Colonization of Arabidopsis roots by Pseudomonas ﬂuorescens primes the plant to produce higher levels of ethylene upon pathogen infection

Plants develop an enhanced defensive capacity against a broad spectrum of plant pathogens after colonization of the roots by selected strains of non-pathogenic, ﬂuorescent Pseudomonas spp. In Arabidopsis thaliana , this rhizobacteria-induced systemic resistance (ISR) functions independently of salicylic acid but requires responsiveness to the plant hormones jasmonic acid and ethylene. Leaves of plants of which the roots are colonized by ISR-inducing Pseudomonas ﬂuorescens WCS417r bacteria show an enhanced capacity to convert the ethylene precursor 1-aminocyclopropane-1-carboxylate (ACC) to ethylene. Here we show that this enhanced ACC-converting capacity leads to a potentiated expression of the ethylene-responsive genes PDF1·2 and HEL after treatment of the leaves with 1 m M ACC, and a signiﬁcantly higher level of ethylene emission after challenge inoculation with the bacterial pathogen Pseudomonas syringae pv. tomato DC3000/ avrRpt2 . P. ﬂuorescens WCS374r bacteria that are unable to induce ISR against P. syringae pv. tomato DC3000 in Arabidopsis likewise enhanced the in vivo ACC oxidase acitivity in Col-0 plants. Moreover, the ISR-compromised mutants jar1-1 and npr1-1 also showed a signiﬁcant increase in their ability to convert ACC to ethylene after treatment of the roots with P. ﬂuorescens WCS417r. These results suggest that the induction of an enhanced ACC-converting capacity is a general response of plants to P. ﬂuorescens bacteria and that this response does not contribute to ISR against P. syringae pv. tomato DC3000 in Arabidopsis . Nevertheless, P. ﬂuorescens strains clearly prime the plant to produce more ethylene upon pathogen infection. The increased capacity for ethylene production might contribute to an enhanced defensive capacity against pathogens that are sensitive to ethylene-dependent defense responses. q 2003 Elsevier Ltd. All rights reserved.


Introduction
Selected root-colonizing Pseudomonas spp. strains have been shown to trigger a plant-mediated resistance response in aboveground plant parts. This type of induced resistance is often referred to as rhizobacteria-mediated induced systemic resistance (ISR) [36]. Phenotypically, rhizobacteria-mediated ISR resembles classic pathogeninduced systemic acquired resistance (SAR), in which noninfected parts of locally infected plants become more resistant to further infection [25]. Pseudomonas fluorescens strain WCS417r has been shown to trigger ISR in several plant species, e.g. carnation [37], radish [16], tomato [6], bean [1] and Arabidopsis thaliana [21,22]. Colonization of Arabidopsis roots by WCS417r protects the plant systemically against different types of pathogens, including the bacterial leaf pathogens Pseudomonas syringae pv. tomato and Xanthomonas campestris pv. armoraciae, the fungal root pathogen Fusarium oxysporum f.sp. raphani, the fungal leaf pathogen Alternaria brassicicola and the oomycete leaf pathogen Peronospora parasitica [21,34,40]. The ability to develop ISR in response to selected strains of rhizosphere bacteria has been documented for several plant species [36] and is dependent on the host/rhizobacterium combination. For example, P. fluorescens WCS374r is able to trigger ISR in radish against F. oxysporum [16] but is unable to do so in Arabidopsis [40]. This suggests that a specific recognition between the plant and the ISR-inducing rhizobacterium is required for the induction of ISR.
Both rhizobacteria-mediated ISR and pathogen-induced SAR are effective against a broad spectrum of pathogens, but are regulated by distinct signaling pathways. Pathogeninduced SAR requires salicylic acid (SA) whereas WCS417r-mediated ISR functions independently of SA [7,19,21]. Analysis of the jasmonic acid (JA)-response mutant jar1-1, the SAR-compromised mutant npr1-1, and a range of ethylene-response mutants of Arabidopsis, revealed that components of the JA and the ethyleneresponse are required for triggering ISR and that this induced resistance response, like SAR, depends on the defense regulatory protein NPR1 [13,23,38]. In Arabidopsis, both JA and ethylene activate specific sets of defenserelated genes and, when applied exogenously, they can confer resistance against certain pathogens, such as P. syringae pv. tomato DC3000 [23,39]. To investigate whether ISR is associated with changes in JA/ethyleneresponsive gene expression, Van Wees et al. [39] monitored the expression of a set of well-characterized JA-and/or ethylene-responsive genes (i.e. LOX1, LOX2, VSP, PDF1·2, HEL, CHI-B, and PAL1) in Arabidopsis plants expressing WCS417r-mediated ISR. None of the genes tested were upregulated in induced plants, neither locally in the roots, nor systemically in the leaves. This suggests that the resistance attained was not associated with major increases in the levels of either JA or ethylene. Indeed, analysis of JA levels in, and ethylene production by, leaves treated with WCS417r and untreated plant parts expressing ISR, revealed no changes in the production of these signal molecules [20]. Therefore, it is assumed that the JA and ethylene dependency of ISR is based on enhanced sensitivity to these hormones, rather than on an increase in their production.
Ethylene has repeatedly been implicated in the regulation of primary resistance responses. In many cases, infection by microbial pathogens is associated with enhanced production of this hormone [2] and a concomitant activation of a large set of defense-related genes [27]. In higher plants, ethylene is produced from methionine via S-adenosyl-L-methionine and 1-aminocyclopropane-1-carboxylate (Met ! SAM ! ACC ! ethylene) [10,11,30]. The last two steps of this biosynthetic pathway are catalysed by ACC synthase and ACC oxidase, respectively. Pathogen infections leading to chlorotic or necrotic symptoms cause an increase in ethylene production with ACC synthase and ACC oxidase activity being increased sequentially [28]. Whereas under normal conditions the conversion of SAM to ACC by ACC synthase is the rate-limiting step, during infections ACC accumulates transiently, indicating that ACC oxidase activity restricts ethylene production. Previously, it was demonstrated that the capacity for converting ACC to ethylene was increased systemically in SAR-expressing tobacco and Arabidopsis plants [5,20], providing a greater capacity for producing ethylene after challenge inoculation. Interestingly, a similar systemic response was observed in plants after treatment of the roots with ISR-inducing WCS417r bacteria [20]. Exogenous application of the ethylene precursor ACC has been shown to induce resistance against P. syringae pv. tomato in Arabidopsis [20,23,39]. Therefore, a faster or greater production of ethylene in the initial phase of infection might contribute to enhanced resistance against this pathogen.
To investigate the role of the WCS417r-induced enhanced capacity to convert ACC to ethylene in ISR, we analysed the expression of the ethylene-responsive genes HEL, encoding an hevein-like protein [24], and PDF1·2, encoding a plant defensin with antimicrobial properties [18], after exogenous application of ACC to leaves of plants that were grown in soil with or without WCS417r bacteria. In addition, we monitored ethylene emission after challenge inoculation with the necrosis-inducing avirulent pathogen P. syringae pv. tomato DC3000/avrRpt2. To assess the significance of the enhanced ACC-converting capacity to ISR, we tested the effect of an ISR-inducing and an ISRnoninducing strain of P. fluorescens on this response using wild-type Col-0 plants and the ISR-response mutants jar1-1 and npr1-1.

Growth conditions of rhizobacteria and plants
Non-pathogenic Pseudomonas fluorescens WCS417r was used for induction of ISR, whereas P. fluorescens WCS374r was used as a non-inducing control strain. Both strains were grown on King's medium B (KB) agar Plate [12] for 24 h at 288C. Subsequently, bacterial cells were collected and resuspended in 10 mM MgSO 4 to a final density of 10 9 cfu ml 21 (OD 600 ¼ 1·0), before being mixed through the soil.
Seeds of wild-type Arabidopsis thaliana accession Col-0 and Col-0 mutants jar1-1 [29] and npr1-1 [3] were sown in quartz sand. Two-week-old seedlings were transferred to 60 ml pots containing a sand-potting soil mixture that had been autoclaved twice for 20 min with a 24 h interval. Before transfer of the seedlings, a suspension of P. fluorescens bacteria (10 9 cfu ml 21 ) was mixed through the soil to a final density of 5 £ 10 7 cfu g 21 . Control soil was supplemented with an equal volume of 10 mM MgSO 4 . Plants were cultivated in a growth chamber with a 8 h day (200 m E m 22 s 21 at 248C) and 16 h night (208C) cycle at 70% relative humidity. Plants were watered on alternate days, and once a week supplied with a modified halfstrength Hoagland's nutrient solution, as described [21].

Determination of ACC-converting capacity
To determine the capacity of leaf tissues to convert ACC to ethylene, rosettes of 5-week-old plants were detached from the roots, weighed, and dipped in a solution containing 1 mM ACC and 0·015% (v/v) of the surfactant Silwet L-77 (Van Meeuwen Chemical BV, Weesp, The Netherlands). Control plants were dipped in 0·015% (v/v) Silwet L-77 only. For the determination of the ACC-converting capacity of young and old leaves, rosettes were dissected into two equal parts, thereby separating the oldest leaves from the younger leaves higher up in the rosette. Subsequently, both parts were weighed separately before dipping in 1 mM ACC, 0·015% (v/v) Silwet L-77. After ACC treatment, excess moisture was removed from the leaves with paper towels. Subsequently, each rosette (part) was placed in a gas-tight serum flask that was incubated under climate chamber conditions. At specific time points, 1 ml gas samples were withdrawn through the rubber seal. The concentration of ethylene was determined by GC as described by De Laat and Van Loon [5].

Pathogen inoculation
The avirulent pathogen, Pseudomonas syringae pv. tomato DC3000 with the plasmid pV288 carrying avirulence gene avrRpt2 [14] was used for challenge inoculation. P. syringae pv. tomato DC3000/avrRpt2 bacteria were cultured overnight at 288C in liquid King's medium B, supplemented with 25 mg l 21 kanamycin to select for the plasmid. Subsequently, bacterial cells were collected by centrifugation and resuspended in 10 mM MgSO 4 to a final density of 10 7 cfu ml 21 . Plants were inoculated by pressure infiltrating a suspension of P. syringae pv. tomato DC3000/ avrRpt2 at 10 7 cfu ml 21 into the leaves of 5-week-old plants (12 leaves per plant).

RNA blot analysis
For RNA blot analysis, leaf tissues were collected and frozen in liquid nitrogen. Total RNA was extracted by homogenizing 0·5 g of frozen leaf tissue in 0·5 ml of extraction buffer (0·35 M glycine, 0·048 M NaOH, 0·34 M NaCl, 0·04 M EDTA, 4% (w/v) SDS). The homogenates were extracted with phenol and chloroform, and the RNA was precipitated using LiCl, as described [26]. Fifteen mg of RNA was denatured using glyoxal and dimethyl sulfoxide [26]. Subsequently, samples were electrophoretically separated on 1·5%-agarose gels and blotted onto Hybond-N þ membranes (Amersham,'s-Hertogenbosch, The Netherlands) by capillary transfer. The electrophoresis buffer and blotting buffer consisted of 10 and 25 mM sodium phosphate (pH 7·0), respectively. RNA blots were hybridized with ACO, HEL, and PDF1·2 gene-specific probes, and a probe for the constitutively expressed b-tubulin (TUB) gene as described previously [39]. Probes for HEL, PDF1·2 and TUB were labeled with a-32 P-dCTP by random-primer labeling using templates that were generated by PCR as described [39]. Probes to detect ACO transcripts were derived from an Arabidopsis ACO cDNA clone [9].

Enhanced ACC-converting capacity in P. fluorescens WCS417r-treated Arabidopsis
Non-infected leaves of tobacco plants that are induced to express SAR after predisposal infection with tobacco mosaic virus (TMV) show an enhanced capacity to convert ACC to ethylene. The magnitude of the induced increase in the ACC-converting capacity was shown to decrease with leaf age [5]. A similar enhancement of the in vivo ACC oxidase activity has been observed in Arabidopsis after colonization of the roots by ISR-inducing WCS417r bacteria [20]. To further investigate this we determined the ACC-converting capacity in the young upper parts and the older lower parts of the rosettes of 5-week-old control and WCS417r-treated Col-0 plants. In control plants grown in soil without rhizobacteria, the constitutive level of ethylene production did not significantly differ between young and old leaves ( Fig. 1; 2 /2 ACC). This suggests that the ACC synthase activity, which is the rate-limiting step in ethylene biosynthesis in non-infected plants [4], does not differ between young and old leaves. After exogenous application of a saturating dose of 1 mM ACC, young leaves of control plants showed a 12-and 45-fold increase in ethylene production at 6 and 24 h after ACC application, respectively ( Fig. 1: 2 /þ ACC). In older leaves, this increase ethylene production was significantly lower (4-and 14-fold at 6 and 24 h after ACC application, respectively), indicating that the capacity to convert ACC to ethylene is higher in young leaves than in old leaves.
In Col-0 plants grown in soil with ISR-inducing WCS417r bacteria, the constitutive level of ethylene production in young and old leaves did not differ from that in control plants (data not shown), confirming previous findings showing that constitutive ethylene emission in WCS417r-treated plants is not enhanced [20]. As in control plants, application of 1 mM ACC to the leaves of WCS417rtreated plants resulted in a considerable increase in the production of ethylene in both young and older leaves ( Fig. 1: 417/ þ ACC). Compared to ACC-treated control plants, this increase in ethylene emission was significantly higher in both young and older leaves. At 6 h after ACC application, ethylene emission was 91% higher in young leaves and 132% higher in old leaves of WCS417r-treated plants in comparison to similarly treated leaves of control plants. At 24 h after ACC application, these differences were 64 and 49%, respectively.

ACO gene expression
To investigate whether the enhanced ACC-converting capacity observed in WCS417r-treated plants is based on increased expression of this gene, we analysed ACO transcript levels in young and old leaves of control and WCS417r-treated plants. Fig. 2 shows that ACO mRNA levels in control and WCS417r-treated plants were similar. Moreover, when normalized for equal levels of TUB mRNA, ACO transcript levels did not differ between young and old leaves. This suggests that the higher ACC-converting capacity of young leaves, and the enhanced ACC-converting capacity of P. fluorescenstreated plants is regulated post-translationally.

Ethylene-responsive gene expression
To investigate whether the enhanced ACC-converting capacity in WCS417r-treated plants results in a potentiated expression of ethylene-responsive genes, we analysed the expression of the defense-related genes PDF1·2 and HEL. Fig. 3 shows that both PDF1·2 and HEL mRNA accumulated after treatment of the leaves with ACC. At 6 h after ACC application, PDF1·2 mRNA was detectable in WCS417r-treated plants, but not yet in control plants. At 24 h after ACC application, PDF1·2 transcript levels were clearly higher in WCS417r-treated plants. Similarly, the expression of the HEL gene was considerably higher in rhizobacteria-treated plants at 24 h after ACC application. These results indicate that the enhanced ACC-converting capacity of WCS417r-treated plants leads to an augmented level of ethylene-responsive gene expression in response to treatment with ACC.

Pathogen-induced ethylene production
Upon infection by necrotizing pathogens, ethylene emission rises due to a sequential increase in the activity of ACC synthase and ACC oxidase, respectively [4,28]. The pathogen-induced increase in ACC synthase activity leads to elevated levels of ACC in the initial phases of the infection process. Due to their enhanced ACC-converting  capacity, WCS417r-treated plants are, therefore, likely to produce more ethylene in the initial phases of infection. To test this hypothesis, we measured ethylene production in control and WCS417r-treated plants at different time points after inoculation with the avirulent pathogen Pseudomonas syringae pv. tomato DC3000/avrRpt2. Fig. 4 shows that WCS417r-treated plants, indeed, produced significantly more ethylene at 6 and 24 h after pathogen infection than similarly inoculated control plants. At 6 h after inoculation, WCS417r-treated plants produced 20% more ethylene than did control plants. At 24 h after inoculation this difference was still 15%. These results indicate that the enhanced ACC-converting capacity of WCS417r-treated plants leads to an enhanced production of ethylene after pathogen infection.

ACC-converting capacity in plants treated with ISR-noninducing rhizobacteria
To investigate whether the elevated capacity to produce ethylene after pathogen infection in WCS417r-treated plants is associated with ISR, we analysed the ACC-converting capacity of wild-type Col-0 plants after colonization of the roots with P. fluorescens WCS374r. WCS374r is a wellcharacterized rhizobacterial strain that has been shown to induce ISR in radish [16] but is unable to do so in Arabidopsis against P. syringae pv. tomato DC3000 [40]. In similarity to the effect of WCS417r, treatment of the roots with WCS374r bacteria resulted in an enhanced capacity to convert ACC to ethylene (Fig. 5(a)), and a significant increase in ethylene production in the first 24 h after inoculation with P. syringae pv. tomato DC3000/avrRpt2 ( Fig. 5(b)). Furthermore, like in WCS417r-treated plants, ACC-induced expression of the ethylene-responsive genes PDF1·2 and HEL was augmented in plants treated with WCS374r bacteria (data not shown). Altogether, these results indicate that the ability of P. fluorescens bacteria to induce an enhanced ACC-converting capacity in aboveground plant parts is not related to the ability of these rhizobacteria to induce ISR against P. syringae pv. tomato DC3000 in Arabidopsis.  3.6. ACC-converting capacity in ISR non-expressing mutants jar1-1 and npr1-1 To further investigate the role of the enhanced ACCconverting capacity in ISR, we examined this response in mutant jar1-1 and npr1-1 plants that are both unable to express WCS417r-mediated ISR against P. syringae pv. tomato DC3000 [20,23,38]. Fig. 6 shows that colonization of the roots by WCS417r results in a similar enhancement of the capacity of the leaves to convert ACC to ethylene in Col-0, jar1-1 and npr1-1 plants. At 6 h after ACC treatment, WCS417r-treated Col-0, jar1-1, and npr1-1 plants emitted 40,39, and 25% more ethylene, respectively, than did ACC-treated control plants. At 24 h after ACC treatment, these percentages were 32, 21, and 17%, respectively. Evidently, the enhanced ACC-converting capacity triggered by P. fluorescens rhizobacteria develops independently of ISR against P. syringae pv. tomato DC3000.

Discussion
Arabidopsis plants of which the roots are colonized by ISR-inducing WCS417r bacteria possessed an enhanced capacity to convert ACC into ethylene, confirming previous preliminary observations [20]. As shown, the enhanced in vivo ACC oxidase activity in WCS417r-treated Arabidopsis is most pronounced in young leaf tissues (Fig. 1) and leads to augmented expression of the ethylene-responsive genes PDF1·2 and HEL after treatment of the leaves with ACC (Fig. 2). The conversion of ACC to ethylene is catalysed by the enzyme ACC oxidase. The expression of the ACO gene in the leaves did not change in response to treatment of the roots with WCS417r (Fig. 2), suggesting that the observed increase in the in vivo ACC oxidase activity is regulated post-translationally.
In response to infection by necrosis-inducing pathogens, plants accumulate ACC as a result of enhanced ACC synthase activity [4,28]. Therefore, the enhanced ACCconverting capacity observed in WCS417r-treated plants provides a greater potential for producing ethylene upon pathogen attack. Indeed, WCS417r-treated plants emitted significantly more ethylene after inoculation with P. syringae pv. tomato DC3000/avrRpt2 (Fig. 4). Ethylene has been shown to play a crucial role in ISR signal transduction. This is evidenced by the fact that ethyleneinsensitive mutants are blocked in their ability to express ISR against P. syringae pv. tomato DC3000 [13,23]. Moreover, the Arabidopsis ISR1 locus, which is essential for expression of ISR against several pathogens [33,35], is involved in ethylene signaling as well [32]. Thus, a faster or greater production of ethylene in the initial phase of infection upon challenge inoculation of ISR-expressing plants might contribute to the enhanced resistance attained. Therefore, we attempted to correlate the ability of WCS417r to enhance the ACC-converting capacity in the leaves with its capacity to induce ISR against P. syringae pv. tomato DC3000. Surprisingly, wild-type Col-0 plants and Arabidopsis mutants jar1-1 and npr1-1, that are both compromised in their ability to express ISR against P. syringae pv. tomato DC3000 [20,23,38], were equally able to enhance their ACC-converting capacity in response to colonization of the roots by WCS417r (Fig. 6). Moreover, strain WCS374r, which is unable to induce ISR against P. syringae pv. tomato DC3000 in Arabidopsis [40], induced an enhanced ACC-converting capacity in the leaves as well, resulting in an increase in ethylene production after challenge inoculation with P. syringae pv. tomato DC3000/ avrRpt2 that was comparable to that observed in WCS417rtreated plants (Fig. 5). Thus, it can be concluded that the enhanced ACC-converting capacity observed in WCS417rand WCS374r-treated plants is a general response of plants to these P. fluorescens bacteria and that this response does not contribute to ISR against P. syringae pv. tomato DC3000 in Arabidopsis.
Our conclusion that the WCS417r-induced enhancement of the ACC-converting capacity does not contribute to the level of ISR against P. syringae pv. tomato DC3000 is strengthened by another observation in this study. In the absence of WCS417r bacteria, 5-week-old wild-type Col-0 plants showed a significantly higher ACC-converting capacity in the young leaves as compared to the older ones (Fig. 1). Assuming that the resulting elevated levels of ethylene emission in response to P. syringae pv. tomato DC3000 infection would contribute to resistance against this pathogen, one would expect that young leaves are more resistant to P. syringae pv. tomato DC3000 than older leaves. However, the opposite is true, because recently Kus et al. [15] demonstrated that resistance to P. syringae pathogens increases with leaf age. Notwithstanding the fact that the enhanced ACCconverting capacity does not contribute ISR against P. syringae pv. tomato DC3000, our results clearly show that P. fluorescens strains prime the plant to produce more ethylene upon infection and to express ethylene-dependent defense-related genes to a higher level upon stimulation with saturating doses of ACC. Therefore, we can assume that this trait can contribute to enhanced resistance against pathogens that are resisted particularly through ethylenedependent defense responses. In Arabidopsis, ethylene has been implicated to function as an important regulator of resistance against necrotrophic pathogens such as Botrytis cinerea [31], Erwinia carotovora [17], and various Pythium spp. [8]. Whether the P. fluorescens-mediated enhanced ACC-converting capacity contributes to resistance against these pathogens needs to be investigated.