Priority Effects in the Apple Flower Determine If the Siderophore Desferrioxamine Is a Virulence Factor for Erwinia amylovora CFBP1430

Abstract

Iron is crucial for bacterial growth and virulence. Under iron-deficiency bacteria produce siderophores, iron chelators that facilitate the iron uptake into the cell via specific receptors. Erwinia amylovora, the causative agent of fire blight, produces hydroxamate-type desferrioxamine siderophores (DFO). The presented study reassesses the impact of DFO as a virulence factor of E. amylovora during its epiphytic phase on the apple flower. When inoculated in semisterile Golden Delicious flowers no difference in replication and induction of calyx necrosis could be observed between E. amylovora CFBP1430 siderophore synthesis (DfoA) or uptake (FoxR receptor) mutants and the parental strain. In addition, mutant strains only weakly induced a foxR promoter-gfpmut2 reporter construct in the flowers. When analyzing the replication of the receptor mutant in apple flowers harboring an established microbiome, either naturally, in case of orchard flowers, or by pre-inoculation of semisterile greenhouse flowers, it became evident that the mutant strain had a significantly reduced replication compared to the parental strain. The results suggest that apple flowers per se are not an iron-limiting environment for E. amylovora and that DFO is an important competition factor for the pathogen in precolonized flowers. IMPORTANCE Desferrioxamine is a siderophore produced by the fire blight pathogen E. amylovora under iron-limited conditions. In the present study, no or only weak induction of an iron-regulated promoter-GFP reporter was observed on semisterile apple flowers, and siderophore synthesis or uptake (receptor) mutants exhibited colonization of the flower and necrosis induction at parental levels. Reduced replication of the receptor mutant was observed when the flowers were precolonized by microorganisms. The results indicate that apple flowers are an iron-limited environment for E. amylovora only if precolonization with microorganisms leads to iron competition. This is an important insight for the timing of biocontrol treatments.

Keywords: Erwinia amylovora; apple flowers; desferrioxamine; iron deficiency; replication; secondary colonization; siderophore mutants.

FIG 1

Schematic representation of the ferrioxamine cluster in E. amylovora CFBP1430. Large gray arrows indicate coding sequences. Small black arrows show the position and direction of primers used for site-directed mutagenesis of the dfoA and foxR genes. A kanamycin resistance cassette (aphT) was inserted into the EcoRV cutting site of the dfoA gene, resulting in the desferrioxamine synthesis mutant EAdfoA. Replacing the 203 pb SalI fragment of the foxR gene by aphT resulted in the ferrioxamine receptor mutant EAfoxR.

FIG 2

Siderophore indication CAS-agar with E. amylovora CFBP1430 parental strain EA^S (A), desferrioxamine synthesis mutant EAdfoA (B), ferrioxamine receptor mutant EAfoxR (C) and complemented receptor mutant EAfoxRco (D). The orange halos are indicative for siderophore production.

FIG 3

LB (A) and KB (B) growth curves of E. amylovora CFBP1430 strains EA^S (solid), desferrioxamine synthesis mutant EAdfoA (dotted), ferrioxamine receptor mutant EAfoxR (dashed) and complemented mutant EAfoxRco (dotdash). Error bars represent standard deviations.

FIG 4

Percentage fraction of flowers in each infection grade. Each column represents 48 evaluated flowers from two independent detached flower assay experiments. Flowers were inoculated onto the hypanthium with E. amylovora CFBP1430 strains EA^S, desferrioxamine synthesis mutant EAdfoA, and ferrioxamine receptor mutant EAfoxR. Evaluation of the flowers infection grade was performed after 4 days of incubation at 26°C according to the following scale: grade 1: calyx green; grade 2: calyx necrotic; grade 3: calyx and pedicel necrotic.

FIG 5

CFU of parental strain E. amylovora CFBP1430 EA^S, corresponding desferrioxamine synthesis mutant (EAdfoA), ferrioxamine receptor mutant (EAfoxR), and complemented ferrioxamine receptor mutant (EAfoxRco) 0 h, 24 h and 72 h p.i. on stigma and hypanthium of detached GD flowers.

FIG 6

CFU and necrosis of GD flower calyxes after spray inoculation of flowering GD trees with E. amylovora CFBP1430 strains EA^S, EAdfoA, and EAfoxR. (A) CFU were determined for 10 flowers per strain 24 h and 48 h p.i. Error bars represent the standard deviation from the mean. (B) Disease symptoms of flowers rated 6 days after inoculation. Error bars represent the standard deviation of three trees. In total 413, 474 and 333 flowers were evaluated for EA^S, EAdfoA and EAfoxR, respectively.

FIG 7

GFPmut2 expression under the control of the EAfoxR promoter in E. amylovora CFBP1430 strains EA^S (EAgfp=A, blue), synthesis mutant EAdfoA (EAdfoAgfp=B, blue), and receptor mutant EAfoxR (EAfoxRgfp=C, blue) reisolated from GD flowers 48 h p.i. on the stigmata. E. amylovora CFBP1430, in red, lacking the reporter construct is the negative control.

FIG 8

CFU of reisolated E. amylovora CFBP1430 strains 48 h p.i. (A) parental strain EA^S (n = 9) and mutant EAfoxR (n = 20) from semisterile GD flowers, greenhouse; (B) parental strain EA^S (n = 10) and mutant EAfoxR (n = 20) from open GD flowers, orchard-P23; (C) parental strain EA^S (n = 20), mutant EAfoxR (n = 20) and complemented mutant strain EAfoxRco (n = 19) from open GD flowers, orchard-P23; (D) parental strain EA^S (n = 10) and mutant EAfoxR (n = 9) from GD balloon flowers, orchard-P23. Error bars represent the standard deviation of the mean. Significant differences between treatments are marked with different letters (P-value < 0.05, one-way ANOVA, Tukey’s multiple-comparison test).

FIG 9

CP values of E. amylovora CFBP1430 strains. (A) EA^S (n = 20), corresponding mutant EAfoxR (n = 20) and complemented mutant strain EAfoxRco (n = 19) reisolated from open GD flowers, orchard-P23 after 48 h p.i.; (B) EA^S (n = 10) and mutant EAfoxR (n = 10) reisolated from GD balloon flowers, orchard-P23, 48 h p.i. The bacterial DNA was extracted from each infected flower and qPCR performed with an amsC (amylovoran synthesis) specific probe. Error bars represent the standard deviation of the mean. Significant differences between treatments are marked with different letters (P-value < 0.05, one-way ANOVA, Tukey’s multiple-comparison test).

FIG 10

CFU of E. amylovora CFBP1430 strains EA^S (n = 20) and corresponding mutant EAfoxR (n = 20) 48 h p.i. onto PW pre-inoculated semisterile GD flowers from the greenhouse. Error bars represent the standard deviation of the mean. Significant differences between treatments are marked with different letters (P-value < 0.05, one-way ANOVA, Tukey’s multiple-comparison test).