Visualization and characterization of Pseudomonas syringae pv. tomato DC3000 pellicles

Cellulose, whose production is controlled by c-di-GMP, is a commonly found exopolysaccharide in bacterial biofilms. Pseudomonas syringae pv. tomato (Pto) DC3000, a model organism for molecular studies of plant–pathogen interactions, carries the wssABCDEFGHI operon for the synthesis of acetylated cellulose. The high intracellular levels of the second messenger c-di-GMP induced by the overexpression of the heterologous diguanylate cyclase PleD stimulate cellulose production and enhance air–liquid biofilm (pellicle) formation. To characterize the mechanisms involved in Pto DC3000 pellicle formation, we studied this process using mutants lacking flagella, biosurfactant or different extracellular matrix components, and compared the pellicles produced in the absence and in the presence of PleD. We have discovered that neither alginate nor the biosurfactant syringafactin are needed for their formation, whereas cellulose and flagella are important but not essential. We have also observed that the high c-di-GMP levels conferred more cohesion to Pto cells within the pellicle and induced the formation of intracellular inclusion bodies and extracellular fibres and vesicles. Since the pellicles were very labile and this greatly hindered their handling and processing for microscopy, we have also developed new methods to collect and process them for scanning and transmission electron microscopy. These techniques open up new perspectives for the analysis of fragile biofilms in other bacterial strains.     

Evidence on antimicrobial properties and mode of action of a chitosan obtained from crustacean exoskeletons on Pseudomonas syringae pv. tomato DC3000

Pseudomonas syringae pv. tomato DC3000 (Pto DC3000) causes bacterial speck of tomato, a widely spread disease that causes significant economical losses worldwide. It is representative of many bacterial plant diseases for which effective controls are still needed. Despite the antimicrobial properties of chitosan has been previously described in phytopathogenic fungi, its action on bacteria is still poorly explored. In this work, we report that the chitosan isolated from shrimp exoskeletons (70 kDa and 78 % deacetylation degree) exerts cell damage on Pto DC3000. Chitosan inhibited Pto DC3000 bacterial growth depending on its concentration, medium-pH, and presence of metal ion (Mg⁺²). Biochemical and cellular changes resulting in cell aggregation and impaired bacterial growth were also viewed. In vivo studies using fluorescent probes showed cell aggregation, increase in membrane permeability, and cell death, suggesting the chitosan antibacterial activity is due to its interaction as a polycation with Pto DC3000 membranes. Transmission electron microscopic analysis revealed that chitosan also caused morphological changes and damage in bacterial surfaces. Also, the disease incidence in tomato inoculated with Pto DC3000 was significantly reduced in chitosan pretreated seedlings, revealing a promising action of chitosan as nontoxic biopesticide in tomato plants. Indeed, a wider comprehensive knowledge of the mechanism of action of chitosan in phytopathogenic bacterial cells will increase the chances of its successful application to the control of spread disease in plants.     

The c‐di‐GMP phosphodiesterase BifA is involved in the virulence of bacteria from the Pseudomonas syringae complex

In a recent screen for novel virulence factors involved in the interaction between Pseudomonas savastanoi pv. savastanoi and the olive tree, a mutant was selected that contained a transposon insertion in a putative cyclic diguanylate (c‐di‐GMP) phosphodiesterase‐encoding gene. This gene displayed high similarity to bifA of Pseudomonas aeruginosa and Pseudomonas putida. Here, we examined the role of BifA in free‐living and virulence‐related phenotypes of two bacterial plant pathogens in the Pseudomonas syringae complex, the tumour‐inducing pathogen of woody hosts, P. savastanoi pv. savastanoi NCPPB 3335, and the pathogen of tomato and Arabidopsis, P. syringae pv. tomato DC3000. We showed that deletion of the bifA gene resulted in decreased swimming motility of both bacteria and inhibited swarming motility of DC3000. In contrast, overexpression of BifA in P. savastanoi pv. savastanoi had a positive impact on swimming motility and negatively affected biofilm formation. Deletion of bifA in NCPPB 3335 and DC3000 resulted in reduced fitness and virulence of the microbes in olive (NCPPB 3335) and tomato (DC3000) plants. In addition, real‐time monitoring of olive plants infected with green fluorescent protein (GFP)‐tagged P. savastanoi cells displayed an altered spatial distribution of mutant ΔbifA cells inside olive knots compared with the wild‐type strain. All free‐living phenotypes that were altered in both ΔbifA mutants, as well as the virulence of the NCPPB 3335 ΔbifA mutant in olive plants, were fully rescued by complementation with P. aeruginosa BifA, whose phosphodiesterase activity has been demonstrated. Thus, these results suggest that P. syringae and P. savastanoi BifA are also active phosphodiesterases. This first demonstration of the involvement of a putative phosphodiesterase in the virulence of the P. syringae complex provides confirmation of the role of c‐di‐GMP signalling in the virulence of this group of plant pathogens.     

Construction and Use of a Nonradioactive DNA Hybridization Probe for Detection of Pseudomonas syringae pv. Tomato on Tomato Plants

Pseudomonas syringae pv. tomato, the causal agent for bacterial speck of tomato, produces the phytotoxin coronatine. A 5.3-kilobase XhoI fragment from the chromosomal region controlling toxin production was cloned into the plasmid pGB2, and the resulting recombinant plasmid, pTPR1, was tested for its ability to serve as a diagnostic probe for P. syringae pv. tomato. In a survey of 75 plant-associated bacteria, pTPR1 hybridized exclusively to those strains that produced coronatine. The detection limit for this probe, which was labeled with the Chemiprobe nonradioactive reporter system, was approximately 4 × 10³ CFU of lesion bacteria. During the 1989 growing season, a total of 258 leaf and fruit lesions from nine tomato fields were screened for P. syringae pv. tomato by using pTPR1 and the culture method of detection. The best agreement between the two methods, 90%, occurred early in the season with samples taken from relatively young (5-week-old) plants. Young plants also had a higher percentage of P. syringae pv. tomato-positive lesions. P. syringae pv. tomato was the only coronatine producer recovered from the nine tomato fields. All 244 P. syringae pv. tomato strains isolated during this study reacted strongly with the probe. The P. syringae pv. tomato population of healthy field tomato leaves was determined by a pTPR1 colony hybridization procedure. Every probe-positive colony that was isolated and characterized was identified as P. syringae pv. tomato. The pTPR1 probe should expedite disease diagnosis and facilitate epidemiological studies of this pathogen. It also should aid in screening transplant seedlings for bacterial speck infestation.     

Pseudomonas syringae lytic transglycosylases coregulated with the type III secretion system contribute to the translocation of effector proteins into plant cells

Pseudomonas syringae translocates virulence effector proteins into plant cells via a type III secretion system (T3SS) encoded by hrp (for hypersensitive response and pathogenicity) genes. Three genes coregulated with the Hrp T3SS system in P. syringae pv. tomato DC3000 have predicted lytic transglycosylase domains: PSPTO1378 (here designated hrpH), PSPTO2678 (hopP1), and PSPTO852 (hopAJ1). hrpH is located between hrpR and avrE1 in the Hrp pathogenicity island and is carried in the functional cluster of P. syringae pv. syringae 61 hrp genes cloned in cosmid pHIR11. Strong expression of DC3000 hrpH in Escherichia coli inhibits bacterial growth unless the predicted catalytic glutamate at position 148 is mutated. Translocation tests involving C-terminal fusions with a Cya (Bordetella pertussis adenylate cyclase) reporter indicate that HrpH and HopP1, but not HopAJ1, are T3SS substrates. Pseudomonas fluorescens carrying a pHIR11 derivative lacking hrpH is poorly able to translocate effector HopA1, and this deficiency can be restored by HopP1 and HopAJ1, but not by HrpH(E148A) or HrpH_1-241. DC3000 mutants lacking hrpH or hrpH, hopP1, and hopAJ1 combined are variously reduced in effector translocation, elicitation of the hypersensitive response, and virulence. However, the mutants are not reduced in secretion of T3SS substrates in culture. When produced in wild-type DC3000, the HrpH(E148A) and HrpH_1-241 variants have a dominant-negative effect on the ability of DC3000 to elicit the hypersensitive response in nonhost tobacco and to grow and cause disease in host tomato. The three Hrp-associated lytic transglycosylases in DC3000 appear to have overlapping functions in contributing to T3SS functions during infection.     

Protection of Tomato Seedlings against Infection by Pseudomonas syringae pv. Tomato by Using the Plant Growth-Promoting Bacterium Azospirillum brasilense

Pseudomonas syringae pv. tomato, the causal agent of bacterial speck of tomato, and the plant growth-promoting bacterium Azospirillum brasilense were inoculated onto tomato plants, either alone, as a mixed culture, or consecutively. The population dynamics in the rhizosphere and foliage, the development of bacterial speck disease, and their effects on plant growth were monitored. When inoculated onto separate plants, the A. brasilense population in the rhizosphere of tomato plants was 2 orders of magnitude greater than the population of P. syringae pv. tomato (10⁷ versus 10⁵ CFU/g [dry weight] of root). Under mist chamber conditions, the leaf population of P. syringae pv. tomato was 1 order of magnitude greater than that of A. brasilense (10⁷ versus 10⁶ CFU/g [dry weight] of leaf). Inoculation of seeds with a mixed culture of the two bacterial strains resulted in a reduction of the pathogen population in the rhizosphere, an increase in the A. brasilense population, the prevention of bacterial speck disease development, and improved plant growth. Inoculation of leaves with the mixed bacterial culture under mist conditions significantly reduced the P. syringae pv. tomato population and significantly decreased disease severity. Challenge with P. syringae pv. tomato after A. brasilense was established in the leaves further reduced both the population of P. syringae pv. tomato and disease severity and significantly enhanced plant development. Both bacteria maintained a large population in the rhizosphere for 45 days when each was inoculated separately onto tomato seeds (10⁵ to 10⁶ CFU/g [dry weight] of root). However, P. syringae pv. tomato did not survive in the rhizosphere in the presence of A. brasilense. Foliar inoculation of A. brasilense after P. syringae pv. tomato was established on the leaves did not alleviate bacterial speck disease, and A. brasilense did not survive well in the phyllosphere under these conditions, even in a mist chamber. Several applications of a low concentration of buffered malic acid significantly enhanced the leaf population of A. brasilense (>10⁸ CFU/g [dry weight] of leaf), decreased the population of P. syringae pv. tomato to almost undetectable levels, almost eliminated disease development, and improved plant growth to the level of uninoculated healthy control plants. Based on our results, we propose that A. brasilense be used in prevention programs to combat the foliar bacterial speck disease caused by P. syringae pv. tomato.     

The plant growth-promoting rhizobacterium Bacillus cereus AR156 induces resistance in tomato with induction and priming of defence response

In a previous study, we demonstrated the ability of the rhizobacterium Bacillus cereus AR156 (AR156) to protect tomato against bacterial wilt caused by Ralstonia solanacearum and root-knot disease caused by Meloidogyne incognita. Here, we investigate the ability of AR156 to promote plant growth and its role in the systemic protection of tomatoes cultivated in greenhouses against bacterial speck disease caused by Pseudomonas syringae pv. tomato DC3000 (DC3000). In our experiments, the AR156 population reached 10⁵–10⁶ CFU/g rhizosphere soil, and remained at that level in the rhizosphere of tomato plants for more than 2 months. In terms of its ability to promote plant growth, AR156 increased the average biomass of the tomato by 47.7%. AR156 also elicited induced systemic resistance against DC3000, significantly reduced bacterial speck disease severity 1.6-fold, and inhibited proliferation of the pathogen by approximately 15-fold. This strain triggered the accumulation of defence-related genes (PR1 and PIN2) in tomato leaves and primed the leaves for accelerated defence-related gene expression upon challenge with DC3000. That suggested simultaneous activation of the salicylic acid and the jasmonic acid dependent signalling pathways by AR156 against DC3000. In conclusion, B. cereus AR156 was found to form robust colonies in the roots of tomato and had some beneficial effects, including biological control of bacterial speck disease via ISR and promotion of plant growth.     

Pseudomonas syringae colonizes distant tissues in Nicotiana benthamiana through xylem vessels

The ability to move from the primary infection site and colonize distant tissue in the leaf is an important property of bacterial plant pathogens, yet this aspect has hardly been investigated for model pathogens. Here we show that GFP‐expressing Pseudomonas syringae pv. syringae DC3000 that lacks the HopQ1‐1 effector (PtoDC3000ΔhQ) has a strong capacity to colonize distant leaf tissue from wound‐inoculated sites in N. benthamiana. Distant colonization occurs within 1 week after toothpick inoculation and is characterized by distant colonies in the apoplast along the vasculature. Distant colonization is blocked by the non‐host resistance response triggered by HopQ1‐1 in an SGT1‐dependent manner and is associated with an explosive growth of the bacterial population, and displays robust growth differences between compatible and incompatible interactions. Scanning electron microscopy revealed that PtoDC3000ΔhQ bacteria are present in xylem vessels, indicating that they use the xylem to move through the leaf blade. Distant colonization does not require flagellin‐mediated motility, and is common for P. syringae pathovars that represent different phylogroups.     

ATG5 is required to limit cell death induced by <Emphasis Type="Italic">Pseudomonas syringae</Emphasis> in <Emphasis Type="Italic">Arabidopsis</Emphasis> and may be mediated by the salicylic acid pathway

Autophagy can be regarded as a protection mechanism to restrict programmed cell death (PCD) induced by pathogen infection during plant innate immunity in the early stages. Autophagy related 5 (ATG5) plays an important role in autophagy in Arabidopsis. We investigated the function of ATG5 in Arabidopsis in the hypersensitive response (HR)-PCD elicited by both virulent and avirulent strains of Pseudomonas syringae pv. tomato bacteria DC3000. Results show that ATG5 plays a vital role in limiting HR induced by P. syringae strains and colocalizes with autophagic bodies during the early phase of bacterial infection. In addition, the P. syringae-induced response is mediated by the salicylic acid (SA) signaling pathway. In summary, ATG5 is required for limiting HR-PCD induced in Arabidopsis by P. syringae strains and may be mediated by SA signaling.     

Identification of a biosynthetic gene cluster and the six associated lipopeptides involved in swarming motility of Pseudomonas syringae pv. tomato DC3000

Pseudomonas species are known to be prolific producers of secondary metabolites that are synthesized wholly or in part by nonribosomal peptide synthetases. In an effort to identify additional nonribosomal peptides produced by these bacteria, a bioinformatics approach was used to "mine" the genome of Pseudomonas syringae pv. tomato DC3000 for the metabolic potential to biosynthesize previously unknown nonribosomal peptides. Herein we describe the identification of a nonribosomal peptide biosynthetic gene cluster that codes for proteins involved in the production of six structurally related linear lipopeptides. Structures for each of these lipopeptides were proposed based on amino acid analysis and mass spectrometry analyses. Mutations in this cluster resulted in the loss of swarming motility of P. syringae pv. tomato DC3000 on medium containing a low percentage of agar. This phenotype is consistent with the loss of the ability to produce a lipopeptide that functions as a biosurfactant. This work gives additional evidence that mining the genomes of microorganisms followed by metabolite and phenotypic analyses leads to the identification of previously unknown secondary metabolites.     

The HopQ1 Effector’s Nucleoside Hydrolase-Like Domain Is Required for Bacterial Virulence in Arabidopsis and Tomato,but Not Host Recognition in Tobacco

Bacterial pathogens deliver multiple effector proteins into host cells to facilitate bacterial growth. HopQ1 is an effector from Pseudomonas syringae pv. tomato DC3000 that is conserved across multiple bacterial pathogens which infect plants. HopQ1’s central region possesses some homology to nucleoside hydrolases, but possesses an alternative aspartate motif not found in characterized enzymes. A structural model was generated for HopQ1 based on the E. coli RihB nucleoside hydrolase and the role of HopQ1’s potential catalytic residues for promoting bacterial virulence and recognition in Nicotiana tabacum was investigated. Transgenic Arabidopsis plants expressing HopQ1 exhibit enhanced disease susceptibility to DC3000. HopQ1 can also promote bacterial virulence on tomato when naturally delivered from DC3000. HopQ1’s nucleoside hydrolase-like domain alone is sufficient to promote bacterial virulence, and putative catalytic residues are required for virulence promotion during bacterial infection of tomato and in transgenic Arabidopsis lines. HopQ1 is recognized and elicits cell death when transiently expressed in N. tabacum. Residues required to promote bacterial virulence were dispensable for HopQ1’s cell death promoting activities in N. tabacum. Although HopQ1 has some homology to nucleoside hydrolases, we were unable to detect HopQ1 enzymatic activity or nucleoside binding capability using standard substrates. Thus, it is likely that HopQ1 promotes pathogen virulence by hydrolyzing alternative ribose-containing substrates in planta.     

Nutritional Similarity between Leaf-Associated Nonpathogenic Bacteria and the Pathogen Is Not Predictive of Efficacy in Biological Control of Bacterial Spot of Tomato

It has been demonstrated that for a nonpathogenic, leaf-associated bacterium, effectiveness in the control of bacterial speck of tomato is correlated with the similarity in the nutritional needs of the nonpathogenic bacterium and the pathogen Pseudomonas syringae pv. tomato. This relationship was investigated further in this study by using the pathogen Xanthomonas campestris pv. vesicatoria, the causal agent of bacterial spot of tomato, and a collection of nonpathogenic bacteria isolated from tomato foliage. The effects of inoculation of tomato plants with one of 34 nonpathogenic bacteria prior to inoculation with the pathogen X. campestris pv. vesicatoria were quantified by determining (i) the reduction in disease severity (number of lesions per square centimeter) in greenhouse assays and (ii) the reduction in leaf surface pathogen population size (log₁₀ of the number of CFU per leaflet) in growth chamber assays. Nutritional similarity between the nonpathogenic bacteria and X. campestris pv. vesicatoria was quantified by using either niche overlap indices (NOI) or relatedness in cluster analyses based upon in vitro utilization of carbon or nitrogen sources reported to be present in tomato tissues or in Biolog GN plates. In contrast to studies with P. syringae pv. tomato, nutritional similarity between the nonpathogenic bacteria and the pathogen X. campestris pv. vesicatoria was not correlated with reductions in disease severity. Nutritional similarity was also not correlated with reductions in pathogen population size. Further, the percentage of reduction in leaf surface pathogen population size was not correlated with the percentage of reduction in disease severity, suggesting that the epiphytic population size of X. campestris pv. vesicatoria is not related to disease severity and that X. campestris pv. vesicatoria exhibits behavior in the phyllosphere prior to lesion formation that is different from that of P. syringae pv. tomato.     

Contribution of the non-effector members of the HrpL regulon, <Emphasis Type="Italic">iaaL</Emphasis> and <Emphasis Type="Italic">matE</Emphasis>, to the virulence of <Emphasis Type="Italic">Pseudomonas syringae</Emphasis> pv. tomato DC3000 in tomato plants

Background

The phytohormone indole-3-acetic acid (IAA) is widely distributed among plant-associated bacteria. Certain strains of the Pseudomonas syringae complex can further metabolize IAA into a less biologically active amino acid conjugate, 3-indole-acetyl-ε-L-lysine, through the action of the iaaL gene. In P. syringae and Pseudomonas savastanoi strains, the iaaL gene is found in synteny with an upstream gene, here called matE, encoding a putative MATE family transporter. In P. syringae pv. tomato (Pto) DC3000, a pathogen of tomato and Arabidopsis plants, the HrpL sigma factor controls the expression of a suite of virulence-associated genes via binding to hrp box promoters, including that of the iaaL gene. However, the significance of HrpL activation of the iaaL gene in the virulence of Pto DC3000 is still unclear.

Results

A conserved hrp box motif is found upstream of the iaaL gene in the genomes of P. syringae strains. However, although the promoter region of matE is only conserved in genomospecies 3 of this bacterial group, we showed that this gene also belongs to the Pto DC3000 HrpL regulon. We also demonstrated that the iaaL gene is transcribed both independently and as part of an operon with matE in this pathogen. Deletion of either the iaaL or the matE gene resulted in reduced fitness and virulence of Pto DC3000 in tomato plants. In addition, we used multicolor fluorescence imaging to visualize the responses of tomato plants to wild-type Pto DC3000 and to its ΔmatE and ΔiaaL mutants. Activation of secondary metabolism prior to the development of visual symptoms was observed in tomato leaves after bacterial challenges with all strains. However, the observed changes were strongest in plants challenged by the wild-type strain, indicating lower activation of secondary metabolism in plants infected with the ΔmatE or ΔiaaL mutants.

Conclusions

Our results provide new evidence for the roles of non-type III effector genes belonging to the Pto DC3000 HrpL regulon in virulence.

     

Regulons of Three Pseudomonas syringae pv. tomato DC3000 Iron Starvation Sigma Factors

Pseudomonas syringae pv. tomato DC3000 contains genes for 15 sigma factors. The majority are members of the extracytoplasmic function class of sigma factors, including five that belong to the iron starvation subgroup. In this study, we identified the genes controlled by three iron starvation sigma factors. Their regulons are composed of a small number of genes likely to be involved in iron uptake.     

Proteolysis of a Negative Regulator of Innate Immunity Is Dependent on Resistance Genes in Tomato and Nicotiana benthamiana and Induced by Multiple Bacterial Effectors

RPM1-interacting protein 4 (RIN4), a negative regulator of the basal defense response in plants, is targeted by multiple bacterial virulence effectors. We show that RIN4 degradation is induced by the effector AvrPto from Pseudomonas syringae and that this degradation in Solanaceous plants is dependent on the resistance protein, Pto, a protein kinase, and Prf, a nucleotide binding site–leucine-rich repeat protein. Our data demonstrate overlap between two of the best-characterized pathways for recognition of pathogen virulence effectors in plants. RIN4 interacts with multiple plant signaling components and bacterial effectors in yeast and in planta. AvrPto induces an endogenous proteolytic activity in both tomato (Solanum lycopersicum) and Nicotiana benthamiana that degrades RIN4 and requires the consensus site cleaved by the protease effector AvrRpt2. The interaction between AvrPto and Pto, but not the kinase activity of Pto, is required for proteolysis of RIN4. Analysis of many of the effectors comprising the secretome of P. syringae pv tomato DC3000 led to the identification of two additional sequence-unrelated effectors that can also induce degradation of RIN4. Therefore, multiple bacterial effectors besides AvrRpt2 elicit proteolysis of RIN4 in planta.     

Production and Comparison of Peptide Siderophores from Strains of Distantly Related Pathovars of Pseudomonas syringae and Pseudomonas viridiflava LMG 2352

The production of peptide siderophores and the variation in siderophore production among strains of Pseudomonas syringae and Pseudomonas viridiflava were investigated. An antibiose test was used to select a free amino acid-containing agar medium favorable for production of fluorescent siderophores by two P. syringae strains. A culture technique in which both liquid and solid asparagine-containing culture media were used proved to be reproducible and highly effective for inducing production of siderophores in a liquid medium by the fluorescent Pseudomonas strains investigated. Using asparagine as a carbon source appeared to favor siderophore production, and relatively high levels of siderophores were produced when certain amino acids were used as the sole carbon and energy sources. Purified chelated siderophores of strains of P. syringae pv. syringae, P. syringae pv. aptata, P. syringae pv. morsprunorum, P. syringae pv. tomato, and P. viridiflava had the same amino acid composition and spectral characteristics and were indiscriminately used by these strains. In addition, nonfluorescent strains of P. syringae pv. aptata and P. syringae pv. morsprunorum were able to use the siderophores in biological tests. Our results confirmed the proximity of P. syringae and P. viridiflava; siderotyping between pathovars of P. syringae was not possible. We found that the spectral characteristics of the chelated peptide siderophores were different from the spectral characteristics of typical pyoverdins. Our results are discussed in relation to the ecology of the organisms and the conditions encountered on plant surfaces.     

The impact of cytokinin on jasmonate-salicylate antagonism in Arabidopsis immunity against infection with Pst DC3000

Cytokinin has long been shown to be an essential modulator of growth and development in plants. However, its implications in plant immunity have only recently been realized. The interaction between jasmonate and salicylate pathways is regarded as a central backbone of plant immune defense. However, the effect of cytokinin on the jasmonate and salicylate mediated balance in plant immunity is still not known. Here, we analyze the impact of cytokinin on the jasmonate-salicylate antagonism in Arabidopsis immunity regarding infection with hemibiotrophic pathogen Pseudomonas syringae pv tomato DC3000 (Pst DC3000). Systems biology analysis of a refined hormone immune pathway model provides insights into the impact of cytokinin on the balance between jasmonate and salicylic acid pathways in Arabidopsis. Targeted experiments validate model simulations monitoring bacterial growth in wild type plants as well as in jasmonate pathway mutants. An integrated analysis shows that CK promotes the SA pathway of plant immunity and does not promote JA-mediated Arabidopsis susceptibility against infection with Pst DC3000. Finally, we discuss these results in the context of an emerging model of auxin-cytokinin antagonism in plant immunity.