Mutation of the Erwinia amylovora argD gene causes arginine auxotrophy, nonpathogenicity in apples, and reduced virulence in pears

Abstract

Fire blight is caused by Erwinia amylovora and is the most destructive bacterial disease of apples and pears worldwide. In this study, we found that E. amylovora argD(1000)::Tn5, an argD Tn5 transposon mutant that has the Tn5 transposon inserted after nucleotide 999 in the argD gene-coding region, was an arginine auxotroph that did not cause fire blight in apple and had reduced virulence in immature pear fruits. The E. amylovora argD gene encodes a predicted N-acetylornithine aminotransferase enzyme, which is involved in the production of the amino acid arginine. A plasmid-borne copy of the wild-type argD gene complemented both the nonpathogenic and the arginine auxotrophic phenotypes of the argD(1000)::Tn5 mutant. However, even when mixed with virulent E. amylovora cells and inoculated onto immature apple fruit, the argD(1000)::Tn5 mutant still failed to grow, while the virulent strain grew and caused disease. Furthermore, the pCR2.1-argD complementation plasmid was stably maintained in the argD(1000)::Tn5 mutant growing in host tissues without any antibiotic selection. Therefore, the pCR2.1-argD complementation plasmid could be useful for the expression of genes, markers, and reporters in E. amylovora growing in planta, without concern about losing the plasmid over time. The ArgD protein cannot be considered an E. amylovora virulence factor because the argD(1000)::Tn5 mutant was auxotrophic and had a primary metabolism defect. Nevertheless, these results are informative about the parasitic nature of the fire blight disease interaction, since they indicate that E. amylovora cannot obtain sufficient arginine from apple and pear fruit tissues or from apple vegetative tissues, either at the beginning of the infection process or after the infection has progressed to an advanced state.

FIG 1

(A) Illustration of the argD gene and flanking genes yhfK and pabA showing the genome nucleotide position of the Tn5 insertion in the argD(1000)::Tn5 mutant and the genomic DNA (gDNA) segment used for complementation. Genome nucleotide positions are numbered according to their homologous positions in the complete genome sequence of E. amylovora CFBP 1430. PCR primer (Table 1) hybridization locations and directions are indicated. (B) PCR analysis and agarose gel electrophoresis were used to verify Tn5 transposon insertion in the argD gene in the argD(1000)::Tn5 mutant. Using primers P1 and P2, a 1,448-bp PCR product was amplified from wild-type gDNA but not from that of the argD(1000)::Tn5 mutant (argD⁻), as expected. When the PCR extension time was increased to 4 min with primers P1 and P2, a 3,449-bp PCR product was amplified from the argD(1000)::Tn5 mutant's gDNA, representing the argD gene interrupted by Tn5. As expected, primer combinations P1 plus RP and FP plus P2 amplified a 303- and a 1,145-bp product, respectively, further confirming the presence of a Tn5 transposon insertion in the argD gene of the argD(1000)::Tn5 mutant. Control PCRs with no gDNA template produced no PCR products.

FIG 2

Auxotrophy of the argD(1000)::Tn5 mutant. Bacterial population densities of the indicated strains growing in M9 minimal medium supplemented with thiamine and nicotinic acid (M9TN medium) and with sorbitol as the carbon source were determined by serial dilution plating at 0 and 48 h after starting the culture. Values are the means of three replicates, and error bars represent standard errors; P = 0.05. Bars with the same letter have no statistically significant difference based on a Tukey test (log₁₀). This experiment was performed three times, with similar results each time.

FIG 3

Map of the pCR2.1-argD plasmid used to complement the argD(1000)::Tn5 mutant. The wild-type argD complementation fragment (Fig. 1) was inserted with the direction of argD transcription opposite to that of the Plac promoter in pCR2.1. Unique restriction sites are indicated in boldface.

FIG 4

Restoration of growth of the argD(1000)::Tn5 mutant in minimal medium by the addition of arginine. Bacterial population densities of the indicated strains growing in M9 minimal medium supplemented with thiamine and nicotinic acid (M9TN medium) with sorbitol as the carbon source and supplemented with 0.1 mg/ml arginine were determined by serial dilution plating at 0 and 48 h after starting the culture. Values are the means of three replicates, and error bars represent standard errors; P = 0.05. There was no statistically significant difference between the wild type and the argD(1000)::Tn5 mutant at any time point, according to a Tukey test (log₁₀). This experiment was performed three times, with similar results each time.

FIG 5

The argD(1000)::Tn5 mutant was not pathogenic in apple trees. Actively growing shoot tips were inoculated with the indicated E. amylovora strains at 10⁸ CFU/ml by shoot tip wound inoculation. Values are the means of three replicates, and error bars represent standard errors; P = 0.05. Bars with the same letter have no statistically significant difference based on a Tukey test (log₁₀). The entire experiment was performed twice, with similar results each time.

FIG 6

Pathogenicity assay and quantitative growth analysis in detached immature apple and pear fruit. (A) Immature ‘Gala' apple fruit halves at 7 days postinoculation (dpi) with the indicated bacterial strains. Tissue necrosis and bacterial ooze are a symptom and a sign, respectively, of fire blight disease in immature apple fruit. (B) Immature ‘Bosc' pear fruit halves at 5 dpi with the indicated bacterial strains. (C) Bacterial populations in immature ‘Gala' apple fruit halves at 7 dpi with the indicated bacterial strains, as determined by serial dilution plating. The same apples shown in panel A were used for the quantitative growth assay. (D) Bacterial populations in immature ‘Bosc' pear fruit halves at 5 dpi with the indicated bacterial strains. In panels C and D, values are the means of three replicates, and error bars represent standard errors; P = 0.05. Bars with the same letter have no statistically significant difference based on the Tukey test (log₁₀). Experiments were performed at least three times with at least five immature apple and pear fruit halves inoculated per strain, with the same results each time.

FIG 7

Fire blight symptoms and quantitative growth analysis in immature apple fruit inoculated with wild-type E. amylovora strain HKN06P1 and a nalidixic acid-resistant derivative (HKN06P1^NalR). (A) Strains HKN06P1 and HKN06P1^NalR produced similar fire blight symptoms in immature apple fruits at 7 days postinoculation (dpi). (B) Strains HKN06P1 and HKN06P1^NalR grew to similar population densities at 7 dpi, as determined by serial dilution plating. Extracts from fruits inoculated with HKN06P1 were plated in LB medium without antibiotic selection, and extracts from fruits inoculated with HKN06P1^NalR were plated on LB medium supplemented with 25 μg/ml nalidixic acid. The immature apple fruits shown in panel A were used for the quantitative growth analysis. Bars with the same letter have no statistically significant difference based on a Tukey test (log₁₀).

FIG 8

Bacterial growth in immature apple fruits after inoculation with a 1:1 mixture of HKN06P1^NalR and the argD(1000)::Tn5 mutant. Bacterial population densities were determined by serial dilution plating; HKN06P1^NalR populations were determined by plating on LB medium supplemented with 25 μg/ml nalidixic acid, while argD(1000)::Tn5 mutant populations were determined by plating on LB medium supplemented with 50 μg/ml kanamycin. Values are the means of four replicates, and error bars represent standard errors; P = 0.05. Bars with the same letter have no statistically significant difference based on a Tukey test (log₁₀). The experiment was performed twice, with similar results each time.

FIG 9

The pCR2.1-argD plasmid was stably maintained in the argD(1000)::Tn5 mutant growing in planta. (A) Plasmid stability assay in immature apple fruits. Immature apple halves were inoculated with the indicated E. amylovora strains, and bacterial populations were determined at 0 and 7 days postinoculation (dpi) by serial dilution plating of the fruit extract on LB plates supplemented with kanamycin, carbenicillin, or nalidixic acid, as indicated. (B) Plasmid preparations from randomly selected carbenicillin-resistant isolates from the experiment whose results are shown in panel A were resolved by agarose gel electrophoresis. Plasmid preparations from wild-type strains carrying pCR2.1 and pCR2.1-argD which had not passed through plants served as controls. A plasmid preparation from wild-type E. amylovora carrying no foreign plasmids was also included. (C) Plasmid stability in apple trees inoculated with the indicated E. amylovora strains. ‘Gala' apple trees were inoculated by shoot tip wounding with the indicated E. amylovora strains, and bacterial populations were determined at 21 dpi by serial dilution plating of shoot tissue extract on LB plates supplemented with kanamycin, carbenicillin, or nalidixic acid, as indicated. (D) Plasmid preparations from randomly selected carbenicillin-resistant isolates from the experiment whose results are shown in panel C were resolved by agarose gel electrophoresis. Plasmid preparations from wild-type strains carrying pCR2.1 and pCR2.1-argD which had not passed through plants served as controls. In panels A and C, values are the means of three replicates, and error bars represent standard errors; P = 0.05. Bars with the same letter have no statistically significant difference based on a Tukey test (log₁₀).

FIG 10

HKN06P1^NalR(pCR2.1), HKN06P1^NalR(pCR2.1-argD), and argD(1000)::Tn5(pCR2.1-argD) strains caused similar fire blight disease progression in apple trees. Actively growing shoot tips of ‘Gala' apple trees were inoculated with the indicated E. amylovora strains at 10⁸ CFU/ml by shoot tip wound inoculation. The extent of shoot necrosis at 7, 14, and 21 days postinoculation (dpi) is shown. Values are the means of three replicates, and error bars represent standard errors; P = 0.05. Bars with the same letter have no statistically significant difference based on a Tukey test (log₁₀). The entire experiment was performed twice, with similar results each time.