Mixed infections have been broadly applied to the study of bacterial pathogens in animals. However, the application of mixed infection-based methods in plant pathogens has been very limited. An important factor for this limitation is the different dynamics that mixed infections have been reported to show in the different types of models. Reports in systemic animal infections have shown that any bacterium has the same probability of multiplying within a mixed infection than in a single infection. However, in plant pathogens, bacterial growth in a mixed infection does not seem to reflect growth in a single infection, as growth interference takes place between the co-inoculated strains. Here we show that growth interference in mixed infection between different Pseudomonas syringae strains is not intrinsic to growth within a plant host, but dependent on the dose of inoculation. We also show that the minimal inoculation dose required to avoid interference depends on the aggressiveness of the pathogen as well as the type of virulence factor that differentiates the co-inoculated strains. This study establishes the basis for the use of mixed infection-based applications to the study of phytopathogenic bacteria. Analysis of the virulence of a type III effector mutant and an hrp regulatory mutant illustrate the increased accuracy and sensitivity of competitive index assays vs. regular growth assays. Several applications of this assay are addressed, and potential implications for this and other mixed infection-based methods are discussed.

In Pseudomonas syringae, the type III secretion system (T3SS) is essential for disease in compatible hosts and for eliciting the hypersensitive response in incompatible hosts. P. syringae pathovars secrete a variable number of type III effectors that form their secretomes. The secretome of Pseudomonas syringae pv. phaseolicola 1448a (Pph1448a) currently includes 22 experimentally validated effectors, one HrpL-regulated candidate for which translocation results have been inconsistent, two translocated candidates for which in planta expression has not been established, one bioinformatically identified candidate, and six candidates that have been experimentally discarded. We analyzed the translocation and/or expression of these and other candidates to complete the Pph1448a effector inventory, bringing this inventory to 27 bona fide effectors, including a new one that does not belong to any of the previously described effector families. We developed a simple process for rapidly making single and double knockout mutants and apply it to the generation of an effector mutant collection that includes single knockouts for the majority of the Pph1448a effector inventory. We also generated two double mutant strains containing effectors with potentially redundant functions and analyzed the virulence of the single and double mutant strains as well as strains expressing each of the effectors from a plasmid. We demonstrate that AvrB4-1 and AvrB4-2, as well as HopW1-1 and HopW1-2, are fully redundant and contribute to virulence in bean plants, thus validating this approach for dissecting the contribution of the Pph1448a type III effector inventory to virulence. We also analyzed the effect that the expression of these four effectors from Pseudomonas syringae pv. tomato DC3000 (PtoDC3000) has during its interaction with Arabidopsis thaliana, establishing that AvrB4-1, but not the others, determines a restriction of bacterial growth that takes place mostly independently of the salicylic acid (SA)-signaling pathway.

Many type III-secreted effectors suppress plant defenses, but can also activate effector-triggered immunity (ETI) in resistant backgrounds. ETI suppression has been shown for a number of type III effectors (T3Es) and ETI-suppressing effectors are considered part of the arms race model for the co-evolution of bacterial virulence and plant defense. However, ETI suppression activities have been shown mostly between effectors not being naturally expressed within the same strain. Furthermore, evolution of effector families is rarely explained taking into account that selective pressure against ETI-triggering effectors may be compensated by ETI-suppressing effector(s) translocated by the same strain. The HopZ effector family is one of the most diverse, displaying a high rate of loss and gain of alleles, which reflects opposing selective pressures. HopZ effectors trigger defense responses in a variety of crops and some have been shown to suppress different plant defenses. Mutational changes in the sequence of ETI-triggering effectors have been proposed to result in the avoidance of detection by their respective hosts, in a process called pathoadaptation. We analyze how deleting or overexpressing HopZ1a and HopZ3 affects virulence of HopZ-encoding and non-encoding strains. We find that both effectors trigger immunity in their plant hosts only when delivered from heterologous strains, while immunity is suppressed when delivered from their native strains. We carried out screens aimed at identifying the determinant(s) suppressing HopZ1a-triggered and HopZ3-triggered immunity within their native strains, and identified several effectors displaying suppression of HopZ3-triggered immunity. We propose effector-mediated cross-suppression of ETI as an additional force driving evolution of the HopZ family.