•Pathogenicity of the Gram-negative plant pathogen Xanthomonas campestris pv. vesicatoria (Xcv) depends on a type III secretion system that translocates a cocktail of > 25 type III effector proteins into the plant cell.
•In this study, we identified the effector AvrBsT as a suppressor of specific plant defense. AvrBsT belongs to the YopJ/AvrRxv protein family, members of which are predicted to act as proteases and/or acetyltransferases.
•AvrBsT suppresses the hypersensitive response (HR) that is elicited by the effector protein AvrBs1 from Xcv in resistant pepper plants. HR suppression occurs inside the plant cell and depends on a conserved predicted catalytic residue of AvrBsT. Yeast two-hybrid based analyses identified plant interaction partners of AvrBs1 and AvrBsT, including a putative regulator of sugar metabolism, SNF1-related kinase 1 (SnRK1), as interactor of AvrBsT. Intriguingly, gene silencing experiments revealed that SnRK1 is required for the induction of the AvrBs1-specific HR.
•We therefore speculate that SnRK1 is involved in the AvrBsT-mediated suppression of the AvrBs1-specific HR.
Plant pathogenic bacteria have evolved sophisticated strategies to exploit their corresponding host organisms. Pathogenicity of most members of the Gram-negative genera Xanthomonas, Ralstonia and Pseudomonas depends on a type III secretion (T3S) system, which spans both bacterial membranes and is associated with an extracellular pilus and a channel-like translocon in the eukaryotic plasma membrane (He et al., 2004). The T3S system translocates a cocktail of type III effector proteins (T3Es) directly into the plant cell cytosol (Ghosh, 2004; Grant et al., 2006; Block et al., 2008). T3Es have diverse functions and target multiple host cellular pathways such as gene expression, hormone signaling, proteasome-dependent protein degradation and defense responses to the benefit of the pathogen (Speth et al., 2007; Block et al., 2008; Göhre & Robatzek, 2008; Lewis et al., 2009). The main target of many T3Es appears to be the plant immune system. Plants defend themselves against microbial invaders by basal defense responses that include expression of pathogenesis-related genes, production of reactive oxygen species and callose deposition into the plant cell wall (Dangl & Jones, 2001; Bent & Mackey, 2007; de Wit, 2007). Basal plant defense is activated upon recognition of pathogen- (or microbe-) associated molecular patterns such as flagellin, lipopolysaccharides and elongation factor EF-Tu by specific receptors (Jones & Dangl, 2006). Furthermore, during coevolution with pathogens, plants have also evolved disease resistance (R) genes that detect individual T3Es. R gene-mediated recognition of T3Es (also designated avirulence (Avr) proteins) is often associated with the hypersensitive response (HR), a rapid programmed cell death at the infection site that restricts bacterial multiplication (Dangl & Jones, 2001). Successful pathogens often deliver T3Es that suppress basal plant defense responses or effector-triggered programmed cell death (Abramovitch & Martin, 2004).
In our laboratory, we study Xanthomonas campestris pv. vesicatoria (Xcv, also termed Xanthomonas euvesicatoria or Xanthomonas axonopodis pv. vesicatoria; Jones et al., 2004), the causal agent of bacterial spot disease of pepper and tomato. The T3S system of Xcv is encoded by the chromosomal hrp (hypersensitive response and pathogenicity) gene cluster, which is essential for disease in susceptible plants and the HR in resistant plants, and translocates c. 30 T3Es into the plant cell (Bonas et al., 1991; Thieme et al., 2007; White et al., 2009). Efficient effector protein translocation depends on the cytoplasmic HpaB protein, a T3S chaperone (Büttner et al., 2004, 2006; Lorenz et al., 2008). T3S chaperones specifically bind to one or several T3S substrates and promote their secretion and/or stability (Feldman & Cornelis, 2003; Wilharm et al., 2007). HpaB from Xcv has a broad substrate specificity and is required for the efficient secretion of all T3Es tested so far (Büttner et al., 2006; J. Stuttmann et al., unpublished). Notably, T3Es from Xcv differ in their HpaB dependency: class A T3Es depend on HpaB for translocation, whereas class B T3Es are translocated even in the absence of HpaB, albeit in reduced amounts (Büttner et al., 2006). The differential contribution of HpaB to the translocation of T3Es suggests that there is a hierarchy in effector protein translocation and that class A and class B T3Es are translocated at a subsequent stage during infection (Büttner et al., 2006).
In this study, we identified the T3E AvrBsT from Xcv as a suppressor of the HR elicited by AvrBs1 in resistant pepper plants. AvrBs1 promotes virulence of Xcv in pepper Capsicum annuum cv Early Cal Wonder (ECW) under field conditions and is recognized in ECW-10R pepper plants by the corresponding resistance gene Bs1 (Ronald & Staskawicz, 1988; Wichmann & Bergelson, 2004). Transient expression studies showed that AvrBs1 induces chlorosis in the nonhost plant Nicotiana benthamiana and decreases the starch content in chloroplasts (Gürlebeck et al., 2009). The C-terminal region of AvrBs1 is homologous to AvrA from Pseudomonas syringae pv. glycinea (Napoli & Staskawicz, 1987), but the function of AvrA and AvrBs1 is unknown (Escolar et al., 2001). AvrBsT belongs to the clan CE of cysteine proteases with homology to YopJ (C55 family; YopP in Yersinia enterocolitica; Ciesiolka et al., 1999). Members of the YopJ/AvrRxv protein family are present in many plant and animal pathogenic bacteria and include AvrRxv, AvrBsT, AvrXv4 and XopJ from Xcv, YopJ from Yersinia spp., AvrA from Salmonella spp. and T3Es from Ralstonia solanacearum, P. syringae and Erwinia spp. (Minsavage et al., 1990; Keen & Buzzell, 1991; Whalen et al., 1993; Galyov et al., 1994; Astua-Monge et al., 2000; Noël et al., 2003; Shrestha et al., 2005; Ma et al., 2006). Most members of the YopJ family from animal pathogens contribute to the host–pathogen interaction; however, their biochemical function and impact on pathogenicity are a matter of debate (Mukherjee et al., 2006; Sweet et al., 2007). Several biochemical activities have been shown for members of the YopJ/AvrRxv family, that is isopeptidase (Orth et al., 2000; Hotson & Mudgett, 2004; Ma et al., 2006), acetyltransferase (Mukherjee et al., 2006), and deubiquitination protease (Sweet et al., 2007), all of which depend on a conserved catalytic triad consisting of histidine, aspartate or glutamate, and cysteine residues. One target of YopJ in animal cells is a mitogen-activated protein kinase kinase, MKK6, which is required for the activation of the transcription factor NF-κB, suggesting that YopJ interferes with signal transduction pathways (Mukherjee et al., 2006). Mutant studies revealed that the Avr activities of the YopJ/AvrRxv family members from plant pathogens depend on the catalytic triad, suggesting that the enzymatic function is required for the recognition by corresponding plant R proteins (Orth et al., 2000; Roden et al., 2004a; Bonshtien et al., 2005; Whalen et al., 2008). Notably, a chimeric protein between the N-terminal region of AvrXv4 from Xcv and the C-terminal catalytic domain of YopP from Y. enterocolitica still exhibits Avr activity, suggesting a conservation of the catalytic activity of YopJ/AvrRxv homologs (Whalen et al., 2008).
Here, we show that AvrBsT from Xcv suppresses the efficient induction of the AvrBs1-specific HR. Protein–protein interaction revealed that AvrBsT interacts with the SNF1-related kinase SnRK1 from pepper in the plant cell cytoplasm. Silencing of the SnRK1 transcript leads to a severe reduction of the AvrBs1-specific HR, suggesting that SnRK1 is involved in the AvrBs1-induced plant immunity.
Materials and Methods
Bacterial strains and growth conditions
For bacterial strains and plasmids used in this study, see Table 1. Plasmids were introduced into Escherichia coli by electroporation and into Xcv and Agrobacterium tumefaciens by conjugation using pRK2013 as helper plasmid in triparental matings (Figurski & Helinski, 1979; Ditta et al., 1980). E. coli strains were grown at 37°C in Lysogeny broth (LB) or Super medium (Qiagen). Xcv strains were cultivated at 30°C in nutrient-yeast-glycerol (NYG) (Daniels et al., 1984) or in minimal medium A (Ausubel et al., 1996), which was supplemented with sucrose (10 mM) and casamino acids (0.3%) and A. tumefaciens strains at 30°C in yeast extract broth (YEB) medium. Antibiotics were added at the following final concentrations: ampicillin, 100 μg ml−1; hygromycin, 50 μg ml−1; kanamycin, 25 μg ml−1; rifampicin, 100 μg ml−1; and spectinomycin, 100 μg ml−1.
Table 1. Bacterial strains and plasmids used in this study
The near-isogenic pepper cvs ECW, ECW-10R and ECW123 (Minsavage et al., 1990; Stall & Minsavage, 1996) and tomato Solanum lycopersicum cv Moneymaker plants were grown at 25°C with c. 65% relative humidity and 16 h light. Xcv strains were hand-inoculated with a needleless syringe into the apoplast of leaves at 2 × 108 cfu ml−1 in 1 mM MgCl2 if not stated otherwise. For in planta growth curves, bacteria were inoculated at a density of 104 cfu ml−1 and bacterial growth was determined as described in Bonas et al. (1991). For in planta transient expression studies and virus-induced gene silencing (VIGS), A. tumefaciens GV3101 was grown overnight in YEB medium, resuspended in inoculation medium (10 mM MgCl2, 5 mM MES, pH 5.3, 150 μM acetosyringone) and inoculated into leaves at concentrations of 8 × 108 cfu ml−1.
Generation of GATEWAY-compatible expression constructs
For the generation of avrBs1, avrBsT, avrBsTC222A and SnRK1 expression constructs, corresponding coding sequences were amplified by PCR and inserted into pENTR/D-TOPO and recombined into pDEST17, pGWB2, pGWB5, pGWB20, pSPYNE and pGWB735/1 using GATEWAY® technology (see Table 1). Primer sequences are listed in Table 2.
Table 2. Primers used in this study
Generation of avrBsT and avrRxv deletion strains
To generate an avrBsT deletion mutant in Xcv, the 4.3 kb PstI fragment from pXV943 containing avrBsT was cloned in pBluescript KS followed by digestion with SalI and religation, resulting in pKSavrBsTLR. A 900 bp deletion and a frameshift in avrBsT (resulting in deletion of amino acids 14–314 out of 350) were introduced by BglII/BsrB1 digestion of pKSavrBsTLR and religation. The resulting insert was cloned into the SalI/XbaI sites of suicide vector pOK1 (Table 1) and introduced into the genome of Xcv strain 75-3 by homologous recombination as described in Huguet et al. (1998). To delete avrRxv, an avrRxv-containing cosmid isolated from a genomic cosmid library of Xcv strain 75-3 (Minsavage et al., 1990) was digested with EcoRI and HindIII and a 6 kb fragment was cloned into pBluescript KS. The first 801 bp (corresponding to amino acids 1–267 out of 374) from avrRxv were deleted by BglII digestion and religation. The fragment containing the avrRxv-flanking regions was cloned into the BamHI/SalI sites of pOK1 and the resulting construct was conjugated into Xcv strain 85-10 to select for strain 85-10ΔavrRxv.
Yeast two-hybrid screening
For yeast two-hybrid (Y2H) screens we used the BD Matchmaker™ Library Construction & Screening Kit (Clontech, Heidelberg, Germany) according to the manufacturer’s instructions. Screens were performed with a tomato cDNA library (Gürlebeck, 2007) and a pepper cDNA library that was generated from a mixture of leaf material of ECW-10R (untreated, inoculated with Xcv 85-10 and 85-10(pDS400), respectively) and of ECW-30R (untreated and inoculated with 85-10(pDSF300)).
RNA analysis and rapid amplification of cDNA ends (RACE)
For 5′-RACE, RNA was isolated from C. annuum ECW and ECW-10R with RNeasy Plant Mini Kit (Qiagen) and RACE was performed using a BD SMART RACE cDNA Amplification Kit (Clontech) according to the manufacturer’s instructions.
For quantitative reverse-transcription PCR (qRT-PCR) analysis, RNA was extracted from 1.9 cm2 leaf tissue using RNeasy Plant Mini Kit (Qiagen). cDNA was synthesized from 4.5 μg RNA using RevertAid™ H Minus First Strand cDNA Synthesis Kit (Fermentas, St Leon-Rot, Germany). qRT-PCR was performed on an iCycler (Bio-Rad) using ABsolute QPCR SYBR® Green Fluorescein Mix (ABgene Limited, Hamburg, Germany) and c. 9 ng of cDNA template as technical triplicates. PCR profiles are available upon request. For transcript abundance comparisons of different genes, the amplification efficiency for each gene was determined using a standard curve plot of a dilution series. Amplification specificity was determined by melting curve analysis. Transcript abundances of the constitutively expressed elongation factor 1A were used to account for differences in cDNA amounts as described in the user bulletin 2 (Applied Biosystems, Foster City, CA, USA).
Bimolecular fluorescence complementation
Bimolecular fluorescence complementation (BiFC) experiments were performed as described previously (Hu et al., 2002) using pSPYNE and pGWB735/1 (Table 1).
Virus-induced gene silencing (VIGS)
Gene silencing in pepper ECW123 plants was performed as described previously (Liu et al., 2002a; Chung et al., 2004). At 21 d after initiation of silencing, plants were inoculated with Xcv at bacterial density of 8 × 108 cfu ml−1. The silencing efficiency was determined by qRT-PCR.
Trypan blue staining and ion leakage measurements
For trypan blue staining, samples of inoculated leaf tissue were harvested 12 and 16 h post-inoculation (hpi), boiled 2 min in trypan blue solution (Koch & Slusarenko, 1990) and incubated overnight. After 2 d of bleaching in chloral hydrate (2.5 g ml−1 water), samples were incubated in 70% glycerol and analyzed by microscopy. For electrolyte leakage experiments, triplicates of 3.1 cm2 infected leaf material were taken 2 and 16 hpi. Leaf discs were placed on the bottom of a 15 ml tube and covered with plastic screen. A total of 7 ml of deionized water was added to each tube and vacuum-infiltrated (1 min). Tubes were placed on to a rotary shaker at 100 rpm for 1 h and conductivity was determined (first reading) with a conductometer (Knick, Berlin, Germany). To determine maximum conductivity of the entire sample, conductivity was measured after boiling the samples for 30 min (Stall et al., 1974).
Protein purification and biochemical assays
His-tagged derivatives of AvrBsT and AvrBsTC222A were expressed in E. coli BL21 and purified using nickel nitrilotriacetic acid (Ni-NTA) columns (Qiagen). Protein amounts were analyzed by SDS-PAGE and Coomassie staining and adsorbance was measured at 280 nm. Purified proteins were dialyzed in reaction buffer (100 mM Tris, pH 8.0, 150 mM NaCl, 1 mM Dithiothreitol (DTT) DTT) and protease activity of equal protein amounts was determined as described in Ma et al. (2006) using RediPlate 96 EnzChek Protease Assay Kit red fluorescence (Invitrogen) in a Tecan SpectraFluor Fluorescence Reader (Tecan Trading AG, Männedorf, Switzerland) at 630 nm.
For transacetylation assays, N-terminally His6 epitope-tagged derivatives of AvrBsT and AvrBsTC222A were purified from E. coli under native conditions. Equal protein amounts were incubated with 5,5′-dithiobis-(2-nitrobenzoic acid) (DTNB) and acetyl-CoA (Williams et al., 1975). The assay is based on a color change to yellow if the disulfide bond of DTNB is cleaved by thiols (e.g. from CoA-SH), resulting in 2-nitro-5-thiobenzoate (NTB). The absorbance was measured at 410 nm in a Tecan SpectraFluor Fluorescence Reader (Tecan Trading AG).
The AvrBs1-specific HR in pepper ECW-10R plants is suppressed in the presence of AvrBsT
Genome sequence and Southern blot analyses revealed that the T3E gene avrBs1 is present in the pepper-pathogenic Xcv strain 85-10 and in the tomato-pathogenic strain 75-3 (Ronald & Staskawicz, 1988; Escolar et al., 2001; Hajri et al., 2009). Notably, however, induction of the AvrBs1-specific HR in resistant ECW-10R pepper plants by strain 75-3 is significantly reduced when compared with strain 85-10, suggesting that AvrBs1 is not efficiently delivered by strain 75-3 or that the AvrBs1-specific HR is suppressed (Fig. 1a). The genome of strain 75-3 is not yet sequenced, but it was reported that strain 75-3 contains AvrBsT, which is absent from strain 85-10 (Ciesiolka et al., 1999; Escolar et al., 2001; Thieme et al., 2005; Hajri et al., 2009). We therefore wondered whether the reduction of the AvrBs1-specific HR was a result of the presence of AvrBsT. To investigate a potential influence of AvrBsT on the HR induction, we deleted avrBsT from the genome of strain 75-3 and inoculated the resulting deletion mutant strain 75-3ΔavrBsT into leaves of ECW-10R plants. In contrast to strain 75-3, strain 75-3ΔavrBsT induced a confluent HR in a similar manner to strain 85-10 (Fig. 1a). To confirm this finding, we introduced the expression construct pDD62avrBsT encoding AvrBsT into strain 85-10. Strain 85-10 (pDD62avrBsT) induced a significantly reduced HR when compared with strain 85-10 carrying the empty vector, suggesting that the presence of AvrBsT leads to a reduced AvrBs1-specific HR (Fig. 1a). We also investigated whether suppression of the AvrBs1-specific HR depends on the residues of the predicted catalytic triad of AvrBsT (H154, E173, C222). These residues are conserved in all YopJ/AvrRxv family members and were shown to be essential for the induction of the AvrBsT-specific HR in N. benthamiana and Arabidopsis thaliana (Orth et al., 2000; Cunnac et al., 2007). Notably, ectopic expression of avrBsTC222A in Xcv strain 85-10 did not interfere with the induction of the AvrBs1-specific HR (Fig. 1a).
Next, we analyzed whether suppression of the AvrBs1-induced HR occurs inside the plant cell. For this, avrBs1 and avrBsT were transiently expressed in leaves of pepper ECW-10R plants via Agrobacterium-mediated T-DNA transfer. As expected, expression of avrBs1 induced a confluent HR (Fig. 1b; Escolar et al., 2001). Coexpression of avrBs1 and avrBsT led to a significant reduction in HR induction, suggesting that the AvrBsT-dependent HR suppression occurs inside the plant cell (Fig. 1b). No HR suppression was observed upon coexpression of avrBsTC222A and avrBs1, which confirms our finding that the predicted catalytic cysteine residue of AvrBsT is crucial for the suppression of the AvrBs1-specific HR.
Analysis of the AvrBsT-mediated HR suppression by trypan blue staining and ion leakage assays
To characterize the suppression of the AvrBs1-specific HR by AvrBsT in more detail, plant cell death was monitored by trypan blue staining of infected tissue. Trypan blue is a vital stain that specifically stains dead cells but is not absorbed by cells with intact plasma membranes (Derenzis & Schechtman, 1973). Xcv strain 85-10 carrying the empty vector or the avrBsT expression construct was inoculated into leaves of pepper ECW-10R plants. Samples from infected leaf tissue were collected 12 hpi, stained with trypan blue and analyzed by transmission light microscopy. In contrast to untreated leaf material that remained unstained, almost all cells were stained by trypan blue in tissue inoculated with strain 85-10 Fig. 1(c). In leaves inoculated with strain 85-10 expressing avrBsT, only a few cells were stained, which is in agreement with the reduced HR induction by this strain Fig. 1(c); see Fig. 1(a). Similar results were obtained 16 hpi (data not shown). Reduced staining of infected plant tissue was also observed after inoculation of strain 75-3ΔavrBsT when compared with strain 75-3 (data not shown).
For the quantitative analysis of the HR suppression, we determined ion leakage in leaves infected with strain 85-10 carrying the empty vector or the avrBsT expression construct. Ion leakage is associated with cell death and leads to conductivity changes of samples in distilled water. As expected for an ongoing HR, conductivity increased in samples inoculated with Xcv strain 85-10 (Fig. 1d). In the presence of AvrBsT, however, ion leakage was reduced by c. 50% (Fig. 1d), which correlates well with the observed HR suppression by AvrBsT.
AvrBsT does not significantly contribute to bacterial virulence in tomato plants
Given the function of AvrBsT as HR suppressor, we wondered whether AvrBsT contributes to bacterial virulence. We therefore analyzed disease symptoms and growth of Xcv strains 75-3 and 75-3ΔavrBsT in leaves of susceptible tomato cv Moneymaker plants. No differences in disease symptoms were detected (data not shown). Furthermore, deletion of avrBsT from Xcv 75-3 did not alter bacterial growth in planta, suggesting that AvrBsT does not significantly contribute to bacterial virulence (Fig. 2a). It remains to be investigated whether the lack of an obvious virulence function of AvrBsT is a result of the presence of functionally redundant effector proteins. As mentioned earlier, AvrBsT belongs to the YopJ/AvrRxv protein family, which includes XopJ and AvrRxv, both present in strain 85-10. Notably, however, deletion mutant derivatives of strain 85-10 lacking xopJ and/or avrRxv induced the HR in ECW-10R pepper plants in a similar fashion to strain 85-10, suggesting that both do not suppress the AvrBs1-specific HR (Fig. 2b). Furthermore, Agrobacterium-mediated coexpression of avrBs1 with xopJ or avrRxv, did not affect the HR in ECW-10R plants (data not shown). Thus, the suppression of the AvrBs1-specific HR is specifically caused by AvrBsT.
AvrBs1 and AvrBsT are class B effectors that are translocated in the absence of the general T3S chaperone HpaB
To confirm the T3S system-dependent translocation of AvrBs1 and AvrBsT into the plant cell, we generated fusion proteins between AvrBs1 and AvrBsT and the reporter protein AvrBs3Δ2, which is a derivative of the type III effector AvrBs3 lacking the translocation signal. However, AvrBs3Δ2 contains the effector domain and is recognized in AvrBs3-responsive ECW-30R pepper plants when fused to a functional T3S and translocation signal (Szurek et al., 2002; Noël et al., 2003). AvrBs3Δ2 reporter constructs were introduced into Xcv strains 85-10hrpG* (85*), the T3S mutant 85-10hrpG*ΔhrcV (85*ΔhrcV), and the hpaB deletion mutant 85-10hrpG*ΔhpaB (85*ΔhpaB) (Büttner et al., 2004). hrpG* strains contain a constitutively active derivative of the key regulator HrpG, which activates hrp gene expression and is essential for the translocation of class B T3Es in the hpaB deletion mutant (Wengelnik et al., 1996, 1999; Büttner et al., 2006). For in vivo translocation assays, bacteria were inoculated into leaves of AvrBs3-responsive ECW-30R plants. Derivatives of strain 85* delivering AvrBs1-AvrBs3Δ2 and AvrBsT-AvrBs3Δ2 induced the HR in ECW-30R, whereas no HR was observed with the T3S mutant 85*ΔhrcV, suggesting that both fusion proteins were translocated by the T3S system (Fig. 3a).
AvrBsT-AvrBs3Δ2 and AvrBs1-AvrBs3Δ2 both induced a partial HR when delivered by strain 85*ΔhpaB, suggesting that translocation occurred even in the absence of the global T3S chaperone HpaB, albeit in reduced amounts (Fig. 3a). This is in agreement with the previous finding that strain 85*ΔhpaB induces a partial AvrBs1-specific HR in ECW-10R plants (Fig. 3a; Büttner et al., 2004). Differences in the HR were not the result of differences in protein stabilities, because similar amounts of AvrBsT-AvrBs3Δ2 and AvrBs1-AvrBs3Δ2 were detected in wild-type and hpaB deletion mutant strains (Fig. 3b). Our results therefore suggest that AvrBs1 and AvrBsT belong to class B of T3Es that are translocated in the absence of the global T3S chaperone HpaB (Büttner et al., 2006). It is therefore tempting to speculate that AvrBs1 and AvrBsT are translocated at the same stage during infection.
Isolation of plant interaction partners of AvrBs1 and AvrBsT
To identify potential plant interaction partners of AvrBs1 and AvrBsT, we used AvrBs1 and the catalytic mutant AvrBsTC222A as baits in Y2H screens. AvrBsTC222A was chosen to ensure that interaction partners of AvrBsT are not proteolytically cleaved or modified. Since AvrBs1 and AvrBsTC222A did not auto-activate the reporter genes, yeast strains expressing these baits were mated with strains containing prey cDNA libraries. The latter were generated from mRNA isolated from leaves of pepper lines ECW and ECW-30R and tomato Solanum lycopersicum cv Moneymaker, respectively. We analyzed c. 108 clones per screen and identified potential interaction partners of AvrBs1 and AvrBsTC222A (Table 3). Potential interactors of AvrBs1 include, for example, proteins with homology to a tonoplast-intrinsic protein, a tyrosine kinase and a DNA-binding protein with basic helix-loop-helix motif. Proteins interacting with AvrBsT showed homology to elongation factor eEF1A, a subunit of the 26S proteasome, an ABC transporter and a sucrose nonfermenting 1 (SNF1)-related kinase 1 (SnRK1) (Fig. 4a). Reproducibility of the interactions was tested by retransforming prey and bait plasmids into yeast. Prey plasmids encoding potential interactors of AvrBs1 and AvrBsTC222A did not activate expression of reporter genes in yeast strains that contained Lamin C, suggesting that the interactions were specific (data not shown). We also tested whether interactors of AvrBs1 interact with AvrBsTC222A and vice versa; however, no common interacting protein was identified.
Table 3. Plant interactors of AvrBs1 and AvrBsT identified in yeast two-hybrid screens
Amino acid identity (%)
aInteractors were isolated from a tomato cDNA-library. eEF1A and the SNF1-homologous protein were also isolated from the pepper cDNA library.
bA potential contribution to the induction of the AvrBs1-specific HR in ECW-10R pepper plants was analyzed by VIGS. No, no reduction in HR induction observed; nd, not determined.
The AvrBsT interactor SnRK1 is required for the efficient induction of the AvrBs1-specific HR
To test whether plant proteins interacting with AvrBsT or AvrBs1 are biologically relevant for the AvrBs1-specific HR, the corresponding genes were silenced in different AvrBs1-responsive pepper cultivars using VIGS. For VIGS, 300–500 bp fragments of the respective coding sequences were cloned into a tobacco rattle virus-silencing vector and expressed in planta using Agrobacterium-mediated T-DNA delivery. As negative control, plants were treated with Agrobacterium carrying the empty silencing vector. Two weeks later leaves were inoculated with Xcv strain 85-10. qRT-PCR analysis revealed that VIGS resulted in a > 70-fold reduction of the SnRK1 transcript at this time-point (Fig. 4b).
Silencing of the SnRK1 transcript led to a severe reduction of the AvrBs1-specific HR induced by strain 85-10 in ECW-10R and ECW123 (Fig. 4c,d). The AvrBs1-specific HR is visible as a white necrosis of the inoculated leaf area 4 d post-inoculation (dpi) (Fig. 4e). The partial necrosis in SnRK1-silenced ECW-10R and ECW123 plants after inoculation of strain 85-10 was presumably the result of residual recognition of AvrBs1, because in virus-treated control plants a similar phenotype was observed after inoculation with strain 85-10 carrying AvrBsT (Fig. 4c,d). Besides, AvrBs1 ECW123 also recognizes AvrBs3. Strain 85-10 carrying AvrBs3 elicited a white necrotic HR in virus-treated ECW123 controls, whereas SnRK1-silenced plants showed a reduced brownish HR (Fig. 4d). The brown necrosis was caused by recognition of AvrBs3 (compare also Fig. 4e), suggesting that SnRK1 is not essential for the Bs3-mediated HR. Therefore SnRK1 does not appear to be a general mediator of plant defense.
SnRK1 was isolated from both pepper and tomato and is homologous to SNF1 from yeast. We determined the complete coding sequence from pepper ECW and ECW-10R plants by 5′-RACE. Both sequences were identical and share 95% DNA sequence identity with Snf1 from tomato (accession number AF143743). Comparison of the corresponding proteins revealed 98% sequence identity. SnRK1 contains predicted serine/threonine kinase, ubiquitin-associated and kinase-associated domains.
SnRK1 and AvrBsT interact in the plant cell cytosol
To investigate the subcellular localization of AvrBs1, AvrBsT and SnRK1 in the plant cell, green fluorescent protein (GFP) fusions were synthesized in N. benthamiana after Agrobacterium-mediated gene delivery. We did not use pepper plants for these experiments because of the very low efficiency of Agrobacterium-mediated transformation of pepper (R Szczesny & U Bonas, unpublished). Since AvrBsT induces the HR in N. benthamiana 24 hpi (Orth et al., 2000; data not shown), GFP fluorescence was analyzed 20 hpi. Confocal laser-scanning microscopy of infected plant tissue revealed that AvrBs1-GFP localized to the plant cell cytoplasm and was nuclear-excluded as reported previously (Fig. 5a; Gürlebeck et al., 2009). By contrast, GFP fusions of AvrBsT and SnRK1 localized to both the cytoplasm and nuclei (Fig. 5a). Detection of GFP fluorescence in the nucleus was not a result of nuclear transport of GFP-containing protein degradation products since all GFP fusion proteins were stably synthesized (Fig. 5b). Our results therefore suggest that AvrBsT and SnRK1 localize to the same cellular compartments in the plant cell.
To confirm the interaction between AvrBsT and SnRK1 in planta, we performed BiFC (Hu et al., 2002). BiFC is observed when translational fusions between the two nonfluorescent portions of the yellow fluorescent protein (YFP) to two potential interaction partners are brought into close proximity. We generated expression constructs encoding YFPC-SnRK1 and AvrBsT-YFPN and co-delivered the corresponding genes into leaves of N. benthamiana by Agrobacterium. Infected leaf material was analyzed by confocal laser-scanning microscopy; however, no fluorescent signal was detected 20 hpi (data not shown; note that AvrBsT induces the HR in N. benthamiana 24 hpi). Because this time-point might be too early to detect BiFC, we repeated the experiment with YFPC-SnRK1 and the catalytic mutant AvrBsTC222A-YFPN, which does not induce the HR in N. benthamiana (Orth et al., 2000; data not shown). Inspection of infected leaf material 48 hpi revealed YFP fluorescence in the plant cell cytoplasm (Fig. 5c). No fluorescence was observed for YFPC-SnRK1 and AvrBs1-YFPN (Fig. 5c), which confirms that SnRK1 is a specific interactor of AvrBsT and does not interact with AvrBs1. Taken together, we conclude that AvrBsT and SnRK1 interact in the plant cell cytoplasm.
AvrBsT does not degrade or transacetylate SnRK1
AvrBsT belongs to the conserved YopJ/AvrRxv protein family, members of which were proposed to act as cysteine proteases (Orth, 2002). To test AvrBsT for proteolytic activity, His6 epitope-tagged derivatives of AvrBsT and AvrBsTC222A were synthesized in E. coli, purified (see the Materials and Methods section; Fig. 6a) and incubated with fluorescence-labeled casein as substrate. AvrBsT led to a weak increase of fluorescence indicative of a protease activity, which was dependent on the predicted catalytic cysteine residue at position 222 (Fig. 6b). We also investigated whether stability of SnRK1 is affected in the presence of AvrBsT. For this, genes encoding AvrBsT, AvrBsTC222A and AvrBs1 (used as control), and a c-Myc epitope-tagged derivative of SnRK1, were expressed under control of the cauliflower mosaic virus 35S promoter in leaves of pepper ECW-10R after Agrobacterium-mediated gene delivery. Leaf material was harvested 1 dpi and analyzed by immunoblotting using a c-Myc specific antibody. Coexpression of SnRK1 with either avrBsT, avrBsTC222A or avrBs1 did not result in any obvious degradation of SnRK1 (Fig. 6c).
We also investigated a possible transacetylase activity of AvrBsT because the AvrBsT homolog YopJ acts as acetyltransferase (Mukherjee et al., 2006). For this, derivatives of SnRK1, AvrBsT and AvrBsTC222A with N-terminal His6 epitope tag were synthesized in E. coli, purified under native conditions (see the Materials and Methods section) and incubated with DTNB and acetyl-CoA (see the Materials and Methods section). With AvrBsT we observed a slight increase in absorbance indicative of the generation of NTB; however, similar results were obtained with the catalytic mutant AvrBsTC222A (Fig. 6d). Taken together, we conclude that AvrBsT does not cleave or transacetylate SnRK1 under the conditions tested.
In this study, we identified the T3E AvrBsT from Xcv as a suppressor of the AvrBs1-specific HR in pepper plants. The HR suppression occurred when both AvrBsT and AvrBs1 were delivered by Xcv or were synthesized in planta after Agrobacterium-mediated gene delivery, suggesting that AvrBsT interferes with plant defense inside the plant cell. In vivo translocation assays revealed that both effectors are translocated by the T3S system even in the absence of the global T3S chaperone HpaB, suggesting that AvrBs1 and AvrBsT belong to class B of T3Es and are presumably translocated at a similar stage during infection. AvrBsT is homologous to members of the YopJ/AvrRxv family, which were predicted to act as cysteine proteases and/or acetyltransferases. Acetyltransferase activity was indeed demonstrated for YopJ from Yersinia spp. (Mukherjee et al., 2006). By contrast, a weak protease activity was shown for the YopJ-homologs HopZ1 and HopZ2 from Pseudomonas (Ma et al., 2006). Here, we show that AvrBsT from Xcv exhibits a weak in vitro protease activity, which depends on the conserved predicted catalytic cysteine residue of AvrBsT at position 222. Interestingly, cysteine 222 is also required for the AvrBsT-dependent HR induction in pepper, N. benthamiana and A. thaliana ecotype Pi-0, which suggests that the induction of defense responses in AvrBsT-responsive plants depends on the predicted enzymatic activity of AvrBsT (Orth et al., 2000; Cunnac et al., 2007). Notably, the AvrBsT-HR in A. thaliana is suppressed by a carboxylesterase (Cunnac et al., 2007).
To our knowledge, AvrBsT is the first member of the YopJ/AvrRxv family known to suppress effector-triggered plant immunity. Notably, we did not observe a suppression of the AvrBs1-specific HR by AvrBsT homologs such as XopJ or AvrRxv from Xcv that are both expressed in Xcv strain 85-10. However, it was previously shown that XopJ suppresses callose deposition in the plant cell wall, which is part of basal plant defense (Bartetzko et al., 2009). Suppression of plant defense responses appears to be one of the major missions of T3Es from plant pathogenic bacteria and is often achieved by modulation of host protein turnover. Several T3Es that interfere with plant immunity act as proteases; for example, the cysteine proteases AvrPphB and AvrRpt2 from P. syringae and XopD from Xcv. AvrPphB cleaves the plant protein kinase PBS1 which is guarded by the R protein RPS5 (Shao et al., 2003; Ade et al., 2007). AvrRpt2, by contrast, degrades RIN4, which is a negative regulator of plant defense (Axtell et al., 2003; Kim et al., 2005). XopD presumably acts on small ubiquitin-like modifier (SUMO)-conjugated proteins in the plant cell nucleus (Hotson et al., 2003). A function as SUMO protease was also proposed for the AvrBsT homolog YopJ; however, no specific host target proteins for de-SUMOylation by YopJ were identified (Orth et al., 2000). Preliminary data suggest that AvrBsT from Xcv does not significantly alter the amounts of SUMO- or ubiquitin-associated proteins (R Szczesny et al., unpublished).
To identify potential plant targets of AvrBsT we performed Y2H screens. Plant interactors of AvrBsT include a subunit of the 26S proteasome and SnRK1. We did not identify 14-3-3 proteins although a member of this large protein family was previously shown to interact with AvrRxv from Xcv (Whalen et al., 2008). It is conceivable that YopJ/AvrRxv family members differ in the choice of their host targets, which might also be reflected by differences in protein localization. Thus, in contrast to AvrRxv, which is in the cytoplasm (Bonshtien et al., 2005), XopJ from Xcv is targeted to the plasma membrane (Thieme et al., 2007), and AvrBsT localizes to both cytoplasm and nucleus (Fig. 5).
Given the finding that several T3Es exploit the host proteasome to modulate protein turnover, for example, by acting as ubiquitin ligase (Speth et al., 2007), it will be interesting to investigate whether AvrBsT associates with the proteasome in planta and modifies its ubiquitination status. Thus far, we have not determined whether the predicted association of AvrBsT with the proteasome is required for the interference with plant immunity. However, our gene silencing experiments revealed that reduced transcript abundance of SnRK1 lead to a severe reduction of the AvrBs1-specific HR. SnRK1 does not interact with AvrBs1 and is therefore presumably indirectly involved in the recognition of AvrBs1 by the cognate, as yet unknown R gene Bs1 or in downstream signaling that leads to the HR-induction. It is unlikely that SnRK1 represents the R protein Bs1 because the SnRK1 open reading frame is identical in susceptible ECW and resistant ECW-10R plants. We cannot exclude that resistance in ECW-10R depends on differences in the SnRK1 promoter region. However, transient expression of SnRK1 under control of the 35S promoter in ECW-10R pepper plants did not induce any detectable cell death reaction (data not shown), suggesting that enhanced transcript abundance of SnRK1 is not sufficient to trigger the HR.
SnRK1 is the plant ortholog of SNF1 from yeast and the AMP-activated protein kinase (AMPK) from mammals that are important regulators of sugar metabolism (Hardie, 2007; Hedbacker & Carlson, 2008; Halford & Hey, 2009). SnRK1 phosphorylates enzymes involved in primary metabolism and is a central regulator of reprogramming gene expression in response to environmental stresses (Halford & Hey, 2009). Similarly to SNF1 and AMPK, SnRK1 is presumably part of a heterotrimeric complex. The AMPK heterotrimer consists of a catalytic α subunit that is homologous to SNF1, and β and γ subunits (Davies et al., 1994; Mitchelhill et al., 1994). Homologs of the β and the γ subunits in yeast are encoded by SNF4 and the related genes SIP1, SIP2 and GAL83. Proteins with homology to SNF4, SIP1, SIP2 and GAL83 have been cloned from A. thaliana, Solanum tuberosum and maize and were shown to interact with SnRK1 (Bouly et al., 1999; Lakatos et al., 1999; Kleinow et al., 2000; Lumbreras et al., 2001). In agreement with this finding, we identified a protein from pepper with homology to GAL83 as interactor of pepper SnRK1 (R Szczesny & U Bonas, unpublished).
Interestingly, there is growing evidence for a role of SNF1 homologs in response to pathogens. First, SnRK1 was found to be induced in citrus plants after infection with X. axonopodis pv. citri (Cernadas et al., 2008). Further-more, the SnRK1 homolog AKIN1 from N. benthamiana is important for resistance against tomato golden mosaic virus and beet curly top virus (Hao et al., 2003). The viral proteins AL2 and L2 interact with AKIN1, which in turn is inactivated (Sunter et al., 2001). It remains to be investigated whether AvrBsT inhibits the function of pepper SnRK1 and thereby suppresses the AvrBs1-specific HR. In any case, the mere interaction of SnRK1 and AvrBsT is not sufficient for the HR-suppression because the catalytic core mutant AvrBsTC222A also binds SnRK1 but does not suppress the AvrBs1-induced HR. This suggests that the predicted catalytic activity of AvrBsT is essential for the HR suppression. Surprisingly, we did not detect degradation of SnRK1 in the presence of AvrBsT. Furthermore, in vitro assays did not reveal an obvious acetylation of SnRK1 by AvrBsT. We cannot, of course, exclude that AvrBsT has an as yet unknown enzymatic function or that degradation or acetylation of SnRK1 occurs under natural conditions but is not detectable in vitro or upon in planta overexpression of SnRK1 and avrBsT. Furthermore, it is possible that AvrBsT degrades another subunit of the predicted SnRK1-containing trimeric complex, for example the GAL83 homolog that interacts with SnRK1.
Taken together, our findings suggest that the effector AvrBsT from Xcv targets pepper SnRK1, a regulator of sugar metabolism, which is required for AvrBs1-induced immunity. Future studies are needed to elucidate whether AvrBsT modifies the SnRK1 complex and, if so, which kind of modification occurs.
We are grateful to M. Schattat for contributing to microscopy studies, to M. B. Mudgett and to T. Lahye for providing plasmids, to P. Römer for contributing to the generation of the pepper Y2H library, to C. Kretschmer for sequencing, and to B. Rosinsky for glasshouse work. This work was supported by grants from the Deutsche Forschungsgemeinschaft (Sonderforschungsbereich 648 ‘Molekulare Mechanismen der Informationsverarbeitung in Pflanzen’) to D.B. and U.B.