Pseudomonas cichorii strain SPC9018 harbours hrp genes that encode a type III secretion system. The hrp mutants have reduced virulence on lettuce and lose virulence on aubergine. Of the hrp genes, the B3 gene (B3) is located next to hrpW, encoding a harpin, and is predicted to encode a small (12 kDa) acidic (pI = 5·21) chaperone. In this study, involvement of the two genes in SPC9018 virulence was analysed. An hrpW-mutant (CWN1) and a B3-deleted mutant (ΔB3) of SPC9018 lacked virulence on aubergine and had reduced virulence on lettuce. Furthermore, both mutants also lost their ability to induce the hypersensitive response (HR) in tobacco leaves. The phenotypes were similar to those of the hrpL-deficient mutant (SPC9018-L). Transformation of CWN1 and ΔB3 with the plasmid containing hrpW and B3, respectively, originating from SPC9018 restored their virulence and ability to induce the HR in tobacco leaves. The population of CWN1 drastically decreased in aubergine following inoculation. In contrast, the population of ΔB3 was retained after inoculation, which was similar to that observed with SPC9018-L. RT-PCR analysis showed that hrpW and B3 are transcribed as a single transcript and its expression is dependent on HrpL. The two-hybrid analysis showed association of HrpW with the B3 protein. Therefore, HrpW can be implicated in the virulence of P. cichorii and its ability to induce the HR in tobacco, with some assistance from the B3 protein, probably acting as a chaperone.
Pseudomonas cichorii causes rot on lettuce. Previous studies have shown that de novo protein synthesis in lettuce leaves is required for the development of disease symptoms (Hikichi et al., 1998; Kiba et al., 2006a). The development of disease symptoms is closely associated with programmed cell death, following heterochromatin aggregation and laddering of genomic DNA in the P. cichorii infected lettuce cells (Kiba et al., 2006a). Kiba et al. (2009) demonstrated that during the development of bacterial rot in lettuce caused by P. cichorii, there may be at least two cell death mechanisms, mitochondria-dependent programmed cell death (PCD) without DNA fragmentation (non-apoptotic/necrosis-like PCD) and other mechanisms with DNA fragmentation (apoptosis-like PCD). The disease symptoms may be synergistically enhanced by both non-apoptotic/necrosis-like PCD and apoptosis-like PCD.
Necrotic leaf spot on aubergine is also known to be caused by P. cichorii. Kiba et al. (2006b) showed that development of necrotic lesions following PCD on leaves of aubergine inoculated with P. cichorii was commonly associated with de novo protein synthesis, generation of intracellular reactive oxygen species and caspase III-like protease activity.
In several Gram-negative phytopathogenic bacteria, the hrp genes are essential determinants for disease development on compatible hosts and for elicitation of the hypersensitive response (HR) on resistant plants (Alfano & Collmer, 1997). The hrp genes encode proteins in the type III secretion system (T3SS). The T3SS allows secretion in the apoplast and injection into the plant cell of a number of bacterial proteins, which is believed to contribute to pathogenesis by controlling plant defences and triggering cell death. Araki et al. (2006) and Hojo et al. (2008) reported that hrp genes exist in the genomic DNA of P. cichorii strains 83-1 and SPC9018. The hrp-deficient mutants originating from SPC9018, the hrpG-mutant, the hrcT-mutant, the hrpL-deficient mutant (SPC9018-L) and the hrpS-deficient mutant, had reduced virulence on lettuce. Furthermore, these mutants lost both their ability to grow vigorously on aubergine leaves and their virulence in aubergine, indicating that virulence of P. cichorii with respect to aubergine is dependent on the hrp genes (Hojo et al., 2008). Therefore, the bacterial proteins involved in virulence are secreted through the T3SS. As yet, there is no information regarding the type III effectors secreted through the T3SS of P. cichorii.
The T3SS-mediated secretion and translocation is thought to be a one-step process, whereby certain type III effectors are translocated directly into host cells through two bacterial membranes and the host plasma membrane. In the case of plant pathogens, effectors must also traverse plant cell walls. This process involves protein secretion machinery similar to the flagellar export system, an extracellular pilus or needle-like appendage, and particular T3SS-secreted proteins called translocators (Mota & Cornelis, 2005). Translocators are required for effector transport across the host membrane, either by forming pores in the membrane (Goure et al., 2005) or by forming a complex with the pore-formers (Holmström et al., 2001). For plant pathogenic bacteria, both HrpF of Xanthomonas campestris pv. vesicatoria (Büttner et al., 2002) and HrpK of P. syringae pv. tomato (Petnicki-Ocwieja et al., 2005) have been characterized as putative pore-forming translocators. In Erwinia amylovora, two harpins, HrpW and HrpN, which belong to the harpin family of T3SS-secreted proteins that are glycine rich and heat stable, are candidate translocators, because they are probably targeted to the plant apoplast (Wei et al., 1992; Kim & Beer, 1998). HrpW has a pectate lyase domain (Kim & Beer, 1998), as does a homologue, HrpW1, in P. syringae (Charkowski et al., 1998), and several similar proteins that have been detected in other plant pathogenic enterobacteria (Bell et al., 2004). The pectate lyase domain has not shown enzymatic activity in either P. syringae or E. amylovora, but HrpW1 of the former was shown to bind to calcium pectate, a major component of plant cell walls (Charkowski et al., 1998).
Substrates of the T3SS lack a single, defined secretion signal, and there is evidence for the existence of several independent secretion signals in which a specific chaperone is necessary for the efficient secretion of some effectors (Aldridge & Hughes, 2001). Chaperones of the T3SS are characterized by small acidic proteins, encoded by genes usually located in the vicinity of other genes encoding their corresponding substrates (Page & Parsot, 2002; Feldman & Cornelis, 2003). Chaperones are required for optimal secretion of their effectors, even if their mode of action remains evasive, and they stabilize some of their effectors by preventing unproductive interactions, maintaining solubility, or preserving secretion-prone conformations (Parsot et al., 2003).
Two genes, encoding HrpW and B3 protein, exist among the hrp genes of SPC9018 (Hojo et al., 2008). The predicted protein sequence of the 52·7 kDa HrpW revealed at least two distinct domains similar to harpins; a harpin-like domain (amino acids 1–309) rich in glutamine, serine and glycine, and a domain from amino acids 252–328 containing six imperfect glycine-rich repeats with many acidic and polar residues. The deduced amino acid sequences of the B3 protein showed that the protein is a small (12 kDa) and acidic (pI = 5·21) protein and has been predicted as a chaperone (Hojo et al., 2008). As yet, there is no information regarding the function of the HrpW and B3 protein. This study analyses the involvement of hrpW and the B3 gene (B3) in virulence of P. cichorii.
Materials and methods
Bacterial strains, culture conditions and plasmids
Bacterial strains and plasmids are listed in Table 1. Pseudomonas cichorii strains were routinely grown in PY-medium (5 g L−1 polypeptone, 2 g L−1 yeast extract, distilled water) at 30°C. Escherichia coli strains were grown in LM medium (Hanahan, 1983) at 37°C. Ampicillin (50 μg mL−1), kanamycin (50 μg mL−1) and tetracycline (30 μg mL−1) were used in selective media.
Table 1. Bacterial strains and plasmids used in this study
2·8-kb BamHI- and EcoRI-digested fragment of p4-57 in pHSG398
1·4-kb blunt-ended fragment containing aphA-3 of pUCK18K in blunt-ended EcoRV and BglII sites of phrpW
2·6-kb blunt ended PstI- and BamHI-digested fragment containing sacB of pUCD800 in blunt-ended XbaI site of phrpWKm
800-bp PCR product from SPC9018 genomic DNA in pHSG398
2·3-kb BamHI- and KpnI-digested PCR fragment from SPC9018 genomic DNA in pAVRE
1·4-kb KpnI-digested fragment containing KmR of pUCK191 in pavrEhrpWchp
2·6-kb BamHI- and PstI-digested fragment containing sacB of pUCD800 in p398-WchpKm
3·4-kb PCR fragment from SPC9018 genomic DNA in pLAFR3
1·7-kb PCR fragment containing hrpW ORF from SPC9018 genomic DNA in pT25
1·0-kb PCR fragment containing hrpZ from SPC9018 genomic DNA in pT25
350-bp PCR fragment containing B3 gene ORF from SPC9018 genomic DNA in pT18
Isolation of genomic DNA, plasmid DNA manipulations and polymerase chain reaction (PCR) were performed according to standard techniques (Sambrook et al., 1989). Pseudomonas cichorii was transformed by electroporation as described by Hojo et al. (2008). Double-stranded DNA sequencing templates were prepared with GenElute™ Plasmid Miniprep kits (Sigma Chemical). Sequences were determined using an Automated DNA Sequencer Model 373 (Applied Biosystems). DNA sequence data were analysed using the dnasis-Mac software (Hitachi Software Engineering). Enzymes including restriction endonucleases (Takara) were used according to the manufacturer’s instructions. The primers used in this study are listed in Table 2.
Table 2. Sequences of all oligonucleotide primers used in this study
aRestriction enzyme sites of KpnI in the primer sequence are underlined.
Creation of an hrpW-mutant
A BamHI- and EcoRI-digested 2·8 kb fragment of p4-57 (Hojo et al., 2008) was ligated between the BamHI and EcoRI sites of pHSG398 (Takara), to create phrpW. A 1·4 kb KpnI- and BamHI-digested fragment of pUC18K (Ménard et al., 1993) containing an aphA-3 cassette (KmR) was blunt-ended by T4 DNA polymerase and ligated between blunt-ended phrpW EcoRV and BglII sites to create phrpWKm. A 2·6 kb blunt-ended PstI- and BamHI-digested DNA fragment containing sacB from pUCD800 (Gay et al., 1985) was ligated at a blunt-ended phrpWKm XbaI site to create phrpWKmsacB. This plasmid was electroporated into SPC9018 cells and the resultant kanamycin- and sucrose-resistant recombinants, CWN1, were selected. To verify correct insertion containing the KmR gene in the hrpW locus of the isolated genetic background, a PCR product resulting from the use of primers hrpW-left and hrpW-right was digested with HincII and a 2·2 kb fragment was observed (data not shown), demonstrating that CWN1 was an hrpW-mutant.
Creation of a B3-deficient mutant
An 800 bp fragment was amplified by PCR from the genomic DNA of SPC9018 using primers avrE-right and Kpn-avrE-left. The 800 bp SacI- and KpnI-digested fragment was ligated between the SacI and KpnI sites of pHSG398 to create pAVRE. A 2·3 kb fragment was amplified by PCR from the genomic DNA of SPC9018 using primers hrpW-left and Kpn-hrpW-right. The 2·3 kb BamHI- and KpnI-digested fragment was ligated into the BamHI and KpnI sites of pAVRE to create pavrEhrpWchp. A 1·4 kb KpnI-digested fragment of pUCK191 containing a KmR cassette was ligated into the KpnI site of pavrEhrpWchp to create p398-WchpKm. The 2·6 kb PstI- and BamHI-digested DNA fragment containing sacB from pUCD800 was ligated into the PstI and BamHI sites of p398-WchpKm to create p398-WchapKmsac. This plasmid was electroporated into SPC9018 cells and the resultant kanamycin- and sucrose-resistant recombinants, ΔB3, were selected. Analysis by PCR with the primers hrpW-left and avrE-right was performed to verify correct insertion of the 4·4 kb fragment containing the KmR cassette in the B3 locus of the isolated genetic backgrounds (data not shown), demonstrating that ΔB3 was a B3-deficient mutant of SPC9018.
The bacteria-infiltrated area in aubergine leaves (0·6 g) 8 h after infiltration with P. cichorii strains was homogenized in RNA extraction buffer (100 mm glycine pH 9·5, 10 mm EDTA, 0·1 m NaCl, 1% [w/v] SDS, 0·1% [w/v] bentonite), extracted with phenol-chloroform-isoamyl alcohol (25:24:1) twice, and then extracted with chloroform-isoamyl alcohol (24:1). The aqueous phase was mixed with one-third the volume of 10 m LiCl and incubated at −20°C for 2 h. After centrifugation at 17 500 g for 30 min, the pellet was washed with 2 m lithium solution (2 m LiCl, 50 mm EDTA), and the pellet was dissolved in TE buffer. Phenol extraction and ethanol precipitation were performed to recover the RNA (Kanda et al., 2003).
Reverse transcription-PCR (RT-PCR) for hrpW and B3
The cDNA fragments corresponding to hrpW, B3 and an operon with hrpW and B3 were synthesized from total RNA (6 μg) using RAV-2 reverse transcriptase (Takara) with the primers W-RT2, Wc-RT2 and Wc-RT2, respectively. The PCR was then carried out using W-RT1 and W-RT2 for the amplification of a 300 bp DNA fragment specific to hrpW, with Wc-RT1 and Wc-RT2 for the amplification of a 300 bp DNA fragment specific to B3, and W-RT1 and Wc-RT2 for the amplification of a 632 bp DNA fragment specific to an operon with hrpW and B3.
Complementation of CWN1 and ΔB3
A 3·4 kb fragment was amplified by PCR from the genomic DNA of SPC9018 using primers hrpW-left and avrE-right. The 3·4 kb BamHI- and SacI-digested fragment was ligated between the BamHI and SacI sites of pLAFR3 (Staskawicz et al., 1987) to create phrpWoperon. Plasmid phrpWoperon was transformed into CWN1 and ΔB3 competent cells, and tetracycline resistant transformants, CWN1-W and ΔB3-W, were created, respectively.
The two-hybrid analysis
A 1·7 and a 1·0 kb DNA fragment were amplified by PCR from the genomic DNA of SPC9018 to include the hrpW and hrpZ open reading frames (ORFs) using primers W-1 and W-2, and Z-1 and Z-2, respectively. The KpnI-digested 1·7 and 1·0 kb fragments were ligated into KpnI-digested pT25 (Karimova et al., 1998), resulting in the creation of phrpW-T25 and phrpZ-T25, respectively.
A KpnI-digested 350 bp PCR product incorporating the B3 ORF using Wc-1 and Wc-2 as primers was ligated into KpnI-digested pT25 (Karimova et al., 1998), resulting in creation of phrpWchaperon-T18.
Escherichia coli DPH1 cells (Karimova et al., 1998) were co-transformed with phrpW-T25 and phrpWchaperon-T18, or phrpZ-T25 and phrpWchaperon-T18, creating ampicillin- and chloramphenicol-resistant clones. These clones were placed on LB-X-gal agar plates containing 40 μg mL−1 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-gal), ampicillin, and chloramphenicol, and incubated for 30 h at 30°C, with blue colonies observed. The transformants containing p25T-zip and p18T-zip (Karimova et al., 1998) were used as the positive clones.
Aubergine (Solanum melongena cv. Senryo no. 2), lettuce (Lactuca sativa cv. Success) and tobacco (Nicotiana tabacum cv. Samsun NN) were grown in pots containing commercial soil (Tsuchitaro, Sumitomo Forestry) at 25°C; light (16 h per day) was supplied at 10 000 lux throughout the experimental period. Five-week-old test plants were inoculated by leaf-infiltration using a 1 mL disposable syringe with 1·0 × 108 CFU mL−1 bacteria in a 20 μL volume. For all assays, inoculum concentrations were determined spectrophotometrically and confirmed by dilution plating. Lettuce plants were coded and inspected for symptoms daily for 7 days post-inoculation. Plants were rated on a 0–3 disease index scale, where 0 = no symptoms; 1 = discolouring; 2 = browning; and 3 = collapse (Hojo et al., 2008). Within each trial, 12 plants of each strain were treated, yielding 60 plants per strain.
Bacterial population in planta
Areas (1 cm2) inoculated with P. cichorii strains were excised from aubergine and lettuce leaves of five plants at 0, 1, 2 and 3 days post-inoculation and ground using a mortar and pestle. Samples (0·1 mL) of the original solution and 10-fold serial dilutions of it were spread onto three plates of selective agar media: PCSM (Uematsu et al., 1982) for SPC9018, PCSM containing kanamycin at 50 μg mL−1 for CWN1 and ΔB3, or PCSM containing kanamycin at 50 μg mL−1 and tetracycline at 30 μg mL−1 for CWN1-W and ΔB3-W. Colonies were counted after 2 days of incubation at 30°C to estimate the population.
Virulence of the hrpW-mutant and the B3-deficient mutant
To analyse involvement of hrpW and B3 in virulence of SPC9018, an hrpW-mutant, CWN1, and a B3-deficient mutant, ΔB3, were created. When lettuce leaves were inoculated with SPC9018, rot symptoms were first observed in the inoculated area at 12–18 h post-inoculation, progressing beyond the inoculated area at 3 days post-inoculation (Fig. 1a). When lettuce leaves were inoculated with CWN1 and ΔB3, rot symptoms caused by the mutants were first observed 1 day post-inoculation and the disease development was delayed compared with those caused by SPC9018 (Fig. 2), suggesting that both mutants reduced their virulence on lettuce leaves, similar to SPC9018-L.
Necrotic lesions were observed in the inoculated area of aubergine leaves with SPC9018 at 1 day post-inoculation and then progressed beyond the inoculated area within 3 days post-inoculation (Fig. 1b). In contrast, both CWN1 and ΔB3 lost their virulence on aubergine, similar to SPC9018-L (Fig. 1b).
To confirm the involvement of hrpW and B3 in P. cichorii virulence, CWN1 and ΔB3 were transformed with a plasmid carrying hrpW and B3 from the SPC9018 genome. The resulting transformants were CWN1-W and ΔB3-W, respectively. Both transformants exhibited virulence against lettuce and aubergine, similar to SPC9018 (Figs 1 & 2). These results suggest that the hrpW and B3 are involved in virulence of SPC9018.
HR-induction of P. cichorii strains in infiltrated tobacco leaves
In tobacco leaves infiltrated with SPC9018, the HR was induced 1 day post-inoculation (Fig. 1c). However, both CWN1 and ΔB3 lost their ability to induce the HR in tobacco leaves, similar to SPC9018-L. CWN1-W and ΔB3-W induced the HR in tobacco leaves 1 day post-inoculation. These results suggest that hrpW and B3 are required for the induction of the HR by SPC9018 on tobacco leaves.
Population of P. cichorii strains in aubergine and lettuce leaves
SPC9018, CWN1-W and ΔB3-W grew vigorously in lettuce leaves, reaching 2·3 × 107, 3·2 × 107 and 2·9 × 107 CFU cm−2, respectively, at 1 day post-inoculation. CWN1 and ΔB3 grew more slowly in lettuce leaves compared with the parent strain, and reached population sizes of 8·4 × 106 and 6·7 × 106 CFU cm−2, respectively, by day 3 post-inoculation (Fig. 3a).
The population of CWN1 in aubergine leaves decreased to 6·9 × 103 CFU cm−2 by day 3 post-inoculation (Fig. 3b). The population of ΔB3 showed little change after inoculation into aubergine leaves, remaining at 9·0 × 104–1·2 × 105 CFU cm−2 by day 3 post-inoculation. In contrast, at 1 day post-inoculation, the parent strain reached its maximum population size of 2·3 × 107 CFU cm−2, and CWN1-W and ΔB3-W reached their maximum densities of 1·7 × 107 and 2·6 × 107 CFU cm−2, respectively. The parental and the mutant strains grew similarly in both PS medium and PCSM medium (data not shown).
Expression of hrpW and B3 in infected aubergine leaves
To observe transcripts of hrpW and B3 in P. cichorii strains infecting aubergine leaves, RT-PCR analysis was performed using W-RT1 and W-RT2, and Wc-RT1 and Wc-RT2, respectively, as primers. When inoculated into aubergine leaves, SPC9018 expressed both hrpW and B3 at 8 h post-inoculation (Fig. 4a).
Reverse transcription-polymerase chain reaction analysis using W-RT1 and Wc-RT2 as primers showed the 632 bp DNA fragment amplification, suggesting hrpW and B3 are transcribed as a single transcript. ΔB3 expressed hrpW. CWN1 expressed neither hrpW nor B3, suggesting that CWN1 is a non-polar mutant. This is supported by virulence of CWN1-W.
The consensus sequence of the hrp box (GGAACC-N15–16-CCANNCA) was identified 135 bp upstream of hrpW (Hojo et al., 2008). When inoculated into aubergine leaves, SPC9018 expressed both genes at 8 h post-inoculation. However, the hrpL-deleted mutant, SPC9018-L (Hojo et al., 2008) did not express either of the genes (Fig. 4b). These results suggest that SPC9018 in infected aubergines expresses hrpW and B3, and the expression is dependent upon hrpL.
Interaction of HrpW with the B3 protein
Escherichia coli DHP1 was transformed with phrpW-T25 including hrpW and phrpWchaperon-T18 including B3, creating CDHP1W. The resulting colony was blue, similar to the transformants for E. coli DHP1 with p25T-zip and p18T-zip used as the positive clones (Fig. 5). The colony of the E. coli DHP1 transformant with phrpZ-T25 including hrpZ and phrpWchaperon-T18 was white, similar to colonies for E. coli DHP1 and the transformant with p25T and phrpWchaperon-T18 used as the negative clones. These results showed that HrpW and the B3 protein interact.
Hojo et al. (2008) demonstrated that the hrp mutants of SPC9018 grew slowly, and the appearance of disease symptoms on infected lettuce leaves was delayed compared with the wildtype strain. Therefore, the hrp cluster plays a role in virulence at the early stages of infection in lettuce leaves, although the hrp genes are not directly implicated in induction of PCD in infected lettuce leaves. It is thought that the putative T3SS-dependent effector proteins may hinder or delay the plant defence response, giving the bacteria time to multiply before inducing PCD in lettuce leaves. On infected lettuce leaves, CWN1 and ΔB3 grew at a slower rate (Fig. 3a), and the appearance of disease symptoms was delayed compared with the wildtype strain (Figs 1a & 2). Furthermore, similar to SPC9018-L, CWN1 and ΔB3 lost their virulence on aubergine (Fig. 1b). Both mutants also lost their ability to induce the HR in infiltrated tobacco leaves (Fig. 1c), which is dependent on the hrp genes. Therefore, hrpW and B3 may be involved in certain functions related to the T3SS in SPC9018.
A few harpins are thought to be translocators of T3SS effectors through the host plasma membrane during plant–bacteria interactions (Büttner & He, 2009). Although it is currently unclear how harpins facilitate effector translocation, the P. syringae harpins HrpW1 and HopAK1 contain a C-terminal pectate lyase-like domain (Kvitko et al., 2007). This domain in HrpW1 binds to calcium pectate, a major plant cell wall component (Charkowski et al., 1998). This evidence suggests that harpins may be involved in modifying the plant cell wall to facilitate initial penetration of the T3SS pilus (Büttner & He, 2009). Kvitko et al. (2007) showed that HrpK1 and harpins are functionally redundant and act at the same step of effector translocation. Notably, harpins are found to be associated with synthetic lipid membranes and to form pores (Lee et al., 2001; Racapéet al., 2005; Engelhardt et al., 2009). In HrpW of SPC9018, the harpin-like domain and six imperfect glycine-rich repeats with many acidic and polar residues are contained. Furthermore, the 253 C-terminal amino acids of HrpW exhibited an identity of 56·3% with Pel from Pectobacterium carotovorum ssp. carotovorum strain PC1. Virulence on aubergine and the ability to induce the HR in tobacco leaves of SPC9018 was dependent on the hrp genes (Hojo et al., 2008). Mutation of hrpW resulted in loss of the bacterial virulence on aubergine (Fig. 1b) and its ability to induce the HR in tobacco leaves (Fig. 1c). Therefore, it is proposed that HrpW might function as the translocator of T3SS effectors.
It has been shown that the secreted HrpJ proteins from E. amylovora and P. syringae are required for efficient T3SSs (Fu et al., 2006; Bocsanczy et al., 2008). Both proteins contributed to the secretion of harpin proteins, suggesting that they may indirectly affect effector protein translocation (Fu et al., 2006; Nissinen et al., 2007; Bocsanczy et al., 2008). In X. campestris pv. vesicatoria, translocation of effector proteins is differentially regulated by the global T3SS chaperone HpaB, which specifically promotes the translocation of a certain class of effector proteins (Büttner et al., 2006). The activity of HpaB is presumably controlled by the secreted regulator HpaA that binds to HpaB in the bacterial cytoplasm and allows secretion of extracellular components of the T3SS. After assembly of the T3SS, secretion of HpaA liberates HpaB and thus activates secretion and translocation of effector proteins (Lorenz et al., 2008). In addition to HpaB, translocation of a certain set of effectors from X. campestris pv. vesicatoria requires HpaH, which is a predicted lytic transglycosylase that might facilitate assembly of the T3SS (Büttner et al., 2007). Hojo et al. (2008) reported that populations of the hrp mutants from SPC9018 lose their virulence in aubergine and show little change after inoculation into aubergine leaves. From the deduced amino acid analysis, the B3 protein is predicted to be a small and acidic protein and a chaperone. The population of ΔB3 showed little change after inoculation into aubergine leaves, similar to the hrp mutants (Fig. 4b). The two-hybrid analysis showed the interaction of HrpW with the B3 protein (Fig. 5). Furthermore, the RT-PCR analysis suggests an operon with hrpW and B3 and their hrpL-dependent expression (Fig. 3). Therefore, the B3 protein may act as a chaperone to assist the function of HrpW, which could be involved in translocation of T3SS effectors into plant cells. This speculation is supported by the loss of virulence on aubergine (Fig. 1b) and the HR induction on tobacco leaves (Fig. 1c) of ΔB3, similar to those of CWN1 and SPC9018-L.
Reboutier et al. (2007) showed that the HrpW from E. amylovora, at subnanomolar concentrations, was able to decrease defence responses triggered by another harpin from this bacteriium, HrpN. This antagonism could be the result of opposed anion channel modulations triggered by HrpW and HrpN. At greater concentrations HrpW, alone or in combination with HrpN, induced cell death. This form of cell death involved strong ion channel activation and shared similarity with apoptosis volume decrease, a form of programmed cell death that has been previously described in animal cells, highlighting the role of ion channels during PCD processes (Reboutier & Bouteau, 2008). Furthermore, ectopic expression of popA, which encodes a harpin, in Ralstonia solanacearum strain OE1-1, immediately after invasion into tobacco plants induces defence responses, leading to a drastic decrease in the bacterial population (Kanda et al., 2003). In this study, the mutation of hrpW leads to a drastic decrease in the bacterial population in aubergine (Fig. 4b). The results might be consistent with the notion of HrpW also inducing a change in the ion channels, leading to induction of host defence in infected aubergine leaves.
Taken together, the HrpW and the B3 protein may be involved in the T3SS functions regarding its virulence. For further understanding of the function of HrpW and B3 protein, it will be necessary to identify the T3SS effectors that are involved in P. cichorii virulence.
The authors thank Dr Ayami Kanda-Hojo for her critical reading. This work was supported by Grants-in-Aid for Scientific Research awarded to YH (20380029), AK (21580057), and KO (22580052) from the Ministry of Education, Science, Sports and Culture, Japan.