Different signaling pathways of induced resistance by rhizobacteria in Arabidopsis thaliana against two pathovars of Pseudomonas syringae

Authors

  • Choong-Min Ryu,

    Corresponding author
    1. Department of Entomology and Plant Pathology, 209 Life Sciences Building, Auburn University, Auburn, AL 36849 USA.
    2. Present address: Plant Biology Division, The Samuel Roberts Noble Foundation, 2510 Sam Noble Parkway, Ardmore, OK 73402 USA
      Author for correspondence: Choong-Min RyuTel: +1 580 224 6742Fax: +1 580 224 6692Email: cryu@noble.org
    Search for more papers by this author
  • Chia-Hui Hu,

    1. Department of Entomology and Plant Pathology, 209 Life Sciences Building, Auburn University, Auburn, AL 36849 USA.
    Search for more papers by this author
  • M. S. Reddy,

    1. Department of Entomology and Plant Pathology, 209 Life Sciences Building, Auburn University, Auburn, AL 36849 USA.
    Search for more papers by this author
  • Joseph W. Kloepper

    1. Department of Entomology and Plant Pathology, 209 Life Sciences Building, Auburn University, Auburn, AL 36849 USA.
    Search for more papers by this author

Author for correspondence: Choong-Min RyuTel: +1 580 224 6742Fax: +1 580 224 6692Email: cryu@noble.org

Summary

  • • The mechanisms by which plant growth-promoting rhizobacteria (PGPR) mediate induced systemic resistance are currently being intensively investigated from the viewpoint of signal transduction pathways within plants.
  • • Here, we determined whether our well-characterized PGPR strains, which have demonstrated induced resistance on various plants, also elicit induced resistance in Arabidopsis thaliana. Nine different PGPR strains were evaluated for their capacity to cause induced resistance on Arabidopsis against two pathovars of Pseudomonas syringae. Six strains significantly reduced severity of P. syringae pv. tomato, whereas seven strains reduced severity of P. syringae pv. maculicola.
  • • From the initial screenings, four strains (90-166, SE34, 89B61 and T4) were selected because of their consistent induced resistance capacity. Elicitation of induced resistance with these strains depended on how disease severity was measured. Three strains (90-166, 89B61 and T4) induced resistance in NahG plants (SA-deficient), indicating a salicylic acid-independent pathway, which agrees with the previously reported pathway for induced resistance by PGPR. However, differences from the reported pathway were noted with strain 89B61, which did not require jasmonic acid or ethylene signaling pathways for induced resistance, and with strain T4, which induced resistance in npr1 plants.
  • • These results indicate that strains 89B61 and T4 induce resistance via a new pathway or possibly a variation of the previously reported pathway. This information will broaden our understanding of ways in which microorganisms can signal physiological changes in plants.

Introduction

Plants have evolved numerous mechanisms to defend themselves against microbial pathogens. Some of these defense mechanisms are constitutive, such as the physical barriers of the cell wall, while others are induced (Agrios, 1997). Induced disease resistance occurs when a plant exhibits an increased level of resistance to infection by a pathogen after prior treatment with an inducing agent. Some selected strains of plant growth-promoting rhizobacteria (PGPR) have been found to activate plant defense via induced systemic resistance (ISR) (Kloepper et al., 1992; van Loon et al., 1998). The process of active resistance in ISR is dependent on activation of the host plant's physical or chemical barriers. Induced systemic resistance develops systemically following colonization of plant roots by PGPR (Wei et al., 1991). By contrast to PGPR, incompatible pathogens trigger systemic acquired resistance (SAR) following the hypersensitive response (HR), which is a plant defense mechanism that induces rapid, localized cell death at the infection site of pathogens, thereby interfering with disease progress (Heath, 2000).

Salicylic acid (SA) is one of the key chemical signals produced in response to pathogen attack on resistant plants and is required for the induction of SAR (Dempsey et al., 1999). Production of SA and induction of SAR are most often exhibited following the HR. Activation of the HR is governed by resistance genes encoding receptors that recognize specific pathogens (Staskawicz et al., 1995). The subsequent induction of SAR results from a complex signal transduction process (Pickett & Poppy, 2001) and leads to accumulation of pathogenesis-related (PR)-proteins. Recently, an SA analogue, benzo(1,2,3)thiadiazole-7-carbothioic acid S-methyl ester (BTH), was commercialized by Syngenta under the name of Bion in Europe and Actigard in USA (Tally et al., 1999). We used this chemical as a positive control in our experiment.

Several approaches to defining the signal pathway for ISR have been undertaken. Induced systemic resistance mediated by Pseudomonas fluorescens WCS417 in Arabidopsis and by Serratia marcescens 90–166 in tobacco was shown to be independent of SA accumulation (Pieterse et al., 1996; Press et al., 1997). By contrast, Pseudomonas aeruginosa 7NSK2 elicited ISR against Tobacco mosaic virus in tobacco and Botrytis cinerea on tomato via an SA-dependent pathway (De Meyer & Hofte, 1997, 1999). However, induced resistance by the same strain against Pseudomonas syringae on Arabidopsis was SA-independent (Ran, 2002). Using mutant lines of Arabidopsis and strain WCS417r, van Loon et al. (1998) and Pieterse et al. (2002) proposed a model pathway for signal transduction in PGPR-mediated ISR. In the proposed pathway, ISR caused by PGPR is dependent on jasmonic acid (JA), ethylene, and the regulatory gene NPR1, while it is independent of SA and does not result in accumulation of PR-proteins. The studies of induced resistance related signaling pathways have used the following signaling mutants of Arabidopsis: jar1 or fad3-2 fad7-2 fad8 for jasmonic acid; ein2 or etr1 for ethylene; and npr1 for the regulatory gene NPR1 (van Loon et al., 1998; Vijayan et al., 1998; Kus et al., 2002; Pieterse et al., 2002).

The objectives of this study were (1) to determine whether PGPR strains that have been reported to induce resistance against several plant pathogens on cucumber, tomato, and tobacco in the greenhouse and field protect A. thaliana against P. syringae, (2) determine if induced systemic protection by PGPR depends on the pathogens used to challenge plants and (3) determine if signal pathways of plants treated with our PGPR are the same as the model proposed by van Loon et al. (1998) and Pieterse et al. (2002) by using plant signaling defective mutants such as NahG for SA, fad3-2 fad7-2 fad8 for jasmonic acid, ein2 for ethylene, and npr1 for the regulatory gene NPR1.

Materials and Methods

PGPR strains and inoculum preparation

Nine different PGPR strains were used: S. marcescens 90–166, Bacillus pumilus SE34, P. fluorescens 89B61, Bacillus pasteurii C9, Paenibacillus polymyxa E681, Bacillus subtilis GB03, Bacillus amyloliquefaciens IN937a, Enterobacter cloacae JM-22, and Bacillus pumilus T4. These strains had previously induced systemic protection in tobacco, pepper, cucumber and tomato against several diseases (Wei et al., 1991, 1996; Kloepper, 1996; Raupach et al., 1996; Zehnder et al., 1999; Yan et al., 2002; Zhang et al., 2002). Pathogens used were P. syringae pv. tomato DC3000 and P. syringae pv. maculicola ES4326 (kindly provided by B. J. Staskawicz, University of California, Berkeley, CA, USA) (Kus et al., 2002).

Before use, the strains of PGPR and pathogens were stored at −80°C in tryptic soy broth (TSB) amended with 20% glycerol. The strains were removed from ultra-cold storage, streaked onto tryptic soy agar (TSA), and incubated at 28°C for 24 h to check for purity. Single colonies were transferred to TSA and incubated for 2 d. Both pathovars of P. syringae were grown on Pseudomonas Agar F (Difco, St Louis, MO, USA). For experimental use, fully grown bacteria were scraped off plates and resuspened into sterilized distilled water (SDW). The bacterial suspensions were adjusted to109 colony forming-units (cfu) ml−1 based on optical density.

Arabidopsis lines and growth conditions

Transgenic NahG (SA deficient) and mutant npr1 (nonexpression of PR proteins) Arabidopsis were obtained from Dr Xinnian Dong, Duke University, Durham, NC, USA (Cao et al., 1994). The mutant line fad3-2 fad7-2 fad8 (jasmonic acid deficient) was provided by Dr John Browse, Washington State University, Pullman, WA, USA (Vijayan et al., 1998). Mutant ein2 (ethylene insensitive) was obtained from Dr Joseph R. Ecker, University of Pennsylvania, Philadelphia, PA, USA (Alonso et al., 1999). All mutant and transgenic lines were derived from the parental A. thaliana ecotype Columbia (Col-0), which was obtained from the Ohio State University Stock Center, Columbus, OH, USA. The Arabidopsis seeds were surface-sterilized with 6% sodium hypochlorite (100% commercial laundry bleach) containing 0.1% Triton X-100, washed four times with SDW, and maintained at 4°C for 2 d to enhance germination. The seeds were then suspended in 0.4% low-melting-point agarose on soil-less media (Speedling, Sun City, FL, USA), hereinafter referred to as potting media. Plants were grown at 23 ± 3°C under a 12-h natural light regime in a greenhouse.

Initial screening of induced resistance of A. thaliana against P. syringae pv. tomato and P. syringae pv. maculicola by PGPR

Two weeks after seeding, one seedling of Col-0 was transplanted into a 10-cm square pot. Five millliters of PGPR suspension was applied to the base of plants in the potting media at 108−109 cfu g−1 soil at the time of transplanting. An additional PGPR treatment (booster) was applied 1 wk after transplanting. A stock solution of benzo(1,2,3)thiadiazole-7-carbothioic acid S-methyl ester (BTH) (Syngenta Research, Triangle Park, NC, USA) at 0.33 mm was freshly prepared in SDW for each experiment. The BTH, a chemical inducer, was used as a positive control. Control treatments consisted of SDW. One week after booster treatment, freshly prepared suspensions of P. syringae pv. tomato and P. syringae pv. maculicola suspensions in SDW containing 200 µl l−1 Tween-20 (Sigma, St Louis, MO, USA) were sprayed onto the leaves. Inoculated plants were placed in a dew chamber (100% humidity) under darkness for 2 d at 27°C and were then transferred to a greenhouse. Seven days after pathogen challenge, disease severity was measured by two methods. First, the ‘percentage disease’ was measured by recording the per cent of total plant leaf surface showing symptoms for each plant from 0 = no symptoms to 100 = most severe with necrotic symptoms. Second, the number of symptomatic leaves per plant was counted. This experiment was designed as a randomized complete block (RCB) with 12 replications and one plant per replication. The experiment was conducted three times.

Spatial separation of PGPR and pathogens

To confirm spatial separation of PGPR and pathogens, one antibiotic-resistant mutant of each strain was used. Spontaneous rifampicin-resistant mutants were screened by growing colonies on TSA amended with 100 µg ml−1 rifampicin (rif-TSA). Isolated colonies with similar growth rates as the wild-type strains were stabilized by growing on rif-TSA for several generations. The rif-mutants of each strain were applied to Arabidopsis seedlings in the potting media as described previously. Four weeks after treatment with rif-resistant PGPR strains, three leaves on each plant were removed and ground with a sterile mortar and a pestle. The dilution plating method was used to isolate rif-resistant colonies on TSA amended with 100 µg ml−1 rif for selection of rif-resistant and 100 µg ml−1 cycloheximide for inhibition of fungal growth. The cfu were counted 48 h after incubation at 27°C.

Induced resistance on NahG transgenic plants by PGPR

Among PGPR strains in the initial screening, four strains – 90-166, SE34, 89B61, and T4 – were selected for further study based on consistent elicitation of ISR. To determine the role of SA in ISR, protection against P. syringae pv. tomato and P. syringae pv. maculicola was assessed on NahG plants. This experiment was designed as a randomized complete block (RCB) with 12 replications and one plant per replication. The experiment was repeated three times.

Induced resistance on npr1, fad3-2 fad7-2 fad8 and ein2 plants by PGPR

To test if PGPR elicit ISR via signaling pathways that are different from the model proposed by van Loon et al. (1998) and Pieterse et al. (2002), protection was assessed on npr1, fad3-2 fad7-2 fad8, and ein2 against P. syringae pv. tomato and P. syringae pv. maculicola. The effect of PGPR on growth of Arabidopsis challenged with the two pathogens was also assessed by measuring foliar fresh weight 3 wk after PGPR inoculation. This experiment was designed as a randomized complete block (RCB) with 12 replications and one plant per replication. The experiment was repeated three times.

Data analysis

Data were subjected to analysis of variance using JMP software (SAS Institute Inc., Cary, NC, USA). Significance of PGPR treatment effects was determined by the magnitude of the F-value at P = 0.05. When a significant F-value was obtained for treatments, separation of means was accomplished using Fisher's protected least significant difference (LSD) at P = 0.05. Results of repeated trials of each experiment outlined above were similar. Hence, one representative trial of each experiment is reported in the Results section.

Results

Initial screening of induced resistance of A. thaliana against P. syringae pv. tomato and P. syringae pv. maculicola by PGPR

Disease severity was decreased by six of the nine PGPR strains against P. syringae pv. tomato and by seven strains against P. syringae pv. maculicola in Arabidopsis Col-0 (Table 1). Six strains (90-166, SE34, 89B61, C9, JM22 and T4) elicited systemic protection against both pathovars (Table 1). Strains GB03 and IN937a did not protect plants against both pathovars, although the two PGPR strains have been reported to induce resistance in cucumber and tomato (Kloepper et al., 1996; Raupach et al., 1996; Zender et al., 1999).

Table 1.  Induced resistance of Arabidopsis thaliana against Pseudomonas syringae pv. tomato and P. syringae pv. maculicola by plant growth-promoting rhizobacteria (PGPR)
TreatmentaDisease severity measured as percentage disease (%)b
PsmcPstc
  • Numbers represent mean of 12 replications per treatment, one seedling per replication. aPGPR were inoculated in the potting media, and a 2-wk-old seedling of Col-0 and NahG transgenic A. thaliana was transplanted into the media. two milliliters of 0.33 mm benzo(1,2,3)thiadiazole-7-carbothioic acid S-methyl ester (BTH) solution was applied by drenching. bPercentage disease was measured by recording the per cent of total plant leaf surface showing symptoms for each plant.

  • c

    Pst, P. syringae pv. tomato DC3000; Psm, P. syringae pv. maculicola ES4326. Pst and Psm were sprayed onto leaves until run-off, 1 wk after PGPR treatment. Different letters indicate significant differences among means using Fisher's protected LSD test at P = 0.05.

90-16652f40bc
SE3463cde41bc
89B6150fg45b
C952ef40b
E68150fg71a
GB0370bcd73a
IN937a75abc71a
JM2239gh26c
T459def36bc
BTH28h27c
Control78ab65a

Spatial separation of PGPR and pathogens

To exclude direct contact between PGPR strains and pathogen, we confirmed that none of the PGPR strains were detected on the rosette leaves where inoculated with pathogen. No rif-mutants of any strain were detected on Arabidopsis leaves (data not shown).

Induced resistance on NahG transgenic plant by PGPR

Among PGPR strains in the initial screening, four strains (90-166, SE34, 89B61 and T4) were selected for further studies because of their consistent induced resistance capacity (data not shown). In the NahG transgenic line, all four strains significantly reduced disease severity of P. syringae pv. maculicola and three strains significantly reduced disease severity against P. syringae pv. tomato, as measured by the percentage disease scale, compared with the control (Table 2). The BTH treatment protected both Col-0 and NahG plants.

Table 2.  Induced resistance on NahG transgenic and Col-0 Arabidopsis thaliana against Pseudomonas syringae pv. tomato and P. syringae pv. maculicola by plant growth-promoting rhizobacteria (PGPR)
TreatmentsaDisease severity measured as percentage disease (%)b
PsmcPstc
Col-0NahGCol-0NahG
  1. Numbers represent mean of 12 replications per treatment, one seedling per replication. aPGPR were inoculated in the soilless mixture at 3 wk-old seedling of Col-0 and NahG transgenic A. thaliana transplanted in the soilless media. 2 mL of 0.33 mm benzo(1,2,3)thiadiazole-7-carbothioic acid S-methyl ester (BTH) solution was applied by drenching. bPercentage disease was measured by recording the percent of total plant leaf surface showing symptoms for each plant. cPst, P. syringae pv. tomato DC3000; Psm, P. syringae pv. maculicola ES4326. Pst and Psm were sprayed onto leaves until run-off, 1 wk after PGPR treatment. Different letters indicate significant differences among means using Fisher's protected LSD test at P = 0.05.

90-16613c40c56b49b
SE3413c60b46b64ab
89B6110c21d55b53b
T426b43c55b58b
BTH13c14c15c23c
Control40a81a83a83a

Induced resistance on npr1, fad3-2 fad7-2 fad8, and ein2 plants by PGPR

All four strains consistently elicited ISR on Col-0 with both methods of assessing disease severity (percentage disease and number of symptomatic leaves) (Figs 1 and 2). To determine signaling pathway of ISR elicited by the four selected PGPR strains, ISR capacity of these PGPR strains was evaluated in the three signaling Arabidopsis mutants, which are npr1 for a regulatory gene NPR1, fad3-2 fad7-2 fad8 for jasmonic acid signaling and ein2 for ethylene signaling. Strain SE34 did not elicit ISR in npr1 or ein2 plants as determined by both methods of assessing disease severity. Strain T4 caused reduction in disease by both pathovars in npr1 and the wild-type Col-0. Protection by the other three strains, 90-166, SE34 and 89B61 varied, depending on mutant lines, method of assessing disease severity and P. syringae pathovars. Plants treated with strain SE34 showed reduction of both percentage disease and number of symptomatic leaves only on ein2 against P. syringae pv. tomato (Fig. 2). Strain 90-166 reduced both percentage disease and number of symptomatic leaves per plant against both P. syringae pathovars on ein2 plants (Figs 1 and 2) but only percentage disease against P. syringae pv. tomato on the fad3-2 fad7-2 fad8 plants (Fig. 2). Strain 89B61 caused a reduction of both percentage disease and number of symptomatic leaves per plant in ein2 and fad3-2 fad7-2 fad8 plants with both pathovars (Figs 1 and 2). The BTH treatment reduced both percentage disease and number of symptomatic leaves per plant with both pathovars in ein2 and fad3-2 fad7-2 fad8 plants, but in npr1 plants it only reduced the number of symptomatic leaves with P. syringae pv. maculicola (Fig. 1) (Table 3).

Figure 1.

Induced systemic resistance of Arabidopsis thaliana against Pseudomonas syringae pv. maculicola by selected plant growth-promoting rhizobacteria (PGPR) strains 1 wk after challenge. (a) percentage disease where 0 = no symptoms; 100 = most severe necrotic symptoms; (b) number of symptomatic leaves per plant. Numbers represent means of 12 replications per treatment, one seedling per replication. The PGPR were inoculated in the potting media containing 3-wk-old seedlings of Col-0, npr1-1, fad3-2 fad7-2 fad8 and NahG transgenic Arabidopsis lines. Two milliliters of 0.33 mm benzo(1,2,3)thiadiazole-7-carbothioic acid S-methyl ester (BTH) solution was applied by drenching. Disease severity was measured by percentage disease and number of symptomatic leaves per plant. Pseudomonas syringae pv. maculicola ES4326 was sprayed onto leaves until run-off, 1 wk after PGPR treatment. Different letters indicate significant differences using Fisher's protected LSD test at P = 0.05. The experiment was conducted three times. Results of repeated trials of each experiment outlined were similar. Hence, one representative trial of each experiment is reported here.

Figure 2.

Induced systemic resistance of Arabidopsis thaliana against Pseudomonas syringae pv. tomato by selected plant growth-promoting rhizobacteria (PGPR) strains 1 wk after challenge. (a) percentage disease where 0 = no symptoms; 100 = most severe necrotic symptoms; (b) number of symptomatic leaves per plant. Numbers represent mean of 12 replications per treatment, one seedling per replication. The PGPR were inoculated in the potting media containing 3-wk-old seedlings of Col-0, npr1-1, fad3-2 fad7-2 fad8, and NahG transgenic Arabidopsis lines. Two milliliters of 0.33 mm benzo(1,2,3)thiadiazole-7-carbothioic acid S-methyl ester (BTH) solution was applied by drenching. Disease severity was measured by percentage disease and number of symptomatic leaves per plant. Pseudomonas syringae pv. tomato DC3000 was sprayed onto leaves until run-off, 1 wk after PGPR treatment. Different letters indicate significant differences using Fisher's protected LSD test at P = 0.05. This experiment was conducted three times. Results of repeated trials of each experiment outlined were similar. Hence, one representative trial of each experiment is reported here.

Table 3.  Summary of induced resistance elicited by several plant growth-promoting rhizobacteria (PGPR) strains in Arabidopsis thaliana
PGPR StrainsSalicylic acid PsmaPstaNPR1 PsmPstJasmonic acid PsmPstEthylene PsmPst
  • a

    Pst, Pseudomonas syringae pv. tomato DC3000; Psm, P. syringae pv. maculicola ES4326.

  • b

    – , independent of this signal based on both disease severity measurements; + +, dependent on this signal based on both disease severity measurements; ±, either one independent of this signal from two disease severity measurements.

90-166b+ ++ ++ +±
SE34+ ++ ++ ++ +±+ +
89B61+ ++ +
T4±+ +±
BTH+ ++ +±

Discussion

Research into how PGPR induce systemic disease resistance provides an understanding of how microorganisms signal physiological changes in plants. Novel signaling mechanisms are revealed by finding differences between reported models of signal transduction and plant responses to pathogens during induced resistance elicited by different microorganisms. Collectively, our results suggest potential novel signal mechanisms of ISR because our results differ from past studies and current models of induced resistance by PGPR.

The results reported here demonstrate that the level of systemic protection elicited in Arabidopsis by PGPR was dependent on the PGPR strain and the challenge pathogen. Six of nine PGPR strains reduced severity of P. syringae pv. tomato, while seven strains reduced severity of P. syringae pv. maculicola (Table 1). Although only one strain, E681, differed in ISR capacity with the two pathovars, this was still unexpected, because ISR is considered to be a broad-spectrum resistance against many pathogens. Our finding that PGPR strain E681 elicits ISR against one pathovar but not against another indicates some specificity in the defensive reactions elicited during ISR for this strain.

Expression profiling using microarray has recently suggested that the response of Arabidopsis to P. syringae pv. maculicola and to P. syringae pv. tomato is mostly similar (Tao et al., 2003). This result agrees with our data. However, there are some exceptions with strain SE34 in NahG and ein2 plants and T4 and BTH treatments in fad3-2 fad7-2 fad8 plants (Table 3). Surprisingly, assessing ISR also depended on the method used to measure disease severity, which has not been reported previously. Previous results showed that PGPR strain P. fluorescens WCS417r elicited ISR in Arabidopsis against the bacterial leaf pathogens P. syringae pv. tomato (Pieterse et al., 1996; van Wees et al., 1997) and Xanthomonas campestris (axonopodis) pv. armoraciae (Ton et al., 2002). These results were based on measuring disease severity as the proportion of leaves with symptoms. Using basically this same measure (number of leaves per plant showing symptoms) in our study, we concluded that ISR resulted in fewer cases by PGPR compared with measuring disease severity with a 0–100% scale. This finding suggests that conclusive evidence of repeatable systemic protection by PGPR might be more accurate when based on more than one method of assessing disease severity.

The role of defense signaling molecules such as SA, JA and ethylene in ISR has been studied with transgenic or insensitive mutant plants (Pieterse et al., 1996; van Wees et al., 1997; Yan et al., 2002; Zhang et al., 2002). NahG plants carry a bacterial nahG gene encoding salicylate hydroxylase that degrades SA to catechol, an inactive form that does not elicit SAR but is involved in nonhost resistance (Dempsey et al., 1999; van Wees & Glazebrook 2003). NahG plants do not totally block salicylic acid accumulation but are enough to interfere SAR and SA-dependent induction of SAR-related genes (Dempsey et al., 1999). However, the precise pathway of SA biosynthesis and signaling is not yet clearly established (Cameron, 2000). Sid1 and Sid2 reported by Wildermuth et al. (2001) are genes that are similar to SA biosynthesis pathways from bacterial origin. Wildermuth et al. (2001) suggested, therefore, that the pathway is located in the plastid. These sid mutants have been little affect on phenylpronanoid pathway (directly associate with SA biosynthasis) than the NahG plants (Cameron, 2000). However, many scientists still use NahG transgenic plants for determining SA signaling pathways. In our studies, experiments with NahG plants showed that strains 89B61 and SE34 induced resistance in tomato against Phytophthora infestans, and strains 90-166, 89B61 and SE34 induced resistance in tobacco against Peronospora tabacina, indicating that these strains do not require SA to protect plants (Yan et al., 2002; Zhang et al., 2002). In this study, we confirmed these results by finding that strains 90-166, 89B61 and T4 systemically protected NahG Arabidopsis against P. syringae pv. maculicola and P. syringae pv. tomato (Table 2). These results are in agreement with the signal pathway model proposed by van Loon et al. (1998) and Pieterse et al. (2002). Results with strain SE34 were different. This strain did not induce resistance in NahG Arabidopsis against P. syringae. pv. tomato and it induced resistance against P. syringae. pv. maculicola at a reduced level compared with the other three PGPR strains. These results indicate that ISR elicited by SE34 is somehow dependent on SA-signaling pathways (Table 3). In our study, ISR elicited by strains 90-166, SE34 and 89B61 required JA and ethylene signaling pathways, based on lack of protection of JA- or ethylene-insensitive tomato lines (Yan et al., 2002). These results are also in agreement with the model proposed by van Loon et al. (1998) and Pieterse et al. (2002). However, our results showing that strain 89B61 protected JA-insensitive fad3-2 fad7-2 fad8 and ethylene-insensitive ein2 mutants (Figs 1 and 2) and that T4 protected npr1 are at variance with the model (Table 3). All previous reports of ISR elicited by PGPR were dependent on NPR1 (van Loon et al., 1998). Our results with strain 89B61 and T4 suggest that they induced resistance via a new pathway or possibly a variation on previously reported pathways.

Acknowledgements

We thank and acknowledge William Fowler for help with preparation of the manuscript and with editing the final version. We also thank and acknowledge X. Dong, J. R. Ecker, John Browse and the Ohio State University Stock Center for providing Arabidopsis seeds and B. J. Staskawicz, for kindly providing bacterial pathogens.

Ancillary