Maladaptation in wild populations of the generalist plant pathogen Pseudomonas syringae

Abstract

Multihost pathogens occur widely on both natural and agriculturally managed hosts. Despite the importance of such generalists, evolutionary studies of host-pathogen interactions have largely focused on tightly coupled interactions between species pairs. We characterized resistance in a collection of Arabidopsis thaliana hosts, including 24 accessions collected from the Midwest USA and 24 from around the world, and patterns of virulence in a collection of Pseudomonas syringae strains, including 24 strains collected from wild Midwest populations of A. thaliana (residents) and 18 from an array of cultivated species (nonresidents). All of the nonresident strains and half of the resident strains elicited a resistance response on one or more A. thaliana accessions. The resident strains that failed to elicit any resistance response possessed an alternative type III secretion system (T3SS) that is unable to deliver effectors into plant host cells; as a result, these seemingly nonpathogenic strains are incapable of engaging in gene for gene interactions with A. thaliana. The remaining resident strains triggered greater resistance compared to nonresident strains, consistent with maladaptation of the resident bacterial population. We weigh the plausibility of two explanations: general maladaptation of pathogen strains and a more novel hypothesis whereby community level epidemiological dynamics result in adaptive dynamics favoring ephemeral hosts like A. thaliana.

Figure 1

Minimum Evolution tree based on a 570 bp fragment of the gyrase B gene showing evolutionary relationships among 42 strains of Pseudomonas syringae. Strains that elicit a host resistance response (HR+) are suffixed with +; those that carry a xenologous T3SS and do not elicit a host resistance response (HR−) are suffixed with −. Bootstrap values (>75%) from analysis of 10,000 replicates are shown above nodes. The tree was rooted with Pseudomonas aeuruginosa strain PA01 as the outgroup. Strains collected from Arabidopsis are marked with an asterisk. All strains collected from A. thaliana are within one of the group 2 clades, and HR-strains carrying a xenologous T3SS are found exclusively in clade 2c.

Figure 2

Mean resistance of Midwest and globally distributed lines of the host A. thaliana when challenged with 12 strains of the pathogen P. syringae collected from Midwestern A. thaliana plants (resident strains) and 18 strains collected worldwide, from a largely unrelated group of 13 agricultural species (non-resident strains). (a). All P. syringae strains. (b). Only strains belonging to P. syringae clade 2 (see Fig. 1). See table S1 for information on host and pathogen sampling. Error bars represent standard errors around the mean.

Figure 3

Growth of 24 resident strains of the pathogen Pseudomonas syringae and 18 nonresident strains in the A. thaliana ecotype Col-0. Resident strains are further divided into strains that have a xenologous T3SS and do not trigger the hypersensitive response (HR) in the host (HR−), and strains that have an ancestral T3SS and do trigger HR (HR+). Bars reflect one standard error of the mean. Each strain is identified by a number found in Table S1.