R alstonia solanacearum type III secretion system effector Rip36 induces a hypersensitive response in the nonhost wild eggplant S olanum torvum

Summary

Ralstonia solanacearum is a Gram‐negative soil‐borne bacterium that causes bacterial wilt disease in more than 200 plant species, including economically important Solanaceae species. In R. solanacearum, the hypersensitive response and pathogenicity (Hrp) type III secretion system is required for both the ability to induce the hypersensitive response (HR) in nonhost plants and pathogenicity in host plants. Recently, 72 effector genes, called rip (Ralstonia protein injected into plant cells), have been identified in R. solanacearum  RS1000. RS1002, a spontaneous nalixidic acid‐resistant derivative of RS1000, induced strong HR in the nonhost wild eggplant Solanum torvum in an Hrp‐dependent manner. An Agrobacterium‐mediated transient expression system revealed that Rip36, a putative Zn‐dependent protease effector of R. solanacearum, induced HR in S. torvum. A mutation in the putative Zn‐binding motif (E149A) completely abolished the ability to induce HR. In agreement with this result, the RS1002‐derived Δrip36 and rip36E149A mutants lost the ability to induce HR in S. torvum. An E149A mutation had no effect on the translocation of Rip36 into plant cells. These results indicate that Rip36 is an avirulent factor that induces HR in S. torvum and that a putative Zn‐dependent protease motif is essential for this activity.

Ralstonia solanacearum is a Gram‐negative soil‐borne bacterium that causes bacterial wilt disease in more than 200 plant species in over 50 plant families, including many economically important crops, such as potato, tomato, tobacco, banana and eggplant, worldwide (Hayward, 1991; Lebeau et al., 2010). The pathogen enters host plants through natural openings or wounds in the root system, penetrates the xylem vessels, colonizes the root cortex and vascular parenchyma, proliferates and spreads throughout the vascular system, and produces a large amount of exopolysaccharide (EPS) which blocks water transport, resulting in the wilting and death of infected plants (Genin and Denny, 2012; Schell, 2000). In addition, R. solanacearum secretes plant cell wall‐degrading enzymes, such as endopolygalacturonase and endoglucanase, which break down pectins and cellulosic glucans in the cell wall (Gonzalez and Allen, 2003; Liu et al., 2005).

Numerous Gram‐negative plant pathogenic bacteria use the Hrp type III secretion system (T3SS) encoded by a cluster of approximately 20 hypersensitive response (HR) and pathogenicity (hrp) genes (Alfano and Collmer, 1997; Galan and Wolf‐Watz, 2006; Tang et al., 2006; Van Gijsegem et al., 2000). Basically, Hrp T3SS acts as a specialized system for the injection of virulence factors, so‐called effector proteins, from pathogens into host plant cytoplasm. Generally, effector proteins modulate host cellular functions to establish infection and promote propagation in plants. Recent studies have unveiled biochemical activities of several T3SS effectors, including proteases, phosphatases, ubiquitin ligases, ribosyltransferases, transcriptional activators and phosphothreonine lyases (Dean, 2011; Deslandes and Rivas, 2012; Grant et al., 2006; Hann et al., 2010; Staskawicz et al., 2001; Zhou and Chai, 2008). Like many other plant pathogenic bacteria, Hrp T3SS is essential for the growth of R. solanacearum in host plants, as Hrp T3SS‐deficient mutants completely lose their pathogenicity, and effector proteins collectively contribute to virulence, as the mutation of a single effector gene produces little or no effect in most cases (Boucher et al., 1987; Cunnac et al., 2004; Kanda et al., 2003; Mukaihara et al., 2004; Zolobowska and Van Gijsegem, 2006). In R. solanacearum strain GMI1000, 74 T3SS effector proteins have been identified by in silico analysis of the whole genome sequence and functional genomic approaches (Cunnac et al., 2004; Poueymiro and Genin, 2009). To date, three effector proteins, namely AvrA, PopP1 and PopP2, have been shown to act as avirulence factors in nonhost plants. For example, AvrA is responsible for HR elicitation in Nicotiana tabacum and N. benthamiana (Carney and Denny, 1990; Poueymiro et al., 2009; Robertson et al., 2004). PopP1, a member of the YopJ/AvrRxv (C55) family of cysteine proteases, is the major HR elicitor in Petunia St40 (Lavie et al., 2002) and N. glutinosa (Poueymiro et al., 2009). PopP2, another member of the YopJ‐like family, confers RRS1‐R‐mediated resistance on Arabidopsis (Deslandes et al., 2003). Mutation of the conserved cysteine residue in the catalytic triad of PopP2 abolishes its ability to induce RRS1‐R‐mediated resistance in Arabidopsis (Tasset et al., 2010). Some of the above‐mentioned effector proteins have been shown to contribute to virulence in host plants. For example, AvrA is required for the early stage of root infection of the legume Medicago truncatula (Turner et al., 2009) and bacterial fitness on tomato leaves (Macho et al., 2010). Macho et al. (2010) have also shown that PopP2 is required for bacterial fitness on eggplant or bean. Thus, effector proteins play important roles in the interaction between R. solanacearum and its host/nonhost plants.

Plants have evolved the ability to recognize potential pathogens by detecting conserved microbial molecules, such as bacterial flagellin, elongation factor and peptidoglycans, called pathogen/microbe‐associated molecular patterns (PAMPs/MAMPs), leading to the activation of the basal defence response, called PAMP‐triggered immunity (PTI) (Monaghan and Zipfel, 2012). Subsequently, pathogens developed effector proteins to suppress PTI for successful infection. Furthermore, plants developed a system for the recognition of bacterial effectors through the sensing of their virulence by the product of the resistance (R) gene. Therefore, effectors are virulence factors which promote disease in compatible interactions, but function as avirulence factors, which induce resistance responses, in incompatible interactions. Understanding the mechanism by which an Avr protein is recognized by the corresponding R protein may lead to the identification of the virulence function of the effector protein and the detection system in plants, and contribute to the engineering of more durable resistance in crops.

In R. solanacearum strain RS1000, 72 T3SS effector proteins, the so‐called Rip (Ralstonia protein injected into plant cells) proteins, have been identified using a calmodulin‐dependent adenylate cyclase (Cya) reporter system, and their translocations into plant cells have been demonstrated (Mukaihara et al., 2010). It has been shown that the wild eggplant Solanum torvum is highly tolerant to most R. solanacearum strains (Clain et al., 2004; Gousset et al., 2005). Hence, this plant species is widely used as a rootstock for eggplant or tomato cultivation. Infiltration of R. solanacearum RS1002, a spontaneous nalixidic acid‐resistant derivative of RS1000, into the leaves of S. torvum Sw. cv. Torubamubiga induced HR, a rapid cell death in the infiltrated area (Fig. 1A). To identify the RS1000 effector protein eliciting HR, we cloned the entire coding region of each rip gene into a binary vector pEl2Ω‐MCS (Fig. 2; Ohtsubo et al., 1999) under the control of the high expression promoter 35S‐Ω. Agrobacterium tumefaciens strain GV3101 was transformed by the resultant plasmids, and then each Rip was transiently expressed in the leaves of S. torvum and tobacco (N. tabacum L. cv. Xanthi NC) via agroinfiltration. We tested 64 RS1000 effectors and found that Rip36 specifically induced HR in S. torvum leaves, but not in tobacco leaves (Fig. 2B–D). Transient expression of AvrA, an avirulence protein in tobacco leaves (Carney and Denny, 1990; Poueymiro et al., 2009; Robertson et al., 2004), specifically induced HR only in tobacco leaves, indicating that the system worked effectively.

Figure 1.

Hypersensitive responses (HRs) induced by inoculation of Ralstonia solanacearum wild‐type (WT) RS1002 and its derivatives in Solanum torvum  Torubamubiga (A, C) and tobacco (B) leaves. Leaves were infiltrated with bacterial suspensions (OD ₆₀₀ of 0.3) of R. solanacearum  RS1002 (WT), RS1662 (Δrip36), RS1668 (rip36E149A), RS1650 (avrA::Gm^r) or RS1273 (ΔhrpY), or the complemented strains RS1672 (Δrip36  Tnrip36 ⁺) or RS1673 (Δrip36  Tnrip36E149A), and incubated under a 16 h light/8 h dark cycle at 28 °C. Photographs show representative results of S. torvum at 1 day post‐inoculation (dpi) (A, C) or tobacco at 2 dpi (B) from three independent experiments, which gave similar results. The broken lines represent the infiltrated area. (D) Summary of the plant responses caused by various strains in S. torvum and tobacco leaves. NT, not tested.

Figure 2.

Effects of Agrobacterium‐mediated transient expression of Rip36, Rip36E149A and AvrA in Solanum torvum  Torubamubiga and tobacco leaves. (A) Construction of a series of pEl2Ω‐MCS plasmids expressing the Rip effector. Leaves were infiltrated with Agrobacterium tumefaciens  GV3101 harbouring no plasmid, an empty plasmid, a plasmid expressing Rip36, Rip36E149A or AvrA at OD ₆₀₀ of 0.5, and incubated under a 16 h light/8 h dark cycle at 28 °C. Photographs show representative results of S. torvum at 2 days post‐inoculation (dpi) (B) or tobacco at 4 dpi (C) from three independent experiments, which gave similar results. The broken lines represent the infiltrated area. (D). Summary of plant responses. HR, hypersensitive response.

Rip36 is a Zn‐dependent protease‐like T3SS effector with a putative Zn‐binding motif (HExxH). Therefore, we next investigated the molecular mechanism of Rip36‐induced HR in S. torvum. To examine whether the HExxH motif of Rip36 is required to induce HR, we changed the essential glutamic acid of position 149 to alanine (HELIH to HALIH) using a QuickChange XL site‐directed mutagenesis kit (Stratagene, La Jolla, CA, USA). Experimental details for DNA construction are described in Supporting Information Text S1. Wild‐type (WT) rip36 and rip36E149A genes were cloned into a binary vector pEl2Ω‐MCS (Ohtsubo et al., 1999) and transiently expressed in leaves of S. torvum and tobacco (N. tabacum L. cv. Xanthi NC) via a transient Agrobacterium‐mediated expression system. We found that Rip36E149A no longer induced HR in S. torvum (Fig. 2). This result indicates that the putative Zn‐dependent protease motif of Rip36 is required for HR elicitation in S. torvum.

To confirm the results obtained from the transient expression analysis in R. solanacearum, we constructed rip36 and avrA mutant derivatives of strain RS1002, as described in Text S1. The R. solanacearum mutants and plasmids used in this study are listed in Table 1.

Table 1.

Bacterial strains and plasmids

Strain	Description	Reference
Ralstonia solanacearum
RS1002	RS1000 Nalr	Mukaihara et al. (2004)
RS1273	RS1002 ΔhrpY	Mukaihara et al. (2004)
RS1650	RS1002 avrA::Gmr	This study
RS1662	RS1002 Δrip36	This study
RS1668	RS1002 rip36E149A	This study
RS1672	RS1662 chr::mini‐Tn5rip36 +	This study
RS1673	RS1662 chr::mini‐Tn5rip36E149A	This study
Agrobacterium tumefaciens
GV3101	Gmr, Rifr	Lamblin et al. (2001)
Escherichia coli
DH5α	F‐ λ‐ ø80dlacZΔM15 Δ(lacZYA‐argF)U169 recA1 endA1 hsdR17(rK ‐mK +) supE44 thi‐1 gyrA relA1	Takara, Kyoto, Japan
S17‐1	thi pro hsdR‐ hsdM+ recA [chr::RP4‐2‐Tc::Mu‐Km::Tn7]	Schäfer et al. (1994)
S17‐1 λpir	λ pir lysogen of S17‐1	Simon et al. (1983)
Plasmid
pEl2Ω‐MCS	Binary vector that can replicate in both E. coli and A. tumefaciens, used for transient expression analysis	Ohtsubo et al. (1999)
pEl2Ω‐MCS_avrA	pEl2Ω‐MCS carrying rip5 at XbaI and SpeI sites	This study
pEl2Ω‐MCS_rip36	pEl2Ω‐MCS carrying rip36 at XbaI and SpeI sites	This study
pEl2Ω‐MCS_rip36E149A	pEl2Ω‐MCS carrying rip36E149A at XbaI and SpeI sites	This study
pARO‐HA‐'Cya	CyaA reporter plasmid	Murata et al. (2006)
pARO‐rip36‐HA‐'Cya	pARO‐HA‐'Cya carrying rip36 at XbaI and XhoI sites	Mukaihara et al. (2010)
pK18mobsacB	Small mobilizable vector, Kmr, sucrose sensitive (sacB)	Schäfer et al. (1994)
pK18‐Δrip36	pK18mobsacB‐derived plasmid that contains a 1.6‐kb chimeric PCR product deleting rip36, Kmr	This study
pK18‐rip36E149A	pK18mobsacB‐derived plasmid that contains a 657‐bp rip36E149A, Kmr	This study
pK18‐avrA	pK18mobsacB‐derived plasmid that contains a 2.8‐kb PCR product containing avrA, Kmr	This study
pK18‐avrA::Gmr	pK18mobsacB‐derived plasmid that contains disrupted avrA by the insertion of a Gmr cassette into pK18‐avrA, Gmr, Kmr	This study
pBSL118	Mini‐Tn5‐derived plasmid vector for insertion mutagenesis, Ampr, Kmr	Alexeyev et al. (1995)
pBSL‐rip36	pBSL118‐derived plasmid that contains a 2.09‐kb rip36, Ampr, Kmr	This study
pBSL‐rip36E149A	pBSL118‐derived plasmid that contains a 2.09‐kb rip36E149A, Ampr, Kmr	This study

In S. torvum, HR was induced by inoculation of strains RS1002 (parental WT) and RS1650 (avrA::Gm^r), but not strains RS1662 (Δrip36) and RS1668 (rip36E149A) (Fig. 1A). HR in tobacco leaves was induced by inoculation of the WT, Δrip36 and rip36E149A strains, but not by the avrA::Gm^r strain (Fig. 1B). This result clearly indicates that Rip36 and AvrA of RS1002 are required for HR in S. torvum and tobacco, respectively. Furthermore, Rip36 abolished the ability to induce HR in S. torvum when the Zn‐binding motif HELIH was changed to HALIH, indicating that the putative Zn‐dependent protease motif is required for HR elicitation in S. torvum. We also confirmed that these HRs were dependent on hrpY, a gene encoding an essential component of the Hrp T3SS, supporting the hypothesis that these T3SS effectors determine HR.

To confirm that Rip36 is a crucial avirulence factor in the induction of HR in S. torvum, we introduced the WT rip36 and rip36E149A genes into a Δrip36 mutant strain, and generated two types of complemented strain, as described in Text S1. The leaves of S. torvum were infiltrated with the resultant complemented strains, RS1672 (Δrip36 Tnrip36 ⁺) and RS1673 (Δrip36 Tnrip36E149A). As expected, HR in S. torvum leaves was restored by the introduction of Tnrip36 ⁺, but not Tnrip36E149A (Fig. 1C). This result further confirmed that Rip36 is an avirulence factor that induces HR in S. torvum and that the Zn‐protease motif is essential for this HR.

To exclude the possibility that the E149A mutation of Rip36 affects the ability to translocate into plant cells via the Hrp T3SS, we constructed R. solanacearum strains expressing the Rip36‐HA‐'Cya and Rip36E149A‐HA‐'Cya fusion proteins, as described previously (Murata et al., 2006; Text S1). An in planta adenylate cyclase assay was carried out as described previously (Murata et al., 2006). RS1273 (ΔhrpY) was used as a negative control because this strain lacks the ability to translocate effectors into plant cells (Mukaihara and Tamura, 2009). The cyclic adenosine monophosphate (cAMP) in S. torvum leaves inoculated with the WT strain expressing Rip36E149A‐HA‐'Cya was increased to a level similar to that in leaves inoculated with the strain expressing Rip36‐HA‐'Cya (Fig. 3A). As predicted, the increase in the cAMP level was completely abolished by the ΔhrpY mutation (Fig. 3A). This result indicates that the E149A mutation in the putative Zn‐binding motif has no effect on the translocation of Rip36 into plant cells.

Figure 3.

Translocation and stability of Rip36 and Rip36E149A. (A) The cyclic adenosine monophosphate (cAMP) levels of Solanum torvum leaves inoculated with Ralstonia solanacearum  RS1002 (hrp ⁺) or RS1273 (ΔhrpY) strains expressing the calmodulin‐dependent adenylate cyclase (Cya) fusion with Rip36 or Rip36E149A. The cAMP level is shown as an average of three replications with standard deviations in parentheses. (B) The stability of Rip36 and Rip36E149A in S. torvum leaves after infiltration. Leaves of S. torvum were infiltrated with RS1002 (lane 1) and the rip36E149A (lane 2) and Δrip36 (lane 3) mutants, and total proteins were prepared from the inoculated leaves at 15 h post‐inoculation. The Rip36 and Rip36E149A proteins were detected by an anti‐Rip36 peptide antibody after sodium dodecylsulphate‐polyacrylamide gel electrophoresis (SDS‐PAGE). Photograph shows a representative result of Western blotting from two independent experiments. An arrow indicates the position of the 20‐kDa marker protein.

Next, we tested the stability of the Rip36E149A protein by immunoblot analysis. Briefly, R. solanacearum strains RS1002 (WT), RS1662 (Δrip36) and RS1668 (rip36E149A) were grown overnight in BG medium (Bacto‐Agar 15 g, glucose 5 g/L) at 28 °C, pelleted, washed with distilled water and then resuspended at an optical density at 600 nm (OD₆₀₀) of 0.3. The leaves of 5–6‐week‐old S. torvum were infiltrated with the bacterial suspensions and incubated at 28 °C in a growth chamber. After 15 h, the infiltrated parts of the leaves were collected, frozen with liquid nitrogen and ground with a mortar and pestle. Then, 1.5 mL of 50 mm Tris‐HCl buffer (pH 7.5) was added per 0.5 mg of leaf tissue. Cell debris was removed by centrifugation twice for 10 min at 10 000 × g, and the supernatant was collected in new tubes. Proteins were then precipitated from the supernatant by adding 5% trichloroacetic acid and incubating for 1 h at 4 °C. Precipitated proteins were pelleted by centrifugation and washed with 70% ethanol, dried, resuspended in 50 μL PBST buffer (137 mm NaCl, 8.1 mm Na₂HPO₄, 2.68 mm KCl, 1.47 mm KH₂PO₄ and Tween‐20, pH 7.4) and investigated by Western blot analysis after sodium dodecylsulphate‐polyacrylamide gel electrophoresis (SDS‐PAGE), as described previously (Nguyen et al., 2012). A rabbit polyclonal anti‐Rip36‐specific peptide antibody was generated using the synthetic peptide (peptide sequence: QGLQARSKYKPRNA, 205–218 amino acids) by Operon Biotechnologies (Tokyo, Japan) by the same method as described previously (Nguyen et al., 2012). Bands corresponding to the expected size of the Rip36 protein were detected in WT and rip36E149A mutant strains, whereas no band was detected in the Δrip36 strain (Fig. 3B). On the basis of these results, it is clear that the point mutation protein Rip36E149A is as stable as WT Rip36, and that the E149A mutation in the Zn‐protease motif has no effect on the translocation and stability of Rip36 in plant cells.

To examine whether Rip36 contributes to the virulence of R. solanacearum on host eggplant (Solanum melongena cv. Senryo‐nigou), RS1002 (WT) and the Δrip36 and rip36E149A mutants [5 × 10⁴ colony‐forming units (cfu)/mL] were infiltrated into eggplant leaves and bacterial growth was measured at 2 and 5 days post‐inoculation (dpi). This quantitative assay is reported to be more sensitive than the disease scoring method (Macho et al., 2010). However, no significant difference in bacterial multiplication was observed between RS1002, Δrip36 and rip36E149A mutant strains at either 2 or 5 dpi (Fig. 4A). This result suggests that Rip36 has no or little effect on the virulence of R. solanacearum in host eggplant, probably because of the functional redundancy of a large effector repertoire, as reported by Mukaihara et al. (2010).

Figure 4.

Effect of rip36 mutation on the growth of Ralstonia solanacearum  RS1002 on host eggplant (A) and nonhost Solanum torvum (B). Leaves were infiltrated with wild‐type (WT) RS1002 or Δrip36 and rip36E149A mutants at 5 × 10⁴ colony‐forming units (cfu)/mL. Bacterial growth was measured 2 and 5 days after inoculation. Error bars indicate the standard deviation measured from six biological replicates. Means of three independent experiments are presented. Asterisks indicate a significant difference from the growth of the RS1002 WT strain in a t‐test (***P < 0.001).

We further measured the bacterial growth on nonhost S. torvum. The leaves of S. torvum were infiltrated with RS1002 (WT) and the Δrip36 and rip36E149A mutants (5 × 10⁴ cfu/mL), and bacterial growth was measured at 2 and 5 dpi. The Δrip36 and rip36E149A mutants increased significantly in S. torvum leaves when compared with the WT RS1002 at both 2 and 5 dpi (Fig. 4B). This indicates that Rip36 is the major avirulence factor in RS1002 that elicits resistance and restricts bacterial growth in S. torvum leaves.

In this study, we have shown that the R. solanacearum T3SS effector Rip36 functions as an avirulence protein to induce HR in S. torvum. We have also shown that the putative Zn‐dependent protease motif of Rip36 is essential for its avirulence function. In the fungal pathogen Magnaporthe grisea, the causal agent of rice blast disease, an AVR‐Pita avirulence protein is known to contain a Zn‐dependent protease motif. The protease motif is essential for the avirulence function of AVR‐Pita because a point mutation in the conserved motif completely abolishes the ability to induce rice blast resistance mediated by the corresponding disease resistance (R) protein Pi‐ta in rice (Orbach et al., 2000). It is also known that AVR‐Pita interacts directly with Pi‐ta, and that a mutation in the protease motif results in the loss of avirulence and disrupts the physical interaction between them. This direct interaction raises the possibility that the Pi‐ta protein may sense the cleavage by AVR‐Pita (Jia et al., 2000). Our results also indicate that the Zn‐protease activity of Rip36, like Avr‐Pita, is required for the induction of HR, although the enzymatic activity has not yet been tested. It has been reported that the Rip36 homologue NleD from enteropathogenic Escherichia coli cleaves c‐JUN N‐terminal kinase by its Zn‐dependent protease activity and inhibits the inflammatory reaction, an animal innate immunity response (Baruch et al., 2011). In R. solanacearum, Rip36 might cleave a particular S. torvum protein to exhibit its virulence function, and the resultant peptide fragment might be specifically recognized by a resistance protein, leading to HR. The Rip36 homologue HopH1 from Pseudomonas syringae pv. tomato DC3000 may contribute to virulence on host plants because the deletion of both hopH1 and hopC1 from this pathogen reduced both lesion formation and growth in Arabidopsis and tomato (Wei et al., 2007). It is interesting that the putative protease motif of HopH1 is required for its virulence function. The mechanism of HR induction by Rip36 might be similar to that of the cysteine protease effectors AvrRpt2 from P. syringae pv. tomato DC3000 (Axtell and Staskawicz, 2003) and AvrPphB from P. syringae pv. phaseolicola (Shao et al., 2002). AvrRpt2 and AvrPphB effectors cleave Arabidopsis RIN4 and PBS1 proteins, respectively, whose cleavage is monitored by the corresponding R proteins RPS2 and RPS5, respectively (Mackey et al., 2003; Shao et al., 2003). To reveal the mechanism of Rip36‐induced HR in S. torvum, the identification of the target protein is now under investigation.

Supporting information

Text S1  Plasmid constructions and generation of Ralstonia solanacearum strains.