Antifungal and plant growth promotion activity of volatile organic compounds produced by Bacillus amyloliquefaciens

Abstract Fusarium wilt of watermelon, caused by F. oxysporum f.sp. niveum (FON), is a devastating disease that causes extensive losses throughout the world. Five bacterial strains (L3, h, β, b, and L) isolated from the watermelon rhizosphere showed antagonistic activity against FON during in vitro tests. Strain L3 produced diffusible and volatile organic compounds (VOCs) which showed the strongest antifungal activity. Arabidopsis thaliana plantlets exposed to VOCs produced by strain L3 showed a 2.39‐fold increase in biomass, 1.40‐fold increase in primary root length, and 5.05‐fold increase in number of lateral roots. Confocal laser scanning microscope showed that the GFP‐labeled strain L3 could colonize along the elongation and differentiation zones of watermelon roots. In greenhouse pot experiments, the biocontrol efficiency of strain L3 against fusarium wilt of watermelon was up to 68.4% in comparison with the control treatment. In addition, inoculation of the strain L3 resulted in a 23.4% increase in plant fresh weight. Based on 16S rDNA sequence analysis, the strain L3 was identified as Bacillus amyloliquefaciens L3. Fourteen VOCs produced by strain L3 were identified through GC‐MS analysis. Of nine VOCs tested, 2‐nonanone and 2‐heptanone were proved to have strong antifungal properties. Acetoin and 2,3‐butanediol were found to promote plant growth. The results suggested B. amyloliquefaciens L3 was a potential biocontrol agent, and that VOCs produced by B. amyloliquefaciens L3 play important roles in the process of biocontrol and plant growth promotion.

one of the leading methods for controlling Fusarium wilts (Sun, Song, & Fu, 2015). However, soil fumigation with chemicals is known to have broad biocidal activity, and detrimental effects were found not only on target pathogens but also on non-target microorganisms and the fumigation chemicals themselves could be extremely dangerous to humans (Yan, Wang, & Li, 2017). The application of biological control agents (BCA) has been recognized as an environmentally friendly and sustainable method to reduce the effects of plant diseases (Raza, Wang, Wang, & Wu, 2016;Raza, Wei, Wei, & Ling, 2016;Saravanakumar, Li, & Yu, 2013).
Rhizosphere soils from healthy plants which survived in a field infested with phytopathogens are a good source for the isolation of biocontrol agents (Huang, Wei, & Tan, 2013). In recent decades, many biocontrol agents have been successfully isolated from rhizosphere niches, such as Bacillus spp., Pseudomonas spp., Trichoderma spp., Streptomyces spp., and effectively used for the control of Fusarium wilt in many different commercial crops (Faheem, Raza, & Wei, 2015). Among these promising BCAs, Bacillus spp. are well-known for their inherent property to produce spores and for their resistance to extreme conditions (Shafi, Tian, & Ji, 2017). Bacillus subtilis SQR9 isolated from the rhizosphere of cucumber, showed strong antagonism against F. oxysporum f.sp. cucumerinum (Cao, Zhang, & Ling, 2011). Akarm, Mahboob, and Javed (2013) showed that B. thuringiensis strain 199 can protect tomato plants against Fusarium wilt. The genus Bacillus spp. were found to be well adapted to the rhizosphere of watermelon. However, relative few biocontrol agents belonging to Bacillus spp. are used to control the Fusarium wilt of watermelon (Raza, Yuan, & Ling, 2015;Zhao, Wang, & Liang, 2018). Thus, more studies are still needed to explore biocontrol agents belonging genus Bacillus to strengthen the range of weapons available for the biocontrol of Fusarium wilt of watermelon.
Bacillus strains exhibit various biological control mechanisms, such as production of a wide spectrum of antibiotics, synthesis extracellular enzymes, competition for nutrients and niches, and induction of systemic resistance in plants against pathogens (Chowdhury, Hartmann, & Gao, 2015;Santoyo, Orozco-Mosqueda, & Govindappa, 2012). Among these mechanisms, previous research has shown that root colonization ability is a prerequisite for biological control agent's activity (Mendis, Thomas, Schwientek, & Salamzade, 2018). Cavaglieri, Orlando, and Rodriguez (2005) reported that the biological control efficiency of plant pathogens was directly related to the root colonization ability of biological agents. Zhang, Wu, He, and Li (2011) demonstrated the critical importance of the colonization of B. subtilis N11 on banana roots to stop the pathogen invading by using a GFP-tagged B. subtilis N11. Therefore, it is essential to evaluate root colonization ability when searching promising BCAs.
In this study, five (L3, h, β, b, and L) potential biocontrol strains of Bacillus were isolated from the healthy watermelon plants which survived in a field heavily infested with FON. The objectives of the study were to: (a) isolate Bacillus with antifungal activity against FON, (b) evaluate antifungal and plant-growth promotion effect of VOCs produced by the isolated strains, (c) evaluate the root colonization ability of the selected biocontrol strain, (d) evaluate the biocontrol efficiency and plant growth effect by performing greenhouse pot experiments, and (e) discover the components and functions of VOCs secreted by the selected strain.

| Microorganisms and growth conditions
The culture of F. oxysporum f. sp. niveum (FON) used in this study was originally isolated from infected stems collected from watermelon fields located in Jiangsu, China by the method described by Chang, Ling, Chen, and Huang (2015). FON was pathogenic to watermelons in in vivo experiments. Stock cultures were maintained on potato dextrose agar (PDA) plates at 4°C. Pre-cultures were established by transferring a stock agar plug containing mycelia onto fresh PDA plates and incubating for 4 days at 28°C.

| Isolation and screening of bacteria for biocontrol activity
The rhizosphere soils were sampled from healthy watermelon plants that had survived in a field with a history of Fusarium wilt in Jiangsu province of China. The serial-dilution-pour method was used to isolate rhizosphere bacteria on nutrient agar (Beef extract 3.0 g/L, peptone 5.0 g/L, sodium chloride 3.0 g/L, and agar 20.0 g/L).
The bacterial strains were further purified and then screened for antagonistic activity toward FON using dual culture technique.
Briefly, a 3 mm agar plug from the edge of a 4-day-old culture of FON on PDA was placed in the middle of fresh PDA plate. The bacterial strains were spot inoculated in the edge of the plate and incubated at 28°C. The PDA plate inoculated only with FON was used as control treatment. Each treatment was repeated three times. The diameter of FON in each plate was then recorded after 5 and 10 days.

| Antagonistic effects of VOCs on growth of FON
The antagonistic effects of VOCs produced by the selected bacterial strains (L3, h, β, L, and b) isolated above, were measured according to the method described by Wu et al. (2014). The Petri dish containing modified Murashige-Skoog (MS) culture medium (0.5% TSB and 2% agar) inoculated with the isolated strains was covered with another Petri dish containing PDA inoculated with a 6 mm diameter plug of FON. Then the two dishes were sealed with Parafilm to obtain a double-plate chamber. The average distance between MS and PDA agar surface was 1.5 cm. The double-plate chamber was incubated at 28°C for 4 days. The double-plate chamber without bacterial strains were used as control. The experiment was repeated three times.

| Plant growth promotion by VOCs
Arabidopsis thaliana Col-0 seeds were surface sterilized in 70% (v/v) ethanol for 30 s and afterwards incubated for 5 min in sodium hypochlorite solution, then rinsed with sterile water for four times. Seeds were placed on Petri dishes containing 0.5 × MS for 2 days at 4°C and then germinated for 2 days on Petri dishes containing 0.5 × MS, 0.8% sucrose and 0.6% bacto agar. Then, five plantlets were transferred to 90 mm diameter Petri dishes containing 0.5 × MS, 0.8% sucrose and 1% bacto agar.
Each of the five selected bacterial strains was pre-cultured in a 60 mm diameter Petri dish containing MS, 0.5% TSB, and 2% bacto agar for 24 hr at 30°C. Then two Petri dishes containing plantlets and bacterial strains were put into a 150 mm diameter plate and sealed with Parafilm and placed vertically in a growth chamber at 22°C, 16 hr light/8 hr dark (Appendix Figure A1). Each treatment had three replicates. Plant fresh weight, primary root length, and lateral root number were measured after 7 days incubation.

| Construction of GFP-tagged L3 and root colonization assay
Based on the results of 2.3-2.5 section, the strain L3 was chosen for further studies. The gfp-marked shuttle vector, pHAPII (GenBank accession number HM151400), was used to construct the GFP-tagged L3 by the method described by Cao et al. (2011). A GFP-tagged L3 mutant that emitted green fluorescence was chosen and used in the root colonization assay.
Watermelon "Sumi 8" were surface-sterilized by immersing in sodium hypochlorite solution (2%, v/v) for 15 min and rinsed five times in sterilized distilled water. The seeds were placed on moist filter papers in a petri dish under sterile conditions for germination at 28°C. After 3 days, the germinated seeds were then sown in soilless culture medium (3:2:3:2 mix of fermentation bed farming material; vermiculite; perlite; peat) at 28°C with 16 hr light/8 hr dark. Then the seedlings with 1-2 leaves were used for the subsequent studies.
A hydroponic system was used for colonization assays as described by Zhang et al. (2011). The watermelon seedlings were collected and gently washed to remove the adhered substrate. The seedlings were placed into 50-ml flasks (one seedling in each flask) containing 50 ml of liquid 1/2 MS medium at 28°C. A 1 ml suspension of GFP-tagged L3 (10 8 cfu/ml) was added to each flask of hydroponic medium. The culture conditions were the same as described above. The root samples were collected at 48 hr and 96 hr.
Part of each root sample (0.1 g) was ground in a mortar with 0.9 ml sterilized distilled water until a fine homogenate was obtained.
The suspension was diluted, then plated on modified LB medium (20 μg/ml, kanamycin) and colony-forming units were counted after incubating at 37 ℃ for 2 days. The patterns of GFP-tagged L3 colonization of watermelon roots were examined by a Confocal Laser Scanning Microscope (Leica Model TCS SP2, Heidelberg, Germany) as described by Huang, Zhang, Yong, and Yang (2012).
Roots without inoculation with the GFP-tagged L3 served as control. Each treatment was repeated three times.

| Biological control and plant growth promotion activity of the strain L3
The ability of strain L3 to suppress Fusarium wilt of watermelon was investigated in an FON-infested soilless growth substrate. Treatments included: a soilless growth substrate inoculated with FON as a control (CK), and soilless growth substrate inoculated with FON and the strain L3 (L3). Each treatment included 30 watermelon seedlings. The spore suspension of FON (10 6 CFU/ml) was first drenched into the growth substrate, followed by a suspension of strain L3 (10 9 CFU/ml).
The final concentration of FON (10 5 CFU/g) and strain L3 (10 8 CFU/g) in the growth substrates. Watermelon seedlings were then transplanted in the substrate trays (450 × 20 cm), and then grown at 16 hr light/8 hr dark at 28°C. Plants were watered as required for plant growth and disease development. Disease incidence was recorded after 28 days when symptoms of Fusarium wilt was starting to appear on the watermelon plants, and continued for another 8 days.
Incidence of Fusarium wilt was calculated by using the formula: The growth substrate samples near to watermelon root rhizosphere were collected at 0, 20, and 36 days after transplanting and stored at −20°C. The population of FON in the growth substrate samples were determined by real-time PCR following the method described by Zhao, Liu, and Ling (2014).
The plant-growth promotion ability of the strain L3, was assessed in pot experiment with a non-infested growth substrate as Disease incidence (%) = Number of diseased plants Total number of plants × 100 described above. Briefly, strain L3 was used as a drench treatment, and control pots received an equal volume of sterile distilled water.
Each treatment consisted of three replicate pots (one plant per pot).
After 20 days of transplanting, plant growth was determined by weighing individual plants.

| Identification of VOCs by SPME-GC-MS
Solid-phase micro-extraction (SPME) technique was used to collect VOCs produced by the strain L3 (Wu et al., 2014

| Verification of synthetic compounds against FON
Among the identified VOCs, nine standard compounds were purchased from the reagent company (Appendix Table A1). The antifungal activity of the standard compounds was assessed using the I-plate system described by Yuan, Raza, Shen, and Huang (2012).
The I-plate was prepared with PDA on one side and inoculated with a FON plug (4 mm). Then 50 μl of each standard compound was added to the other side. The I-plate was sealed with Parafilm and incubated at 28°C. The I-plates added with methanol or distilled water were used as control. The colony diameter of FON was recorded after 4 days incubating. The experiment was repeated three times.

| Plant growth promotion activities of synthetic compounds
The plant growth promotion activities of the nine compounds were measured by the modified method described by Ryu et al. (2003) and Zou et al. (2010). Briefly, three 2-day-old germinated Arabidopsis thaliana Col-0 seedlings were transferred to the one side of the Iplate containing 0.5 × MS, 0.8% sucrose and 1% bacto agar. Then, the nine synthetic compounds were diluted separately in alcohol, and 20 μl of the resulting suspension was applied to a sterile filter paper disk on the other side of the I-plate. A total of 10 μg, 100 μg, 500 μg and 1,000 μg doses of each synthetic compounds were tested. Each treatment was repeated for three times. The fresh weight of the Arabidopsis thaliana Col-0 seedlings was measured after 10 days.

| Characterization of strain L3
The biochemical and physiological characteristics of strain L3 were tested according to the method described by Wu et al. (2014).
The genomic DNA of the strain L3 was extracted by using the E.Z.N.A. Bacterial DNA kit (OMEGA Ltd.). The primer pair (F：5′-AGAGTTTGATCCTGGCTCAG-3′; R：5′-AAGTCGTAACAAGGTA-3′) was used to amplify the 16S rRNA gene of the strain L3. Then the 16S rRNA gene was sequenced and Blast searched against the NCBI database. The sequences of its close relatives were used to construct a neighbor-joining phylogenetic tree using MEGA 4.0.

| Statistical analysis
The data were assessed with one-way ANOVA. Duncan's multiplerange 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).

| Isolation of bacteria for biocontrol activity
Among the bacteria isolated from the soil samples, five strains that showed antagonistic activity against FON were named as L3, h, β, L, and b (Figure 1), respectively. After 5 days of incubation, the antifungal activities of these five strains were similar. The inhibition rates were 48.71%, 49.35%, 43.37%, 43.80%, and 47.86%, respectively.
After 10 days of incubation, the inhibition rates of these strains were 58.12%, 51.21%, 44.67%, 45.08%, and 44.91%, respectively. Strain L3 showed the strongest antifungal ability out of the five strains.

| Antagonistic effects of VOCs on growth of FON
All five strains showed significant inhibitory effects against FON mycelial growth during the bioassay without direct contact, although the VOCs produced by these strains could not completely inhibit the FON mycelia growth (Figure 2a). The morphology of the mycelial growth of FON was found abnormal in the VOCs treatment. Overall, mycelial growth inhibition was higher in the strain L3 treatment, the mycelial growth inhibition rate was 19.1% after 5 days incubation (Figure 2b).

| Growth promotion of A. thaliana with VOCs
VOCs produced by five antagonistic strains (L3, h, β, L, and b) promoted growth of A. thaliana (Figure 3). VOCs emitted by the strain L3 enhanced fresh weights by 2.39-fold and the number of lateral roots of A. thaliana plants by 5.05-fold, compared to the control. However, no statistically significant effects were found on the primary root length.

| Root colonization ability of the strain L3
The strain L3 was found to form thick biofilm in a static culture medium (Appendix Figure A2). In a hydroponic system, the GFPtagged cells could easily be distinguished from the background fluorescence in different parts of watermelon roots (Figure 4). The population of the strain L3 colonized on the watermelon's root was approximately 4.65 × 10 6 CFU per gram of root after two days of incubation. However, the population of the strain L3 decreased to 1.47 × 10 6 CFU per gram of root after 4 days incubation.

| Biological control and growth promotion activity of the strain L3
In the greenhouse experiment, the symptoms of Fusarium wilt of watermelon appeared 28 days after transplanting watermelon seedlings in the infested substrate and disease incidence increased rapidly during the next 8 days ( Figure 5). Disease incidence of the L3 treatment (22.5%) was far lower than the control treatment (71.2%).
The relative biocontrol efficiency of the strain L3 was up to 68.4%.
Based on real-time PCR, the population of FON were similar in both treatments during the first 20 days ( Figure 6). The populations of FON significantly increased in both two treatments after 36 days of planting in the infested substrate. The population of FON was significantly lower in the L3 treatment (2.90 × 10 5 copies/g) compared with the CK treatment (5.84 × 10 5 copies/g). These results corresponded to the disease incidence described above.
In the pot experiment, the fresh weights of watermelon plants treated with strain L3 were significantly higher than the control treatment 20 days after transplanting in the non-infested substrate. The strain L3 treatment showed an increase in 23.4% relative to the control treatment.

| Identification of VOCs by SPME-GC-MS
Fourteen specific VOCs peaks were identified from the strain L3 (Appendix Figure A3). The identified VOCs included eight ketones

| Plant growth promotion activities of synthetic compounds
Of the nine pure compounds, visual inspection revealed that plant shoot and root growth, when compared with the control treatment 1.88-fold by 2,3-butanediol (500 μg), compared to ethanol treatment respectively (Figure 10).

| Identification of strain L3
Bio-chemical tests showed that strain L3 was Gram-positive.
Microscopic examination revealed that the strain L3 cells are motile, rod-shaped, and can form spores when grown on NB culture medium. The phylogenetic tree of the 16S rRNA gene sequencing of strain L3 revealed that it was clustered closely to B. amyloliquefaciens, with a sequence similarity score of 99% in NCBI sequence alignment (Appendix Figure A4). Thus, the strain L3 was identified as B. amyloliquefaciens after considering all taxonomic characteristics.

| D ISCUSS I ON
The application of biological control agents (BCA) and grafted plants are considered sustainable control approaches to manage Fusarium wilt of watermelon (Ge, Liu, Nwet, & Zhao, 2016;Keinath & Hassell, 2014). Isolation of BCA against phytopathogens, is usually based on initial screening using the dual-culture plate assay (Hermosa, Grondona, & Iturriaga, 2000). However, the isolation method was mainly considered the ability of the isolated strains In addition to the antifungal ability of VOCs, our results clearly showed the plant growth promotion activity of VOCs produced by the isolated strains. Among the five strains, the strain L3 also showed the best plant growth promotion activity. A 2.39-fold increase in biomass, 1.40-fold increase in primary root length, and 5.05-fold increase in number of lateral roots of A. thaliana were detected after exposure to the VOCs produced by strain L3. Our results were in accordance with previous reports (Ryu et al., 2003;Vacheron, Desbrosses, & Bouffaud, 2013). However, Park, Dutta, F I G U R E 6 Effect of Bacillus amyloliquefaciens strain L3 on the number of FON in the rhizosphere soil of watermelon at 0, 20, and 36 days after transplanting. CK: control, inoculated with FON, L3: inoculated with FON and the strain L3. "*" above the columns indicate a significant difference at p < 0.05 F I G U R E 7 Effect of Bacillus amyloliquefaciens strain L3 on the fresh weight of watermelon seedling grown for 20 days in a greenhouse. CK: control treatment, L3: strain L3 was inoculated in the soilless substrate (10 8 cfu/g). "*" above the columns indicate a significant difference at p < 0.05 Ann, and Raaijmakers (2015) reported the VOCs released by P. fluorescens SS101 increased the biomass of A. thaliana up to 8.8-fold.
The proportionately smaller increase in biomass observed in this study could be due to the larger volume of the culture plate and lower concentration of available VOCs to the seedlings in the test chamber. Overall, after considering the antifungal and plant growth promotion activity of the five isolated strains, the strain L3 was chosen for the further analysis.
Biofilm formation and root colonization ability have been recognized as essential factors for BCA to survive in the rhizosphere battlefield against phytopathogens (Bhattacharyya & Jha, 2012;Raaijmakers, Paulitz, & Steinberg, 2009). Based on the fact that the strain L3 could form a thick biofilm in a static culture medium, we expected the strain L3 could ensure good colonization of watermelon roots. Previous studies have shown the elongation and differentiation zones of the plant root, as well as in the lateral roots and the junctions between the roots are preferred by PGPR strains . In our study, the GFP signals of GFP-tagged B. amyloliquefaciens L3 were found on the root tips, primary roots, and lateral root junctions (Figure 4). These results are consistent with previous studies Ji, Lu, Gai, & Zheng, 2008 (Minuto, Spadaro, & Garibaldi, 2006;Wang, Yuan, & Zhang, 2013 (Yuan et al., 2012). In this study, a total of 14 volatile substances produced by strain L3 were identified. The components of VOCs produced by strain L3 partly overlapped with other Bacillus spp., which mainly included alcohol, ketones, and alkanes derivatives (Raza, Wang et al., 2016;Raza, Wei et al., 2016). Among the identified VOCs, 2-heptanone, 1-hexano,2-ethyl-, and 2-nonanone exhibited 100% inhibition of mycelial growth of FON. However, 1-hexano,2-ethyl-was produced in low quantity. Acetoin was the main VOC released by strain L3, but demonstrated low antifungal activity. Considering their antifungal activity and relative peak area, 2-nonanone and 2-heptanone were potential candidates as effective antifungal VOCs produced by strain L3.
In previous research, acetoin and 2,3-butanediol has been used in foods, cigarettes, cosmetics, detergents, and chemical synthesis (Xiao & Lu, 2014). Acetoin is a precursor of 2,3-butanediol and can be bio-transformed by plants and microorganisms to 2,3-butanediol stereoisomers (Javidnia, Faghih-Mirzaei, & Miri, 2016). In this study, we found that ace- In conclusion, our results demonstrated that B. amyloliquefaciens L3 isolated from watermelon's rhizosphere soil could be an excellent candidate for the development of biocontrol agents. We also reported here that VOCs produced by B. amyloliquefaciens L3 play important roles in disease control and in growth promoting processes.
However, further studies are still needed to understand the role of VOCs produced by B. amyloliquefaciens L3 under field conditions.

CO N FLI C T O F I NTE R E S T S
The authors declare that there is no conflict of interests.

AUTH O R S CO NTR I B UTI O N
YC Wu and Y Ma designed and supervised the study. CG Li was involved in the greenhouse pot experiment. JY Zhou was involved in verifying plant growth promotion activities of VOCs. All of the other experiments were performed by YC Wu.

E TH I C S S TATEM ENT
None required.

DATA ACCE SS I B I LIT Y
All data are provided in full in the results section of this article. F I G U R E A 1 Schematic representation of the experimental setup for Arabidopsis thaliana growth promotion by VOCs F I G U R E A 2 Biofilm formation of the strain L3 in static culture medium (LB) F I G U R E A 3 SPME chromatography-mass spectrometry (GC-MS) profile of volatile organic compounds (VOCs) produced by the strain L3. Red arrow means the specific VOCs produced by strain L3 F I G U R E A 4 Phylogenetic analysis of strain L3 and related species using the neighbor-joining approach. Bootstrap values obtained with 1,000 resamplings are indicated as percentages at all branches. The GenBank accession number for each microorganism is shown in parentheses after the species name