Erwinia amylovora Auxotrophic Mutant Exometabolomics and Virulence on Apples

Abstract

The Gram-negative bacterium Erwinia amylovora causes fire blight disease of apples and pears. While the virulence systems of E. amylovora have been studied extensively, relatively little is known about its parasitic behavior. The aim of this study was to identify primary metabolites that must be synthesized by this pathogen for full virulence. A series of auxotrophic E. amylovora mutants, representing 21 metabolic pathways, were isolated and characterized for metabolic defects and virulence in apple immature fruits and shoots. On detached apple fruitlets, mutants defective in arginine, guanine, hexosamine, isoleucine/valine, leucine, lysine, proline, purine, pyrimidine, sorbitol, threonine, tryptophan, and glucose metabolism had reduced virulence compared to the wild type, while mutants defective in asparagine, cysteine, glutamic acid, histidine, and serine biosynthesis were as virulent as the wild type. Auxotrophic mutant growth in apple fruitlet medium had a modest positive correlation with virulence in apple fruitlet tissues. Apple tree shoot inoculations with a representative subset of auxotrophs confirmed the apple fruitlet results. Compared to the wild type, auxotrophs defective in virulence caused an attenuated hypersensitive immune response in tobacco, with the exception of an arginine auxotroph. Metabolomic footprint analyses revealed that auxotrophic mutants which grew poorly in fruitlet medium nevertheless depleted environmental resources. Pretreatment of apple flowers with an arginine auxotroph inhibited the growth of the wild-type E. amylovora, while heat-killed auxotroph cells did not exhibit this effect, suggesting nutritional competition with the virulent strain on flowers. The results of our study suggest that certain nonpathogenic E. amylovora auxotrophs could have utility as fire blight biocontrol agents.IMPORTANCE This study has revealed the availability of a range of host metabolites to E. amylovora cells growing in apple tissues and has examined whether these metabolites are available in sufficient quantities to render bacterial de novo synthesis of these metabolites partially or even completely dispensable for disease development. The metabolomics analysis revealed that auxotrophic E. amylovora mutants have substantial impact on their environment in culture, including those that fail to grow appreciably. The reduced growth of virulent E. amylovora on flowers treated with an arginine auxotroph is consistent with the mutant competing for limiting resources in the flower environment. This information could be useful for novel fire blight management tool development, including the application of nonpathogenic E. amylovora auxotrophs to host flowers as an environmentally friendly biocontrol method. Fire blight management options are currently limited mainly to antibiotic sprays onto open blossoms and pruning of infected branches, so novel management options would be attractive to growers.

Keywords: Erwinia amylovora; amino acid; auxotroph; fire blight; glycolysis; nucleotide; parasitism; tricarboxylic acid cycle.

FIG 1

Genomic map showing names and positions of genes identified in this study as having an auxotrophic mutant phenotype in Erwinia amylovora strain HKN06P1. Numbers around the circle indicate base pair positions in the E. amylovora CFBP 1430 genome.

FIG 2

Metabolic pathway map. Positions of disruptions identified in the auxotrophic mutant screen are indicated with red Xs. CoA, coenzyme A.

FIG 3

Erwinia amylovora auxotrophic mutant disease severity in apple fruitlets, growth in fruitlet medium, and correlations. (A) Disease severity in apple fruitlets at 7 days after inoculation. The disease severity ratio (dsr) represents the ratio of disease severity in fruitlets inoculated with the mutant divided by the disease severity of fruitlets inoculated with wild type. A dsr of 1 indicates that the disease severity of the mutant equals that of the wild type, and values less than 1 indicate that the mutant caused less severe disease than did the wild type. The dsr values represent averages of all available alleles for each gene; error bars indicate the standard error. Affected metabolic and biosynthetic pathways for the various gene categories are indicted. Asterisks and triangles indicate statistically significant differences from the wild type using Student's t test, with P values of <0.01 and <0.05, respectively. (B) Growth in fruitlet medium, as measured by optical density at 600 nm after 20 h; error bars indicate standard error. The prototroph is a fully virulent E. amylovora strain carrying a Tn5 transposon in a nongenic region. Asterisks indicate statistically significant difference from the prototroph using Student's t test, P < 0.0001; n = 3 per strain. (C) Scatter plot correlation of growth in fruitlet medium with dsr including all mutants. Fuchsia points represent cysteine auxotrophs; green points represent glycolysis and PTS mutants. All other mutants are indicated by blue points. (D) Scatter plot correlation of growth in fruitlet medium with dsr, not including the cysteine, glycolysis, and PTS mutants.

FIG 4

Disease severity caused by auxotrophic Erwinia amylovora mutants in 'Gala' apple tree shoots. Eight selected Erwinia amylovora mutants were inoculated on 'Gala' apple tree shoots by shoot tip wounding, and the percentage of total shoot length showing necrosis was measured over the course of 3 weeks. Error bars indicate standard error; n = 20 shoots per strain. Within each time point, bars sharing the same letter are not statistically different according to an ANOVA (α = 0.05). The entire experiment was performed twice with similar results; results from a representative experiment are shown.

FIG 5

Hypersensitive responses in tobacco. Suspensions of the indicated Erwinia amylovora strains at an OD₆₀₀ of 0.1 in 10 mM MgCl₂ were infiltrated into Nicotiana tabacum cv. ‘Glurk’ leaf segments. Tissue collapse was photographed at 48 h postinoculation. MgCl₂, 10 mM MgCl₂ without bacteria; n.i., not infiltrated. The experiment was performed twice. Four leaves were infiltrated in each experiment, with similar results on all leaves. Photo shows a representative leaf.

FIG 6

Exometabolomic heatmap of fruitlet medium before and after culturing with Erwinia amylovora. The indicated E. amylovora strains were cultured for 20 h in fruitlet medium. Each column represents a distinct biological replicate. Wild-type+Tet represents the supernatant of wild-type E. amylovora grown in the presence of 2 μm tetracycline as a bacteriostatic. Values are the log₂ fold change compared to uncultured (fresh) fruitlet medium. Clustering was performed in R using a Ward clustering algorithm and Pearson distance measures. The four metabolites highlighted in red are directly related to predicted pathway disruptions in the four auxotrophic mutants grown for the metabolomics assay.

FIG 7

Addition and depletion of all detected amino acids in spent fruitlet medium from the indicated Erwinia amylovora strains. Data are derived from the same data set as those shown in Fig. 6. Mutant strains were designated grower, partial grower, or nongrower based on their previously observed ability to grow in fruitlet medium. Depletion and addition are shown relative to amino acid levels in uncultured fruitlet medium. Error bars indicate the standard error.

FIG 8

Wild-type Erwinia amylovora growth on detached apple flowers pretreated with nonvirulent bacterial suspensions. Flowers were pretreated by spraying with control 10 mM MgCl₂ or the indicated bacterial suspensions 4 h prior to inoculation with wild-type E. amylovora at 10³ CFU per flower. Wild-type E. amylovora populations per flower were determined at 72 h postinoculation. (A) Growth of wild-type E. amylovora on apple flowers pretreated with an E. amylovora argD mutant at 10¹⁰ CFU ml⁻¹. (B) Growth of wild-type E. amylovora on apple flowers pretreated with an E. amylovora argD mutant at 10⁹ CFU ml⁻¹. (C) Representative apple flowers at 72 h following the indicated pretreatment. (D) Growth of wild-type E. amylovora on apple flowers pretreated with heat-killed E. amylovora argD mutant cells at 10¹⁰ CFU ml⁻¹. (E) Growth of wild-type E. amylovora on apple flowers pretreated with E. coli SM10(λpir) at 10¹⁰ CFU ml⁻¹. Error bars indicate standard error. Statistical significance was determined using Student’s t test, n = 4 flowers per treatment per experiment (Expt).