Abstract
Strain B-FS01, isolated from rape (Brassica napus) stem infected by Slerotinia sclerotiorum and identified as Bacillus subtilis, exhibited predominantly antagonistic activities against Fusarium moniliforme Sheldon ATCC 38932. Antifungal active compounds (AAC) were isolated and purified from the cultures of strain B-FS01 against ATCC 38932. The HPLC/electron spray ionization/collision-induced dissociation mass spectrum of AAC revealed a cluster of fengycin homologues containing fengycins A, fengycins B and a new type of fengycin. Further toxic assay of AAC in vitro against F. moniliforme indicated that AAC could strongly inhibit the growth of both mycelia and spores. In addition, treatment with AAC significantly modified the maize seed infection by ATCC 38932.
Introduction
A biological species of mating populations of Gibberella fujikuroi species complex, population A Fusarium moniliforme Sheldon (synonym Fusarium verticilliodes; Gibberella moniliforme), is a phytopathogenic filamentous fungus that has been reported to infect maize kernels (Nireberg & O'Donnell, 1998). Fusarium moniliforme usually grows as an endophyte on maize vegetative and reproductive tissues, normally causing no disease symptoms in the plant (Bacon & Hinton, 1996). However, under the conditions of bad weather, insect and fungal attack, and other environmental stresses, it causes seedling blight, stalk rot and ear rot (Nelson et al., 1993). During the biotrophic and/or endophytic association with maize, as well as saprophytic growth, F. moniliforme produces fumonisins, which occur widespread and pose potentially adverse effects on human and animal health (Riley et al., 1994; Marasas, 1995; Dutton, 1996). Fumonisins disrupt sphingolipid biosynthetic pathways in both animal and plant cells, with profound consequences on cellular metabolism (Wang et al., 1991; Merrill et al., 2001).
The intercellular nature of F. moniliforme makes chemical control highly unlikely (Bacon et al., 2001). Recently, biological control by antagonistic organisms has been found to be a reliable management of phytopathogenic fungi (Shoda, 2000). Some endophytes such as Bacillus subtilis RRC101 (Bacon et al., 2001) and Trichoderma species (Yates et al., 1999) display antagonistic activities to F. moniliforme and great promise for reducing fumonisin accumulation during the endophytic (vertical transmission) growth phase. However, in the case of maize or maize-based products in storage, use of a zoetic biocontrol agent is impractical because F. moniliforme has already been well established by the time the conditions are fit for growth of the agent (Klich et al., 1994). Cereals, when treated with living organisms, are hard to gain public acceptance. Indeed, some bacteria, such as Pseudomonas and Bacillus spp., produce antibiotics, which play a major role in disease suppression (Fravel, 1988; Dowling & O'Gara, 1994; Eckart, 1994; Cook et al., 1995; Eshita et al., 1995). Direct use of these antibiotics is one of attractive control strategies to make up for the defaults of living organisms. Bacillus subtilis strains produce a broad spectrum of bioactive peptide antibiotics with a high potential for biotechnological and pharmaceutical applications (Steller & Vater, 2000). The families of lipopeptides surfactin (Kakinuma et al., 1969; Kowall et al., 1998), fengycin (Vanittanakom et al., 1986) and iturin (iturin, mycosubtilin, bacillomycin) (Peypoux et al., 1978; Desjardins et al., 1992; Hiraoka et al., 1992) are a prominent class of such antibiotics. Iturin A has been reported to show efficacy in controlling some fungi in stored grain (Klich et al., 1994), but not other lipopeptide antibiotics.
In this study, an antagonistic bacterial strain B-FS01 isolated from rape stem infected by Slerotinia sclerotiorum was identified as B. subtilis and characterized to function against F. moniliforme and some other pathogenic fungi. By chromatographic, mass spectrometer and other chemical analysis, antifungal active compounds (AAC) from the culture of B-FS01 were isolated and identified as fengycin homologues. The efficacy of AAC for control of F. moniliforme in vitro and in maize seeds was also determined. The newly identified active compounds may offer an alternative for controlling the growth of F. moniliforme and accumulation of released toxin in maize.
Materials and methods
Bacterial strain
Strain B-FS01 was isolated by NA culture (0.3% beef extract, 1% peptone, 0.5% NaCl and 1.5% agar) in September 2002 from rape stem infected by S. sclerotiorum at the Experimental Station of Jiangsu Academy of Agricultural Sciences, Nanjing, China. The strain was then cultured in the Czapek medium (1 L of distilled water containing 30 g glucose, 0.5 g MgSO4·7H2O, 0.5 g KCl, 1 g KH2PO4, 2 g NaNO3 and 0.01 g FeSO4) for purification of antibiotics. The NA medium was supplemented with 20% glycerol for storage of strain B-FS01 at −80°C.
Fungal strains
The test fungi Botrytis cinerea, Phytophora capsici, Alternaria alternate, Colletotrichum musae, Rhizoctonia solani, Fusarium oxysporumf. sp. niveum and Fullvia vulva were provided by the Institute of Plant Protection, Jiangsu Academy of Agricultural Sciences. Physalospora piricola and S. sclerotiorum were provided by the College of Plant Protection, Nanjng Agricultural University. Aspergilis flavus (CGMCC 3.2890), Fusarium graminearum (Royal Dental College of Arhus 946) and F. moniliforme Sheldon ATCC 38932 were purchased from China General Microbiological Culture Collection Center (CGMCC).
Identification of strain B-FS01
Strain B-FS01 was identified by the method described previously (Sneath, 1986). Observation was performed on B-FS01 under a transmission electron microscope (H-600, Hitachi) at a magnification of 12 000 times. Total DNA of B-FS01 was isolated using the bacterial genomic DNA extraction Kit (TianWei Co., China). PCR amplification of partial 16S rRNA gene fragments was carried out using primers as described by Delong (1992). The 16S rRNA gene fragment was sequenced in the Boya Gene Company (Shanghai, China). The sequence was analyzed by Ribosomal Database Project II, and a phylogenic tree topology was performed with the closest 15 sequences by the cluster x software (Thompson et al., 1994).
Determination of antimicrobial activity of strain B-FS01
The antifungal activity of B-FS01 strain against phytopathogenic fungi was determined by dual-culture assay. B-FS01 was inoculated as a line on one edge of a 90 mm-diameter Petri plate containing potato dextrose agar (PDA) medium (20% potato, 2% dextrose and 1.5% agar) and incubated at 30°C. After 24 h, a 6-mm diameter actively growing fungus was inoculated at the center. Dual-inoculated plates, with fungus alone as control, were incubated at 28°C with a 12-h photoperiod. The inhibition zone between the two cultures was measured 7 days after inoculation for A. flavus and 3 days after inoculation for the other tested fungi.
Purification of AAC
B-FS01 was grown in the Czapek medium at 30°C for 48 h. Cells were removed from the culture by centrifugation (15 000 g, 20 min). The supernatant was brought to different saturation with ammonium sulfate [(NH4)2SO4] at 4°C and then the precipitate in the mixture between 25% and 60% saturation of (NH4)2SO4 was collected by centrifugation (10 000 g, 20 min) and dissolved in 50 mM sodium phosphate buffer (pH 7.5). After centrifugation at 10 000 g for 10 min the insoluble part was discarded. The supernatant was then desalted on a Sephadex G-25 medium column (20 mm × 40 cm). The first peak was collected, boiled for half an hour, centrifuged at 10 000 g for 10 min and then applied to a DEAE-52 column (16 mm × 15 cm) previously equilibrated with 50 mM Tris-HCl buffer (pH 7.5) containing 0.05 M NaCl. Antifungal fractions were obtained from the elution with 0.5–0.7 M NaCl in the same buffer, concentrated by ultrafiltration with a PM10 membrane (Amico) and placed onto a Sephadex G-100 gel column (15 mm × 80 cm). The column was equilibrated with 10 mM ammonium acetate (NH4Ac) buffer, and eluted with the same buffer at a flow rate of 0.5 mL min−1. Fractions containing AAC were collected, concentrated and lyophilized. Through these purification steps of AAC, fractions were determined by the absorbance at 280 nm and the anti-F. moniliforme fractions underwent further processing. Reverse-phase HPLC was performed for the final purification of AAC. The lyophilized active collection was dissolved in 10 mM NH4Ac water solution, applied to a Cosmosil C18-AR-II column (4.6 mm ID × 250 mm; 300 A) with a mobile phase: 20 mM NH4Ac water solution (A), methanol (B). A linear gradient from 45% to 95% methanol was applied in 30 min at a flow rate of 0.6 mL min−1. This step was repeated several times, and peaks with the same retention time were collected for determination of activity. The final active collection was dried, weighed and dissolved in 10 mM NH4Ac buffer for the following experiments.
Analysis of AAC by HPLC/electrospray ionization (ESI)/collision-induced dissociation (CID) MS
HPLC/ESI/CID MS analysis of AAC was performed with Surveyor-LCQ DECA XP Plus of Thermo Finnigan (Thermo Electron Corporation, San Jose, CA). The electrospray source was operated at a capillary voltage of 32 V, a spray voltage of 5 kV and a capillary temperature of 275°C. For the CID experiment, helium was used as the collision gas and the collision energy was set at 35%.
Inhibition of radial growth of F. moniliforme by AAC
To assay antagonistic activity against hyphal growth, an agar with mycelia of F. moniliforme was placed in the center of a Petri dish (diameter 6 cm) containing 10 mL of PDA with different concentrations (104, 52, 26, 13, 6.5, 3.25 and 1.625 µg mL−1, respectively) of AAC, and incubated at 26°C for 72 h. Radial growth was measured, and the inhibitory activity of ACC was expressed as a percentage of the growth on the untreated ACC medium. The IC50 (50% inhibitory concentration of AAC) value was calculated. The tests were carried out with three replicates.
In vitro spore growth assay
Fresh spores were harvested from cultures of plugs containing mycelia of F. moniliforme in the 6% mung bean liquid medium for 3 days and suspended in sterile distilled water with 0.01% Tween-20. The activity of AAC against spore growth was determined in triplicate using disk diffusion on agar plate. The test plates (90 mm in diameter) contained 17.5 mL of PDA seeded with spores of F. moniliforme, at a concentration of 104 mL−1. Doubly diluted AAC-containing solutions were applied to oxford plates (6 mm in diameter). The concentration, which developed a zone of inhibition of at least 6 mm diameter, was considered as the minimal inhibitory concentration (MIC).
Effects of AAC on maize infection with F. moniliforme
Shelled maize seeds were used for the experiment. Seeds were surface sterilized by 75% ethanol for 40 s and placed in culture flasks (200 seeds flask−1). The prepared seeds were sprayed with 10 mL AAC extraction (containing 26 mg AAC) flask−1, inoculated with a spore suspension of F. moniliforme, giving about 10 000 conidia flask−1. After incubation at 26°C for 7 days, the seeds were sampled. The control seeds were sprayed with 10 mL sterile water only. Fifty kernels from each replicate (n=3) were plated and incubated at 25°C on M3S1B agar medium (Griffin & Garren, 1974). Three to 4 days after plating, hyphal tips of fungal colonies were observed under a binocular microscope (Wilson et al., 1975). Infection of seeds by F. moniliforme was determined by the morphology and production of microconidia in chains under a microscope (Nelson et al., 1983). Percentage of infection was then calculated by measuring the proportion of infected seeds.
Results
Identification of strain B-FS01
Strain B-FS01 grew well on an NA plate and formed gray to ash, wrinkled colonies. Under the microscopes, the strain was observed to be a gram-positive, short-rod-like bacterium surrounded with growing flagellum responsible for swimming motility. A spot of c. 2.0 × 7.7 µm was observed under the phase-contrast light microscope. These salient features, as well as its physiological and biochemical properties (e.g. optimal growth temperature of 25–28°C; growth in 7% NaCl; and production of caseinase, catlase, tyrosinase, oxidase and acid form mannitol) were similar to those of B. subtilis. However, they differed from B. subtilis in its ability to hydrolyze lecithin and inability to form acid from xylose.
To confirm the above observation, a 16S rRNA gene fragment from strain B-FS01 was isolated, sequenced and subjected to a blast search and phylogenic analysis. The sequence (GenBank accession number DQ520955) contained 1510 nucleotides. The blast results indicated that the sequence had a high similarity to those of six B. subtilis strains, seven unidentified Bacillus strains, an unidentified strain and a Paenibacillus strain (Fig. 1). Phylogenic tree analysis showed that the strains closest to B-FS01 were B. subtilis strain AY887082A and AY775778. They shared more than 99% similarity. Therefore, the strain B-FS01 identified in the current study was most likely B. subtilis.
A phylogenic tree of strain B-FS01 was generated by the neighbor-joining method. The numbers at the nodes represent bootstrap values; those before the strain names are GenBank accession number. Scale bar — 0.001 nucleotide substitution per site (knuc value).
Antifungal activities of strain B-FS01
When tested in vitro against 12 fungal species of phytopathogens, B-FS01 inhibited the mycelial growth of most fungi (Table 1), but was not able to inhibit the growth of A. flavus. It is noted that strain B-FS01 exhibited strongly antagonistic activity against F. moniliforme Sheldon ATCC 38932.
Antifungal activity of strain B-FS01 against 12 fungal pathogens including Fusarium moniliforme
Values are the means ± SDs of triplicate measurements.
NI, no inhibition.
Antifungal activity of strain B-FS01 against 12 fungal pathogens including Fusarium moniliforme
Values are the means ± SDs of triplicate measurements.
NI, no inhibition.
Purification and analysis of AAC
To purify AAC, B-FS01 was cultivated in Czapek medium containing no proteinaceous components that might interfere with the purification procedure. The culture filtrates showed high antifungal activities in agar diffusion assays. Owing to the thermal stability of AAC, most of the proteins from culture filtrates were removed by boiling the sample for 30 min. Finally, apparent homologues of AAC were obtained through a C18-medium-column by HPLC (Fig. 2).
HPLC chromatogram of active fractions collected through a Sephadex G-100 gel column. Column: Cosmosil C18-AR-II column, 4.6 mm ID × 250 mm, 300 A, 5 µm particles. Mobile phase: 20 mM NH4Ac water solution (A), methanol (B). Detector: 215 nm. Components at retention time between 40 and 45 min were collected for further identification by HPLC/ESI-MS/MS.
The HPLC/ESI mass spectrum of AAC purified from the culture of B. subtilis B-FS01 revealed a cluster containing several molecules that were observed at [M+H] m/z=1434.0, 1435.9, 1450.6, 1464.0, 1479.0, 1492.9, 1506.0 and 1519.9 (Table 2). The m/z of these peaks showed high similarity to the fengycin homologues (Steller & Vater, 2000; Vater et al., 2002; Kim et al., 2004). Each of these ions was selected as a precursor ion for further CID analysis. The results showed that the appearance of productions of the precursor ions had regularities that appeared in the CID spectra of fengycin homologues (Wang et al., 2004; Sun et al., 2006): product ions of m/z 1080 (ions with neutral losses of fatty acid–Glu) and 966 (ions with neutral losses of fatty acid–Glu–Orn) were found in the CID spectra of precursor ions of m/z 1434.0,1435.9, 1450.6, 1464.0 and 1492.9, as well as fengycins A (Table 2); product ions of m/z 1108 (ions with neutral losses of fatty acid–Glu) and 994 (ions with neutral losses of fatty acid–Glu–Orn) were found in CID spectra of precursor ions of m/z 1506.0 and 1519.9, as well as fengycins B (the substitution of Ala for Val in the lactone ring of fengycin A) (Table 2). The CID spectra of a molecular ion of m/z of 1479.0 (Fig. 3) contained product ions of m/z 980 and 1094. Although the two product ions were different from fengycins A and fengycins B, they could still be explained as neutral losses of fatty acid–Glu and fatty acid–Glu–Orn of a new fengycin type. The above results could allow to conclude that the identified AAC was the a fengycin homologues containing fengycins A, fengycins B and a new fengycin type.
Analysis of AAC from strain B-FS01 by HPLC/ESI/CID MS
Containing one double bond in the fatty acid chain. ?, Identified in this study.
Abu probably was amino acid at position 6 of the newly identified fengycin type.
Analysis of AAC from strain B-FS01 by HPLC/ESI/CID MS
Containing one double bond in the fatty acid chain. ?, Identified in this study.
Abu probably was amino acid at position 6 of the newly identified fengycin type.
CID spectrum of a precursor ion of m/z 1479.0.
In vitro anti-F. moniliforme activity
AAC strongly inhibited the mycelial growth of F. moniliforme. A dose–response curve was measured for F. moniliforme, and the IC50 value of AAC against F. moniliforme was 20 µg mL−1. In addition, MIC of AAC against the growth of F. moniliforme spores was found to be 0.78 µg per plate in agar diffusion assays (Fig. 4). Growth inhibition zones caused by AAC were stable for a few weeks, when the test plates were incubated at room temperature.
Growth of Fusarium moniliforme Sheldon ATCC 38932 spores in response to 12.48 (1), 6.24(2), 3.12(3), 1.56 (4), 0.78(5) and 0.39(6) µg of AAC per plate, respectively.
Effects of AAC on maize infection with F. moniliforme
The activity of AAC against F. moniliforme infection of maize seed was assayed. As shown in Fig. 5, the infection percentage of maize treated with AAC was 38–39% 1 week after inoculation, while the control maize seeds (without ACC treatment) showed 90–100% infection of F. moniliforme. However, more treated seeds were infected over time. The protective effect of the AAC extract was no longer observed after 3 weeks
Effect of AAC on percentage infection of maize seeds by ATCC 38932. Each data point represents the mean ± SD of three replicates. ▲, percentage infection of maize seeds sprayed with sterile water; ▪, percentage infection of maize seeds sprayed with AAC.
Discussion
Bacillus subtilis is well known for controlling fungal and its bacterial diseases and potential to produce antibiotics with an amazing variety of structures (Shreiber et al., 1988; Stein, 2005). In this study, the strain B-FS01 was identified and it was found that it is B. subtilis. This identification could be supported by the result of a previous experiment, in which the flagellin central domain sequence of strain B-FS01 (GenBank accession number EF362756) was determined and the identity to the members of DB9011 type subgroup of B. subtilis defined by Asano et al., (2001) was found. In the DB9011 subgroup, most strains showed antifungal activity. Similarly, the strain B-FS01 in the present study has a potent activity against the strain F. moniliforme Sheldon ATCC 38932 and other phytopathogens. These results indicate that strain B-FS01 might have a broad spectrum of antifungal activities.
The biocontrol potential of B-FS01 related to the strong antagonism was developed in vitro by the strain and its cell-free culture filtrate against F. moniliforme. This suggests that antibiosis, because of the production of AAC by strain B-FS01, could play a major role in the inhibition of the disease. By multiple-chromatography approaches, AAC could be isolated from the cell-free culture of B-FS01. HPLC/ESI/CID mass spectral analysis demonstrates that clustered molecules are fengycins A, fengycins B and a novel fengycin type. Shu et al., (2002) reported that the FenE2 module could activate Abu as well as Ala and Val in vitro, but no fengycin type containing Abu has been reported. In this study, the 14 Da difference in molecular weight between Abu and Ala or Val just corresponds to the speciality of regular ions in the CID spectrum of this novel fengycin type in the AAC, which suggests that the amino acid at position 6 of this fengycin type was Abu. Besides fengycins, no other antifungal substance was found. With the observation of HPLC/ESI spectra, fengycins A were absolutely dominant in the quantity of fengycin homologues (Table 2), suggesting that they may play important roles in anti-F. moniliforme activity. It has been reported that the biological activity of iturins, which are nonribosomal antifungal lipopeptide antibiotics produced by B. subtilis as well as fengycin, depended on the composition of the peptide cycle and the nature and length of lipid chain (Peypoux et al., 1978; Quentin et al., 1982; Maget-Dana & Peypoux, 1994). Whether fengycin homologues used the same characters as iturins requires more investigations.
Fengycins have antifungal activity against some filamentous fungi in vitro (Vanittanakom et al., 1986). But to date, no report has been available on the antifungal application of fengycins in practice. Thus, it is necessary to carry out intensive research on an individual target pathogen for their actual application. The capacity of AAC for inhibiting the growth of F. moniliforme was determined. The results showed that AAC had strong anti-F. moniliforme activity in vitro against the growth of both mycelia and spores. It was interesting to find that AAC was able to modify the F. moniliforme infection of maize seeds in the first 2 weeks. However, the activity could not last too long. The reason was likely the degradation or interaction of the AAC with other substances secreted by germinating seedlings. It is noted that fengycin homologues are less toxic for erythrocytes than iturin A (Vanittanakom et al., 1986), and its hemolytic activity is 40-fold less than that of surfactin (Deleu et al., 2005a). Hence, AAC might be one of the ideal candidates to control maize infection by F. moniliforme.
One possible mechanism for antifungal activity of fengycins is that they interact with sterol and phospholipid molecules in membrane and thus disturb the structural properties of target membranes (Vanittanakom et al., 1986; Deleu et al., 2005b). This could also be supported by the results of intensive staining of AAC-treated hyphae of F. moniliforme with propidium iodide, where the damage of toxin-mediated cytoplasmic membrane occurred (data not shown). However, no staining of spores was observed. These results indicate that another mechanism might be responsible for the activity against the growth of spores.
The methods that were used to extract and purify fengycins for this study are much too complex and labor intensive for commercial application. According to the method of Kim et al., (2004), the methanol extraction from precipitates of B-FS01cultures at pH 2.0 mainly contained fengycin homologues and surfactin-like homologues, which are distinguished by its exceptional emulsifying, foaming, anti-viral and antimycoplasma activities (Peypoux et al., 1999). Such extractions of fengycins could be considered for agricultural practice. A sound method for fengycins application is also crucial to control postharvest contamination caused by F. moniliforme efficiently.
Acknowledgement
L.B.H and Z.Q.S contributed equally to this work.
References
Author notes
Editor: Andre Klier
