1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

As in many other Gram-negative phytopathogenic bacteria, the Hrp type III secretion system is essential for the pathogenicity of Ralstonia solanacearum on host plants. The expression of most of the type III effector genes previously isolated from R. solanacearum is co-regulated with those of hrp genes by an AraC-type transcriptional activator, HrpB. In order to isolate type III-related pathogenicity genes, we screened hrpB-regulated genes in R. solanacearum. Using a transposon-based system, we isolated 30 novel hpx (hrpB-dependent expression) genes outside the hrp gene cluster. Most of the hpx genes contain a PIP (plant-inducible promoter) box-like motif in their putative promoter regions. Seven hpx genes encoded homologues of known type III effectors and type III-related proteins found in other animal and plant pathogens. Four encoded known enzymes, namely, glyoxalase I, Nudix hydrolase, spermidine synthase and transposase. Interestingly, six hpx genes encoded two types of leucine-rich repeat (LRR) protein. Products of the remaining genes did not show any significant homology to known proteins. We also identified two novel hrpB-regulated genes, hpaZ and hpaB, downstream of hrpY in the hrp cluster. The hpaB gene of R. solanacearum, but not hpaZ, was required for both the pathogenicity and ability to induce hypersensitive reaction on plants. We show that a hpaB null mutant still produces Hrp pili on the cell surface although it shows a typical Hrp-defective phenotype on plants.


  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

Many Gram-negative pathogens of plants and animals utilize the type III secretion system (TTSS) to interact with their hosts (Hueck, 1998; Galán and Collmer, 1999; Cornelis and Van Gijsegem, 2000). Some of the TTSS components show significant homology to those of a flagellar basal body, suggesting that the two systems have a common ancestor (Van Gijsegem et al., 1995; Aizawa, 2001; Gophna et al., 2003). Indeed, the direct observation of type III secretion machinery complexes purified from animal pathogens shows their striking morphological similarities to the flagellar basal body (Kubori et al., 1998; Tamano et al., 2000). In animal pathogens, a short needle-like substructure extends from a type III secretion apparatus that penetrates the bacterial outer and inner membranes similar to a flagellum from the flagellar basal body (Kubori et al., 1998). On the other hand, a long pilus-like filament, called the Hrp pilus, probably extends from type III secretion complexes in plant pathogens (Roine et al., 1997; Van Gijsegem et al., 2000; Jin and He, 2001). Various results indirectly suggest that virulence proteins, the so-called type III effector proteins, are injected from pathogens into host eukaryotic cells via such needle- or pilus-like substructures of the type III secretion apparatus (Van Gijsegem et al., 2000; Wei et al., 2000; Jin and He, 2001). In representative animal and plant pathogens, the direct translocation of some effector proteins into the host cytosol via TTSS has been demonstrated (Sory and Cornelis, 1994; Casper-Lindley et al., 2002). At present, bacterial effector proteins are considered to function inside a host cell to stimulate or suppress various host cellular functions to benefit pathogenic infection (Staskawicz et al., 2001)

In most Gram-negative plant pathogenic bacteria, except for Agrobacterium, TTSS is present and its components are encoded within a large hrp (hypersensitive reaction and pathogenicity) gene cluster (Cornelis and Van Gijsegem, 2000). The finding that hrp-defective mutants completely lose both their pathogenicity and ability to induce hypersensitive reaction (HR) on plants provides good indirect evidence of the translocation of effector proteins into plant cells (Alfano and Collmer, 1997). Genes encoding effector proteins that elicit HR on non-host plants have been isolated as avr (avirulence) genes from various phytopathogenic bacteria (Leach and White, 1996). However, in many cases, mutation in a single avr gene causes no significant changes in the pathogenicity of a pathogen, probably because of redundancy. It is important to determine the exact number of effector proteins possessed by a certain plant pathogen and to identify their functions in host cells. To this end, attempts to catalogue type III effector genes in phytopathogenic bacteria, such as Xanthomonas campestris and Pseudomonas syringae, are under way using functional screening systems (Noël et al., 2001; Boch et al., 2002; Fouts et al., 2002; Guttman et al., 2002) or by analysing whole-genome DNA sequences (Petnicki-Ocwieja et al., 2002; Salanoubat et al., 2002; Zwiesler-Vollick et al., 2002).

Ralstonia solanacearum is a soil-born vascular pathogen that causes bacterial wilt diseases in many plant species, including economically important crops, such as tomato, potato, banana, tobacco and eggplant (Hayward, 1991). In R. solanacearum GMI1000, the hrp gene cluster is organized in seven transcription units that comprise at least 20 hrp, hrc and hpa genes (Van Gijsegem et al., 1995). The transcription of hrp units 1, 2, 3, 4 and 7 is dependent on an AraC-type transcriptional activator, HrpB, encoded by the hrpB gene in the hrp cluster (Arlat et al., 1992; Genin et al., 1992). As in many other Gram-negative phytopathogenic bacteria, the multiplication of R. solanacearum in host plants is completely dependent on Hrp TTSS (Boucher et al., 1987). R. solanacearum Hrp-secreted proteins PopA, PopB and PopC have been identified in an operon at the left end outside the hrp cluster (Arlat et al., 1994; Guéneron et al., 2000). The expression of the popABC operon is also dependent on hrpB (Arlat et al., 1994). Because the disruption of popABC showed no significant effect on the pathogenicity of R. solanacearum on host plants (Arlat et al., 1994), it is considered that other unknown virulence effector genes are present in this pathogen.

In many phytopathogenic bacteria, the expression of type III effector genes is co-regulated with those of hrp genes by common regulatory proteins (Alfano and Collmer, 1997). This strongly suggests that unknown effector genes are also controlled by hrpB in R. solanacearum. It is known that R. solanacearum genes regulated by hrpB, such as hrp, hrc and popABC genes, have a common characteristic in gene expression: the expression is specifically activated in a hrpB-dependent manner in a minimal medium, which probably mimics the environmental conditions in planta, but repressed in a nutrient-rich medium (Arlat et al., 1992). In this study, we screened R. solanacearum genes exhibiting such an expression characteristic using a transposon-based system with β-galactosidase (LacZ) and β-galactoside permease (LacY) activities. We identified two hrpB-regulated genes, hpaZ and hpaB, in the hrp cluster and 30 novel hpx (hrpB-dependent expression) genes in other loci. This is the first functional screening of genes belonging to the hrpB regulon in R. solanacearum. We also examined the effect of null mutation in hpaZ, hpaB and most of the hpx genes on the pathogenicity and the hypersensitivity of R. solanacearum.


  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

Screening of h rpB-dependent transcription units using a transposon-based system

Using the plasmid pTNP104, we constructed a library of mini-Tn5lacZYA insertions in R. solanacearum strain RS1085, a Nalr EPS derivative of the wild-type strain RS1000. In RS1085, the high production of extracellular polysaccharides (EPSs) and the formation of fluid colonies are completely eliminated by the ΔepsP mutation. Because this strain can form more than 100 single colonies on a plate, we can easily apply any genetic manipulations to the strain as compared with the wild type. Mini-Tn5lacZYA contains a promoterless lacZYA operon and a streptomycin/spectinomycin-resistant (Smr/Spr) gene for the selection of transposition events (Fig. 1A). Because R. solanacearum has neither β-galactosidase activity nor β-galactoside permease activity, when pTNP104 was conjugatively transferred into RS1085 and the cells were selected on hrp-inducing lactose minimal plates, only an exoconjugant, in which lacZYA was inserted under an active promoter, could form a colony. The frequency of the appearance of Lac+ Smr/Spr colonies was ≈10−7 per recipient cell. Lac+ Smr/Spr colonies were once patched on BG plates in which hrp induction is impossible, and then replica-plated on hrp-inducing glucose minimal and BG plates containing X-Gal. Approximately 7800 colonies were screened and 50 of these were blue only on the hrp-inducing glucose minimal plates similar to colonies of RS1087, which has the popA–lacZYA fusion as a control strain (part of the results are shown in Fig. 2). To examine whether promoter activities of these 50 fusions are dependent on hrpB, Δ(hrpB–hrpD) mutation was introduced into each fusion. In 48 of these 50 fusions, induction in the minimal medium was completely abolished by the Δ(hrpB–hrpD) mutation as observed in RS1088, a Δ(hrpB–hrpD) derivative of RS1087, and was complemented by the plasmid pRS200 carrying wild-type hrpB (Fig. 2). This shows that in these 48 fusions, promoter activities are regulated by hrpB. These fusions were designated M1 to M48 and used in further studies.


Figure 1. Structure of transposons. The mini-Tn5lacZYA (A) and mini-Tn5′lacZY (B) transposons were used for promoter and ORF screenings respectively. The filled boxes indicate inner (I) and outer (O) ends of IS50, and the thick line between them indicates a transposon sequence. The promoterless lacZYA or 5′-truncated ′lacZY genes are indicated by filled arrows in each panel. Open arrows indicate antibiotic resistance genes. An open circle represents the origin (ori) of ColE1 used for plasmid rescue. The thin line shows the genome of R. solanacearum. A hooked arrow from a filled circle indicates the direction of transcription from an active promoter. A hatched box represents an open reading frame encoded within the transcription unit.

Download figure to PowerPoint


Figure 2. Expression of hrpB-dependent transcription units on hrp-inducing and hrp-repressing plates. Mutants with mini-Tn5lacZYA fusion (M21 and M25) and popA–lacZYA insertion (RS1087), their ΔhrpB derivatives and ΔhrpB derivatives harbouring hrpB+ plasmid pRS200 are shown. Mutants are first patched on a hrp-repressing BG plate and then replica-plated on hrp-inducing glucose minimal and hrp-repressing BG plates respectively. β-Galactosidase activity is monitored using X-Gal after 48 h (hrp-inducing medium) and 24 h (hrp-repressing medium) incubations.

Download figure to PowerPoint

Identification of hpx genes

Each of the chromosomal DNAs from M1 to M48 was purified and the DNA sequence surrounding the transposon was determined as described in Experimental procedures. Of 48 transposons, 20 were inserted into known hrpB-regulated genes, such as hrpF, hrcV, hpaP, hrcQ, hrcR, hrpV, hrpW, hrpX, hrpY and popC. Five of six transcription units known to be regulated by hrpB were detected (Table 1). This result indicates that our selection strategy can detect hrpB-dependent promoters. Two transposons were inserted into the uncharacterized region between the hrpY and hrpG genes thereby trapping two novel open reading frames (ORFs) (Fig. 3) (DDBJ databank Accession No. AB167955). The first ORF encoded a homologue (46.3% identity, 74.7% similarity) of the HrpY protein, a major component of a Hrp pilus filament, from R. solanacearum (Van Gijsegem et al., 2000) and the second ORF encoded a homologue (49.0% identity, 85.4% similarity) of the HpaB protein, a hrp-associated protein, from Xanthomonas axonopodis pv. glycines (Kim et al., 2003). The first and second ORFs were, respectively, designated hpaZ and hpaB (Fig. 3; Table 1).

Table 1. hrp and popABC operons detected by screening.
Transcription unitGene (fusiona)
  • a

    . Mutants isolated from mini-Tn5lacZYA fusion are shown in two-digit numbers and those from mini-Tn5′ lacZY fusion in three-digit numbers.

  • b

    . In which the transposon is inserted in the opposite direction to the transcriptional direction of hrcC and probably traps readthrough transcription from the popABC promoter.

hrp unit 2hrpF (M27, M30) hrpH (M117) hrpK (M188)
hrp unit 3hrcV (M23, M33, M37, M223) hpaP (M38, M168)
hrp unit 4hrcQ (M42, M165, M211, M212) hrcR (M35, M107) hrpV (M2, M20, M31, M36, M43, M187, M205) hrpW (M45) hrpX (M47)
hrp unit 7hrpY (M175) hpaZ (M46, M134, M220) hpaB (M18)
popABCpopA (M162) popC (M137, M152, M173, M199, M201, M225) hrcC (M32b)

Figure 3. Genetic organization and restriction map of the hrp right end region. Filled arrows represent the new genes identified in this study. Open and filled flags represent the transcriptional and translational fusions of the mini-Tn5lacZYA and mini-Tn5′lacZY transposons respectively. Flag heads show the transcriptional direction of the fusion. Numbers on the flag denote the transposon or mutant itself. A hooked arrow from a filled circle indicates the direction of hrp transcription unit 7 from the PIP box. Thick lines indicate the DNA fragments used in complementation or marker-exchange experiments. Dotted lines represent the position of the deletion in each fragment. Thin arrows indicate polymerase chain reaction (PCR) primers. B, BamHI; E, EcoRI; O, EcoO109I; H, HindIII; K, KpnI; P, PstI; S, SalI; Sm, SmaI; X, XhoI. Restriction sites generated by PCR are indicated by asterisks. Enzyme sites treated with Klenow are shown in parentheses.

Download figure to PowerPoint

The other 26 transposons were inserted outside the hrp cluster, and 14 novel hrpB-regulated genes were identified. These genes were designated hpx. We also sequenced the region upstream or downstream of the hpx genes. From this analysis, we identified four additional hpx genes, namely, hpx4, hpx6, hpx13 and hpx16, respectively, encoded upstream or downstream of hpx3, hpx5, hpx14 and hpx15, whose expressions were also regulated by hrpB. In total, 18 hpx genes were isolated by the promoter screening strategy (Table 2).

Table 2.  Novel hrpB-regulated genes in Ralstonia solanacearum.
GeneDDBJ Accession No.Transcriptional fusionaTranslational fusionbProbable PIP-box motifProtein size (aa)Corresponding ORF in GMI1000 (% identity of protein product)Similar type III-related protein, organism [Accession No.] (blast score, E-value) or comments
  • a

    . Mutants isolated from mini-Tn5lacZYA fusion.

  • b

    . Mutants isolated from mini-Tn5′ lacZY fusion.

  • c

    . Probably comprise an operon with lrpC.

  • d

    . Not predicted as an ORF in GMI1000.

  • NC, not calculated.

hpx1AB177893M5 M44 Yes299RSp0882 (100)No homology
hpx2AB177894M39M193 M194Yes481RSp0879 (98.3)No homology
hpx3AB177895M41 Yes125RSp0848 (100)No homology
hpx4AB177896 M102 M114Yes1335RSp0847 (98.3)No homology, similar to Hpx10
M116 M160
M161 M166
hpx5AB167955M29M190 M204Yes648RSp0842 (100)HpaF, Xanthomonas axonopodis pv. glycines[AAP34359] (97, 1e-18), R. solanacearum PopC-like leucine-rich repeat (LRR) motif (type II)
hpx6AB177897 M154Yes232RSp0841 (100)Hpa2, Xanthomonas oryzae pv. oryzae[AAF61278] (123, 2e-27)
hpx7AB177898M26 M40M113 M128No286RSp0825 (99.3)Probably spermidine synthase
M183 M186
hpx8AB177899M10 Yes228RSp1460 (98.7)No homology
hpx9AB177900M6 Yes878RSp1384 (99.0)No homology
hpx10AB177901M11 M22M104 M195Yes1156RSp1024 (98.2)No homology, similar to Hpx4
hpx11AB177902M3 M13 M24 Yes455RSc0245 (99.3)HolPtoQ, Pseudomonas syringae pv. tomato[AAO54411] (331, 2e-89), probably inocine-uridine-preferring nucleoside hydrolase
hpx12AB178008M48 Yes129dNo homology
hpx13AB178009  Yes621RSc1356 (80.9)LRR motif (type I), hypervariation in LRR domain
hpx14AB178009M21 Yes647RSc1357 (97.2)LRR motif (type I)
hpx15AB178010M25M147 M153Yes462RSc 1800 (99.0)LRR motif (type I)
hpx16AB178010  c538RSc 1801 (98.4)LRR motif (type I)
hpx17AB178011M34M118Yes720RSc 1815 (57.0)AvrBs3 effector family, Xanthomonas spp., large deletion (525 aa) in central repeat domain
hpx18AB178012M4M145Yes306RSc2101 (99.3)No homology
hpx19AB178013 M182Yes180RSp0693 (100)Probably glyoxalase I
hpx20AB178014 M180Yes1035RSp0672 (97.8)LRR motif (type I)
hpx21AB178015 M126 M157Yes411RSp1022 (96.2)No homology
hpx22AB178016 M135No328RSp0206 (98.8)Probably transposase
hpx23AB178017  Yes519RSp1277 (99.4)HopPtoA1, P. syringae pv. tomato[AAO54894] (119, 2e-25)
hpx24AB178018 M109 M115Yes1702RSp1281 (99.2)HolPtoR, P. syringae pv. tomato[AAO54417] (201, 8e-50)
hpx25AB178019 M189 M209Yes641RSp0304 (97.5)HopPtoD1, P. syringae pv. tomato[AAO54410] (233, 1e-59)
M210 M230
hpx26AB178020 M150 M207 M208Yes474RSp1130 (97.3)Probably Nudix hydrolase
hpx27AB178021 M125Yes830RSp0732 (99.5)No homology
hpx28AB178022 M139Yes809RSc2359 (98.6)No homology
hpx29AB178027 M164YesNCRSc1349 (NC)IpaH, Shigella flexneri[AAA26526], IS1424 insertion
hpx30AB178023 M132 M151Yes2498RSc 1839 (96.8)No homology

Further screening of hpx genes using an ORF detection system

For most of the above-mentioned hpx genes that were isolated, only a single insertion was found (Table 2). This suggests that not all the hpx genes were identified by the above promoter screening strategy. Therefore, we further screened hpx genes using another transposon. The transposon mini-Tn5′lacZY contains N-terminal-truncated lacZ and lacY as an operon (′lacZY), and can detect an ORF by in frame ′lacZY fusion (Fig. 1B). When mini-Tn5′lacZY was transposed into RS1085 from pTNP105, the frequency of the appearance of Lac+ Kanr colonies was ≈10−8 per recipient cell. Approximately 24 000 colonies were screened as described above among which 77 showed hrpB-dependent expression. Sequence analysis revealed that 61 of these were independent insertions. Twenty transposons were fused to known hrp and pop genes (Table 1), and 21 were trapped for nine hpx genes, namely, hpx2, hpx4, hpx5, hpx6, hpx7, hpx9, hpx10, hpx15 and hpx17, isolated by the promoter screening strategy (Table 2). The other 21 transposons were inserted into new loci, and 11 new hpx genes were identified (Table 2). An additional hpx gene, hpx23, was identified during the sequence analysis of the flanking region of hpx24. Taking together the results of the two independent mutagenesis, 30 novel hpx genes were isolated in this study.

We measured the β-galactosidase activity of in frame ′lacZ fusion in hpx genes in a hrp-inducing medium and compared it between the wild-type and Δ(hrpB–hrpD) backgrounds. Although the amount of translation products in cells was varied among hpx genes, all of the fusions showed hrpB-dependent induction as expected (Table 3).

Table 3. hrpB-dependent expression of the translation product of hpx genes in a minimal medium.
GeneFusionβ-Galactosidase activityaFold induction
  • a

    .β-Galactosidase activities (Miller's unit ± SD) were measured as described in Experimental procedures and are averages of four experiments.

hpx1hpx1′-′lacZ36.6 ± 3.30.7 ± 0.2 52.2
hpx2M19376.1 ± 9.91.5 ± 0.2 50.7
hpx3hpx3′-′lacZ21.3 ± 1.50.7 ± 0.1 30.4
hpx4M10210.9 ± 1.00.9 ± 0.1 12.1
hpx5M190 318 ± 65.60.5 ± 0.1636
hpx6M15447.4 ± 1.81.0 ± 0.1 47.4
hpx7M183 459 ± 20.40.6 ± 0.1765
hpx8hpx8′-′lacZ 5.8 ± 0.20.6 ± 0.1  9.6
hpx9hpx9′-′lacZ 3.4 ± 0.30.5 ± 0.1  6.8
hpx10M104 6.6 ± 1.80.6 ± 0.1 11.0
hpx11hpx11′-′lacZ20.8 ± 1.00.6 ± 0.1 34.6
hpx12hpx12′-′lacZ 4.8 ± 0.40.6 ± 0.2  8.0
hpx13hpx13′-′lacZ26.0 ± 1.70.7 ± 0.3 37.1
hpx14hpx14′-′lacZ 190 ± 31.41.2 ± 0.2158
hpx15M147 5.1 ± 0.20.5 ± 0.1 10.2
hpx16hpx16′-′lacZ 7.2 ± 0.30.7 ± 0.2 10.2
hpx17M11810.7 ± 1.21.5 ± 0.4  7.1
hpx18M14594.8 ± 7.01.5 ± 0.2 63.2
hpx19M182 453 ± 48.63.8 ± 0.5119
hpx20M180 9.1 ± 0.60.4 ± 0.1 22.7
hpx21M12626.8 ± 6.10.6 ± 0.1 44.6
hpx22M13520.6 ± 2.82.0 ± 0.2 10.3
hpx23hpx23′-′lacZ11.6 ± 0.80.6 ± 0.1 19.3
hpx24M109 165 ± 33.00.2 ± 0.1825
hpx25M189 309 ± 81.23.0 ± 0.3103
hpx26M150 296 ± 41.01.2 ± 0.2246
hpx27M12527.3 ± 7.71.4 ± 0.2 19.5
hpx28M13929.9 ± 4.31.3 ± 0.6 23.0
hpx29M164 5.2 ± 0.30.5 ± 0.2 10.4
hpx30M132 4.4 ± 1.00.5 ± 0.2  8.8

Sequence features of hpx genes

Most of the known hrpB-regulated operons in R. solanacearum contain a perfect or imperfect plant-inducible promoter (PIP) box motif (consensus TTCGC-N15-TTCGC) in the promoter region of their leading genes (Table 4). In all of the hpx genes isolated, except for hpx7 and hpx22, a PIP box-like sequence was identified in their putative promoter regions (Tables 2 and 4).

Table 4.  Probable PIP box motif sequences identified in the putative promoter regions of hrpB-regulated genes.
GenePositionPIP box and −10 sequencea
  • a

    . Position indicates position of the first nucleotide (T) of the first TTCGC motif of a PIP box relative to the start codon of an ORF. N means the number of spacer nucleotides between the first and second TTCGC motifs of a PIP box or between the second motif of a PIP box and a −10 sequence (consensus 5′-YANNRT-3′). Conserved nucleotides are shown in bold.

  • b

    . A leading gene in the known hrpB-regulated operon is shown: hrpK (hrp unit 2), hrcU (hrp unit 3), hrcQ (hrp unit 4), hrpY (hrp unit 7) and popA (popABC operon).

Known genesb
Novel hrpB-regulated genes

We compared sequences of hpx genes with published sequences in GenBank. As summarized in Table 2, seven hpx genes, namely, hpx6, hpx11, hpx17, hpx23, hpx24, hpx25 and hpx29, were found to encode putative type III effectors or type III-related proteins from other animal and plant pathogens. Four hpx genes, namely, hpx7, hpx19, hpx22 and hpx26, encoded known enzymes. Interestingly, six hpx genes, namely, hpx5, hpx13, hpx14, hpx15, hpx16 and hpx20, were found to encode proteins resembling the eukaryotic leucine-rich repeat (LRR) protein, these were divided into two types on the basis of the LRR consensus motif. The other 13 hpx products did not show any significant sequence homology to known proteins.

We also compared amino acid sequences of products of hpx genes with corresponding genes from R. solanacearum strain GMI1000. As shown in Table 2, most of the hpx products showed more than 96% identity to those from GMI1000. In a few hpx genes, various types of mutation, such as hyper-point mutation (hpx13), intragenic deletion (hpx17) or insertion of an IS sequence (hpx29), were observed in these genes.

Phenotypes of hpaZ, hpaB and hpx mutants on plants

We identified two new hrpB-regulated genes, hpaZ and hpaB, downstream of hrpY (Fig. 3). Because the role of these two genes in Hrp TTSS has not yet been characterized in R. solanacearum (although they have also been identified by the genome sequencing of GMI1000; Salanoubat et al., 2002), we examined the phenotypes of mutants on plants. For pathogenicity tests, Tn fusions were transduced into the EPS-producing (EPS+) strain RS1002. RS1252, which has Tn-M46 insertions in hpaZ, elicited HR on non-host tobacco plants (Fig. 4A) and caused diseases on host eggplants (Fig. 4B) as the wild type. Other mutants with a Tn insertion in hpaZ showed the same results (data not shown). On the other hand, RS1253, which has a Tn-M18 insertion in hpaB, lost both ability to induce HR (Fig. 4A) and ability to cause diseases (data not shown). The ΔhpaB mutant, RS1211, showed the same phenotypes (Fig. 4A and B). To confirm this result, we carried out a complementation analysis of the ΔhpaB mutant using a transposon vector, because the IncQ plasmid used in hrpB complementation tests was unstable in R. solanacearum when the transformants were inoculated into plants (data not shown). The loss of hypersensitivity and pathogenicity of the ΔhpaB mutant was completely recovered by a transposon carrying the 1.5 kb EcoRI fragment containing hrpY, hpaZ and hpaB with a PIP box, a probable recognition sequence of HrpB, but not by a transposon carrying the 1.3 kb SmaI–EcoRI fragment without the PIP box (Figs 3 and 4). These results indicate that hpaB but not hpaZ is essential for both hypersensitivity and pathogenicity on plants. In addition, the complementation analysis showed that hpaZ and hpaB are co-transcribed with hrpY as the hrp transcription unit 7 (Fig. 3). We also monitored the growth of the ΔhpaB mutant in the stem of host eggplants. The ΔhpaB mutant, as well as the ΔhrpY mutant, did not multiply efficiently in the host plants (Fig. 4C). This indicates that the loss of pathogenicity in the ΔhpaB mutant was caused by a growth deficiency in the host plants.


Figure 4. HR and pathogenicity tests. A. Response of non-host tobacco leaf infiltrated with a suspension (≈5 × 108 cfu ml−1) of R. solanacearum mutants. Sites: 1, RS1002 (wild type); 2, RS1227 (hrpY::Tn-M175); 3, RS1252 (hpaZ::Tn-M46); 4, RS1253 (hpaB::Tn-M18); 5, RS1211 (ΔhpaB); 6, RS1211/Tn hrp unit 7; 7, RS1211/Tn hrp unit 7 ΔPIP; and 8, distilled water (DW). The photograph was taken 36 h after infiltration. B. Disease symptoms of leaves inoculated with R. solanacearum mutants by the leaf-cutting method (see Experimental procedures). R. solanacearum strains used in this experiment are the same as those shown in (A). The photographs were taken 5 days after inoculation. C. Growth of R. solanacearum mutants in stem of eggplants. A set of five plants were inoculated with 105 cells of R. solanacearum strain: RS1002 (wild type), filled circle; RS1211 (ΔhpaB), filled square; RS1273 (ΔhrpY), filled triangle. Similar results were obtained in three independent trials. Bars indicate standard deviations.

Download figure to PowerPoint

We further examined the effect of Tn insertions in hpx genes on plants; however, we could not obtain mutants that fail to induce HR on non-host tobacco plants or markedly attenuate their pathogenicity on host eggplants (data not shown).

Hrp pili are still produced on the cell surface of a hpaB deletion mutant

Ralstonia solanacearum hpaB mutants were found to exhibit a typical Hrp-defective phenotype, that is, the loss of HR induction and pathogenicity on plants (Fig. 4), suggesting that hpaB encodes an essential component of the Hrp type III secretion apparatus. In contrast to R. solanacearum, however, hpaB mutants of X. axonopodis pv. glycines and X. campestris pv. vesicatoria were reported to lose their pathogenicity but retain the ability to elicit HR on non-host plants, which was the basis for the designation of hpa (hrp-associated) (Büttner and Bonas, 2002; Kim et al., 2003). To investigate the role of hpaB in Hrp TTSS in R. solanacearum, the production of Hrp pili in various mutant derivatives of EPS strain RS1085 was examined by transmission electron microscopy (TEM). When cells were grown in the hrp-inducing medium, three types of filamentous extracellular appendage were observed on the cell surface of wild-type strain RS1085. One was thick and 13–15 nm in diameter (Fig. 5A), and is considered to be the flagellum (Tans-Kersten et al., 2001) because it disappeared in the fliC mutant (data not shown). Other filaments were thin, and divided into two types. The first type of filament was 6–7 nm in diameter, and formed long and straight bundles (Fig. 5A). The other type was thinner, and is considered to be the type IV pilus (Kang et al., 2002) because it disappears in the pilA mutant (data not shown) but is still produced in a hrp-defective mutant (Fig. 5B). On the cell surface of the RS1085 fliC pilA derivative (RS1275), only long, straight and bundle-formed filaments were observed (Fig. 5C and D). Such appendages were not observed in the hrpY mutant (Fig. 5E) nor other hrp-defective mutants such as hrcC or hrcQ (data not shown). From these results and the diameter of the filaments, we referred to the thin extracellular filaments as the Hrp pili reported by Van Gijsegem et al. (2000). In the ΔhpaB mutant, bundle-like pili were also observed on the cell surface (Fig. 5F and G). However, no such structures were observed in the ΔhpaBΔhrpY double mutant (Fig. 5H). This result indicates that the ΔhpaB mutant still produces Hrp pili on the cell surface as the wild type.


Figure 5. Transmission electron microscopy (TEM) of extracellular appendages of R. solanacearum strains cultivated on hrp-inducing medium. hrp+ strain RS1085 (A and B), fliC pilA strain RS1275 (C and D), RS1275 ΔhrpY (RS1276) (E), RS1275 ΔhpaB (RS1277) (F and G) and RS1275 ΔhpaBΔhrpY (RS1278) (H). The boxed square region in (C) and (F) are magnified and shown in (D) and (G) respectively. In (A) and (B), extracellular structures that we refer to as flagella (arrowheads), Hrp pili (arrow) and type IV pili (double arrowheads) are shown with marks. Bars in the micrographs, except (A), (B), (D) and (G), represent 1 µm. Bars in (A), (B), (D) and (G) represent 100 nm.

Download figure to PowerPoint


  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

In this study, we performed genomewide screening of hrpB-regulated genes in R. solanacearum by using a transposon-based system. We identified 30 novel hpx genes and two new hrp genes, namely, hpaZ and hpaB (Tables 1 and 2). Because our system uses β-galactosidase (LacZ) and β-galactoside permease (LacY) activities in selection, only an exoconjugant with a lacZYA transcriptional fusion under the control of an active promoter or with a ′lacZY in frame fusion into an ORF forms a colony on selective plates (Fig. 1). This system enables the large-scale screening of target genes, because only a small percentage of transconjugants could detect active genes when only β-galactosidase (LacZ) activity was used in selection (data not shown). However, our screening is not saturated yet, because one half of the known hrp and hrc genes were not detected by the screening (Table 1) and one half of the hpx genes were trapped by a single insertion (Table 2). Nevertheless, adding the above 32 genes to known hrp, hrc and pop genes, hrpB seems to control a large virulence regulon composed of more than 60 genes in R. solanacearum.

A perfect or imperfect PIP box motif (consensus TTCGC-N15-TTCGC) was found in putative promoter regions of most hpx genes as in known hrpB-dependent genes (Tables 2 and 4). The PIP box motif was first identified in X. campestris pv. vesicatoria as a probable recognition sequence of the hrp-regulatory protein HrpX (Fenselau and Bonas, 1995) and often found upstream of HrpX-dependent genes, such as the hrp, avr and recently isolated xop genes (Wengelnik and Bonas, 1996; Nöel et al., 2001). HrpX is also an AraC-type transcriptional activator and highly homologous to HrpB of R. solanacearum (Wengelnik and Bonas, 1996). Therefore, HrpB, similar to HrpX, presumably recognizes PIP box motif-like sequences as the target motif. The interaction of HrpB with these motifs should be elucidated genetically and biochemically in the future. As exceptions, we found no PIP box-like motifs upstream of hpx7, which was independently isolated in our screen six times, and hpx22 (Table 2). In these cases, HrpB may indirectly regulate the expression of hpx7 and hpx22.

Recently, Cunnac et al. (2004) have identified the 114 PIP box containing genes in the genome of R. solanacearum GMI1000 from computer analysis. Although their hrpB dependency was under investigation, they included most of the hpx genes identified in our study. However, the hpx gene that has no PIP box as mentioned above (hpx7 and hpx22), or that has an imperfect PIP box (hpx17 and hpx18) and a perfect PIP box far upstream of the start codon (hpx19), was missed in the in silico screen. This suggests that our functional screen is useful in the analysis of unsequenced bacteria.

It is noteworthy that five hpx genes encoded known type III effector homologues (hpx17, AvrBs3 effector family of Xanthomonas species (spp.); hpx23, HopPtoA1 of P. syringae pv. tomato; and hpx25, HopPtoD1 of P. syringae pv. tomato) and candidate effector homologues (hpx11, HolPtoQ of P. syringae pv. tomato; and hpx24, HolPtoR of P. syringae pv. tomato) isolated from other plant pathogens (Table 2). Although these effector homologues have been identified in the whole-genome DNA sequencing of R. solanacearum strain GMI1000 (Genin and Boucher, 2002), we found that they are controlled by HrpB. These effector homologues presumably act as type III effector proteins also in R. solanacearum and translocate into host plant cells via Hrp TTSS in which they exhibit their virulence functions. In P. syringae, an alternative sigma factor is used to control Hrp TTSS-related genes whereas an AraC-type transcriptional activator is used in both R. solanacearum and X. campestris (Alfano and Collmer, 1997). The above-mentioned effectors, widely distributed among phytopathogenic bacteria overcoming such a difference between their regulatory systems, may have an important role in pathogenicity. The product of one of the hpx genes, hpx29, was found to share homology with Shigella flexneri type III effector IpaH (Hartman et al., 1990). Whereas our R. solanacearum strain RS1000 contains IS insertion in this gene, this type of pathogenicity gene may be utilized commonly in both animal and plant pathogens.

Most of the known effector proteins act inside host eukaryotic cells, in which they modulate host cellular functions in order to successfully infect. Thus, one of the important characteristics of effector proteins is that they contain eukaryotic protein motifs. In this study, we identified six hpx genes encoding two types (I and II) of LRR protein (Table 2). The LRR motif is a typical protein motif commonly observed in eukaryotic proteins but is unusual in bacterial proteins (Kobe and Deisenhofer, 1994). Only three groups of bacterial LRR proteins have been reported to date, the IpaH/YopM effector family from Shigella and Yersinia, internalins from Listeria, and PopC-like LRR proteins from R. solanacearum and Xanthomonas spp.; these three groups were shown to be pathogenicity determinants (Kobe and Deisenhofer, 1994; Guéneron et al., 2000; da Silva et al., 2002). Interestingly, the type I LRR protein was found to belong to a novel fourth group of bacterial LRR proteins (data not shown), and the type II LRR protein has a PopC-type motif. In the genome of R. solanacearum GMI1000, moreover, two copies of a gene encoding the type I LRR protein are present in other loci (data not shown). The amplification of copy number is also one of the characteristics of effector genes, as noted in the avrBs3 gene family of Xanthomonas spp. (Leach and White, 1996).

Our screening revealed that products of two hpx genes, hpx4 and hpx10, share high homology with each other (20.3% identity, 50.6% similarity) (Table 2). In the genome of GMI1000, three copies of this type of gene were found in other loci, two of which contained a PIP box-like sequence in upstream regions (data not shown). Furthermore, hpx30 is also a member of another gene family in R. solanacearum. Five genes similar to hpx30 were found in the genome of GMI1000, three of which have a PIP box-like sequence (data not shown). Thus, members of these families are also potential effector genes.

As mentioned above, our genomewide screening showed that more than 20 candidate effector genes are present in the genome of R. solanacearum. Moreover, some of the effector genes are considered to be not regulated by HrpB, such as the popP1 gene (Lavie et al., 2002). R. solanacearum has a broad host range and causes diseases in more than 100 plant species (Hayward, 1991). The large number of virulence effector genes present in R. solanacearum may be one of the causes.

We identified four hpx genes encoding known catalytic enzymes, namely, spermidine synthase (hpx7), glyoxalase I (hpx19), transposase (hpx22) and Nudix hydrolase (hpx26) (Table 2). These proteins, except for transposase, may function in this bacterium not as effectors but to protect the pathogen from damage caused by oxidative stresses or toxic compounds in plants during infection. It would be interesting to clarify whether these enzyme homologues are secreted outside the cell via Hrp TTSS.

It is possible that other hpx genes products of which show no significant homology to known proteins encode novel type III effector proteins. As in many other avr genes of phytopathogenic bacteria, null mutation in hpx genes tested in this study showed no particular changes in pathogenicity on 4-week-old eggplants. However, it should be considered that each hpx gene may have a functional redundancy and may be required for the full virulence of R. solanacearum. To examine this possibility, we are now constructing R. solanacearum mutants that have a single disruption or multiple disruptions in hpx genes and comparing in detail their pathogenicity with that of the wild type using 8-week-old eggplants, which are more tolerant to bacterial wilt than 4-week-old eggplants. Our recent study revealed that both types of LRR protein are required for the full virulence of R. solanacearum (Y. Murata and T. Mukaihara, in preparation).

Sequence comparison between hpx genes from our Japanese isolate RS1000 (race 1 biovar 4) and strain GMI1000 (race 1 biovar 3) suggested that the two R. solanacearum strains are closely related to each other in the DNA level (Table 2). In addition, within the sequenced range, the genomic context of hrp genes and the flanking region of hpx genes were the same as GMI1000 (data not shown). The accumulation of mutations, such as hyper-point mutation (hpx13), large deletion (hpx17) and IS insertion (hpx29), was observed in several hpx genes in RS1000 (Table 2), suggesting that a certain selective pressure is present in candidate effector genes in the divergence of R. solanacearum strains.

Some of the hpx products were highly expressed in cells under the hrp-inducing condition (Table 3). These include the homologues of HolPtoR (Hpx24) and HopPtoD1 (Hpx25), one of the type I LRR proteins (Hpx14), the PopC-like type II LRR protein (Hpx5), Hpx2, Hpx18 and three enzyme homologues as mentioned above. It would be interesting to clarify the relationship between the amount of product and the contribution to pathogenicity of hpx genes or to examine whether such a difference in the amount of hpx products is common among R. solanacearum pathogenic variants.

We identified two new hrpB-regulated genes, hpaZ and hpaB, that are encoded downstream of hrpY and co-transcribed with hrpY as the hrp transcriptional unit 7 (Table 1; Fig. 3). hpaZ was found to encode a homologue of HrpY, the major component of a Hrp pilus in R. solanacearum (Van Gijsegem et al., 2000), but not required for pathogenicity and hypersensitivity on plants (Fig. 4). This indicates that HpaZ is not an essential component of Hrp TTSS of R. solanacearum. On the other hand, hpaB was found to be required for pathogenicity on host eggplants and induction of HR on non-host tobacco plants (Fig. 4). hpaB is also conserved in the hrp gene cluster of Xanthomonas spp. (Büttner and Bonas, 2002; Kim et al., 2003) and the two genes share a high homology (49.0% identity and 85.4% similarity), suggesting that the two genes have the same function. In contrast to R. solanacearum, however, it was suggested that hpaB is required for pathogenicity on host plants but not for induction of HR on non-host plants (Büttner and Bonas, 2002; Kim et al., 2003). The difference in the effect of hpaB null mutation on plant responses between the two bacterial species suggests that HpaB is not the core component of the Hrp secretion apparatus.

We partially characterized the effect of hpaB mutation on Hrp TTSS in R. solanacearum using TEM. We observed that Hrp-dependent pili are still produced on the cell surface of a hpaB mutant (Fig. 5). hpaB is the first hrp gene that is required for the Hrp function but dispensable for the production of Hrp pili in R. solanacearum. The mutant phenotype suggests that HpaB may be involved in certain Hrp subfunctions in secretion steps, for example, the translocation of effector proteins into plant cells as a so-called translocon protein or the secretion of a specific Hrp substrate protein as its chaperone protein. To clarify the exact role of HpaB in Hrp TTSS in R. solanacearum, we are now examining the secretion of several effector proteins in hpaB mutants.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

Bacterial strains and media

The bacterial strains used in this study are listed in Table 5. All of the R. solanacearum strains are derivatives of RS1000 that was renamed from the Japanese local isolate 8216 (race 1 biovar 4). RS1002 is a spontaneous Nalr derivative of RS1000. Escherichia coli S17-1 (Simon et al., 1983) was used as a donor for oriT/mob plasmids in conjugation experiments. JM109 was used for plasmid construction. The BG medium (Boucher et al., 1985) was used for the growth of R. solanacearum strains at 28°C and also used as the hrp-repressing medium. One-fourth strength of the M63 glucose minimal medium (Miller, 1992) was used as the hrp-inducing medium (Arlat et al., 1992), except for β-galactosidase assay. The E. coli strains were grown at 37°C in the LB medium (Miller, 1992). Antibiotics were used at the following concentrations: ampicillin (Ap), 50 µg ml−1; gentamicin (Gm), 20 µg ml−1; kanamycin (Km), 20 µg ml−1; nalidixic acid (Nal), 30 µg ml−1; spectinomycin (Sp), 40 µg ml−1; and streptomycin (Sm), 50 µg ml−1.

Table 5.  Bacterial strains.
R. solanacearum
RS1000Wild typeMAFF730103 = Japanese local isolate 8216 (race 1 biovar 4)
RS1002RS1000 NalrThis study
RS1016RS1002 Δ(hrpB–hrpD)::GmrThis study
RS1085RS1002 ΔepsPThis study
RS1087RS1085 popA–lacZYAThis study
RS1088RS1085 popA–lacZYAΔ(hrpB–hrpD)::GmrThis study
RS1096RS1085 hpx13′-′lacZThis study
RS1097RS1085 hpx14′-′lacZThis study
RS1099RS1085 hpx16′-′lacZThis study
RS1101RS1085 hpx1′-′lacZThis study
RS1103RS1085 hpx3′-′lacZThis study
RS1105RS1085 hpx11′-′lacZThis study
RS1106RS1085 hpx9′-′lacZThis study
RS1107RS1085 hpx8′-′lacZThis study
RS1108RS1085 hpx23′-′lacZThis study
RS1111RS1085 hpx12′-′lacZThis study
RS1211RS1002 ΔhpaBThis study
RS1212RS1085 ΔhpaBThis study
RS1227RS1002 hrpY::Tn-M175This study
RS1252RS1002 hpaZ::Tn-M46This study
RS1253RS1002 hpaB::Tn-M18This study
RS1254RS1085 ΔhrpYThis study
RS1255RS1085 ΔhpaBΔhrpYThis study
RS1265RS1085 fliC::pRS845This study
RS1267RS1085 ΔhrpY fliC::pRS845This study
RS1268RS1085 ΔhpaB fliC::pRS845This study
RS1269RS1085 ΔhpaBΔhrpY fliC::pRS845This study
RS1273RS1002 ΔhrpYThis study
RS1275RS1085 fliC::pRS845 pilA::GmrThis study
RS1276RS1085 ΔhrpY fliC::pRS845 pilA::GmrThis study
RS1277RS1085 ΔhpaB fliC::pRS845 pilA::GmrThis study
RS1278RS1085 ΔhpaBΔhrpY fliC::pRS845 pilA::GmrThis study
E. coli
JM109recA1 supE44 endA1 hsdR17 gyrA96 relA1Δ(lac-proAB) F′[traD36 proAB+lacIqlacZΔM15]Yanisch-Perron et al. (1985)
S17-1thi pro hsdRhsdM+recA[chr::RP4-2-Tc::Mu-Km::Tn7]Simon et al. (1983)

Construction of lacZ fusion transposons

Promoter-probing transposon mini-Tn5lacZYA (Fig. 1A) was constructed from pRS415 (Simons et al., 1987), pHRP315 (Parales and Harwood, 1993) and pBSL204 (Alexeyev et al., 1995) as follows. A PstI–EcoRI cassette containing the Sm/Spcr gene from pHRP315 was introduced into pRS415. The resulting plasmid pQV111 was linearized with SmaI and inserted into pBSL204 to yield pTNP104. ORF-detecting transposon mini-Tn5′lacZY (Fig. 1B) was constructed from placZY2 (Jain, 1993) and pBSL118 (Alexeyev et al., 1995). placZY2 was linearized with BamHI and SmaI and then inserted into pBSL118 to yield pTNP105. The backbones of both pTNP104 and pTNP105 are the pBSL plasmids (Alexeyev et al., 1995).

Screening of hrpB-regulated genes in R. solanacearum

Isolation of transcriptional and translational fusions. R. sol-anacearum RS1085 and E. coli S17-1 harbouring plasmids carrying mini-Tn5 derivatives were mated on a nitrocellulose membrane filter on BG plates at 28°C overnight. Then, samples were spread and incubated on 1/4 M63 lactose minimal plates containing Nal, Sm and Sp (for mini-Tn5lacZYA) or Nal and Km (for mini-Tn5′lacZY) at 28°C. Antibiotic-resistant and Lac+ transconjugants appeared after 5 days (in mini-Tn5lacZYA) or 8 days (in mini-Tn5′lacZY) of incubation. Lac+ colonies were once patched on BG plates using the head of a toothpick and incubated at 28°C for 2 days. Then, colonies were replica-plated on the BG and 1/4 M63 glucose minimal plates, each containing X-Gal (100 µg ml−1). The expression profile of fusions on each plate was evaluated by comparing their β-galactosidase activities after overnight incubation (in BG) and 2 days (in minimal medium) of incubation at 28°C.

Construction of ΔhrpB mutants and complementation by a hrpB+ plasmid.  The ΔhrpB derivatives of transconjugants or plasmid-based fusions were constructed by the transformation of RS1016 chromosomal DNA into transconjugants or fusions. To complement for the ΔhrpB strains derived from mini-Tn5lacZYA fusions, the Kmr plasmid pRS200 carrying the wild-type hrpB gene was used. pRS200 was constructed as follows. The 2.2 kb NspI fragment containing the wild-type hrpB gene was subcloned from pRS103 into pMCL200 (Nakano et al., 1995), yielding pRS142. The hrpB fragment was excised with BamHI and KpnI and inserted into pDSK519 (Keen et al., 1988) to produce pRS200. For the complementation of ΔhrpB mutants from mini-Tn5′lacZY fusions, the Smr/Spr IncQ plasmid pRS765 carrying the wild-type hrpB gene was used. pRS765 was constructed by ligating pRS142 with the EcoRI–PstI fragment carrying the oriV oriT/mob rep region from RSF1010 (Scholz et al., 1989).

Transformation of R. solanacearum genomic loci by electroporation

To prepare R. solanacearum competent cells, 0.5 ml of the BG medium overnight culture was centrifuged, and the resulting cell pellet was washed twice with 500 µl of 10% glycerol solution and finally resuspended in 100 µl of 10% glycerol. R. solanacearum chromosomal DNA (2 µg) purified with a Genomic DNA Extraction kit (QIAGEN) was mixed with 100 µl of competent cells and electroporated using a cell porator (BIO-RAD Gene Pulser) under the conditions of 2.0 kV, 200 Ω and 25 µF. The cells were resuspended in the BG medium, incubated for 1 h at 28°C and spread on selection plates.

Complementation by transposon vectors

A 1.5 kb EcoRI fragment containing the hrpY, hpaZ and hpaB genes was subcloned from pRS103 into pHSG398 to yield pRS783. pRS783 was linearized with BamHI and inserted into pBSL118. The resulting pRS784 was transformed into S17-1 and the transposon was transposed into the genome of R. solanacearum mutants by conjugation. pRS786, in which the transposon carries a 1.3 kb SmaI–EcoRI fragment containing hrpY, hpaZ and hpaB but lacks the PIP box element, was constructed similarly.

Cloning of hpx genes

Each chromosomal DNA from Tn insertion mutants was digested and self-ligated for the transformation of JM109. Because each transposon contains the ColE1-type origin, when chromosomal DNA was digested with an appropriate restriction enzyme, the R. solanacearum DNA region surrounding the transposon can be rescued in E. coli. In mini-Tn5lacZYA, EcoRI or KpnI was used to clone the upstream region of transposon, and SmaI or SacI was used for the downstream rescue. In mini-Tn5′lacZY, EcoRI, KpnI or SmaI was used for the upstream rescue and BamHI or NotI for the downstream rescue.

β-Galactosidase assay

The β-galactosidase assay was performed as described by Miller (1992) with some modifications. Cells of R. solanacearum strains were cultured in the BG medium for 16 h at 28°C and collected by centrifugation. The cell pellets were washed twice in distilled water (DW) and resuspended in 1:100 1/4 M63 glutamate minimal medium. The cultures were incubated for 16 h at 28°C and used for the assay.

Plant tests

All plants used in this study were grown on plastic pots containing commercial soil in a greenhouse. To test for HR, R. solanacearum strains were cultured in the BG medium for 16 h at 28°C and centrifugated. The pellets were washed twice in DW and resuspended in DW to yield an OD600 of 0.5 (≈5 × 108 cfu ml−1). The bacterial solution was infiltrated using a needle-less syringe into 4-week-old tobacco (cultivar Xanthi) leaves, and the induction of HR was examined at 36 h after inoculation.

For pathogenicity tests on eggplants (Solanum melongena cultivar senryo-nigou), a leaf-cutting method was applied as follows. A pair of scissors was dipped into a bacterial suspension (≈5 × 108 cfu ml−1), and one-third of a 4-week-old eggplant leaves was cut from the edge. Disease symptoms appearing on the leaves were examined for 5 days after inoculation.

For the tests on the recovery of bacterial cells from inoculated plants, 4-week-old eggplants were cut at the stem above the fourth leaf and were inoculated by depositing a 10 µl droplet containing 105R. solanacearum cells in DW to the cutting surface. The inoculated plants were grown at 28°C in a greenhouse and a 1-cm-long sample of the stem below the inoculation site was obtained daily. The sample was macerated in 1 ml of sterile DW and allowed to stand for 3 h. Then, the number of bacteria in the ooze leaking from the vascular system of the stem was determined by the plating method. Each R. solanacearum strain was tested in three independent experiments using a set of five plants.

Observation of Hrp pili by TEM microscopy

Bacterial cultures were diluted with DW, and the diluent was deposited on collodion membrane-coated carbon grids, stained three times with 0.4% uranyl acetate filtered twice through a 0.2 µm membrane, left to dry, and examined under a Hitachi H-7500 electron microscope (Hitachi) at 80 KV.


  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

This work was supported in part by a Grant-in-Aid for Scientific Research on Priority Areas (A) (No. 15028218) to T.M. from the Ministry of Education, Culture, Sports, Science and Technology, Japan.


  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

Appendix A1. Construction of R. solanacearum mutants.

MMI_4328_sm_AppA1.doc29KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.