†Present address: Bureau for Economic Growth, Agriculture & Trade, US Agency for International Development, 1300 Pennsylvania Avenue, NW, Washington, DC 20523-2110, USA.
The conserved Xanthomonas campestris pv. vesicatoria effector protein XopX is a virulence factor and suppresses host defense in Nicotiana benthamiana
Article first published online: 26 JAN 2005
The Plant Journal
Volume 41, Issue 6, pages 801–814, March 2005
How to Cite
Metz, M., Dahlbeck, D., Morales, C. Q., Sady, B. A., Clark, E. T. and Staskawicz, B. J. (2005), The conserved Xanthomonas campestris pv. vesicatoria effector protein XopX is a virulence factor and suppresses host defense in Nicotiana benthamiana. The Plant Journal, 41: 801–814. doi: 10.1111/j.1365-313X.2005.02338.x
- Issue published online: 26 JAN 2005
- Article first published online: 26 JAN 2005
- Received 19 October 2004; revised 7 December 2004; accepted 9 December 2004.
- virulence effector protein;
- innate immunity;
- Agrobacterium-mediated transient expression;
- enhanced disease susceptibility
Nicotiana benthamiana leaves display a visible plant cell death response when infiltrated with a high titer inoculum of the non-host pathogen, Xanthomonas campestris pv. vesicatoria (Xcv). This visual phenotype was used to identify overlapping cosmid clones from a genomic cosmid library constructed from the Xcv strain, GM98-38. Individual cosmid clones from the Xcv library were conjugated into X. campestris pv. campestris (Xcc) and exconjugants were scored for an altered visual high titer inoculation response in N. benthamiana. The molecular characterization of the cosmid clones revealed that they contained a novel gene, xopX, that encodes a 74-kDa type III secretion system (TTSS) effector protein. Agrobacterium-mediated transient expression of XopX in N. benthamiana did not elicit the plant cell death response although detectable XopX protein was produced. Interestingly, the plant cell death response occurred when the xopX Agrobacterium-mediated transient expression construct was co-inoculated with strains of either XcvΔxopX or Xcc, both lacking xopX. The co-inoculation complementation of the plant cell death response also depends on whether the Xanthomonas strains contain an active TTSS. Transgenic 35S-xopX-expressing N. benthamiana plants also have the visible plant cell death response when inoculated with the non-xopX-expressing strains XcvΔxopX and Xcc. Unexpectedly, transgenic 35S-xopX N. benthamiana plants displayed enhanced susceptibility to bacterial growth of Xcc as well as other non-xopX-expressing Xanthomonas and Pseudomonas strains. This result is also consistent with the increase in bacterial growth on wild type N. benthamiana plants observed for Xcc when XopX is expressed in trans. Furthermore, XopX contributes to the virulence of Xcv on host pepper (Capsicum annuum) and tomato (Lycopersicum esculentum) plants. We propose that the XopX bacterial effector protein targets basic innate immunity in plants, resulting in enhanced plant disease susceptibility.
Phytopathogenic bacteria in the genus Xanthomonas have evolved a highly efficient protein secretion system (type III secretion system, TTSS) to directly deliver effector proteins of bacterial origin to the plant host cell (Buttner and Bonas, 2003). There is direct biochemical evidence for TTSS-dependent translocation of the Xanthomonas effectors AvrBs2 and AvrBs3 into host plant cells (Casper-Lindley et al., 2002; Szurek et al., 2002). Documented cases show that effector proteins target host proteins to suppress or modulate host defense signaling pathways (Buell et al., 2003; Cornelis, 2002). It has been hypothesized that phytopathogenic bacteria contain many effector proteins that either act independently or in concert to target the host functions leading to disease (Alfano and Colmer, 2004; Buell et al., 2003). Although the elimination or post-translational modification of these effector-targeted host proteins has only been documented in a few cases (Axtell and Staskawicz, 2003; Mackey et al., 2003), a new paradigm is beginning to emerge that suggests that this may be a common occurrence.
Plants have evolved a countermeasure to the delivery of virulence effector proteins via the TTSS. This molecular surveillance system directly or indirectly detects the presence of bacterial effector proteins by monitoring the activity of these proteins on their respective host targets (Dangl and Jones, 2001; Ellis et al., 2000; Shao et al., 2003). The surveillance system consists of a ‘super family’ of plant disease resistance (R) genes that encode structurally related proteins involved in effector protein detection (Meyers et al., 2003). Active recognition of effector proteins leads to the induction of a suite of biochemical host defense mechanisms that result in pathogen inhibition and the expression of plant disease resistance, which also coincides with the phenotype of hypersensitive cell death response (HR) (Hammond-Kosack and Jones, 1997).
Site-directed gene ‘knock-out’ studies have shown that loss of certain effector proteins, many of them encoded by avirulence (avr) genes, can compromise pathogen growth and virulence in a host. Several avr genes from phytopathogenic bacteria have been demonstrated to be important for pathogen fitness or virulence. These include avrBs2 of Xanthomonas campestris (Kearney and Staskawicz, 1990) (Gassmann et al., 2000; Swords et al., 1996) pthA of X. citri (Gabriel, 1999), avrRpm1 of Pseudomonas syringae pv. maculicola (Ritter and Dangl, 1995), avrRpt2 of P. syringae pv. tomato (Chen et al., 2000), avrPto of P. s. pv. tomato (Chang et al., 2000), and avrXa7 of X. oryzae pv. oryzae (Yang et al., 2000).
In nature, the ability of a plant pathogen to incite disease on a particular host plant is the rare exception. Most pathogens are actively recognized by a wide range of plant species that resist pathogen infection. This phenomenon has been referred to as non-host resistance (Agrios, 1997). Non-host resistance may involve R protein recognition of non-host pathogen elicitor determinants and in some cases an associated high titer non-host cell death response is visually similar to the avr-Resistance HR (Alfano and Colmer, 2004; Peart et al., 2002). Our initial objective in investigating the molecular basis of non-host resistance was to identify and characterize non-host pathogen elicitor determinants required for associated non-host cell death response to high titer inoculations. We have employed X. campestris pv. vesicatoria (Xcv), the causal agent of bacterial spot disease on pepper and tomato, and X. campestris pv. campestris (Xcc), the causal agent of black rot disease of crucifers, as our model system. Nicotiana benthamiana high titer inoculation of Xcc elicits only a mild chlorosis response that is visually distinct from the Xcv-associated cell death response. This visual phenotype was used to screen a cosmid library of the genome of Xcv that had been conjugated into Xcc. In this study we report the identification of XopX (Xanthomonas outer protein) (Noel et al., 2002), an Xcv bacterial effector protein that allows Xcc to elicit an Xcv cell death-like response when inoculated on N. benthamiana. Sequence analysis of XopX revealed a novel methionine-rich protein that is nearly ubiquitous in strains of Xcv and widespread in other X. c. pathovars. Unexpectedly, characterization of XopX reveals that there is no XopX activation of non-host resistance. Based on our results, we propose that the XopX bacterial effector protein targets basic innate immunity in plants by suppressing host defense function, which in turn allows for greater pathogen growth and an altered visual high titer inoculation response by N. benthamiana.
Isolation and molecular characterization of xopX
To identify possible pathogen elicitor determinants that may play a role in the cell death response, the X. campestris pv. vesicatoria (Xcv) strain GM98-38 (Gassmann et al., 2000) was chosen because high titer (4 × 108 CFU ml−1) inoculation elicits a strong visible plant cell death response in N. benthamiana (Figure 1a). To isolate the gene(s) required for inducing the high titer cell death response, a genomic library of Xcv strain GM98-38 was constructed in the wide host range cosmid vector pLAFR3. Individual clones were conjugated from Escherichia coli (DH5α) to X. campestris pv. campestris 8004 (Xcc) (Turner et al., 1984), a phytopathogenic bacterium that only induces a high titer mild chlorosis response on N. benthamiana (Figure 1a).
Approximately 700 Xcc transconjugants were tested for their ability to induce a cell death response on N. benthamiana and two cosmid clones, pL3-9·22 and pL3-16·9, were isolated. The inserts in these cosmids were found to contain overlapping fragments of genomic DNA based on an EcoRI restriction digest pattern.
The cosmid pL3-9·22 was mutagenized with a modified pTn3-HoHo1 (Stachel et al., 1985) transposon (pTn3-gus) in E. coli and conjugated back into Xcc and screened for loss of cell death response activity on N. benthamiana (Figure 1a). Four of the 80 insertion events disrupted the cell death response activity of the cosmid pL3-9·22. Restriction mapping showed that all four insertions resided within a single 4.7 kb EcoRI fragment. A sub-clone containing this EcoR1 fragment, pL6-E4.7 was shown to contain cell death response activity (Figure 1b). Further sub-cloning delimited the cell death response activity and the sites of transposon insertion to a 2.8 kb SmaI-EcoRI fragment. The 2.8 kb subclone, pL6-SE2.8, was the minimal DNA fragment that conferred cell death response when expressed in Xcc (Table 1).
|Strain||Strain or strains alone||Strain + pL6-SE2.8||Strain + pL6-540stpA||Strain + pL6-511stpB||Strain + pL6-361stpB||Strain + pV-SE2.8HA||Strain + pM-14orfB·stpA:HA|
|X.c. pv. campestris|
|Xcc 8004||(−)||(+)20 h||(+)20 h||(−)||(+)20 h||(+)20 h|
|X.c. pv. vesicatoria|
|GM98-38||(+)40 h||(+)20 h|
|GM98-38-1||(+)48 h||(+)20 h|
|GM98-38(xopX:Tn3)||(−)||(+)20 h||(+)20 h||(−)|
|GM98-38-1(xopX:Tn3)||(−)||(+)20 h||(+)20 h||(−)|
|85* (xopX183:cya)||(+)26 h|
|C58C1 + GM98-38(xopX:Tn3)||(−)||(+)20 h|
|C58C1 + Xcc 8004||(−)||(+)20 h|
|C58C1 + 85*ΔhrcV||(−)||(−)|
The minimal 2.8 kb of Xcv DNA from pL6-SE2.8 was sequenced and this revealed two large overlapping potential open reading frames, orfA and orfB (Figure 2). All four transposon insertions disrupted both orfA and orfB and were not helpful for discriminating which open reading frame contained cell death response activity (Figure 2). Site-directed nucleotide mutations were engineered that created a stop codon in orfA without altering the amino acids coded for in orfB, and a stop codon in orfB without altering orfA (Figure 2). These mutations were subcloned as pL6-537stpA and pL6-511stpB, respectively. Only the orfB mutation blocked cell death response activity and so we conclude that orfB is used for translation of a novel Xcv bacterial effector protein XopX that allows Xcc to elicit cell death response on N. benthamiana. As orfB has three potential start codons, the second start codon was replaced with a stop codon (Figure 2). The bacterial expression construct with this mutation, pL6-361stpB, still has full cell death response activity on N. benthamiana, thus we conclude that the third start codon with a potential ribosome binding site at minus 9 nucleotides is the start of XopX translation (Table 1, Figure 2). As expected for the parental Xcv strain GM98-38, marker-selected double homologous recombination of transposon-mutagenized xopX into the genome copy of xopX disrupts cell death response on N. benthamiana (Table 1). It is interesting to note that strains expressing XopX from multi-copy plasmids elicit the cell death response faster.
An unusual feature of the predicted amino acid sequence of XopX (Genbank accession number AY621073) is the high methionine content of 24 amino acids (>3%) (Figure 2). Computer blast searches (Altschul et al., 1990) identified highly conserved homologs in the X. axonopodis pv. citri (Xac) and Xcc genome sequences with 92.7 and 77.1% amino acid sequence identity with XopX, respectively (Genbank accession numbers AE011681.1 and AE012150.1). No significant homology was found to other sequences in the database that could help predict the function of XopX. The predicted size of the XopX protein encoded by orfB was 74 kDa. To confirm this hypothesis, an HA epitope (Wilson et al., 1984) tag was introduced to the C-terminus of XopX. Immunoblot analysis of Xcc pVSE2.8:HA extracted protein with an anti-HA antibody reveals the predicted 80 kDa epitope-tagged protein (Figure 5c). The XopX:HA-tagged protein has wild type cell death response activity on N. benthamiana when expressed in trans from pV-SE2.8:HA in Xcc and Xcv (Table 1).
To evaluate the conservation of xopX in Xanthomonas, 14 additional Xcv field isolates including 16-82, 75-3, 81-23, 82-8, 84-1, 85-10, 88-5, 90-62, 92-14, 93-1, 206, P28 and 89-1 (R.E.Stall) were analyzed for the presence of xopX hybridizing sequences and 13 strains contained an identical band to Xcv GM98-38 as revealed by DNA gel-blot hybridization (Figure 3a). With the exception of Xcv strain 89-1, all the isolates contain a 4.7 kb EcoRI fragment that hybridized with Xcv xopX probe. Twelve additional X.c. pathovars including aberrans, alfalfa, begonia, carotae, fragariae, holcicola, icanae, malvacearum, pruni, raphani and vitians (R.E.Stall) were examined by DNA gel-blot hybridization of EcoRI-digested genomic DNA (Figure 3b). Nine of the 12 Xc pathovars contain strong homology to xopX. Interestingly, Xcc 8004 has strong homology to xopX (Figure 3b), but does not have XopX cell death response activity on N. benthamiana (Figure 1a).
Evidence for the delivery of XopX to plant cells via the TTSS of Xcv
To demonstrate xopX TTSS dependence, xopX was mutagenized in Xcv strains 85-10, 85* (constitutive hrp expression mutation) and 85*ΔhrcV (TTSS-defective mutant that can no longer deliver TTSS-dependent effectors to the plant host) (Bogdanove et al., 1996; Fenselau et al., 1992; Wengelnik et al., 1999) by marker-selected double homologous recombination of transposon-mutagenized xopX. The xopX cell death response activity on N. benthamiana of 85-10 and 85* is eliminated in the homologous recombination mutants (Table 1) similar to the GM98-38 homologous recombination mutant. Neither the 85*ΔhrcV or the xopX homologous recombination mutants have xopX cell death response activity on N. benthamiana (Table 1) which demonstrates xopX TTSS dependence.
To demonstrate TTSS-dependent delivery of XopX into the host plant cell as a bacterial effector, translational fusions of XopX were made to the calmodulin-dependent adenylate cyclase domain (Cya) of Bordetella pertussis (Casper-Lindley et al., 2002; Sory et al., 1995). Because we were unsure of the effects on delivery of XopX by adding the large Cya reporter to full-length XopX, we made one construct containing the putative N-terminal secretion-translocation sequence of xopX and another with a full-length C-terminal fusion. These Cya reporter fusions were exchanged into the genomic copy of xopX by single homologous recombination. Xcv strains 85*(xopX183:cya), 85*ΔhrcV(xopX183:cya) and GM98-38(xopX183:cya) are Cya translational fusions that include the first 183 codons of xopX and have been exchanged by single homologous recombination into the genomic copy of xopX. This genomic integration event creates a fusion of the putative secretion-translocation sequence of xopX to Cya and also maintains the wild type copy of xopX. Furthermore, the Xcv strain GM98-38 (xopX719:cya) is a Cya translational fusion to the last codon of xopX that has been exchanged by single homologous recombination into the genomic copy of xopX. This genomic integration event creates a full-length xopX translational fusion with Cya and disrupts the wild type copy of xopX (see Experimental procedures for construction details). All Cya fusion strains except 85*ΔhrcV(xopX183:cya) still have xopX cell death response activity on N. benthamiana (Table 1). A non-translocated control Cya reporter fusion with the first 184 codons of RecA (Guttman and Greenberg, 2001) was constructed by single homologous recombination into the genomic copy of RecA to create 85*(RecA:cya) and GM98-38(RecA:cya). This genomic integration event also mutates the RecA gene, which impairs the strains ability to repair UV damage and grow after exposure to short-wave UV.
To assay for active Cya translational fusion constructs, bacterial extracts were reacted in the presence of exogenous calmodulin and ATP for 5 min and then cAMP levels were measured. There is an increase in cAMP levels for all Cya translational fusion constructs compared with the corresponding wild type Xcv strain (Figure 4c). To assay XopX delivery to the plant cell, leaf extracts were obtained from 8-h post high titer (4 × 108 CFU ml−1) leaf inoculations and then assayed for cAMP. Only strains GM98-38(xopX183:cya), GM98-38(xopX719:cya) and 85*(xopX183:cya) have an increase in cAMP levels resulting from translocation of the Cya reporter protein into N. benthamiana (Figure 4a). In addition Xcv has a similar level of translocation activity for the non-host (N. benthamiana) and the host (pepper) interactions The absence of an increase in cAMP levels for 85*ΔhrcV(xopX183:cya) demonstrates XopX delivery to the plant cell is TTSS dependent (Figure 4b).
Agrobacterium-mediated transient expression of XopX in N. benthamiana
As delivery of the XopX protein to the plant cell via Xcc results in a cell death response, we investigated whether XopX expression in planta would elicit a cell death response. A xopX:HA epitope-tagged construct was made and delivered to plant cells employing Agrobacterium-mediated transient expression in N. benthamiana. The binary construct pM-14orfB-stpA:HA has the 35S promoter driving xopX orfB with the 537stpA mutation and the NOS terminator. Unexpectedly, transient expression of XopX did not elicit a cell death response (Table 1, Figure 5a). This is in contrast to Agrobacterium-mediated transient expression of most other effector proteins like avrBs2 (Tai et al., 1999), which give rise to the hypersensitive cell death response in the plants that contain the cognate resistance protein, Bs2. Immunoblot analysis of extracted protein from Agrobacterium-mediated transient expression in N. benthamiana 24 h post-inoculation with an anti-HA antibody reveals the predicted 80 kDa epitope-tagged protein (data not shown). However, a cell death response is observed if the Agrobacterium delivering 35S-xopX:HA is co-inoculated with either XcvΔxopX or Xcc which both lack XopX activity (Table 1, Figure 5a). Co-inoculation with 85*ΔhrcV did not elicit a cell death response demonstrating that even when XopX is expressed in the plant cell this cell death response was dependent on a pathogen with an intact TTSS (Table 1). This suggests that the TTSS or another TTSS-dependent factor is necessary for the XopX-associated cell death response activity.
Transgenic XopX N. benthamiana plants have an enhanced susceptibility to both Xanthomonas and Pseudomonas non-host pathogens
The Xcc co-inoculation requirement for complementation of XopX Agrobacterium-mediated transient cell death response prompted us to transform pM-14orfB-stpA:HA into N. benthamiana and select stable transgenic plants expressing the xopX gene in order to evaluate xopX activity on a wider range of N. benthamiana pathogen interactions. Although there is no apparent alteration in normal plant growth of these transgenic N. benthamiana plants, the inoculation of Xcc that lacks a functional xopX gene can induce a cell death response (Figure 5b). In planta growth assays on a XopX-expressing N. benthamiana line have an increased bacterial growth for a wide range of strains that lack xopX; Xcc8004, XcvGM98-38(xopX:Tn3), Xcv89-1 and P. syringae pv. tomato strain DC3000 (Ps tomDC3000; Cuppels, 1986; Figure 6a,b). Although in planta non-host bacterial growth is increased it does not increase to the population levels of the N. benthamiana virulent pathogen P. syringae pv. tabaci (Ps tab; Figure 6b). In addition, transgenic xopX N. benthamiana have larger and more pronounced lesions 10 days post-inoculation with a low titer (1 × 103 CFU ml−1) of both Ps tomDC3000 and Xcv89-1 (Figure 6c).
We propose that XopX bacterial effector protein targets basic innate immunity in plants, which in turn allows for greater pathogen virulence as defined by increased lesions and in planta pathogen growth. These results are consistent with the observed increase in virulence of Xcc expressing xopX on N. benthamiana in planta growth assays (Figure 1c). XopX targeting of basic innate immunity and the resulting increase in pathogen virulence is also consistent with the associated plant cell death response to high titer inoculations.
Immunoblot analysis of extracted protein, from the stable N. benthamiana 35S-xopX:HA expressing line at 0 and 8 h post-inoculation with the non-xopX-carrying Xcc strain, probed with an anti-HA antibody, also reveals the predicted 80 kDa epitope-tagged protein (Figure 5c).
Contribution of xopX to virulence of Xcv in host interactions
Bacterial pathogens use the TTSS to deliver effector proteins that are essential for successful infection. As XopX contributes to virulence for bacterial non-host interactions and is highly conserved in strains of Xanthomonas and dependent on TTSS for delivery, we investigated its potential role as a virulence factor in host pepper and tomato interactions. The homologous recombination of transposon-mutagenized xopX in strains GM98-38(xopX:Tn3) without avrBs2 and GM98-38-1(xopX:Tn3) with avrBs2 were used for in planta bacterial growth assays on pepper and tomato host plants (Figure 7b,c, respectively). On both pepper and tomato hosts, the Xcv native copies xopX and avrBs2 each contributed to virulence. The reduction in virulence associated with loss of xopX was similar to the reduction with loss of avrBs2, as seen in previous studies (Kearney and Staskawicz, 1990). Reduced in planta growth for strains lacking xopX indicates that it is important for full virulence on host plants. Moreover, mutation of both genes had an additive detrimental effect on bacterial virulence (Figure 7b,c).
Both xopX mutant strains GM98-38(xopX:Tn3) without avrBs2 and GM98-38-1(xopX:Tn3) with avrBs2 no longer induce cell death response in N. benthamiana (Figure 7a; Table 1). Both mutant strains were complemented back to cell death response on N. benthamiana in trans with plasmid subclones of xopX (Table 1).
As both xopX marker exchange mutant strains GM98-38(xopX:Tn3) and GM98-38-1 (xopX:Tn3) block both orfA and orfB it was necessary to use double homologous recombination to incorporate the 537stpA (orfA) and 511stpB (orfB) stop codon mutants into Xcv GM98-38. Only the orfB mutation blocked cell death response in non-host N. benthamiana and also resulted in reduced virulence on host pepper (Figure 7a,d).
The initial objective of these studies was to isolate a bacterial effector protein from Xcv, a pathogen of tomato and pepper, that could elicit a non-host resistance reaction on N. benthamiana. The strategy that we adopted to clone a ‘non-host’ effector gene was similar to the strategy that has been successfully employed to isolate other effector (avr) genes that are recognized by genetically defined disease resistance genes. The strategy takes advantage of the visually distinct disease resistance-associated HR to high titer (4 × 108 CFU ml−1) inoculation of bacterial pathogens expressing the corresponding avr gene. Nicotiana benthamiana leaves also have a visible plant cell death response to high titer inoculation of Xcv. The strategy was successful in that screening of the Xcv GM98-38 cosmid library in Xcc exconjugants resulted in the identification of two cosmids that that convert the mild Xcc chlorosis response to the Xcv visual plant cell death response on N. benthamiana. Molecular characterization of these cosmids revealed that they contained overlapping DNA fragments bearing the same 2.8 kb piece of DNA, and encoded the 74 kDa XopX protein. Further experiments demonstrated that XopX is delivered in a TTSS-dependent manner that is similar to classical avr genes. Interestingly, the visual cell death response phenotype was not elicited by Agrobacterium-mediated xopX expression, which is in contrast to Agrobacterium-mediated avr expression on resistant plants. Cell death response can be elicited if the Agrobacterium-mediated xopX construct is co-inoculated with Xcc or xopX-deficient Xcv. The cell death response is not observed when a hrcV TTSS-deficient strain of Xanthomonas is used, suggesting that TTSS or another TTSS-dependent factor is necessary for the cell death response. Contrary to our expectation that xopX high titer visual cell death response would be associated with non-host resistance, stable transgenic N. benthamiana lines expressing xopX displayed enhanced susceptibility, not resistance, to bacterial infection. There was enhanced susceptibility to not only Xcc but also to a wide range of other non-xopX expressing Xanthomonas and Pseudomonas strains. Therefore, xopX and the associated cell death response are unlike classical avr genes and their associated HR, in that it does not appear to be responsible for compromising pathogen virulence on a recognition-competent host. We propose that XopX bacterial effector protein targets basic innate immunity in plants, which in turn allows for greater pathogen virulence. This is also consistent with the increase in bacterial growth on N. benthamiana observed for Xcc with in trans XopX expression. XopX also contributes to the virulence of Xcv on host pepper and tomato plants, as site-directed xopX mutants of Xcv were less virulent in both of these two susceptible hosts. XopX targeting of basic innate immunity and the resulting increase in pathogen virulence is also consistent with the associated plant cell death response to high titer inoculations.
The biochemical function of XopX is currently unknown as homology searches have not identified proteins that have characterized enzymatic or molecular activities. However, our data demonstrate that XopX is an important virulence effector protein that is delivered by the TTSS of Xanthomonas to the plant cell and suppresses host defense. The contribution of xopX to pathogen virulence is complementary to another known Xcv virulence factor avrBs2 as evidenced by the additive contributions of the two genes to pathogen virulence in the absence of the pepper Bs2 resistance. Each virulence factor appears to make distinct, incremental assaults on the plant host, XopX doing so in part at least by suppressing host defenses. The importance of xopX to pathogen fitness is underscored by its high conservation among plant pathogenic Xanthomonas species.
The understanding of effector protein function as it relates to bacterial virulence and suppression of host defense is emerging as one of the most fascinating and significant areas of molecular plant pathology (Buell et al., 2003). The fact that any one phytopathogenic bacteria may contain as many as 40 effector proteins suggests that there are many ways in which a bacterium may establish a parasitic relationship with its host (Buell et al., 2003; Cornelis, 2002) (Schechter et al., 2004). The elucidation of the molecular basis of effector protein function will undoubtedly play an important role in our ultimate understanding of the molecular basis of bacterial pathogenesis. Furthermore, it has become evident that plants possess molecules that are the targets of bacterial effector proteins in either a direct or indirect manner. The identification and characterization of these interacting partners will be paramount to uncovering the molecular basis of recognition and the expression of plant disease.
Growth conditions for bacterial strains
Xanthomonas campestris strains were grown on NYG medium at 28°C (Daniels et al., 1984). Pseudomonas syringae strains were grown on Pseudomonas agar (Difco, Franklin Lakes, NJ, USA) at 28°C. Agrobacterium tumefaciens strains were grown on Luria agar at 28°C. All strains are referred in the Results section when introduced. Antibiotics were used at the following concentrations: ampicillin (Ap), 100 μg ml−1; rifampin (Rif), 100 μg ml−1; spectinomycin (Sp), 100 μg ml−1; chloramphenicol (Cm), 25 μg ml−1; tetracycline (Tc), 10 μg ml−1; and kanamycin (Km), 25 μg ml−1. Plasmids were introduced into Xc (Turner et al., 1984) and A. tumefaciens (Sambrook et al., 1989) by triparental matings using the helper plasmid pRK2013 (Figurski and Helinski, 1979) or a chloramphenicol-resistant helper plasmid pRK600 derived from pRK2013.
Plant material and plant inoculations
Nicotiana benthamiana, tomato cv. VF36, and pepper line ECW-0 (bs2/bs2) were grown in a greenhouse or a growth chamber and bacterial suspensions were inoculated as previously described (Tai et al., 1999). Bacterial suspensions of 4 × 108 CFU ml−1 were infiltrated into the intercellular spaces of fully expanded N. benthamiana leaves and cell death response reactions recorded over a 3-day period. Bacterial suspensions of 105 CFU ml−1 were used to start in planta bacterial growth assays for recording the bacterial growth as CFU cm−2 of leaf for several days (Tai et al., 1999). In planta growth assays have three replicated samples averaged with standard deviation error bars indicated for each sample point.
Recombinant DNA methods
Genomic DNA was isolated from bacterial strains using a modified previously described protocol (Tai and Tanksley, 1990). DNA gel electrophoresis and blotting to membrane was performed on restriction-digested DNA samples as described in Sambrook et al. (1989). Blotting was performed under alkali conditions to Hybond N+ membrane (Amersham Pharmacia, Piscataway, NJ, USA). Labeled probe was made from the 440 bp PCR product using primers 5′-ACTCTGG- CCGCGTGCCGCCAA-3′ and 5′-CGGCCGCGGCAAATCCCAGGT-3′ located between codons 148 and 313, and hybridized to blots, and detected using the Gene ImagesTM chemiluminescent detection system (Amersham Pharmacia, cat. no. RPN 3691) by recommended procedures.
Construction of Xcv GM98-38 library and subcloning xopX
A genomic library was made from the Xcv GM98-38 strain which gives a high titer (4 × 108 CFU ml−1) cell death response on N. benthamiana. Xcv genomic DNA was partially digested with Sau3A and fragments in the 20–25 kb size range were isolated by gel electrophoresis. DNA was ligated with BamHI-digested pLAFR3 and then packaged into λ phage particles (Stratagene, La Jolla, CA, USA). Phage was then used to deliver the resulting cosmids into E. coli DH5α and single colonies were selected with tetracycline 10 μg ml−1 (Staskawicz et al., 1984). Tri-parental mating was used to transfer the Xcv library into the strain Xcc 8004, which does not give a cell death response on N. benthamiana. Tn3-gus transposon mutagenesis of a pLAFR3 cosmid was used to further localize xopX activity (Huguet and Bonas, 1997). pTn3-gus is a modified pTn3-HoHo1 (Stachel et al., 1985) transposon. The EcoRI, ClaI sites flanking the lacZ gene of pTn3-HoHo1 were replaced with an EcoRI, BamHI, HindIII, and ClaI linker so that the BamHI, HindIII, kanamycin gene from pBRNeo (Southern, 1975) could be cloned in to make pTn3-HoKm. The HindIII site of pRAJ275 (Jefferson et al., 1986) was filled in and an EcoR1 site introduced so that the promoterless gus gene could now be excised as an EcoRI fragment and cloned into the EcoRI site of pTn3-HoKm to make pTn3-gus. The xopX containing cosmid pL3-9·22 was mutagenized with pTn3-gus as previously described by Stachel et al. (1985). These mutagenized pL3-9·22 were then moved back into Xcc 8004 by tri-parental mating and then screened for loss of ability to elicit a cell death response on N. benthamiana. The transposons were localized in mutagenized cosmids by restriction mapping using standard protocols (Ausubel et al., 1994). Restriction mapping showed all four TN3-Ho gus insertions resided within a single 4.7 kb EcoRI fragment of pL3-9·22. This fragment was subcloned into the EcoRI site of both pBluescript KS+ and pLAFR6 to make pB-E4.7 and pL6-E4.7, respectively. The portion of the transposon-tagged 4.7 kb EcoRI fragment plus the left inverted repeat to the EcoRI site of Tn3-gus was subcloned from the four pL3-9·22 mutants into pBluescript KS+ for sequencing and localizing the insertion sites. The smallest subclone, pB-ES2.8, was made by re-ligating SmaI-cleaved pB-E4.7 to remove a 1.8 kb SmaI fragment. The 2.8 kb EcoRI, BamHI fragment from pB-SE2.8 was then subcloned into EcoRI, BamHI cleaved pLAFR6 to make pL6-SE2.8. The sequence of the 2.8 kb SmaI, EcoRI fragment with xopX activity was completed using the ABI Prism FACS kit (Tai et al., 1999). Double-stranded pB-ES2.8 was sequenced using the T3 and T7 primers. Moreover the pBluescript KS+ subclones of the four TN3-gus insertion junctions were sequenced. New primers were designed based on these sequence data and used to complete the full sequence of the 2.8 kb SmaI, EcoRI fragment (Figure 2).
Single codon stop mutations to identify xopX open reading frame
Sequencing of the 2.8 kb SmaI, EcoRI fragment with xopX activity revealed two large overlapping open reading frames orfA and orfB. The phagemid pB-SE2.8 was transformed into E. coli CJ236 so that single-stranded DNA template of pB-ES2.8 was induced with M13K07 phage and used as template to synthesize double-stranded plasmid DNA with site-directed mutations (Ausubel et al., 1994). The primers 537stpA 5′-CCGCTCCACCCTA GCCTGGCTGC-3′ and 511stpB 5′-ACGCCACATAGCCGACGCAAAC -3′ were used to make pB-537stpA which puts a stop codon in only orfA and pB-511stpB which puts a stop codon in orfB only (Figure 2). The 2.8 kb EcoRI, BamHI fragments from pB-537stpA and pB-511stpB were then subcloned into EcoRI, BamHI cleaved pLAFR6 to make pL6-537stpA and pL6-511stpB, respectively. In addition there are two possible start sites for orfB so a stop codon, 361stpB, was introduced in orfB (Figure 2) by overlap PCR using primers 5′-TGGCTGGCGCGCCTGCTACATTTGAACAAGCCTGCC-3′ and 5′-GCAGGCGCGCCAGCCAGTA -3′ as described (Mudgett et al., 2000). A 2-kb overlap-PCR product was cloned in to pCR2.1TOPO to make pCR-SK361stp. The 1.2 kb NotI fragment from pB-SE2.8 was then replaced with the 1.2 kb NotI fragment from pCR-SK361stp to make pB-361stp. The 2.8 kb EcoRI, BamHI fragments from pB-361stpB was then subcloned into EcoRI, BamHI cleaved pLAFR6 to make pL6-361stpB.
Site-directed mutagenesis of genomic xopX in Xcv
Homologous recombination mutagenesis into the genome of Xcv isolates GM98-38 (with mutant avrBs2) and GM98-38-1 (with restored avrBs2) (Gassmann et al., 2000) was accomplished by first conjugating in the Tn3-gus mutagenized xopX cosmid pL3-9·22/2-46 and selecting for the cosmid marker with Tc, 10 μg ml−1 and for the Tn3-gus insertion with Km, 25 μg ml−1. These strains were then grown overnight on Km, 25 μg ml−1 only and re-plated as a mix to a fresh plate the next day. This cycling of the Xcv strains was repeated for 7–14 days and then plated for single colonies. Up to 1000 CFU were screened for double homologous recombination by replica plating onto Km, 25 μg ml−1 and Tc, 10 μg ml−1 and identified as Tc-sensitive and Km-resistant mutants. These marker exchange mutants were verified by testing for a lack of cell death response on N. benthamiana and by Southern analysis (data not shown) of EcoRI-digested genomic DNA, which shows a different alteration in the wild type EcoRI 4.7 kb fragment for each insertion event (data not shown). These strains were used in standard in planta bacterial growth assays (Tai et al., 1999) for recording the bacterial growth of Xc as CFU cm−2 of leaf for several days. Restoration of xopX activity in these exchange mutants with pL6-SE2.8 was verified by testing for cell death response on N. benthamiana. Although pL6-540stpA and pL6-511stpB show that only orfB confers xopX cell death response activity on N. benthamiana they cannot be used for testing the virulence function of orfA and orfB. As double homologous recombination mutation in GM98-382x2-46 blocks both orfA and orfB it was necessary to introduce mutations 540stpA and 511stpB into the genomic copy of xopX. The 2.8 kb SpeI, XhoI fragments from pB-540stpA and pB-511stpB were then subcloned into the XbaI, XhoI-cleaved suicide construct pMsacB to make pMsacB-540stpA and pMsacB-511stpB, respectively. To construct pMsacB the PstI sacB-sacR fragment from pSD800 (Gassmann et al., 2000) was cloned into the PstI site of pLVC18L (Casper-Lindley et al., 2002) to make pLVC18LsacB. The pBluescript KS+ linker from PCR with primers T7 and T3 was cloned into the SmaI site of pLVC18LsacB to make pMsacB. To introduce mutations 540stpA and 511stpB into the genomic copy of xopX the constructs pMsacB-540stpA and pMsacB-511stpB were first conjugated into GM98-38 and selected on Tc 10 μg ml−1 for genomic recombination rescue to the suicide constructs. These single recombination events were cycled overnight without Tc selection for 1 week and then plated on NYGA with 5% sucrose added. The strains GM98-38(540stpA) and GM98-38(511stpB) were isolated by PCR amplifying a number of single colonies for xopX and sequencing to find resolution events that introduced the 540stpA and 511stpB mutations into the genomic copy of xopX. These strains were used in standard in planta bacterial growth assays (Tai et al., 1999) for recording the bacterial growth of Xc as CFU cm−2 of leaf for several days.
Translational XopX fusions with Cya
XopX translational fusions with Cya were made by first cloning directionally into gateway compatible pENTR/D-TOPO (Invitrogen, Carlsbad, CA, USA) the two xopX PCR amplified products of an 879-bp fragment including the promoter region and the first 183 codons, with primer pairs 27Entry 5′-CACCTGGCAATGCGCTGCAACGA-3′ and 906Entry 5′-ATCGCGCAGCTGCGTGATCTGT-3′, and a 577-bp fragment including codons 527 to the last codon 719, with primer pairs 1937Entry 5′-CACCGTCAGGACGCCATGGAACGCT-3′ and 2514Entry 5′-GGACGAAGGCACAGT-3′. These two fragments were then transferred from this entry vector to the Gateway-compatible destination vector pDDesCya using the recommended protocol for an LR reaction (Invitrogen) to make the two suicide reporter constructs pDDesCya(xopX1−183:cya) and pDDesCya(xopX527−719:cya). Cloning the EcoRI, KpnI linker from pBluescript KS+ into EcoRI made the suicide Cya reporter Gateway-compatible destination vector pDDesCya, KpnI-digested pLVC18L (Casper-Lindley et al., 2002) to make pLVC18L2. This construct was then digested with XhoI, EcoRI and the XhoI, EcoRI and the cya-reporter fragment from pUC19(cya) (Casper-Lindley et al., 2002) cloned in to make pLVC18L2(cya). The attR1-ccdB-attR2 B destination vector cassette (Invitrogen) was cloned in the correct orientation into the SmaI site of pLVC18L2(cya) to make pDDesCya such that an LR reaction with a Gateway-compatible pENTR/D-TOPO(PCR clone ending on a target orf codon) will be switched for the ccdB cassette in the pDDesCya in the correct orientation and with the last codon of the clone in frame with Cya reporter. The two suicide reporter constructs pDDesCya(xopX1−183:cya) and pDDesCya(xopX527−719:cya) were conjugated into GM98-38 and selected on Tc 10 μg ml−1 for genomic recombination rescue of the two suicide constructs as single integration events to create the Cya reporter strains GM98-38(xopX183:cya) and GM98-38(xopX719:cya). Reporter construct pDDesCya(xopX1−183:cya) was also marker exchanged into Xcv strains 85* and 85*hrcV to create the Cya reporter strains 85*(xopX183:cya) and 85*hrcV(xopX183:cya). A non-translocated suicide reporter construct pDDesCya(RecA:cya) was made by cloning the PCR product from primers RecA Entry N 5′-CACCAGATCGAAAAGCAATTCGGCA-3′ and Rec A Entry C 5′-TGGAGCGCTTGATGTTGCCGGT-3′ first into pENTR/D-TOPO and then into pDesCya by LR reaction. Genomic single recombination rescue of this construct was used to make the strain GM98-38(RecA:cya). The previously published protocol (Casper-Lindley et al., 2002) for Cya activity assay and cAMP extraction was used to measure levels of cAMP in bacterial-plant interactions at 8 h post-inoculation.
Epitope-tagged xopX and Agrobacterium-mediated transient expression
An HA epitope with a C-terminal stop and Spe1 site was translationally fused to the last codon of xopX by PCR with the primers 2514HA 5′-CTATGCGTAGTCTGGTACGTCGTACGGATAACTAGTGGACGAAGGCACAGTGCTGGCT-3′ and 2027F 5′-TCGCAGACAAGAGCCTGATCGAA-3′. This 423 bp PCR fragment was cloned into pCR2.1-TOPO and screened for the correct orientation so that a KpnI digest cuts the single site in xopX and the vector site to excise a 410-bp C-terminal HA-tagged KpnI fragment which is then used to replace the untagged C-terminal KpnI fragment from pB-SE2.8 to make pB-SE2.8:HA. The 3.0 kb BamHI, EcoRI fragment from pB-SE2.8:HA was cloned into the BamHI, EcoRI sites of pVSP61 to make pV-SE2.8:HA. For Agrobacterium-mediated transient expression of xopX with an HA epitope the primers 344F 5′-GCAGGCTTGTTCAAATGATG-3′ and 1789R 5′-ACTTGGGGGCGATGTCGTGGTT were used to PCR amplify 1.45 kb fragment beginning 14 bp in front of orfB start codon and extending past the single SacI site in xopX. The PCR template pB537stpA was used so that orfA stop was included in the 1.45 kb fragment cloned into pCR2.1-TOPO and screened for the primer 344F on the XbaI side of the vector to make pCR-344·1789. The 0.95 kb SacI fragment from pB-SE2.8:HA replaces the 0.15 kb fragment from pCR-344·1789 to make pCR-14orfB·stpA:HA. The 2.3 kb XbaI, BamHI, fragment from pCR-14orfB·stpA:HA was cloned into the XbaI, BamHI, sites of pMD1 to make the 35S promoter Agrobacterium binary construct pM-14orfB·stpA:HA. This binary xopX expression construct was conjugated into the Agrobacterium strain C58C1 as previously described (Bendahmane et al., 2000; Tai et al., 1999). Agrobacterium C58C1 pM-14orfB·stpA:HA was suspended in induction media at 6 × 108 CFU ml−1 3 h prior to infiltration into N. benthamiana leaves. Leaf samples were collected 24 h post-infiltration for protein isolation.
Generation and analysis of stable transgenic N. benthamiana plants
For plant transformation binary construct pM-14orfB·stpA:HA was conjugated into Agrobacterium strain LBA4404 and standard procedures were used (Tai et al., 1999) to make a stable transgenic N. benthamiana-expressing 35S-xopX:HA line. This line was selected for its 3:1 single locus segregation of 35S-xopX:HA from 12 independent primary transformants. A T3 homozygous 35S-xopX:HA line was selected and that gave a comparable cell death response phenotype to that reported for non- transgenic N. benthamiana inoculated with Xcc-expressing xopX. The T4 self cross of this T3 line was used for analysis.
Immunoblot analysis of proteins from Xcc pV-SE2.8:HA and stable transgenic N. benthamiana expressing xopX:HA co-inoculated with Xcc at 0 and 8 h post-infiltration were transferred to nitrocellulose from SDS/PAGE gels and probed with mouse monoclonal anti-HA-HRP conjugate (16B12; Covance, Berkeley, CA, USA) antisera as previously described (Mudgett et al., 2000). Total protein extracts from Xcv and Agrobacterium-inoculated plants were isolated and quantified for equal loading using the EZ protein protocol (Martinez-Garcia et al., 1999).
We thank members of the Staskawicz lab for insightful discussions and comments on this manuscript.
This work was funded by Department of Energy Grant DE-FG03-88ER139.
- 1997) Plant Pathology. San Diego: Academic Press. (
- 2004) Type III secretion system effector proteins: double agents in bacterial disease and plant defense. Annu. Rev. Phytopathol. 42, 385–414. and (
- 1990) Basic local alignment search tool. J. Mol. Biol. 215, 403–410. , , , and (
- 1994) Current Protocols in Molecular Biology. New York: John Wiley & Sons. , , , , , and (
- 2003) Initiation of RPS2-specified disease resistance in Arabidopsis is coupled to the AvrRpt2-directed elimination of RIN4. Cell, 112, 369–377. and (
- 2000) Agrobacterium transient expression system as a tool for the isolation of disease resistance genes: Application to the Rx2 locus in potato. Plant J. 21, 73–81. , , and (
- 1996) Unified nomenclature for broadly conserved Hrp genes of phytopathogenic bacteria. Mol. Microbiol. 20, 681–683. , , et al. (
- 2003) The complete genome sequence of the Arabidopsis and tomato pathogen Pseudomonas syringae pv. tomato DC3000. Proc. Natl Acad. Sci. USA, 100, 10181–10186. , , et al. (
- 2003) Common infection strategies of plant and animal pathogenic bacteria [Review]. Curr. Opin. Plant Biol. 6, 312–319. and (
- 2002) Direct biochemical evidence for type III secretion-dependent translocation of the AvrBs2 effector protein into plant cells. Proc. Natl Acad. Sci. USA, 99, 8336–8341. , , and (
- 2000) avrPto enhances growth and necrosis caused by Pseudomonas syringae pv. tomato in tomato lines lacking either Pto or Prf. Mol. Plant Microbe Interact. 13, 568–571. , , , and (
- 2000) The Pseudomonas syringae avrRpt2 gene product promotes pathogen virulence from inside plant cells. Mol. Plant Microbe Interact. 13, 1312–1321. , , , and (
- 2002) Yersinia type III secretion: send in the effectors [Review]. J. Cell Biol. 158, 401–408. (
- 1986) Generation and characterization of Tn5 insertion mutants in Pseudomonas syringae pv. tomato. Appl. Environ. Microbiol. 52, 323–327. (
- 2001) Plant pathogens and integrated defence responses to infection [Review]. Nature., 411, 826–833. and (
- 1984) Isolations of mutations of Xanthomonas campestris pv. campestris showing altered pathogenicity. J. Gen. Microbiol. 130, 2447–2455. , , , and (
- 2000) Structure, function and evolution of plant disease resistance genes. Curr. Opin. Plant Biol. 3, 278–284. , and (
- 1992) Determinants of pathogenicity in Xanthomonas campestris pv. vesicatoria are related to proteins involved in the secretion in bacterial pathogens of animals. Mol. Plant Microbe Interact. 5, 390–396. , and (
- 1979) Replication of an origin-containing derivative of plasmid RK2 dependent on a plasmid function provided in trans. Proc. Natl Acad. Sci. USA, 76, 1648–1652. and (
- 1999) Why do pathogens carry avirulence genes? Physiol. Mol. Plant Pathol. 55, 205–214. (
- 2000) Molecular evolution of virulence in natural field strains of Xanthomonas campestris pv. vesicatoria. J. Bacteriol. 182, 7053–7059. , , , , and (
- 2001) Functional analysis of the type III effectors AvrRpt2 and AvrRpm1 of Pseudomonas syringae with the use of a single-copy genomic integration system. Mol. Plant Microbe Interact. 14, 145–155. and (
- 1997) Plant disease resistance genes [Review]. Annu. Rev. Plant Physiol. Plant Mol. Biol. 48, 575–607. and (
- 1997) Hrpf of Xanthomonas Campestris pv. vesicatoria encodes an 87-kDa protein with homology to Nolx of Rhizobium fredii. Mol. Plant Microbe Interact. 10, 488–498. and (
- 1986) B-Glucuronidase from Escherichia coli as a gene-fusion marker. Proc. Natl Acad. Sci. USA, 83, 8447–8451. , and (
- 1990) Widespread distribution and fitness contribution of Xanthomonas campestris avirulence gene avrBs2. Nature (Lond), 346, 385–386. and (
- 2003) Arabidopsis RIN4 is a target of the type III virulence effector AvrRpt2 and modulates RPS2-mediated resistance. Cell, 112, 379–389. , , , and (
- 1999) A simple, rapid and quantitative method for preparing Arabidopsis protein extracts for immunoblot analysis. Plant. J. 20, 251–257. , and (
- 2003) Genome-wide analysis of NBS-LRR-encoding genes in Arabidopsis [Review]. Plant Cell 15, 809–834. , , , and (
- 2000) Molecular signals required for type III secretion and translocation of the Xanthomonas campestris AvrBs2 protein to pepper plants. Proc. Natl Acad. Sci. USA, 97, 13324–13329. , , , , , and (
- 2002) Two novel type III-secreted proteins of Xanthomonas campestris pv. vesicatoria are encoded within the hrp pathogenicity island. J. Bacteriol. 184, 1340–1348. , , and (
- 2002) Ubiquitin ligase-associated protein SGT1 is required for host and non-host disease resistance in plants. Proc. Natl Acad. Sci. USA, 99, 10865–10869. , , et al. (
- 1995) The avrRPm1 gene of Pseudomonas syringae pv. maculicola is required for virulence on Arabidopsis. Mol. Plant Microbe Interact. 8, 444–453. and (
- 1989) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor: Cold Spring Harbor Laboratory Press. , and (
- 2004) Pseudomonas syringae type III secretion system targeting signals and novel effectors studied with a Cya translocation reporter. J. Bacteriol. 186, 543–555. , , , and (
- 2003) Cleavage of Arabidopsis PBS1 by a bacterial type III effector. Science, 301, 1230–1233. , , , , and (
- 1995) Identification of the YopE and YopH domains required for secretion and internalization into the cytosol of macrophages, using the cyaA gene fusion approach. Proc. Natl Acad. Sci. USA, 92, 11998–12002. , , and (
- 1975) Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98, 503. (
- 1985) A Tn3-lacZ transposon for the random generation of beta-galactosidase gene fusions: application to the analysis of gene expression in Agrobacterium. EMBO J. 4, 891–898. , , and (
- 1984) Cloned avirulence gene of Pseudomonas syringae pv. glycinea determines race-specific incompatibility on Glycine max (L.) Merr. Proc. Natl Acad. Sci. USA, 81, 6024–6028. , and (
- 1996) Spontaneous and induced mutations in a single open reading frame alter both virulence and avirulence in Xanthomonas campestris pv. vesicatoria avrBs2. J. Bacteriol. 178, 4661–4669. , , , and (
- 2002) Type III-dependent translocation of the Xanthomonas AvrBs3 protein into the plant cell. Mol. Microbiol. 46, 13–23. , , and (
- 1990) A rapid and inexpensive method for isolation of total DNA from dehydrated plant tissue. Plant Mol. Biol. Rep. 8, 297–303. and (
- 1999) Expression of the Bs2 pepper gene confers resistance to bacterial spot disease in tomato. Proc. Natl Acad. Sci. USA, 96, 14153–14158. , , , , , , and (
- 1984) Behaviour of the transposons Tn5 and Tn7 in Xanthomonas campestris pv. campestris. Mol. Gen. Genet. 195, 101–107. , and (
- 1999) Mutations in the regulatory gene hrpG of Xanthomonas campestris pv. vesicatoria result in constitutive expression of all hrp genes. J. Bacteriol. 181, 6828–6831. , and (
- 1984) The structure of an antigenic determinant in a protein. Cell, 3, 767–778. , , , , and (
- 2000) The virulence factor AvrXa7 of Xanthomonas oryzae pv. oryzae is a type III secretion pathway-dependent nuclear-localized double-stranded DNA- binding protein. Proc. Natl Acad. Sci. USA, 97, 9807–9812. , , and (