The non-protein amino acid 3,4-dehydro-l-proline (DHP) significantly reduced the incidence of fire blight infection on immature pear fruits infected with wildtype Erwinia amylovora. DHP also inhibited biofilm formation in both streptomycin-sensitive and -resistant strains of E. amylovora and induced dispersal of preformed biofilms in the streptomycin-sensitive strain. The investigations shed light on the hitherto undiscovered ability of DHP to inhibit bacterial biofilms and its potential as a chemical control option for fire blight.
Bacterial diseases of plants cause severe economic losses to the agricultural industry all around the world. Fire blight of apple, pear and other members of the Rosaceae family is one such devastating bacterial disease caused by the Gram-negative bacterium Erwinia amylovora (Johnson & Stockwell, 1998). Fire blight was first detected in the USA in the late 1700s. Since then, the disease has spread to several other parts of the world including New Zealand and Europe. Severe outbreaks of this disease necessitate the removal of whole orchards resulting in significant economic loss for fruit growers. Spray applications of streptomycin and copper compounds are the two most widely used chemical control options registered for fire blight in most countries affected by this disease. However, these options have significant drawbacks such as resistance development and phytotoxicity, which necessitates the need to develop novel antibacterial compounds with different modes of action. Streptomycin can no longer be used in geographical regions where the pathogen has already developed resistance to this antibiotic (Loper et al., 1991). Several European countries have prohibited the use of clinically relevant antibiotics such as streptomycin on plants (McManus et al., 2002). Recent findings that biofilm formation is implicated in the pathogenicity of E. amylovora (Koczan et al., 2009) have added another dimension to this problem, necessitating the development of antimicrobial compounds having the ability to inhibit the usually drug resistant biofilms. The paucity of chemical control options for fire blight prompted a screening of compounds for potential activity against E. amylovora. This study describes the hitherto undiscovered potential of the non-protein amino acid 3,4-dehydro-l-proline (DHP) as a chemical control option for fire blight of pome fruit trees, based on its ability to prevent fire blight infection in immature pear fruit and to inhibit E. amylovora biofilms.
Materials and methods
Chemical reagents and bacterial cultures
3,4-Dehydro-proline (l and dl) was purchased from Fluka; boc-3,4-dehydro-l-proline from Novabiochem. Live/dead stain and Luria broth were purchased from Invitrogen. All other chemicals and solvents were purchased from Sigma-Aldrich. Erwinia amylovora strains 1501 (Ea1501) and ICMP1540 were obtained from the International Collection of Microorganisms from Plants (Landcare Research, New Zealand). The streptomycin resistant strain of E. amylovora, Str4Ea, was a gift from Dr Joel Vanneste, Plant and Food Research, New Zealand. The bacterial strains were stored at −80°C and transferred to King's B medium slants for routine use following standard protocols. Immature pear fruits were obtained from the Newstead commercial orchards, Hamilton, New Zealand.
In vitro antibacterial activity
Erwinia amylovora strain 1501 was stored at −80°C and transferred to King's B medium slants for routine use following standard protocols (King et al., 1954). For plate overlays, E. amylovora was grown to mid-log phase (c. 8 h) in EA medium which consisted of (per litre) 11·5 g K2HPO4, 4·5 g KH2PO4, 0·2 g MgSO4, and 0·3 g l-asparagine, 0·05 g l-nicotinic acid and 20 g d-glucose autoclaved separately and mixed with the inorganic solution immediately prior to inoculation. The liquid culture was grown to mid-log phase in this medium for 8 h at 28°C on an orbital shaker. The mid-log phase cultures of E. amylovora (5 mL) were mixed with an equal volume of warm (c. 50°C) 2% agar and poured onto HS plates (10 mL of HS medium with 2% agar per plate) prepared according to Hoitink & Sinden (1970). After overlay with E. amylovora, four 5 mm diameter wells were cut into the plates equidistant from each other. The compounds for bioassay were dissolved in milli-Q water at 10 mg mL−1 and 1 mg mL−1 concentrations. Aliquots (10 μL) of each solution were added into the respective wells and the plates incubated at 28°C overnight. Streptomycin (10 or 1 mg mL−1) was added to one of the wells to serve as the positive control. After overnight incubation at 28°C, inhibitory activity was detected as clear zones of no bacterial growth immediately surrounding the respective wells and in comparison to controls. The bioassay was repeated several times with three replicates of each compound and concentration present per assay to confirm reproducibility of results.
Activity against E. amylovora infection on immature pear fruits
An immature fruit assay reported in the literature (Beer & Rundle, 1983; Vanneste et al., 1996) was used to assess the ability of DHP to either inhibit fire blight infection altogether or reduce the incidence of the onset of infection on cores of the flesh of immature pear fruits in comparison to that of streptomycin in the laboratory. An overnight culture of wildtype E. amylovora strain ICMP1540 grown in Luria broth was incubated at 28°C for 8 h in fresh LB with shaking to ensure that the culture was in exponential phase. It was then diluted in sterile water to the required bacterial concentrations ranging from 6·3 × 102 to 1·1 × 106 colony-forming units (CFU) mL−1 for the experiments.
Williams' Bon Chretien immature pear fruits (Pyrus communis) between 2·5 and 3 cm in diameter, obtained from the Newstead commercial orchards, Hamilton, New Zealand, were surface disinfected with 1·5% sodium hypochlorite and rinsed three times with sterile distilled water and left under airflow in a sterile cabinet to remove excess water. The pears were sliced under sterile conditions and cores of flesh removed using a sterile cork borer #1 (c. 2 mm diameter and 5 mm depth). Cores were taken only from the flesh of the fruit; vascular tissues were avoided. Ten cores were evenly placed on a Petri dish which itself was placed on a wet paper towel in a large sterile plastic tray. Separate Petri dishes of fruit cores were used for each treatment using DHP, streptomycin and water. Drops (10 μL) of DHP (10 mg mL−1), streptomycin (100 mg mL−1) or water were placed on the top of each of the 10 cores used per treatment in the respective dishes. The drops were allowed to dry for about 20 min at room temperature. Each Petri dish of pear cores treated with the different samples was inoculated with E. amylovora (10 μL). Cores treated with water or streptomycin prior to inoculation were used as positive and negative controls, respectively. In addition, 10 cores were treated only with water (without inoculation). Four inoculum levels of E. amylovora ranging from 102 to 105 CFU mL−1 were used, with three trials per level. After inoculation, each tray was bagged to act as a humidity chamber and incubated at 27°C. The cores were examined daily for occurrence of the common fire blight symptoms, characterized by necrosis and production of exudates.
Inhibition of biofilm formation: live/dead staining
Biofilms of E. amylovora were prepared in 12-well plates using an overnight culture of E. amylovora 1501 grown at 28°C in 0·5 × Luria–Bertani (LB) medium with sufficient numbers of test samples, as well as positive and negative controls included in each plate. The positive control comprised the overnight bacterial culture (42 μL) in sterile LB medium (1008 μL) and was present in duplicate in each plate. The negative control comprised the sterile LB medium (1050 μL) in one well of each plate. To test the effect of the inhibitory compounds on biofilm formation, test compound solutions were added to the labelled wells immediately after addition of bacteria. Test wells contained the overnight bacterial culture (42 μL), sterile LB medium (988 μL) and the test compounds (20 μL of a stock solution in sterile water to achieve a final concentration of 1 μg μL−1). Each test sample was present in triplicate in every plate. To provide a surface for biofilm formation which was amenable to microscopy at higher magnification, circular coverslips (16 mm diameter, 1 mm thick) were inserted into each well and the plates incubated statically, for either 24 or 48 h. At each time point, the coverslips were removed from the wells, rinsed with sterile water to remove any planktonic cells and dried at 50°C for 40 min. A solution of standard live/dead bacterial stain, BacLight (Invitrogen) was used to stain the biofilms. A mix was prepared (3:1000 dilution) using 1·5 μL each of the SYTO 9 (green-fluorescent nucleic acid stain) and propidium iodide (red-fluorescent nucleic acid stain) solutions, and 997 μL sterile water. Sixty microlitres of this BacLight solution was dispensed onto each of the twelve coverslips and incubated in the dark for 25 min at room temperature, after which excess stain was washed off with sterile water and the coverslips air-dried. Finally, dried coverslips were inverted onto a drop of mounting oil (supplied with the BacLight kit) on a fresh glass slide for examination of the biofilms at ×630 or ×1000 magnification using either a Leica DMR microscope with a Leica DC500 digital camera, or the Zeiss Axioplan 2 microscope with a Zeiss Axiocam HRc digital camera.
Effect on preformed biofilms – crystal violet staining
The ability of DHP to disperse preformed biofilms of E. amylovora 1501 was also investigated by crystal violet staining using the 12-well plate format. Streptomycin was used as the control. Biofilms were allowed to form in 12-well plates as described above, except that the compounds (DHP or streptomycin) were not added at the time of inoculation. Instead, these were added to ‘48 hour’ preformed biofilms, incubated for a further 24 h, and stained with 1% crystal violet following published procedures to visualize biofilm growth coverage (Kjærgaard et al., 2000). Briefly, after an initial 48 h of biofilm formation in 12-well plates, the supernatant from each well was removed carefully, the wells gently washed twice with sterile water (1 mL) and supplemented with fresh sterile LB medium alone (control wells) or fresh medium plus the inhibitory compounds in sterile water (20 μL to achieve a final concentration of 1 μg μL−1) and incubated at 28°C for a further 24 h. These treated preformed biofilms were then subjected to crystal violet staining, imaging and semiquantitative estimation of the biofilm biomass as described below. The supernatant from each well was discarded carefully, taking care to ensure that the biofilm attached to the bottom of the plate was not dislodged, and 1 mL of a 1% crystal violet solution was added. After 5 min, the dye was removed by decantation and the wells were gently rinsed two times with 1 mL sterile water. Individual wells in the plates were examined for biofilm formation at ×400 magnification under the microscope (Zeiss Axiovert 100). Crystal violet-stained biofilms at the bottom of each well were solubilized with 96% ethanol and the absorbance of this ethanol solution measured at 560 nm for a semiquantitative estimation of biofilm biomass.
In vitro antibacterial activity
In an effort to find novel chemical control options for fire blight, several simple organic molecules like lactones, furanones and ketones, as well as amino acids and their analogues, were screened against E. amylovora in vitro and further investigations conducted on the lead compound identified in the compound screening. Data from the screening experiment are summarized in Table 1. The lactones and furanones were chosen for screening because of their structural similarity to acyl homoserine lactones, the quorum-sensing signalling molecules in Gram-negative bacteria (Teplitski et al., 2000; Suga & Smith, 2003). Two cyclic ketones (rows 36 and 37 in Table 1) that were readily available in the laboratory were also included in the bioassay for comparative purposes. The amino acids were chosen because of their potential to be used in designed antimicrobial peptides which are attractive alternatives to conventional antibiotics to combat the growing problem of antibiotic resistance (Sang & Blecha, 2008). These included the protein amino acids proline and histidine and several analogues of proline including 3,4-dehydro-l-proline (DHP) in its protected and free forms and azetidine-2-carboxylic acid (Azc), as well as two dialkyl glycines (Table 1, numbers 30 and 31). Preliminary observations from this research have been published by De Zoysa et al. (2011). DHP, Azc and 5-nitro-2-furaldehyde (Table 1, numbers 8, 9, 28 and 27) showed strong in vitro inhibitory activity against E. amylovora. The in vitro activity was most pronounced in DHP. In order to gain insight into the structure activity relationships of DHP, the compound library was extended by including 3-pyrroline (the core ring structure of DHP), pyrrolidine (the saturated version of 3-pyrroline) and several substituted analogues of pyrrolidine in the bioassay (Table 1, numbers 10–24). None of these compounds inhibited the growth of E. amylovora. The ability of the non-protein amino acid DHP to prevent fire blight infection in vivo (in immature pear fruits infected with the bacterium) and to disperse preformed biofilms of E. amylovora has also been evaluated, and the results from these investigations are presented in this paper.
Table 1. Compounds screened against Erwinia amylovora
In vivo activity of DHP on immature pear fruits
Visual examination of the fruits for production of exudates, which is the characteristic symptom of fire blight, was used as an indicator to determine the number of infected fruits in an immature pear fruit assay (Beer & Rundle, 1983; Vanneste et al., 1996). Twelve different concentrations of E. amylovora ranging from 6·3 × 102 to 1·1 × 106 CFU mL−1 were used in this study. The results of these immature pear fruit assays are summarized in Figure 1. Both l- and dl-DHP were tested for in vivo activity and were able to prevent E. amylovora infection from spreading in the fruit slices (Fig. S1). DHP showed the same efficacy as streptomycin to control fire blight infection on cores of immature pear fruits inoculated with E. amylovora concentrations ranging from 102 to 103 CFU mL−1. However, at higher E. amylovora concentrations, the efficacy of DHP to prevent fire blight infection in immature pear fruits was lower than that of streptomycin. While no fruit cores treated with streptomycin showed any necrosis or exudate formation, between one and three cores treated with DHP and inoculated with the higher E. amylovora concentrations were infected. However, it is to be noted that the concentration of DHP used in these experiments was 10 times lower than that of streptomycin (10 mg mL−1 against 100 mg mL−1 streptomycin), which could explain the observed lower efficacy of DHP at the higher bacterial concentrations. It is worth noting that DHP did reduce the incidence of disease in pear fruits to the same extent as streptomycin at the lower concentrations of E. amylovora used, despite being present at 10 times lower concentration than streptomycin.
Effect on E. amylovora 1501 and Str4Ea biofilms
After live/dead staining, the biofilm cells on the circular coverslips were visualized with fluorescence microscopy. Representative fluorescence images of the biofilms after 48 h incubation in the absence and presence of DHP and streptomycin are shown in Figure 2. Comparison of the control biofilm image with that of the DHP treated ones show that DHP, at 1, 2 and 4 μg μL−1, is able to disperse the Ea1501 biofilms (Fig. 2a vs 2c–e). Comparison of images in Figure 2c–e also shows a dose-dependent inhibition of Ea1501 biofilms by DHP. Even though some green (live) cells were detected at 1 μg μL−1 DHP, no live cells were detected at 2 and 4 μg μL−1 DHP. No live cells were detected when 1 μg μL−1 of streptomycin was used (Fig. 2b), indicating that streptomycin, at 1 μg μL−1, completely inhibits biofilm formation in Ea1501. DHP, at 1 μg μL−1, also disrupted the biofilms formed by the streptomycin-resistant strain of E. amylovora (Fig. 2h). Higher DHP concentrations were not tried with this isolate of the bacterium. As would have been expected, streptomycin did not show a noticeable effect in dispersing the Str4Ea biofilms (Fig. 2f vs 2g). Even though the streptomycin-treated Str4Ea biofilms showed changes in the overall architecture, the presence of several green live cells indicate that these are indeed resistant to streptomycin (Fig. 2g).
Effect on preformed biofilms
To understand the effects of DHP and streptomycin on preformed biofilms of Ea1501, the biofilms of E. amylovora were allowed to form for 48 h and then treated with 1 μg μL−1 DHP or streptomycin in fresh medium (test wells) or fresh medium alone (control wells) for another 24 h in a 72 h long experiment. These treated biofilms were subjected to crystal violet staining, imaging and semiquantitative estimation of biofilm biomass by measuring the OD560 of the bound dye solubilized in 96% ethanol (Figs 3 and 4). Representative crystal violet stained images shown in Figure 3 clearly indicate that both DHP and streptomycin at 1 μg μL−1 were able to disperse preformed biofilms of E. amylovora (P0 vs P1 and PS). The data from the semiquantitative estimation (Fig. 4) indicated that both DHP and streptomycin significantly dispersed the preformed biofilms of Ea1501 after a 24 h treatment (P <0·05).
Use of antimicrobial peptides (AMPs), which are part of the innate immunity in all forms of life, are less likely to lead to resistance development in bacteria than are conventional antibiotics; therefore AMPs are attractive alternatives to the latter group of compounds. This laboratory is focused on developing novel AMPs for the management of bacterial diseases of fruit trees including fire blight. Mitchell et al. (2008) have previously reported on a novel naturally occurring antimicrobial peptide with broad-spectrum antibacterial activity. Characteristics of this AMP include the presence of the non-protein amino acid homoserine in the sequence and an unusual modification to the beta carboxyl of the C-terminal aspartate. The use of non-protein amino acids in short synthetic peptides helps to overcome the protease susceptibility of these, otherwise attractive, compounds with potential as antimicrobials (Owens & Heutte, 1997; Güell et al., 2011). This current study has identified the non-protein amino acid 3,4-dehydro-l-proline as a lead compound for fire blight control demonstrated by its in vivo activity against E. amylovora infection on immature pear fruits and effect on E. amylovora biofilms. In addition to DHP, several other organic compounds and non-protein amino acids were also screened against E. amylovora in this study. None of the tested compounds, to the best of the authors' knowledge, have been investigated as potential chemical control options for fire blight. The lack of activity of 3-pyrroline, pyrrolidine as well as several derivatives of pyrrolidine and proline indicate that the complete dehydro amino acid structure is essential for the activity against E. amylovora.
Experiments conducted on immature pear fruit tissues allowed the comparison of different compounds for their ability to reduce disease incidence in a controlled environment. The results of the experiments on immature pear fruits show that DHP is able to prevent the spread of infection by E. amylovora in fruit tissue. Orchard studies involving both compounds at the same concentration will be necessary to conclusively compare the efficacy of DHP and streptomycin in fire blight control.
In addition to testing DHP against E. amylovora, the compound was tested for in vitro activity against other bacteria, such as Pseudomonas aeruginosa and Staphylococcus aureus. However, no inhibition of these bacteria was observed (data not shown), indicating that the antibacterial activity of DHP may be species-specific.
Biofilm formation has been implicated in the pathogenicity of E. amylovora (Koczan et al., 2009). Chemicals that interfere with biofilm formation in E. amylovora are currently unknown. The potential of DHP to inhibit the formation of biofilms and disperse preformed biofilms in E. amylovora was investigated using live/dead and crystal violet staining in experiments for the first time. The crystal violet assay has been used as a semiquantitative method to estimate biofilm bacterial biomass (Kjærgaard et al., 2000).
DHP inhibited biofilm formation in E. amylovora 1501 and Str4Ea when added at inoculation (Fig. 2). DHP also exhibited the ability to disperse preformed biofilms in Ea1501 (Figs 3 and 4). Biofilm bacteria reside in protected environments enclosed within extracellular polymer matrices that present a strong barrier to external stimuli and are impervious to most drugs. The results of these investigations on the biofilm inhibitory properties of DHP could be indicative of DHP penetrating the extracellular matrix of E. amylovora. The ability to penetrate the exopolysaccharides (EPS) of biofilms is essential for treating chronic infections caused by biofilms. dl-DHP has been reported to have inhibitory activity against certain bacterial species including Escherichia coli (Smith et al., 1962). However, DHP has never been reported to have any activity against the fire blight pathogen. There are no literature reports on the biofilm inhibitory properties of DHP either. The findings on the antibiofilm properties of DHP in this study have opened up a novel application hitherto undiscovered for DHP. This is also the first report where DHP has been investigated as a possible treatment option for fire blight.
This study establishes the ability of 3,4-dehydro-l-proline to be developed as a potential chemical control option for fire blight. The findings that DHP has strong antibacterial and biofilm inhibitory activities against the streptomycin resistant strain of E. amylovora make it an attractive solution to control the disease where the resistant strains of E. amylovora have been prevalent, particularly in the USA and New Zealand. A suitable formulation of DHP itself, or antibacterial peptides incorporating DHP, would also serve as attractive alternatives to conventional antibiotics such as streptomycin, which is prohibited for agricultural use in Europe. To the best of the authors' knowledge, this is also the first report on the ability of the antibiotic streptomycin to inhibit biofilm formation in any bacterial pathogens.
This work was supported by a Uniservices Grant 1078476. G. H. De Zoysa thanks the Faculty of Science for a Masters scholarship. The authors thank Dr J. Vanneste and his team at Plant and Food Research, New Zealand for the immature pear fruit assay.