (Kvitko et al., 2007). Structurally, HopP1 lacks cysteine residues, and carries soluble lytic transglycosylase (LT) domain at its C terminus (Oh, Kvitko, Morello, & Collmer, 2007). Although LT generally acts on PGNs, the LT of HopP1 is unique, as it can be translocated by the T3SS into plant cells (Choi, Kim, Lee, & Oh, 2013). In non‐host plants, HopP1 elicits the HR (Kvitko et al., 2007). In the current study, HopP1 induced the production of C8 MIPVs in lima bean plants, and these volatiles travelled to the receiver tobacco plants, subsequently inducing SAR. Although our results revealed that the bacterial effector HopP1 is responsible for the emission of C8 volatiles, the underlying mechanism remains largely unknown. One possible scenario is that HopP1 is associated with the Arabidopsis transcriptional regulator abscisic acid repressor 1 (ABR1), an ethylene response factor (ERF). A recent study using the yeast two hybrid assay showed that HopP1, HopAF1 and HopQ1‐1 of Pto DC3000 can interact with ABR1, which is known as an effector hub in plants (Schreiber, Hassan, & Lewis, 2021). In addition, another study proposed that ABR1 regulates the co‐expression of lipid metabolism–related genes such as AAPT during post‐harvest chilling injury of peach (Zhu et al., 2019). Collectively, these studies suggest that docking of HopP1 to the ABR1 hub upregulates lipid metabolism, a key contributor to the biosynthesis of aliphatic molecules (Dudareva, Klempien, Muhlemann, & Kaplan, 2013), to increase C8‐volatile emission. However, other effectors, namely HopAF1 and HopQ1‐1, did not suppress pathogen population size (Figure 3), suggesting that HopP1 is also involved in other yet to be characterized mechanisms that eventually boost the abundance of C8 VOC molecules. Interestingly, it is noteworthy that in contrast to the immune‐activating effectors HopC1, HopI and HopE, HopY elicited greater susceptibility to pathogens (Figure 3). Further investigation is needed to determine how some effector proteins enhance, while others suppress, the plant immune system.

To understand inter‐specific communication, it is important to identify the infochemicals responsible for altering the immunity in receiver organisms. Our GC–MS analysis of the Pto‐inoculated emitter plant demonstrated quantitative differences in volatile profiles, revealing greater accumulation of C8 aliphatic molecules, namely, 1‐octen‐3‐ol, 3‐octanone, and 3‐octanol, in these plants than in non‐inoculated plants. These three VOCs are the products of the oxidative breakdown of linoleic acid and belong to the large family of oxylipins (Assaf, Hadar, & Dosoretz, 1997). Oxylipins are oxygenated natural products derived from fatty acids that act as signalling molecules for intra‐ and intercellular communication in various organisms including plant and fungi (Tsitsigiannis & Keller, 2007). A recent study showed that oxylipins function as xylem‐resident signals regulating Trichoderma virens‐mediated ISR in maize (Zea mays L.) (Carella, 2020; Wang, Borrego, Kenerley, & Kolomiets, 2020). Consistent with this finding, C8 compounds appear to function as interactive interspecific messengers between plants spaced far apart. In a previous study, 1‐octen‐3‐ol directly suppressed the appearance of disease symptoms caused by Botrytis cinerea, a necrotrophic fungal pathogen of Arabidopsis (Kishimoto, Matsui, Ozawa, & Takabayashi, 2007). In another study, 1‐octen‐3‐ol demonstrated an inhibitory effect on both Gram‐negative and Gram‐positive bacteria under in vitro conditions, with a minimum inhibitory concentration (MIC) of 2.0 mg/ml (Al‐Fatimi, Wurster, & Lindequist, 2016). Although we cannot exclude a direct inhibitory effect of 1‐octen‐3‐ol on Pta, its effective concentration was considerably lower than the MIC reported by Al‐Fatimi and coworkers (Al‐Fatimi et al., 2016).

From our finding, effector‐mediated changes in phytohormone levels seem to be the major driver affecting the emission of plant volatiles. Previously, Huang and colleagues (J. Huang et al., 2003) demonstrated that tobacco plants infected with non‐host pathogens such as P. syringae pv. tomato and P. syringae pv. maculicola emitted increased amounts of MeSA and methyl jasmonate (MeJA) (J. Huang et al., 2003). Another study showed that JA is associated with the biosynthesis of volatile compounds but not via SA biosynthesis (Assaf et al., 1997). JA phytohormone is regulated by JASMONATE ZIM‐DOMAIN (JAZ) proteins. The CMV 2b effector protein, which affects the JA pathway of the host plant, inhibits JAZ proteins and represses JA signalling, which in turn alters plant volatile emissions that attract aphids, a CMV vector (Assaf et al., 1997). These studies have revealed the mechanism underlying viral effector‐induced emission of specific volatiles, and the ecological role of these volatiles in inter‐kingdom interactions. However, both bacterial effector‐mediated volatile emission and its role and mechanism in inducing resistance in adjacent organisms are still poorly understood. In the current study, knockdown of SA biosynthesis and signalling genes including SABP2, NbSIPK, NbICS and NbNPR1 in tobacco (receiver plant) did not protect against Pta, whereas silencing of COI1, a JA receptor, increased disease resistance (Figure 5d and Figure S3). These results suggest that SA biosynthesis and signalling pathways, rather than the JA biosynthesis pathway, represent the main routes responsible for airborne SAR in receiver plants.

In this study, we investigated whether volatile compounds emitted from SAR‐expressing lima bean plants affect the resistance phenotype of neighboring tobacco plants. We used a non‐host pathogen Pto (biological inducer) to induce the emission of MIPVs from lima bean (emitter) plants and to consequently provoke interspecific immunity in tobacco (receiver) plants. Our results showed for the first time that bacterial effectors (particularly HopP1) can also modulate phytovolatiles that traverse the air, resulting in SA‐dependent systemic resistance in neighbouring plants. Nonetheless, it remains unknown why pathogen‐challenged plants evolved an airborne siren, and how neighbouring plants perceive MIPVs to militarize themselves against probable pathogen attack without any physical contact. In addition, investigation of the effects of MIPVs induced by co‐inoculation with saprophytic microbes, such as PGPR and pathogens, on inter‐ and conspecific plant–plant interactions will also be interesting. Pathogenic bacteria such as Pto evolved T3SS effectors to disarm the plant defense systems, leading to pathogen susceptibility, for example, in tomato (Y. J. Kim et al., 2002). By contrast, in lima bean, the non‐host pathogen Pto enhances pathogen resistance via specific effectors that induce the production of MIPVs, which in turn trigger plant immunity in the surrounding vegetation. This suggests that a prudent use of the strength of pathogens for particular effector‐driven emission of MIPVs could be employed as one of the promising strategies for the development of a biological agent that can restrain pathogen outbreak in agricultural crops.

In conclusion, we showed that the HopP1 effector of Pto elicits the production of volatile compounds, 1‐octen‐3‐ol, 3‐octanone and 3‐octanol, in the emitter plant, which in turn activates disease resistance in neighbouring plants (Figure 5a,b). Therefore, bacterial effectors play important roles in plant–plant interactions by eliciting MIPVs. The entire mechanism of bacterial type III effector‐dependent airborne immune activation is depicted in Figure 6. Understanding the molecular basis of environmental influences on disease resistance will help us decipher plant–plant interactions mediated by the signalling of volatile compounds in the field and natural ecosystems. Detailed studies dissecting each step in the production of MIPVs in the emitter plant and their mode of action in receiver plants will help further uncover the mechanism underlying interspecific airborne communication.

Figure 6

Proposed mechanism of bacterial type III effector‐dependent airborne immune activation. A lima bean (emitter) plant was challenged with the non‐host pathogen, Pto, which delivered the effector protein HopP1 into the plant cells through the T3SS. The HopP1 protein potentially interacts with plant protein(s) involved in the emission of C8 plant volatiles. Unknown proteins (or enzymes) generate C8 volatiles, including 1‐octen‐3‐ol, which travel to tobacco (receiver) plants through air. The perception of C8 plant volatiles by tobacco plants activates the expression of pathogenesis‐related (PR) genes and subsequently induces systemic acquired resistance (SAR) against Pta through the salicylic acid (SA)‐dependent signalling pathway, thereby suppressing Pta growth. Arrows indicate a multi‐organism interaction cascade that leads to disease resistance in neighbouring plants [Colour figure can be viewed at wileyonlinelibrary.com]

CONFLICT OF INTEREST

The authors declare that the research was conducted in the absence of any commercial or financial relationship that could be construed as a potential conflict of interest.

Appendix S1. Supporting information.

Click here for additional data file.