Biocontrol Traits and Antagonistic Potential of Bacillus amyloliquefaciens Strain NJZJSB3 Against Sclerotinia sclerotiorum, a Causal Agent of Canola Stem Rot

Abstract

Bacillus amyloliquefaciens strain NJZJSB3 has shown antagonism of several phytopathogens in vitro, especially Sclerotinia sclerotiorum. Both the broth culture and cell suspension of strain NJZJSB3 could completely protect the detached leaves of canola (Brassica napus) from S. sclerotiorum infection. In pot experiments, the application of strain NJZJSB3 cell suspension ($10^8CFU/ml$) decreased the disease incidence by 83.3%, a result similar to commercially available fungicide (Dimetachlone). In order to investigate the potential biocontrol mechanisms of strain NJZJSB3, the nonvolatile antifungal compounds it produces were identified as iturin homologs using HPLC-ESI-MS. Antifungal volatile organic compounds were identified by gas chromatography-mass spectrometry. The detected volatiles toluene, phenol, and benzothiazole showed antifungal effects against S. sclerotiorum in chemical control experiments. Strain NJZJSB3 also produced biofilm, siderophores and cell-wall-degrading enzymes (protease and ${\beta}$-1,3-glucanase). These results suggest that strain NJZJSB3 can be a tremendous potential agent for the biological control of sclerotinia stem rot.

Introduction

Sclerotinia sclerotiorum (Lib.) de Bary is a soil-borne plant pathogen that infects over 400 plant species at all growth stages and their harvested products [1]. Millions of dollars have been lost, owing to this pathogen, in canola production every year [27]. Canola (Brassica napus) is a valuable crop worldwide, with about 7 million hectares in China. The incidence of sclerotinia stem rot (SSR) of canola in China was estimated to be 10-20% on average, and reached up to 80% in some instances [22]. It is very difficult to control this pathogen because of its wide host range, high potential virulence, and the survival of sclerotia in soil. Conventional methods, such as crop rotation, cultivar resistance, and moisture regulation have not been entirely effective. The application of fungicides has been a useful method for controlling S. sclerotiorum, but it has problematic environmental and ecological consequences [6].

Biological control agents (BCAs) provide an effective alternative means of controlling S. sclerotiorum [23]. The BCAs Coniothyrium minitans [37], Ulocladium atrum [17], Trichoderma spp. [18], and Pseudomonas spp. [29] have reportedly been effective in controlling Sclerotinia diseases. As antagonists, Bacillus spp. have also been reported to be effective for the biocontrol of multiple soil-borne plant diseases [30], including S. sclerotiorum. B. subtilis strain SB24 significantly reduced the mycelial growth of S. sclerotiorum and suppressed sclerotia formation on the surface of canola leaves [42]. The foliar application of Bacillus amyloliquefaciens strains has been shown to be a promising strategy for the management of SSR in soybean plants [3]. The biological control mechanisms of Bacillus spp. include the production of antibiotics, volatile organic compounds (VOCs), hydrolytic enzymes, and siderophores. The non-ribosomally synthesized cyclic lipopeptide antibiotics produced by Bacillus spp. have displayed a broad spectrum of antimicrobial activities [21]. Hydrolytic enzymes, including protease, β-1,3-glucanases, cellulases, xylanase, lipase, amylase, and chitinase, destroy the components of the cell wall such as chitin, β-glucans, and proteins [39]. Siderophores drastically reduce the availability of ferric ions to certain rhizosphere microflora and inhibit the growth of pathogens [20]. The antimicrobial activities of VOCs produced by Bacillus strains are also well studied [4, 10, 12]. The production and types of VOCs from different Bacillus strains vary from strain to strain and mainly depend on the growth medium formula.

Bacillus amyloliquefaciens strains have shown several antifungal mechanisms to control plant diseases. However, the key biocontrol mechanisms of different biocontrol agents vary greatly. This study was conducted to evaluate the antagonistic efficiency of B. amyloliquefaciens NJZJSB3 on SSR as compared with a chemical fungicide. In addition, possible biocontrol mechanisms including, the production of antifungal lipopeptides, extracellular degradation enzymes, siderophores, and VOCs, were studied.

 

Materials and Methods

Isolation and Identification of Antagonistic Strain

Antagonistic strain B. amyloliquefaciens NJZJSB3 was isolated from a forest soil sample (Tzu-chin Mountain, Nanjing, China) by dilution plate technique. Its 16S rRNA gene was sequenced and BLAST searched against the NCBI database. The sequences of its close relatives were downloaded to construct a neighbor-joining phylogenetic tree using MEGA 4.0. Morphological and physiological properties were determined by the method described by Holt et al. [15].

In Vitro Inhibition of S. sclerotiorum and Other Phytopathogens by NJZJSB3

The antagonistic activity of strain NJZJSB3 was evaluated against several phytopathogens in vitro using the method reported by Bordoloi et al. [7] Briefly, a 6-mm agar plug of S. sclerotiorum was placed in the middle of a PDA plate, and strain NJZJSB3 was inoculated between the edge of the plate and the plug. After 5 days of incubation at 25℃, the inhibition zone was measured. S. sclerotiorum, the causal agent of stem rot disease of canola, was provided by Plant Protection College, Nanjing Agricultural University, Nanjing, China. The antagonistic activity of strain NJZJSB3 against other fungal pathogens, including F. oxysporum f. sp. cucumerinum J. H. Owen, Verticillium dahliae Kleb, F. oxysporum f. sp. niveum, Rhizoctonia solani, F. oxysporum f. sp. cubense, and F. oxysporum Schl. f. sp. dioscoreae Wellman, was also determined by the same procedures. All the pathogens were grown on PDA plates and stored at 4℃.

The antifungal activity of cell-free supernatants of strain NJZJSB3 was evaluated against the target fungus S. sclerotiorum using the agar diffusion method. For the mycelium growth inhibition assay, strain NJZJSB3 was cultured in Landy medium (100 ml) and incubated at 30℃ and 170 rpm. After 3 days, the suspension was centrifuged at 12, 000 rpm for 20 m in and the supernatant was passed through a 0.22 μm Millipore syringe filter, and then diluted with sterile distilled water (SDW) 10-, 20-, 40-, and 60-fold. A total of 200 μl of diluted filtrates was spread on the PDA plates and 6-mm agar plugs of S. sclerotiorum were placed in the center of each plate. For the sclerotia germination assay, after spreading 200 μl of the filtrates at the original concentration on the PDA plates, 10 surface-sterilized sclerotia were placed on them. All of the treatments had three replicates.

Biocontrol of S. sclerotiorum by Strain NJZJSB3 on Detached Canola Leaves In Vivo

The fifth true leaves from the top of canola plants were detached, washed in a flow of water, rinsed twice with SDW, slightly air dried and then treated with (i) SDW; (ii) culture liquid of strain NJZJSB3 in nutrient broth (109 CFU/ml); or (iii) strain NJZJSB3 cell suspension in SDW (109 CFU/ml). The culture liquid and cell suspension of strain NJZJSB3 were applied to detached leaves by spraying, until runoff, with a hand-pump sprayer. After slightly air drying, a 6-mm agar plug of S. sclerotiorum was detached from the pre-culture PDA plate (25℃ for 3 days) and placed on each canola leaf. The liquid culture of strain NJZJSB3 was prepared as described previously. The liquid culture was centrifuged at 6, 000 rpm for 5 min and washed twice with SDW, and then resuspended in SDW as a cell suspension. All treatments had 10 replicates and were placed in a growth chamber with a 16-h photoperiod at 25℃ and 8 h dark period at 20℃ with 75% relative humidity. The diameter of each lesion was recorded after 3 days.

Biocontrol Efficacy of Strain NJZJSB3 in Pot Experiment

For the evaluation of the biocontrol efficacy of strain NJZJSB3 in a pot experiment, canola seedlings were planted in sterilized soil at 25℃ until the 6-leaf age. Fifteen plants were selected for each treatment. Pots were organized in a completely randomized design. The fifth and sixth true leaves from the top were treated with: (A) SDW; (B) Dimetachlone (1 g/l); or (C) strain NJZJSB3 cell suspension (108 CFU/ml in SDW). Both S. sclerotiorum inoculation and cell suspension of strain NJZJSB3 preparation methods were the same as previously described. All treated leaves were covered with a wet paper product to prevent accidental dryness and to provide high relative humidity for the onset of pathogen/antagonist activities. After 3 days, the disease incidence was calculated.

Evaluation of Extracellular Enzyme Production and Biofilm Formation by Strain NJZJSB3

The qualitative assays for β-1,3-glucanase, protease, and siderophore production were performed according to the reported methods [23, 24, 31, 32]. Briefly, the β-1,3-glucanase activity was determined on a solid synthetic medium (glucan, 5 g; NaNO3, 2 g; K2HPO4, 1 g; KCl, 0.5 g; MgSO4·7H2O, 0.5 g; FeSO4·7H2O, 0.01 g; Congo-red, 0.05 g; distilled water, 1,000 ml; pH, 7.0). The protease activity was tested on skim milk agar plates by identifying the presence or absence of a clear zone around the colony. Siderophore production was assessed on chrome azurol S (CAS) agar plates by observing the color change from blue to orange. All of the plates were incubated at 30℃ for 3 days.

The biofilm formation ability of strain NJZJSB3 was also determined using the method described by O'Toole et al. [25]. The growth medium was 1/10 LB medium containing 0.15 mM ammonium sulfate, 100 mM potassium phosphate, 34 mM sodium citrate, 1 mM MgSO4, and 0.1% (w/v) glucose. Strain NJZJSB3 was pre-cultured at 30℃ until reaching an optical density at 600 nm (OD600) of 0.7-0.8. The cells were then diluted to an OD600 of 0.01 using fresh medium and transferred to 24-well PVC microtiter plates for growth at 37℃ for 24 h. Fresh culture medium was used as a control.

Optimum Culture Medium for Maximum Growth and Antagonistic Activity of NJZJSB3

Various culture media, such as potato dextrose broth, nutrient broth, tryptic soya broth, King’s B modified medium, Landy medium, and yeast peptone glucose medium (50 ml each), were inoculated with 1 ml of overnight culture of NJZJSB3. The growth of NJZJSB3 was measured by optical density (OD600) after 24 h at 30℃, and the antagonistic activity of cell-free supernatants was observed after 48 h by disc diffusion assay as described above.

Extraction and Identification of Antifungal Lipopeptides

Strain NJZJSB3 was grown in 50 ml of Landy medium at 30℃ and 170 rpm. After 3 days, lipopeptides were precipitated from cell-free supernatants by adding 6 N HCl to a final pH of 2.0 and stored at 4℃ overnight. The crude precipitates were collected by centrifugation at 12,000 rpm for 30 min at 4℃ and washed twice with distilled water (pH 2.0), dissolved in distilled water (pH 8.0), and subsequently filtered through a 0.22 μm filter. Then, the crude lipopeptides were subjected to the disc diffusion assay for antifungal activity and subsequently re-acidified to pH 2.0 with 6 N HCl. The precipitated yellowish-white residues were collected by centrifugation at 12,000 rpm for 15 min and dissolved in 1 ml methanol [21]. The molecular weights of the lipopeptides were detected by HPLC-ESI-MS operated in positive-ion mode (1200 series and ESI-MS, 6410 Triple Quad LC/MS; Agilent, USA), using a C18 column (50 mm × 2.1 mm, 1.8 μm). The mobile phase A was water with 0.05% acetic acid, and mobile phase B was acetonitrile. The flow rate was maintained at 0.4 ml/min with a gradient for 70 min (65% (v/v) acetonitrile for 3 min, 65%-0% (v/v) acetonitrile from 3-60 min, 0%-65% (v/v) acetonitrile from 60-62 min, 65%-0% (v/v) acetonitrile from 62-70 min).

Identification of Volatile Compounds and Assay of Their Antagonism

A solid-phase micro-extraction (SPME) technique with a fiber coated with divinylbenzene-carboxene-PDMS (DCP, 50/30 μm) was used in this experiment to collect the VOCs. Strain NJZJSB3 was inoculated into 60 ml of modified MS medium containing 1.5% agar, 1.5% sucrose, and 0.4% TSB [11] in a 150 ml vial and incubated at 28℃ with the SPME fiber inserted into the headspace. After 3 days, the vial was treated at 50℃ for 30 min in hot water. The SPME fibers were directly injected into the GC-MS injector for desorption at 220℃ for 1 min. The GC-MS used an RTX-5MS column (30 m, 0.25 mm inside diameter, 0.25 μm) equipped with Trace DSQ (Finnigan) detection. Each run was performed for 27.5 min. The initial oven temperature of 38℃ was held for 3 min, first ramped up at a rate of 10℃ /min to 180℃, and then at a rate of 40℃/min to 240℃, and held for 4 min. The mass spectrometer was operated in the electron ionization mode at 70 eV with a source temperature of 220℃, and a continuous scan from m/z 50 to m/z 500 was used. The mass spectra of VOCs were compared with those in the National Institute of Standards and Technology database [40]. Standard compounds were purchased (Sigma-Aldrich) to confirm the corresponding peaks identified through GC-MS and were used to test for antifungal activity by placing 200 μl of each compound on sterile filter paper discs placed in divided plates, while the other side of plates contained S. sclerotiorum mycelial plugs on the PDA medium [40].

Statistical Analysis

Disease incidence was calculated as the percentage of infected canola leaves over the total number of leaves in each treatment after 3 days. Disease incidence was calculated using the following equation: Disease incidence = number of infected leaves/ total number of leaves in each treatment × 100%. Data were assessed with one-way ANOVA. Duncan’s multiple-range test was applied when one-way ANOVA revealed significant differences (p ≤ 0.05). All statistical analyses were performed with the SPSS ver. 13.0 statistical software (SPSS, Chicago, IL, USA).

 

Results

Strain Identification and Inhibition Spectrum Assay

The 16S rRNA gene sequences showed that strain NJZJSB3 is closely related to B. amyloliquefaciens, with a similarity score of 98.2% in NCBI sequence alignment (Fig. 1). Microscopic examination revealed that strain NJZJSB3 cells are motile, rod-shaped, and have the ability to form spores when grown in LB medium. This strain showed strong antagonistic activity against S. sclerotiorum (Fig. 2A). The culture filtrate of the strain could also inhibit the sclerotia germination (Fig. 2B) and mycelial growth of S. sclerotiorum. The inhibition effect on the growth of S. sclerotiorum mycelia was 80.4% when the culture filtrate was diluted 60-fold (Fig. 3).

Fig. 1.Phylogenetic tree of B. amyloliquefaciens strain NJZJSB3 with closely related species after BLAST search of 16S rRNA sequences. Bootstrap values obtained with 1,000 resamplings are indicated as percentages at all branches.

Fig. 2.(A) Effect of B. amyloliquefaciens NJZJSB3 on S. sclerotiorum growth, and (B) Effect of fermentation broth filtrate on sclerotia germination.

Fig. 3.Effect of liquid culture cell-free filtrate of B. amyloliquefaciens NJZJSB3 on the mycelial growth of S. sclerotiorum after dilution up to 10-, 20-, 40-, and 60-fold. All values are the mean of three replicates. Bars with different letters indicate significant differences among the four treatments, as defined by Duncan’s test (p ˂ 0.05).

Strain NJZJSB3 significantly inhibited the growth of phytopathogens in vitro (Fig. 4). The inhibition rate of NJZJSB3 was highest (74%) against Rhizoctonia solani and lowest (46%) against F. oxysporum Sch. f. sp. dioscoreae Wellman. Strain NJZJSB3 showed 70.0%, 56.0%, 66.4%, and 63.8% antagonistic activity against F. oxysporum f. sp. cucumerinum J. H. Owen, Verticillium dahliae Kleb, F. oxysporum f. sp. niveum, and F. oxysporum f. sp. cubense, respectively. The broad-spectrum antimicrobial activity against these plant pathogens indicated that strain NJZJSB3 can be a potential biocontrol agent for the diseases caused by these pathogens.

Fig. 4.Demonstration of inhibitory effects of B. amyloliquefaciens NJZJSB3 on phytopathogens: (A) F. oxysporum f. sp. cucumerinum J. H. Owen, (B) Verticillium dahliae Kleb, (C) F. oxysporum f. sp. niveum, (D) Rhizoctonia solani, (E) F. oxysporum f. sp. cubense, and (F) F. oxysporum Schl. f. sp. dioscoreae Wellman. The inhibition rate was 70.0%, 56.0%, 66.4%, 74.0%, 63.8%, and 46.0% , respectively.

Control of S. sclerotiorum by Strain NJZJSB3

In the growth chamber, the liquid culture and cell suspension in SDW of strain NJZJSB3 showed evident suppression of S. sclerotiorum on detached canola leaves after 3 days, compared with SDW (Fig. 5). All of the canola leaves treated with SDW had high disease severity with an average lesion diameter of 27.1 mm, while no lesion zone was formed after treatment with culture liquid and cell suspension of NJZJSB3. However, in the pot experiment, the symptoms of water-soaked lesions were found both in leaves treated with strain NJZJSB3 cell suspension (Fig. 6A) and SDW (Fig. 6C), whereas Dimetachlone-treated leaves did not show disease symptoms (Fig. 6B). The infection levels between treatment with SDW and strain NJZJSB3 cell suspension were obviously different. Application of strain NJZJSB3 cells suspension provided 80 ± 5.77% control on canola plants, which was almost similar to the control obtained with commercially used fungicide.

Fig. 5.Effects of culture (A) and cell suspension (B) of B. amyloliquefaciens NJZJSB3 on the suppression of S. sclerotiorum on canola detached leaves. SDW (C) was used as the control. The average diameter of lesion zone appearing in a treatment was 27.1 mm, while no disease-affected leaves appeared in treatments A and B.

Fig. 6.Effects of cell suspension (A) of B. amyloliquefaciens NJZJSB3 and Dimetachlone (B) on the suppression of S. sclerotiorum on canola leaves. SDW (C) was used as control. Infected plant leaves exhibited watersoaked lesions as indicated by red arrows, and the incidence rates of treatments A, B, and C were 100%, 0%, and 16.7%, respectively.

Production of Extracellular Enzymes and Biofilm Formation by NJZJSB3

The appearance of clear zones surrounding strain NJZJSB3 colonies indicated that the strain secreted siderophores (Fig. 7A), protease (Fig. 7B), and β-1,3-glucanase (Fig. 7C). However, chitinase activity was not detected. These enzymes hydrolyze β-glucans and proteins, which are responsible for the integrity of fungal cell walls. The ability of strain NJZJSB3 to form biofilm was investigated in vitro. After incubation for 24 h at 37℃ in 24-well PVC microtiter plates, a thick biofilm was found compared with control culture medium (Fig. 7D).

Fig. 7.Production of siderophores and hydrolytic enzymes, and biofilm formation by B. amyloliquefaciens NJZJSB3. Siderophore production on CAS assay (A), protease production on skim milk agar (B), β-1,3-glucanase production on synthetic medium (C), and biofilm formation in microtiter plate containing LB medium (D).

Optimum Culture Medium for the Growth and Antagonistic Activity of Strain NJZJSB3

Different culture media significantly affected the growth and antagonistic activity of strain NJZJSB3 (Fig. 8). Culturing in Landy medium produced the highest optical density (OD600 = 3.52), followed by PDA (OD600 = 3.04), YPG (OD600 = 3.22), KB (OD600 = 2.25), TSB (OD600 = 1.98), and NA (OD600 = 1.43) media. However, a higher biomass production does not necessarily mean a higher production of antagonistic substances; thus, the antagonistic activity of the supernatants was further tested. Supernatant from Landy medium produced the largest inhibition zone (diameter = 34.67 mm), followed by that from YPG (31.1 mm), PDA (33.0 mm), KB (28.3 mm), NA (28.0 mm), and TSB (19.7 mm) media. Considering both the biomass production and antagonistic activity of the supernatant, Landy medium was chosen for the following studies.

Fig. 8.Effects of different culture media on the growth (A) and antagonistic activity (B) of B. amyloliquefaciens NJZJSB3. All values are the mean of three replicates. Bars with different letters indicate significant differences among the six culture media, as defined by Duncan’s test (p ˂ 0.05).

Identification of Antifungal Compounds Produced by Strain NJZJSB3

The production of antimicrobial compounds is one of the important antagonistic mechanisms. In this experiment, the nonvolatile antifungal compounds were isolated and the precipitated lipopeptides were analyzed by HPLC-ESI-MS. Five peaks corresponding to the elution time at 6.94 min (m/z = 1,030.6), 10.22 min (m/z = 1,044.7), 12.51 min (m/z = 1,058.8), 15.76 min (m/z = 1,072.8), and 17.82 min (m/z = 1,086.8) were identified (Fig. 9). The mass spectrum of these peaks showed that the molecular masses of these five compounds had a difference of 14 Da (-CH2) in their molecular masses (m/z), which belonged to the iturin homologs [28, 35].

Fig. 9.Mass spectrometry total ion chromatogram of the antifungal compounds from NJZJSB3. (A) Peaks eluted at 6.94 min (m/z = 1,030.6), 10.22 min (m/z = 1,044.7), 12.51 min (m/z = 1,058.8), 15.76 min (m/z = 1,072.8), and 17.82 min (m/z = 1,086.8) showed antifungal ability and are indicated by the arrows. Mass spectroscopic analysis of antifungal compounds produced by B. amyloliquefaciens NJZJSB3 revealed active peaks with different molecular weights of 1,030.6, 1,044.7, 1,058.8, 1,072.8, and 1,086.8 (B). Mass spectroscopy was performed in positive-ion mode.

The antifungal activity of the VOCs produced by strain NJZJSB3 was 40.26% compared with the control (Fig. 10). Five volatile compounds identified by GC-MS (Fig. S1) were toluene, phenol, benzothiazole, heptadecane, and 6,10,14-trimethyl-2-pentadecanone (Table 1). As the chemical nature of the VOCs was responsible for the antifungal activity [12], the standard chemicals were purchased to confirm their antifungal effects. Standards of toluene, phenol, and benzothiazole exhibited antifungal activity against S. sclerotiorum. Owing to the unavailability of a standard reagent, the antifungal activity of 6,10,14-trimethyl-2-pentadecanone was not determined.

Fig. 10.B. amyloliquefaciens NJZJSB3 antifungal volatile activity on divided plates. Mycelial plug growth was inhibited in the presence of the bacteria streaked in the different compartments (B), compared with the control (A). The inhibition rate of mycelial growth was 40.26% in the presence of the bacteria streaked in the different compartments compared with the control.

Table 1.Y = inhibited S. sclerotiorum growth; N = did not inhibit S. sclerotiorum growth; NS = could not purchase compound to check antifungal effect.

 

Discussion

Isolation of new antagonistic strains is necessary to improve biological control methods and control plant diseases. In this study, we isolated a strain from the forest soil and identified it as Bacillus amyloliquefaciens NJZJSB3 based on 16S rRNA gene sequence analysis. The antagonist assay results showed that strain NJZJSB3 has broad-spectrum antimicrobial activity against several plant pathogens, such as F. oxysporum f. sp. cucumerinum J. H. Owen, Verticillium dahliae Kleb, F. oxysporum f. sp. niveum, Rhizoctonia solani, F. oxysporum f. sp. cubense, F. oxysporum Schl. f. sp. dioscoreae Wellman, and S. sclerotiorum. Many B. amyloliquefaciens strains have shown broad-spectrum antimicrobial activity in vitro and have been successfully used to control soil-borne diseases under pot and field conditions. For example, the application of B. amyloliquefaciens W19 reduced the Fusarium wilt of banana plants [35]. B. amyloliquefaciens strain B94 was used as a biocontrol agent to suppress R. solani and other fungal plant pathogens [38]. B. amyloliquefaciens strain NJN-6 is an important plant growth-promoting rhizobacterium that can produce secondary metabolites antagonistic to several soil-borne pathogens [41]. Based on the antimicrobial activity examined in the present study, strain NJZJSB3 seems to have potential for use as a biocontrol agent against these phytopathogens.

In this study, strain NJZJSB3 was further evaluated for its potential to control the stem rot disease of canola caused by S. sclerotiorum in vitro and in vivo. The results showed that both the broth culture and cell suspension of strain NJZJSB3 could protect detached canola leaves from lesions in the presence of S. sclerotiorum at high humidity for 3 days. In the greenhouse experiment, strain NJZJSB3 at a concentration of 108 CFU/ml provided 83.3% control of sclerotinia stem rot of canola. Another study showed that the severity of S. sclerotiorum of canola was reduced by 77% with the application of B. subtilis strain NJ-18 at a concentration of 1.0 × 107 CFU/ml [36]. The higher concentration of strain NJZJSB3 than that used by Yang et al. [36] may have been a factor contributing to better disease control. B. amyloliquefaciens strains have been reported to be effective in the control of sclerotinia stem rot of other different plants. B. amyloliquefaciens at a concentration of 106 CFU/ml provided over 80% control on squash, eggplant, and tomato seedlings [1]. Zhang and Xue [42] reported that soybean leaves treated with the cell suspension (5 × 108 CFU/ml) and liquid culture of B. subtilis strain SB24 significantly reduced the SSR severity over 15 days, compared with untreated plants. Further studies are required to evaluate the effects of strain NJZJSB3 in the biocontrol of SSR of canola and other plants at low concentrations.

Knowledge of antagonistic metabolites excreted by strain NJZJSB3 helped us to understand its biocontrol mechanisms. A number of species in genus Bacillus are capable of producing a wide arsenal of antimicrobial substances inhibitory to phytopathogens, including lipopeptides and macrolactins, volatile compounds, and hydrolytic enzymes [2, 26]. B. amyloliquefaciens, which can produce lipopeptides, displayed a strong in vitro antifungal activity against a wide variety of yeast and fungi [26]. However, the major lipopeptide-like compounds produced by strains of B. amyloliquefaciens were different. We identified five lipopeptide compounds produced by strain NJZJSB3 (Fig. 8) that belong to the iturin homologs [5]. The iturin homologs penetrate into the cytoplasmic membrane by the hydrophobic tail, followed by auto-aggregation resulting in pore formation, which causes cellular leakage. The production of these compounds explains why strain NJZJSB3 could effectively inhibit the mycelial growth of several phytopathogens.

In most fungi, chitin and noncellulosic β-glucans are the most abundant skeletal or microfibrillar components, whereas proteins and β-glucans are the main cementing components [19]. Hydrolytic enzymes produced by biocontrol agents, such as chitinases, glucanases, and protease, are important mechanisms involved in the biocontrol of plant pathogens [13, 16]. In this study, we found that strain NJZJSB3 secreted protease and β-1,3-glucanase on agar plates, but the chitinase activity was not detected. Previous studies have demonstrated the lysis of fungal cell walls by either microbial chitinase or β-1,3-glucanase in vitro. The β-1,3-glucanase produced by B. amyloliquefaciens MET0908 showed strong antifungal activity against plant pathogens in a watermelon pot assay. The absence of chitinase production by strain NJZJSB3 would not affect the hydrolytic activity against the fungal cell wall by β-1,3-glucanase and protease. Strain NJZJSB3 is also able to produce siderophores and biofilm in LB medium. Siderophores produced by a microorganism can bind iron with high specificity and affinity, making the iron unavailable for other microorganisms, and thereby limiting their growth. The siderophore-producing bacterium Bacillus subtilis CAS15 has a biocontrol effect on Fusarium wilt and promotes the growth of pepper [39]. Siderophores produced by strain NJZJSB3 can be an important antagonistic trait in the biocontrol of SSR. It was reported that the capability of Bacillus strains to protect plants from phytopathogens is mediated in part by the formation of biofilm on the plant roots [9]. In this study, strain NJZJSB3 also showed the ability to form biofilm in vitro. We sprayed a cell suspension of strain NJZJSB3 on canola leaves to prevent SSR infection and achieved a good biocontrol effect. However, whether strain NJZJSB3 formed biofilm on the surface of canola leaves is still unknown and needs to be studied in the future.

Volatile compounds produced by microbial antagonist strains that exhibit strong inhibitory activity against plant pathogens have received much attention [40]. VOCs produced by S. platensis F-1 can reduce the severity of leaf blight of canola caused by S. sclerotiorum [34]. The chemical nature of the VOCs is responsible for the antifungal activity [12]. In this study, five VOCs were produced by strain NJZJSB3. When these volatile compounds were tested against S. sclerotiorum in the pure form, toluene, phenol, and benzothiazole showed antifungal effects against S. sclerotiorum. Phenol is known for its toxic effects on cells, and has been used as an antiseptic in clinical applications for a long time [8]. The benzothiazoles have been reported not only for the inhibition of fungal mycelial growth [12], but are also involved in clinical applications because of their antitumor properties [31]. The results for the antifungal activity and production levels of these VOCs infer that benzothiazole is a promising candidate as an effective antifugal VOC produced by strain NJZJSB3. Owing to their easy evaporation and antifungal properties, VOCs produced by strain NJZJSB3 might play a positive role in the biocontrol of S. sclerotiorum as a long-distance control mechanism.

In conclusion, Bacillus amyloliquefaciens strain NJZJSB3, isolated and identified in this study, can effectively suppress the S. sclerotiorum infection of canola plants both in vitro and in vivo. The potential biocontrol mechanisms of B. amyloliquefaciens NJZJSB3 were explored, including the production of antifungal lipopeptides, VOCs, and cell-walldegrading enzymes. However, the biocontrol efficiency of B. amyloliquefaciens NJZJSB3 in field experiments should be further studied. B. amyloliquefaciens strain NJZJSB3 can be a potential agent for the biological control of SSR.