MYB72 Is Required in Early Signaling Steps of Rhizobacteria-Induced Systemic Resistance in Arabidopsis

Colonization of Arabidopsis roots by non-pathogenic Pseudomonas fluorescens WCS417r bacteria triggers a jasmonate/ethylene-dependent induced systemic resistance (ISR) that is effective against a broad range of pathogens. Microarray analysis revealed that the R2R3-MYB-like transcription factor gene MYB72 is specifically activated in the roots upon colonization by WCS417r. Here we show that T-DNA knockout mutants myb72-1 and myb72-2 are incapable of mounting ISR against the pathogens Pseudomonas syringae pv. tomato , Hyaloperonospora parasitica , Alternaria brassicicola and Botrytis cinerea , indicating that MYB72 is essential to establish broad-spectrum ISR. Overexpression of MYB72 did not result in enhanced resistance against any of the pathogens tested, demonstrating that MYB72 is not sufficient for the expression of ISR. Yeast two-hybrid analysis revealed that MYB72 physically interacts in vitro with the ETHYLENE INSENSITIVE3 (EIN3)-LIKE3 transcription factor EIL3, linking MYB72 function to the ethylene response pathway. However, WCS417r activated MYB72 in ISR-deficient, ethylene-insensitive ein2-1 plants. Moreover, exogenous application of the ethylene precursor 1-aminocyclopropane-1-carboxylate (ACC) induced wild-type levels of resistance in myb72-1 , suggesting that MYB72 acts upstream of ethylene in the ISR pathway. Collectively, this study identified the transcriptional regulator MYB72 as a novel ISR signaling component that is required in the roots during early signaling steps of rhizobacteria-mediated ISR. a level of auto activation of reporter in


INTRODUCTION
The soil environment that is influenced by plant roots, the rhizosphere, is a nutrient-rich habitat providing niches for numerous micro-organisms. Amongst these, many fungi and bacteria with properties beneficial to plants are present (Marx 2004;Pozo et al. 2004). Some plant-beneficial bacteria, e.g. Bacillus and fluorescent Pseudomonas species (Kloepper et al. 2004;Weller 2007), have been reported to protect plants against pathogenic micro-organisms through different mechanisms, such as competition for nutrients, secretion of antibiotics and lytic enzymes, and stimulation of the plant's defensive capacity (Bakker et al. 2007). The latter phenomenon is commonly referred to as induced systemic resistance (ISR, Van Loon et al. 1998). ISR has been demonstrated in many plant species, e.g. bean, carnation, cucumber, radish, tobacco, tomato and the model plant Arabidopsis thaliana, and is effective against a broad spectrum of plant pathogens, including fungi, bacteria, viruses and even insect herbivores (Van Loon et al. 1998).
The ability of plants to develop ISR in response to root colonization by Pseudomonas bacteria depends on the host -rhizobacterium combination (Van Loon et al. 1998;Pieterse et al. 2002). The non-pathogenic, rhizobacterial strain Pseudomonas fluorescens WCS417r has been shown to trigger ISR in several plant species, and has served as a model strain to study ISR in Arabidopsis (Pieterse et al. 2002). Colonization of Arabidopsis roots by WCS417r triggers ISR against the bacterial leaf pathogens Xanthomonas campestris pv. armoraciae and Pseudomonas syringae pv. tomato DC3000 (Pst DC3000), the fungal leaf pathogen Alternaria brassicicola, the oomycete leaf pathogen Hyaloperonospora parasitica and the fungal root pathogen Fusarium oxysporum f.sp. raphani (Pieterse et al. 1996;Van Wees et al. 1997;Ton et al. 2002b). Protection against these pathogens is characterized by a reduction in disease severity as well as an inhibition of pathogen growth.
Phenotypically, ISR resembles systemic acquired resistance (SAR) that develops upon primary infection with a necrotizing pathogen (reviewed in Durrant and Dong 2004). Although rhizobacteriamediated ISR and pathogen-induced SAR are both effective against a broad spectrum of pathogens, their signal transduction pathways are distinct. The onset of SAR is accompanied by local and systemic increases in endogenous levels of salicylic acid (SA) and the transcriptional reprogramming of a large set of genes, including genes encoding pathogenesis-related (PR) proteins (Van Loon et al. 2006). Some PR-proteins possess in vitro anti-microbial activity and are thought to contribute to the enhanced resistance state of SAR. Transduction of the SA signal requires functional NPR1, a regulatory protein that was identified in Arabidopsis through genetic screens for mutants impaired in their defense response to SA or its functional analogs (Dong 2004). Plants that carry a mutation in the NPR1 gene accumulate normal or even higher levels of SA after pathogen infection, but are impaired in their ability to transcriptionally activate PR genes and to mount a SAR response. Although some rhizobacterial strains can activate the SA-dependent SAR pathway (De Meyer and Höfte 1997), the large majority of the reported resistance-inducing fluorescent Pseudomonas spp. strains have been shown to trigger ISR in a SA-independent manner (Van Loon and Bakker 2005). WCS417r-mediated ISR functions independently of SA as well, as demonstrated by observations that Arabidopsis genotypes impaired in SA accumulation or biosynthesis (i.e. NahG, eds5-1, sid2-1) were still able to develop wild-type levels of ISR upon colonization of the roots by WCS417r (Pieterse et al. 1996;2002;Ton et al. 2002a). Analysis of the jasmonic acid (JA)-response mutants jar1-1 and coi1-1, a range of ethylene (ET)-response mutants, and the SAR-compromised mutant npr1-1, revealed that components of the JA-and the ET-response are required for triggering ISR and that this induced resistance response, like SAR, requires NPR1 (Pieterse et al. 1998;Knoester et al. 1999;Van Wees et al. 2000, M.J. Pozo and C.M.J. Pieterse, unpublished results). However, the ISR and the SAR signaling pathways diverge downstream of NPR1 because, unlike SAR, ISR is not marked by the transcriptional activation of PR genes (Pieterse et al. 1996;Van Wees et al. 1997;Van Wees et al. 1999).
In order to identify genes that mark the onset of ISR, the transcriptome of Arabidopsis was surveyed in roots and leaves upon colonization of the roots by ISR-inducing WCS417r bacteria (Verhagen et al. 2004). Systemically in the leaves, no consistent changes in gene expression were observed in response to effective colonization of the roots by WCS417r, indicating that, in contrast to SAR, the onset of WCS417r-mediated ISR in the leaves is not associated with a major reprogramming of the transcriptome. However, after challenge inoculation of the induced plants with Pst DC3000, 81 genes showed a potentiated expression in the leaves, suggesting that these genes were primed to respond faster and/or more strongly upon pathogen attack. The majority of the primed genes appeared to be regulated by JA and/or ET signaling. Priming of pathogen-induced genes allows the plant to react more effectively to a subsequent invader, which might explain the broadspectrum effectiveness of rhizobacteria-mediated ISR (Conrath et al. 2002;Conrath et al. 2006). In contrast to constitutive activation of defense responses, priming does not require major metabolic changes when no pathogens are present. Therefore, it forms a low-cost defense strategy whilst acting against a broad spectrum of attackers (Van Hulten et al. 2006;Pieterse and Dicke 2007).
Whereas in the leaves no changes in gene expression were evident before challenge inoculation, roots responded to colonization by ISR-inducing WCS417r bacteria with significant changes in the expression of 97 genes (Verhagen et al. 2004). To investigate the biological role of the root-specific, WCS417r-inducible genes in the onset of ISR, we systematically started to analyze T-DNA insertion mutants of these genes. In this study, we demonstrate that the WCS417r-responsive gene MYB72, encoding a R2R3-MYB-like transcription factor protein, functions as an essential component during the early steps of the ISR signaling cascade in Arabidopsis. MYB72 is a member of the large R2R3-MYB gene family of which 125 members have been identified in Arabidopsis (Kranz et al. 1998;Stracke et al. 2001;Yanhui et al. 2006). R2R3-MYB transcription factors are implicated in the regulation of various plant processes, although the function of most of them is still unknown (Stracke et al. 2001).
To investigate the role of these root-specific, WCS417r-induced genes in ISR signaling, we systematically analyzed knockout mutants of these genes for their ability to express WCS417r-ISR against Pst DC3000. A mutant with a T-DNA insertion in the MYB72 gene, which is specifically upregulated in the roots upon colonization by WCS417r (Verhagen et al. 2004, Fig. 1A), was identified as being defective in the activation of ISR and was subjected to further detailed studies. Figure 1B shows that knockout mutant myb72-1 (SAIL_713G10) was unable to mount ISR against Pst DC3000 in response to colonization of the roots by WCS417r.
Previously, rhizobacterial strain Pseudomonas putida WCS358r and a crude cell-wall preparation of WCS417r were demonstrated to trigger the ISR signaling pathway in Arabidopsis, resulting in a similar level of induced protection against Pst DC3000 as ISR induced by live WCS417r bacteria (Van Wees et al. 1997). To find out whether ISR triggered by these inducers is also blocked in myb72-1, roots of Col-0 and myb72-1 plants were treated with killed WCS417r cells, or with living WCS358r bacteria, and tested for the expression of ISR. Col-0 plants treated with crude WCS417r cell wall material or living WCS358r bacteria both showed similar levels of protection against Pst DC3000 to that induced by live WCS417r cells ( Figure 1B). Knockout mutant myb72-1 was unable to mount ISR in response to any of the inducers, confirming that MYB72 is required for the onset of ISR by these activators.
To investigate whether the impaired ISR response of myb72-1 was caused by insufficient root colonization by the rhizobacterial strains, the number of rifampicin-resistant WCS417r and WCS358r bacteria per gram of root fresh weight was determined. No significant differences in the extent of root colonization between Col-0 and myb72-1 plants were observed ( Figure 1C). Thus, the inability of myb72-1 to express WCS417r-mediated ISR was not caused by reduced root colonization.
To confirm that the ISR-minus phenotype of knockout mutant myb72-1 was caused by disruption of the MYB72 gene, a second, independent T-DNA insertion mutant, designated myb72-2 (SALK_052993), was tested for its ability to express WCS417r-ISR. Figure 1D shows that myb72-2, like myb72-1, was unable to mount ISR against Pst DC3000, indicating that a functional MYB72 gene is required for the onset of WCS417r-ISR against this pathogen in Arabidopsis. Verification of the predicted T-DNA insertion sites in the myb72-1 and myb72-2 knockout mutants is described in

Mutant myb72-1 Is Not Impaired in SAR or Resistance Induced by MeJA or ACC
To investigate the effect of the myb72-1 mutation on pathogen-induced SAR, we compared the levels of rhizobacteria-mediated ISR and pathogen-induced SAR in this mutant. SAR was induced 3 days prior to challenge inoculation with virulent Pst DC3000 by infiltrating three lower leaves with avirulent Pst DC3000(avrRpt2). Wild-type Col-0 plants developed significant levels of protection against Pst DC3000 in response to induction of ISR and SAR (Figure 2A). In contrast to ISR, SAR was expressed to wild-type levels in myb72-1, indicating that the ability to develop SAR was not altered in this mutant.
Similarly, chemical induction of SAR by exogenous application of SA resulted in similar levels of protection against Pst DC3000 in Col-0 and myb72-1 ( Figure 2B), confirming that myb72-1 is not impaired in SAR.
Like rhizobacteria-mediated ISR, exogenous application of methyl JA (MeJA) or the ET precursor 1-aminocyclopropane-1-carboxylate (ACC) triggers an enhanced level of resistance against Pst DC3000 (Van Wees et al. 1999). To examine the effect of the myb72-1 mutation on resistance induced by these chemicals, Col-0 plants were pretreated with MeJA or ACC at 7 and 4 days before challenge inoculation with virulent Pst DC3000. Figure 2B shows that myb72-1 developed wild-type levels of protection against Pst DC3000 in response to both chemicals, indicating the myb72-1 mutation has no effect on the ability to express enhanced resistance in response to ACC or MeJA.
These findings suggest that MYB72 operates upstream of JA and ET in the ISR signaling pathway.

MYB72 Is Required for ISR Against a Broad Spectrum of Pathogens
WCS417r-mediated ISR is effective against a broad spectrum of pathogens (Pieterse et al. 1996;Van Wees et al. 1997;Ton et al. 2002b). To examine whether MYB72 is required for the onset of broad-spectrum ISR, we tested the ability of myb72-1 to express ISR against the biotrophic oomycete H. parasitica and the necrotrophic fungi A. brassicicola and B. cinerea. Figure 3A shows that WCS417r-ISR and BTH (benzothiadiazole)-induced SAR resulted in a relatively moderate, but statistically significant, level of protection of Col-0 plants against H. parasitica. In similarity to the bioassays with Pst DC3000, mutant myb72-1 plants failed to develop ISR against this pathogen, whereas induction of SAR resulted in wild-type levels of induced resistance. To test the effectiveness of ISR against A. brassicicola, ISR bioassays were performed in the genetic background of the camalexin-deficient mutant pad3-1. In contrast to wild-type Col-0 plants, pad3-1 is susceptible to A. brassicicola infection and is routinely used in assays to test for induced resistance against this pathogen (Thomma et al. 1999;Ton et al. 2002b;Spoel et al. 2007). Figure 3B shows that induction of ISR in pad3-1 significantly reduced disease symptoms caused by A. brassicicola infection, whereas the pad3-1/myb72-1 double mutant failed to mount ISR against this pathogen. Similarly, Col-0 plants, but not myb72-1, expressed statistically significant levels of ISR against B. cinerea ( Figure 3C).

WCS417r-Induced Priming for Enhanced Callose Deposition Is Blocked in myb72-1
Induced resistance is often associated with priming for enhanced deposition of callose-containing papillae at sites of attempted pathogen attack or tissue injury (Kohler et al. 2002;Ton et al. 2005). Figure 4 shows that induction of ISR by WCS417r, or SAR by exogenous application of BTH, resulted in a significant decrease of the success rate of H. parasitica spores to penetrate the leaves of wildtype Col-0 plants due to enhanced callose depositions in the epidermal cell layer around the entry sites. Treated mutant npr1-1 plants that are blocked in their ability to express both ISR and SAR, did not show this enhanced callose deposition at the sites of attempted pathogen attack and consequently did not display lower penetration success rates of the pathogen, confirming that priming for enhanced callose deposition is associated with ISR and SAR. Priming for enhanced callose deposition was normally expressed in myb72-1 upon induction of SAR with BTH, resulting in a significant decrease in successful spore penetration ( Figure 4). However, unlike BTH, WCS417r did not enhance callose deposition in myb72-1 in response to H. parasitica attack. Hence, although myb72-1 is not impaired in its ability to be primed for enhanced formation of callose-containing papillae, this defense-related trait is not expressed in myb72-1 in response to colonization of the roots by WCS417r bacteria. Together, these results indicate that MYB72 plays an important role in the onset of this primed defense response during ISR. To test whether WCS417r-induced expression of MYB72 in the roots requires ET sensitivity, MYB72 transcript accumulation was examined in the ET-insensitive, ISR-minus mutant ein2-1.

MYB72 Functions Upstream of or in Parallel with ET Signaling in the ISR Pathway
Colonization of the roots by WCS417r bacteria activated MYB72 equally in both Col-0 and ein2-1 ( Figure 5B). These results indicate that MYB72 either acts upstream of ET signaling, or is co-required with components from the ET signaling pathway during the onset of ISR.

MYB72 Is Required but not Sufficient for ISR
To investigate whether MYB72 is not only required but also sufficient for the onset of ISR, transgenic plants that constitutively express MYB72 (35S:MYB72) were generated and tested for enhanced disease resistance. Seven independent homozygous T3 lines (OX1 to OX7) were phenotypically characterized. All transgenic lines displayed a phenotype that was similar to the parental Col-0 line the 35S:MYB72 lines confirmed constitutive expression of MYB72 in all lines, be it to varying levels ( Figure 6A). Bioassays for induced resistance assays were performed with Col-0, the empty vector control and the seven 35S:MYB72 transgenic lines. Figure 6B shows that the level of basal resistance against Pst DC3000, H. parasitica, and B. cinerea in lines OX2 and OX7 was not significantly enhanced compared to from that in the EV control line. Resistance assays with the other OX-lines yielded similar results (data not shown), indicating that ectopic expression of MYB72 is not sufficient for the onset of ISR.

MYB72 physically interacts with EIL3 in vitro
If MYB72 is essential but not sufficient for the onset of ISR, then additional components are likely to be co-required. Transcription factors usually exert their action in a complex with other proteins.
Earlier, a systematic search for proteins that physically interact with MYB transcription factors was initiated by members of the EU-funded REGIA (Regulatory Gene Initiative in Arabidopsis) consortium.
Using the ProQuest yeast two-hybrid system (Invitrogen), this screen revealed that MYB72 physically interacts with the ETHYLENE INSENSITIVE3 (EIN3)-like protein EIL3 (At1g73730; data not shown).
To confirm the interaction of MYB72 with EIL3, the full-length coding regions of MYB72 and EIL3 were isolated, fused to the DNA binding domain (BD) and transcription activation domain (AD) of GAL4, and tested in a yeast two-hybrid assay. Figure  Although EIL3 mRNA levels slightly raised in the roots upon colonization by WCS417r, this rise was not statistically significant. Co-expression of MYB72 and EIL3 makes the MYB72-EIL3 interaction in planta theoretically feasible. However, future studies on the interaction of MYB72 and EIL3 in planta should shed light on the occurrence and significance of this interaction in the onset of ISR.
The EIL3 paralogs EIN3, EIL1 and EIL2 have been demonstrated to function as key transcription factors of ET-regulated gene expression and to act as positive regulators of ET signaling (Stepanova and Ecker 2000). Since MYB72 acts upstream of ET signaling or is co-required with components of the ET signaling pathway during the onset of ISR ( Figure 5), we investigated whether the ISR-minus phenotype of the myb72-1 knockout mutant is caused by a reduced sensitivity to ET. The "triple response" is a reaction of etiolated seedlings to ET, and is commonly used as a reliable marker for ET sensitivity (Guzmán and Ecker 1990). Etiolated Col-0, ET-insensitive ein2-1 and myb72-1 seedlings were grown in the dark on MS-agar plates with or without ACC. Ten days after germination, Col-0 seedlings grown on a concentration range of ACC showed a typical ET-induced growth inhibition of the hypocotyl and root, both characteristics of the triple response ( Figure 7B). As expected, the triple response was not apparent in the ET-insensitive ein2-1 seedlings. In contrast, in mutant myb72-1 seedlings, the triple response was indistinguishable from that in wild-type Col-0 plants. These results demonstrate that the absence of a functional MYB72 protein does not affect ET sensitivity. Hence, the inability of myb72-1 plants to mount ISR is not caused by an inability to react to ET in this mutant.

DISCUSSION
Colonization of the roots of Arabidopsis by non-pathogenic fluorescent Pseudomonas bacteria, such as WCS417r and WCS358r, leads to an enhanced level of resistance against a broad spectrum of pathogens in foliar tissues (Pieterse et al. 2002). Genes whose expression is changed in the roots upon colonization by ISR-inducing rhizobacteria are potentially involved in the onset of ISR. Here, we demonstrate the role of one of these genes, MYB72, in the onset of rhizobacteria-mediated ISR.

Stress Signaling
MYB72 is a member of a large class of genes that contain one or more MYB domains (Stracke et al.

MYB72 Is Required in Early ISR Signaling
Previously, Knoester et al. (1999)  would develop normal levels of ISR in the leaves after application of WCS417r to the roots. However, this was not the case. Thus, ET signaling is required locally at the site of application of the inducer, and may be involved in the generation or translocation of the systemically transported signal. Here we demonstrated that WCS417r-induced expression of MYB72 is not regulated by ET, because mutant ein2-1 plants accumulated normal levels of MYB72 transcripts in the roots upon treatment with WCS417r ( Figure 5B). In addition, the expression of MYB72 was not activated upon treatment of the roots with ACC ( Figure 5A). All together these results demonstrate that MYB72 either acts upstream of ET, or is co-required with components from the ET signaling pathway during the onset of ISR in the roots.

MYB72 Is not Sufficient for the Onset of ISR
Although MYB72 is required for the onset of ISR, overexpression of the MYB72 gene did not result in enhanced disease resistance ( Figure 6). Hence, another signaling component is likely to be co-  (Chao et al. 1997;Solano et al. 1998). If EIL3 would function similarly as EIN3, EIL1 and EIL2 in ET signaling, then physical interaction with MYB72 may facilitate the so far unidentified ET-signaling event that is co-required with MYB72 in the roots for the onset of WCS417r-mediated ISR in the leaves. Besides a potential role in ET signaling, EIL3 (also called SLIM1 for SULFUR LIMITATION1) was recently shown to function as an important transcriptional regulator in the response of Arabidopsis to sulfur deprivation (Maruyama-Nakashita et al. 2006). Collectively, these data highlight that both MYB72 and EIL3 are part of the signaling network involved in the plant's response to biotic and abiotic stresses. Physical interaction between MYB72 and EIL3 in planta and its significance for the onset of ISR remains to be demonstrated and will be the subject of future study.

Model for ISR Signal Transduction
The identification of MYB72 as an important signaling component in the roots for the systemic onset of ISR adds a new factor to ISR signal transduction. Figure 8 summarizes our current understanding of the ISR signaling pathway. The local onset of WCS417r-ISR in the roots requires responsiveness to ET (Knoester et al. 1999), and is associated with an ET-independent activation of the MYB72 gene.
MYB72 is required but not sufficient for the onset of ISR. Hence, MYB72 is assumed to act in concert with another signaling component. Since EIL3 interacts with MYB72 in vitro, EIL3 is a potential candidate in this respect. Systemically in the leaves, expression of ISR requires responsiveness to both JA and ET and is dependent on NPR1 (Pieterse et al. 1998 Van der Ent et al. page: 15 to SAR) (Verhagen et al. 2004). Instead, ISR-expressing plants are primed to express a specific set of JA/ET-responsive genes faster and to a higher level upon pathogen infection (Van Wees et al. 1999;Hase et al. 2003;Verhagen et al. 2004). This enhanced defensive capacity allows the plants to respond faster and/or more strongly to attackers that trigger JA/ET-dependent defense responses, without major metabolic changes in the absence of an intruder (Conrath et al. 2006). Therefore, ISR forms a low-cost defense strategy that is active against a broad spectrum of attackers (Van Hulten et al. 2006).

Cultivation of Rhizobacteria and Pathogens
Non-pathogenic, rifampicin-resistant Pseudomonas fluorescens WCS417r and Pseudomonas putida WCS358r bacteria were used for induction of ISR (Van Wees et al. 1997). Both strains were grown for 24 h at 28°C on King's medium B (KB) agar plates (King et al. 1954), as described previously (Pieterse et al. 1996). Crude cell wall material of WCS417r was prepared according to Van Wees et al., (1997). Colonization of the rhizosphere of wild-type and mutant plants by rifampicin-resistant WCS417r and WCS358r bacteria was examined at the end of each ISR bioassay as described (Pieterse et al. 1996).
An avirulent strain of Pseudomonas syringae pv. tomato DC3000, carrying the avirulence gene avrRpt2 (Pst DC3000(avrRpt2)), (Kunkel et al. 1993) was used for SAR induction. Pst DC3000 Hyaloperonospora parasitica strain WACO9 was maintained on susceptible Col-0 plants as described by Koch and Slusarenko (1990). Sporangia were obtained by washing leaves that were densely covered by sporangiophores in distilled water, collected by centrifugation, and resuspended in water to a final density of 5x10 4 cfu.ml -1 .
Botrytis cinerea strain B0510 was grown on half-strength PDA plates containing penicillin (100 ppm) and streptomycin (200 ppm) for 2 weeks at 22°C. Spores were collected and resuspended in half-strenght potato dextrose broth (Difco Laboratories, Detroit, USA) to a final density of 5.5x10 5 spores.ml -1 . After a 3-h incubation period, the spores were used for inoculation of plants as described (Thomma et al. 1998

Construction of Transgenic Plants
A fusion product of the Cauliflower mosaic virus 35S promoter and the coding region of MYB72 was created by double-joint PCR as described by Yu et al. (2004). The resulting 35S:MYB72 fusion product was cloned into the binary vector pGreenII229 (Hellens et al. 2000). A derivative of the pGreenII229 vector was used as the empty vector (EV) control. Correct construction of the plasmids was verified by DNA sequencing. Subsequently, the binary vectors were transferred to Agrobacterium tumefaciens strain C58(pMP90) (Konzc and Schell 1986), after which Arabidopsis Col-0 plants were transformed according to the floral dip method (Clough and Bent 1998

Induction Treatments
Induction of ISR with living rhizobacteria was performed by mixing ISR-inducing rhizobacteria through the soil as described above. For tests with killed cells, a crude cell wall preparation of WCS417r bacteria (in 10 mM MgSO 4 ) was mixed through the soil in a similar manner, using the equivalent of the number of live bacteria introduced to the soil (5x10 7 cfu.g -1 ). Seven days before challenge inoculation, a similar amount of the crude cell wall material was applied to each plant as a soil drench as described previously (Van Wees et al. 1997).
Biological induction of SAR was performed three days before challenge inoculation by pressure infiltrating three lower leaves with a suspension of Pst DC3000(avrRpt2) bacteria at 10 7 cfu.ml -1 , as described (Pieterse et al. 1996).

Disease Resistance Assays
Pst DC3000 bioassays were performed essentially as described by Pieterse et al., (1996). Plants were challenged when five weeks old by dipping the leaves for 2 s in a solution of 10 mM MgSO 4 , 0.015% (v/v) Silwet L-77 containing 2.5x10 7 cfu.ml -1 Pst DC3000 bacteria. Four days after challenge, disease severity was assessed by determining the percentage of diseased leaves per plant. Leaves were scored as diseased when showing necrotic or water-soaked lesions surrounded by chlorosis.
The disease index was calculated by determining the proportion of leaves with disease symptoms per plant (n=20).
H. parasitica bioassays were performed as described by Ton et al., (2002b path 425 nm). Callose-inducing spores were scored as unsuccessful penetrations, while those that did not trigger a callose response were scored as successful penetrations.
A. brassicicola bioassays were performed as described (Ton et al. 2002b). Because Arabidopsis Col-0 is resistant to A. brassicicola, whereas the camalexin-deficient mutant pad3-1 (Zhou et al. 1999) is susceptible (Thomma et al. 1999;Ton et al. 2002b), the role of MYB72 was investigated in the double mutant pad3-1/myb72-1,created through genetic crossing. When 5 weeks old, homozygous pad3-1 and pad3-1/myb72-1 plants (n=20) were challenge inoculated with A. brassicicola by applying 3-µL droplets of water containing 1x10 6 spores.ml -1 onto the second, third and fourth true leaf pair of each plant. At 5 days after challenge, disease severity was determined. Disease rating was expressed on the basis of symptom severity: I, no visible disease symptoms; II, non-spreading lesion; III, spreading lesion without tissue maceration; IV, spreading lesion with tissue maceration and sporulation of the pathogen.
B. cinerea inoculations were performed with 5-week-old Col-0 and myb72-1 plants (n=20) by applying 5-µL droplets of the spore suspension onto fresh needle-prick wounds on the second, third and fourth true leaf pair of each plant as described (Thomma et al. 1998). At 5 days after challenge, disease severity was determined. Disease ratings were expressed as the percentage of leaves showing spreading lesions.

Gene Expression Analysis
Extraction of total RNA, RNA gel-blot analysis with a gene-specific probe for MYB72 (At1g56160) was performed as described (Van Wees et al., 2000;Verhagen et al., 2004).
Quantitative real-time PCR (Q-PCR) was performed essentially as described by Czechowski et al. (2004). To check for genomic DNA contamination, a PCR with primers designed on EIL2 (At5g21120; EIL2-F and EIL2-R) was carried out. Efficiency of cDNA synthesis was assessed by Q-PCR, using primers of the constitutively expressed gene UBI10 (At4g05320; UBI10-F and UBI10-R). Primers for MYB72 (At1g56160; MYB72-F4 and MYB72-R4), EIL3 (At1g73730; EIL3-F and EIL3-R) and the ETresponsive gene EBF2 (At5g25350; EBF2-F and EBF2-R) were designed and checked as described by Czechowski et al. (2004). Nucleotide sequences of all primers are given in Supplementary Table 1

Yeast Two-Hybrid Assays
Constructs for yeast two-hybrid analyses were generated using vectors pDEST TM 32 and pDEST TM 22 (Invitrogen, Breda, the Netherlands) for protein fusions to the GAL4 DNA-binding domain (BD) or transcriptional-activation domain (AD), respectively. Full-length coding regions of MYB72 and EIL3 cDNA were introduced in both vectors using the GATEWAY TM technology (Invitrogen), following the manufacturer's instructions. Clones containing the BD:MYB72, AD:MYB72, BD:EIL3, and AD:EIL3 fusions were checked by sequence analysis and subsequently used in the yeast two-hybrid assay. AD and BD plasmids were transformed into the a and α mating types of Saccharomyces cerevisiae strain PJ69-4, using a lithium acetate/polyethylene glycol protocol described by Gietz and Woods (2006). PJ69-4 carries ADE2, HIS3, URA and LacZ reporters for reconstituted GAL4 activity (James et al. 1996). Transformants were selected on yeast selective drop-out medium (SD) lacking either leucine (leu) for selection of the BD vectors, or tryptophan (-trp) for selection of the AD vectors. Opposite mating types were co-cultured overnight on nutrient-rich YAPD medium (Sigma-Aldrich Chemie BV, Zwijndrecht, the Netherlands) at 30°C. Diploids harboring both plasmids were selected on SD medium lacking both leucine and trypthophan (-leu, -trp) and used in the yeast two-hybrid assay.

ACC-Induced Triple Response
Seeds of Arabidopsis were surface sterilized for 5 min in 5% (v/v) sodium hypochlorite, washed in 70% (v/v) ethanol, and air-dried. Seeds were subsequently distributed evenly on 1.0% (w/v) agar medium (pH 5.7) containing 0.5% (w/v) Murashige and Skoog (MS) salts (Duchefa bv, Haarlem, The Netherlands), 0.5% (w/v) sucrose, and different concentrations of filter-sterilized ACC, which was added from a 10 mM stock solution. The effect of ACC-derived ET on hypocotyl and primary root length in etiolated seedlings was determined essentially according to Guzmán and Ecker (1990). After pre-germination in the dark for 2 days at 4°C, seedlings were grown for an additional 3 to 7 days at 20°C in darkness after which the triple response was monitored. ISR was induced by growing the plants for 3 weeks in soil containing living ISR-inducing WCS417r or WCS358r bacteria, or crude cell wall material of WCS417r (CW WCS417r). Five-week-old plants were challenge inoculated with virulent Pst DC3000. Four days after challenge inoculation, the percentage of diseased leaves was assessed and the level of induced protection calculated on the basis of the reduction in disease symptoms relative to challenged, non-induced plants. The absolute proportions of diseased leaves in the control treatment was 53.9 (Col-0) and 50.7% (myb72-1). Asterisks indicate statistically significant differences compared to non-induced control plants (Students t-test, α =0.05; n=20). (C) Numbers of rifampicin-resistant WCS417r or WCS358r bacteria (log10 of the number of colony forming units (cfu).ml -1 ) in the rhizosphere of the plants at the end of the bioassay.
In the rhizosphere of non-induced plants, no rifampicin-resistant bacteria were detected (detection limit 10 3 cfu.g -1 root fresh weight (FW)). (D) Quantification of WCS417r-induced protection against Pst DC3000 in wild-type Col-0 and knockout mutants myb72-1 and myb72-2. The level of induced protection was calculated as described above (B). The absolute proportions of diseased leaves in the control treatments was 78.0 (Col-0), 73.4% (myb72-1), 79.1 (Col-0), and 74.1% (myb72-2). Asterisks indicate statistically significant differences compared to non-induced control plants (Students t-test, For details on Pst DC3000 bioassays see legend to Figure 1. Asterisks indicate statistically significant differences compared to non-induced control plants (Students t-test,

Figure 4. Mutant myb72-1 Is Impaired in WCS417r-Mediated Priming for Enhanced Callose Deposition at H. parasitica Infection Sites
Induced resistance against H. parasitica is associated with enhanced deposition of callosecontaining papillae at sites of attempted penetration, resulting in a reduction of the number of spores that successfully penetrate into Arabidopsis leaves. Two days after challenge with H. parasitica, successful penetration of H. parasitica spores was quantified in leaves of Col-0, npr1-1 and myb72-1 plants. Leaves of plants of which the roots were pre-treated with water (Ctrl), WCS417r (ISR) or BTH (SAR) were stained with calcofluor/aniline-blue and analyzed by epifluorescence microscopy (UV). The inset shows a representative example of a germinating H. parasitica spore (s) triggering callose deposition (c) in the underlying epidermal cell. Callose-inducing spores were scored as unsuccessful penetrations, while those that did not trigger a callose response were scored as successful penetrations. The figure depicts the percentage of spores that successfully penetrated the host cell. Asterisks indicate statistically significant differences compared with the non-induced control treatments (Chi-square, α =0.05; n = 150).   Colonization of the roots by ISR-inducing P. fluorescens WCS417r leads to a local, ETindependent activation of the transcription factor gene MYB72. Downstream of, or in parallel with MYB72, a so far unidentified ET signaling component is required in the roots for the onset of broad-spectrum ISR in the leaves. Since EIL3 interacts with MYB72 in vitro, EIL3 is a potential candidate in this respect. Systemically, the ISR signal transduction cascade requires responsiveness to both JA and ET, and is dependent on NPR1. Finally, induction of ISR is associated with priming for enhanced expression of a large set of JA-and ET-responsive genes that becomes apparent only after pathogen attack. This allows the plant to react more effectively to an invading pathogen, which may explain the broad-spectrum characteristic of rhizobacteriamediated ISR.