Inhibitory effect of Xenorhabdus nematophila TB on plant pathogens Phytophthora capsici and Botrytis cinerea in vitro and in planta

Entomopathogenic bacteria Xenorhabdus spp. produce secondary metabolites with potential antimicrobial activity for use in agricultural productions. This study evaluated the inhibitory effect of X. nematophila TB culture on plant pathogens Botrytis cinerea and Phytophthora capsici. The cell-free filtrate of TB culture showed strong inhibitory effects (>90%) on mycelial growth of both pathogens. The methanol-extracted bioactive compounds (methanol extract) of TB culture also had strong inhibitory effects on mycelial growth and spore germinations of both pathogens. The methanol extract (1000 μg/mL) and cell-free filtrate both showed strong therapeutic and protective effects (>70%) on grey mold both in detached tomato fruits and plants, and leaf scorch in pepper plants. This study demonstrates X. nematophila TB produces antimicrobial metabolites of strong activity on plant pathogens, with great potential for controlling tomato grey mold and pepper leaf scorch and being used in integrated disease control to reduce chemical application.

However, the action modes of these antimicrobial compounds on fungal and oomycete pathogens are still unknown.
The inhibitory effect of Xenorhabdus spp. culture on some fungal and oomycete pathogens has been determined 9,22,[24][25][26][27][28][29] . However, there are no studies on the potential inhibitory effect of X. nematophila culture systematically from in vitro to in planta. This study determined the inhibitory effect of X. nematophila TB culture toward a wide range of plant pathogenic fungal and oomycete pathogens in vitro, with special focus on the inhibitory effect of X. nematophila TB culture on B. cinerea causing grey mold in tomato plants and P. capsici causing leaf scorch in pepper plants.

Results
In vitro effect of cell-free filtrate on mycelial growth of different pathogens. The cell-free filtrate of X. nematophila TB culture showed a wide range of inhibitory effect from 15% to 100% on the mycelial growth of all the fungal and oomycete pathogens tested ( Table 1). In particular, the cell-free filtrate of X. nematophila TB culture exhibited high inhibitory effect (.90%) on Botrytis cinerea, Phytophthora capsici, Alternaria solani, Bipolaria maydis, Bipolaris sorokinian, Dothiorella gregaria, Exserohilum turcicum, Physalospora piricola, Rhizoctonia cerealis and Sclerotinia sclerotiorum. Among these, B. cinerea and P. capsici were selected as the pathogens for subsequent studies (Fig. 1).
In vitro effect of methanol extract on mycelial growth of B. cinerea and P. capsici. The concentrations of methanol extract from 6.25 to 400 mg/mL all showed inhibitory effect on the mycelial growth of B. cinerea and P. capsici, and there was a liner relationship between log10-transformed concentration of methanol extract and probittransformed inhibitory rate on the mycelial growth of B. cinerea (y 5 3.0229 1 0.9086x, R 2 5 0.9908) and P. capsici (y 5 2.9523 1 1.2056x, R 2 5 0.9428) (Fig. 2). The methanol extract showed stronger inhibitory effect on P. capsici than B. cinerea when its concentration was higher than 12.5 mg/mL where methanol extract at 1000 mg/mL (relative to 78 mL cell-free filtrate) showed an inhibitory rate of 87.5% (probit 5 6.15) and 66.6% (probit 5 5.43) on the mycelial growth of P. capsici and B. cinerea, respectively. The EC 50 of methanol extract on P. capsici and B. cinerea was 49.94 and 149.92 mg/mL, respectively.
In vitro effect of methanol extract on spore germination of B. cinerea and P. capsici. The concentrations of methanol extract from 6.25 to 400 mg/mL all showed inhibitory effect on the spore germination of B. cinerea and P. capsici, and there was a liner relationship between log10-transformed concentration of methanol extract and probit-transformed inhibitory rate on the spore germination of B. cinerea (y 5 2.3760 1 1.3221x, R 2 5 0.9976) and P. capsici (y 5 1.2270 1 2.0763x, R 2 5 0.9950) (Fig. 3). The methanol extract showed stronger inhibitory effect on P. capsici than B. cinerea when its concentration was higher than 25 mg/mL where methanol extract at 1000 mg/mL showed an inhibitory rate of 95.5% (probit 5 6.70) and 82.3% (probit 5 5.93) on the spore germination of P. capsici and B. cinerea, respectively. The EC 50 of methanol extract on P. capsici and B. cinerea was 65.63 and 96.54 mg/mL, respectively.
Effect of methanol extract and cell-free filtrate on detached tomato fruits infected with B. cinerea. There was a significant effect (P , 0.001) of treatments (viz. methanol extract at 250, 500 and 1000 mg/mL, cell-free filtrate and chemical control) on detached tomato fruits infected with B. cinerea, and at each treatment, there was no significant difference (P . 0.05) between therapeutic effect   and protective effect (Fig. 4). The methanol extract at 1000 mg/mL exhibited 77.35% therapeutic effect and 71.57% protective effect on detached tomato fruits infected with B. cinerea, which were similar to the therapeutic effect and protective effect of the cell-free filtrate and the chemical control (50% Sumilex, 10003). For both the therapeutic effect and protective effect treatment, tomato fruits sprayed with the methanol extract (1000 mg/mL), the cell-free filtrate and 50% Sumilex (10003) showed only small legions while fruits sprayed with water showed large lesions (Fig. 5).
Effect of methanol extract and cell-free filtrate on tomato plants infected with B. cinerea and pepper plants infected with P. capsici.
There was a significant effect (P , 0.001) of treatments (viz. methanol extract at 250, 500 and 1000 mg/mL, cell-free filtrate and chemical control) on tomato plants infected with B. cinerea, and there was no significant difference (P . 0.05) between therapeutic effect and protective effect at each concentration of the methanol extract and the cell-free filtrate (Fig. 6). Both the therapeutic effect and protective effect of the methanol extract at 1000 mg/mL and the cell-free filtrate were higher than 70%, which were similar to the therapeutic effect and protective effect of the chemical control (50% Sumilex, 10003). There was also a significant effect (P , 0.001) of treatments (viz. methanol extract at 250, 500 and 1000 mg/mL, cell-free filtrate and chemical control) on pepper plants infected with P. capsici, and there was no significant difference (P . 0.05) between the therapeutic effect and protective effect at each concentration of the methanol extract and the cell-free filtrate (Fig. 7). At 1000 mg/mL, both the therapeutic effect and protective effect of the methanol extract and the cell-free filtrate were higher than 70%, which were stronger than the chemical control (25% Metalaxyl, 5003) in both therapeutic effect and protective effect that were lower than 61%.

Discussion
The cell-free filtrate of X. nematophila TB culture exhibited a wide range of inhibitory effect on plant pathogenic fungi and oomycetes in vitro. Previous studies have reported the variation in antimicrobial activities of different Xenorhabdus spp. strains to fungal pathogens 9,18,24,27,[35][36][37][38][39] . B. cinerea and P. capsici are the frequently reported pathogens associated with tomato grey mold and pepper leaf scorch, respectively, which result in marked agricultural economic losses annually in China 40,41 . This study found that the cell-free filtrate of X. nematophila TB culture showed strong inhibitory effect on the growth of B. cinerea and P. capsici. And thus, we focused on these two pathogens for both in vitro and in planta studies.
The methanol extract of X. nematophila TB culture not only inhibited the mycelial growth but also inhibited the spore germination of B. cinerea and P. capsici. The inhibitory effect of the methanol extract on the mycelial growth of P. capsici is stronger than B. cinerea, with an EC 50 of 49.94 and 149.92 mg/mL, respectively. Similarly, the inhibitory effect of the methanol extract on the spore germination of P. capsici is stronger than B. cinerea, with an EC 50 of 65.63 and 96.54 mg/mL, respectively. A previous study found that the inhibitory effect of the methanol extract of X. bovienii YL002 culture on both the mycelial growth and spore germination of P. capsici was weaker than on B. cinerea 33 . This difference may be due to different antimicrobial compounds produced by the two species of Xenorhabdus (X. nematophila and X. bovienii). X. bovienii mainly produces indoles and dithiolopyrrolones that exhibited both antifungal and antibacterial activity, but the antifungal activity of indoles is stronger than the antibacterial activity 16,22 . It has been reported that Phytophthora is resistant to most common antifungal compounds, but are sensitive to antibacterial compounds 42 . X. nematophila produces antimicrobial peptides (xenocoumacins, xenortides, xenematide and cyclolipopeptide), nematophin and benzylineacetone. Xenocoumacins, the major antibiotics produced by X. nematophila, can inhibit the growth of P. capsici and B. cinerea and the sporangia production of P. infestans 16,29 . Xenematide only exhibited a noticeable antibacterial effect 14 . Cyclolipopeptide exhibited antifungal activity on B. cinerea, Phytophthora myc and P. oryzae 12 . Nematophin showed significant inhibitory effect on B. cinerea and P. infestans 15 . Benzylineacetone is active against Gram-negative bacteria 13 . Thus, the methanol extract of X. nematophila TB culture showed stronger inhibitory effect on P. capsici than on B. cinerea while the methanol extract of X. bovienii YL002 culture showed stronger inhibitory effect on B. cinerea than on P. capsici.
The active compounds produced by Xenorhabdus spp. may act on P. capsici and B. cinerea by different cellular action mechanisms. P. capsici is an oomycete and differs from the fungi in cell wall composition that contains or produce sterols 43 . The active compounds xenocoumacins can act on P. capsici by reducing the activity of transporters involved in transport systems of P. capsici cell membranes 44 . B. cinerea is an ascomycete pathogen and spores of B. cinerea are considered to be the main source of dispersal of inoculum, and their germination and adhesion on plant surfaces represent crucial steps preceding host penetration and colonization 45,46 . The active  compounds produced by X. nematophlia TB may affect the cellular process required for the germination of B. cinerea spores by multisite inhibitors or interfere with respiration, and thus B. cinerea cannot grow and reproduce normally to penetrate plant host. It has been found that B. cinerea mycelia treated with the active compounds from X. nematophlia YL001 culture showed morphological and structural alterations such as swollen beads, septa malformation, cell wall dissolution and cell liquid seepage 47 .
This study found that the in vitro and in planta effect of the methanol extract of X. nematophila TB culture on B. cinerea and P. capsici showed difference. The inhibitory effect of the methanol extract on both the mycelial growth and spore germination of P. capsici was stronger than on B. cinerea in vitro, but there was no obvious difference between its effect on tomato plants infected with B. cinerea and pepper plants infected with P. capsici in planta,. Traditionally, many studies on the antimicrobial activities of Xenorhabdus spp. are determined by in vitro assays that can give an indication of the inhibitory potency quickly, while there are few studies on plant pathogens in planta 48 . Since in vitro results do not always represent the antimicrobial activities of Xenorhabdus in planta situation, the potency determined in in vitro assays may be overestimated or underestimated compared with in planta situation, where pathgen infection is a well-regulated phenomenon that requires cross talk between host and pathogen through signals probably located on the external surfaces of cells 49 . The disturbance of cell membranes of pathogens by antimicrobial compounds apparently could lead to interference with such signals, which could eventually result in the failure of infection and development of symptoms on plants.
This study demonstrated that the application of methanol extract of the cell-free filtrate of X. nematophila TB culture could control tomato grey mold and pepper leaf scorch. The inhibitory effect of the methanol extract on the severity of grey mold and leaf scorch increased with the increasing concentration of methanol extract. Detached tomato fruits, tomato plants and pepper plants treated with the methanol extract 24 h prior to inoculation prevented the further  www.nature.com/scientificreports SCIENTIFIC REPORTS | 4 : 4300 | DOI: 10.1038/srep04300 establishment and expansion of the pathogen on the fruit or leaf surface. This suggests that the methanol extract can retain its antimicrobial activity for at least 24 h after being applied to fruits or leaves. Moreover, the methanol extract exhibited similar efficiency on both plant diseases tested in this study. Thus, the methanol extract of the cell-free filtrate of X. nematophila TB culture may be used to control plant diseases in agricultural production system. The methanol extract not only has strong activity on plant pathogens in vitro but also has strong control effect on plant diseases in planta. These justify the necessity of additional work to determine the effect of the methanol extract on tomato grey mold caused by B. cinerea and pepper leaf scorch caused by P. capsici in the field. Further work will also be conducted on using nonpolar solvents to extract active compounds and on identifying active compounds from X. nematophila TB culture to find alternate ways to reduce the application of chemical fungicides for sustainable agriculture production.
In summary, the cell-free filtrate and methanol extract of the cellfree filtrate of X. nematophila TB culture are of strong inhibitory effect on the plant pathogens B. cinerea and P. capsici both in vitro and in planta. X. nematophila TB culture has the great potential for controlling grey mold in tomato and leaf scorch in pepper and can be used in the integrated control of these pathogens with the objective of reducing the amount and number of chemical fungicide applications.

Methods
Bacterium strain and culture conditions. X. nematophila TB was isolated from its nematode symbiont S. carpocapsae TB obtained from the soil collected from Taibai Mountain, Qinling, China 28 . This strain was firstly identified to be X. nematophila according to its morphological characteristics 28 , and confirmed to be X. nematophila by molecular identification conducted by PCR amplification and cloning of 16S rRNA gene, and the 16S rRNA gene sequence was then compared with other sequences in GenBank using the BLAST algorithm 30 . The 16S rRNA gene sequence of X. nematophila TB was deposited in GenBank (Accession number: EU124383).
X. nematophila occurs in two phases where phase I exhibits stronger antibiotic activity than Phase II 31 , and thus was used in this study. This strain is deposited in the Agricultural Culture Collection Institute, Northwest A&F University, China. Glycerinated stocks of this strain store at 270uC were used as starting material for subculture. To ensure the presence of phase I, this strain was subcultured onto plates with NBTA media [nutrient agar (NA, g/L) consisting of peptone 10, beef extract 3, NaCl 5 and agar 15 supplemented with 0.04% triphenyltetrazolium chloride (w/v) and 0.025% bromothymol blue (w/v)] and incubated at 28uC in darkness. The NBTA media can differentiate the phase I and phase II of this bacterium where phase I is of green colony while phase II is of red colony after 3 to 5 days' culture. Seed culture of X. nematophila TB was prepared by inoculating a loopful of phase I colony growing on NBTA plate into a 250 mL flask containing 50 mL fresh NB (NA without agar), which was adjusted to a final pH of 7.2, and then cultivated at 28uC in darkness on an Eberbach rotary shaker at 150 rpm for 18 to 24 h, during which time the optical density (600 nm) and pH readings were approximately 2.0 and 7.0, respectively.
Cell-free filtrate and methanol extract of cell-free filtrate of X. nematophila TB culture. Batch cultures of X. nematophila TB were carried out in four 5 L fermenters (Eastbio, China). Each fermenter was equipped with one six-blade disk turbine impeller, pH probes (Mettler-Toledo GmbH, Switzerland), dissolved oxygen (DO) probes (Mettler-Toledo GmbH, Switzerland), a thermometer and foam. Seed culture (300 mL) was transferred into each fermenter containing 3.5 L autoclaved fermentation medium (g/L: glucose 6.13, peptone 21.29, MgSO 4 ?7H 2 O 1.50, (NH 4 ) 2 SO 4 2.46, KH 2 PO 4 0.86, K 2 HPO 4 1.11 and Na 2 SO 4 1.72). The pH of the medium was adjusted to 7.0 using 2.0 mol/L NaOH and 2.0 mol/L HCl. The fermenters were incubated at 28uC with the aeration rate of 2.5 L/min and the agitation speed of 300 rpm. After 72 h, the culture was transferred to centrifuge  bottles, and centrifuged for 20 min (13200 g, 4uC) to get cell-free filtrate and stored at 4uC until required.
The methanol extract of the cell-free filtrate of X. nematophila TB culture was prepared based on methods described 24,32,33 . Briefly, D101 polymeric adsorbent resin (10 g, Bengbu Tianxing Ion-Resin Co. Ltd., China) were suspended in 100 mL sterile distilled water, then treated with 1% sterile HCl and 1% sterile NaOH with one hour each. After three washes with sterile distilled water, the pH was adjusted to 7.5 and kept at 4uC for 24 h. The cell-free filtrate was mixed with activated D101 polymeric adsorbent resin at 1520 and incubated for 24 h. The resin slur was separated by a G3 glass filter, washed with distilled water and 25% methanol, and then placed on the top of the column filled with activated D101. After washing the column with distilled water, methanol was pumped onto the column at 500 mL/min and the eluate was collected in 200 mL aliquots. The methanol extract was dried at 40uC and stored at 4uC until required, and the average yield of the methanol extract of the cell-free filtrate was 12.8 g/L.
Effect of cell-free filtrate on the mycelial growth of different pathogens. Twentysix fungal and oomycete pathogens as listed in Table 1 were used in this test. These pathogens were obtained from the Agricultural Culture Collection Institute, Northwest A&F University, China, and were selected based on their importance associated with plant diseases in Shaanxi Province, China. These pathogens were subcultured onto fresh autoclaved potato dextrose agar (PDA) plates and maintained at 25uC in darkness. To determine the effect of cell-free filtrate of X. nematophila TB culture on these pathogens, cell-free filtrate was mixed with autoclaved PDA that had been cold down to about 70uC at 159, and then poured onto 9-cm Petri dish plates (10 mL mixture per plate). One mycelial disk (0.6 3 0.6 cm) from the edge of 3 to 5day-old-colony of each pathogen growing on PDA was put onto the center of each plate. For each pathogen, there were three replicates (one plate per replicate), and the control plates for comparison were PDA only. The plates were maintained at 25uC in darkness. After 7 days, the colony diameter of each plate was measured, and the inhibitory rate was determined as described 33 . This experiment was repeated once under the same conditions.
Effect of methanol extract on the mycelial growth of B. cinerea and P. capsici. A stock solution of the methanol extract (4000 mg/mL) of the cell-free filtrate of X. nematophila TB culture was prepared by dissolving the dried methanol extract in distilled water. Two-fold dilutions were made from the filter-sterilized stock solution. For each dilution, 1 mL was thoroughly mixed with 9 mL PDA and poured into a Petri dish plate. The final concentrations of the methanol extract in PDA were 6.25, 12.5, 25, 50, 100, 200 and 400 mg/mL. One mycelial disk (6-mm-diameter) from the edge of 3 to 5 day-old-colony of each pathogen growing on PDA was put onto the center of each plate. For each concentration, there were three replicates (one plate per replicate), and the control plates for comparison were PDA only. The plates were maintained at 25uC in darkness. The colony diameter of each plate was measured and the inhibitory rate for each concentration on each pathogen was determined after 7 days. This experiment was repeated once under the same conditions.
Effect of methanol extract on the spore germination of B. cinerea and P. capsici. Spores of B. cinerea and P. capsici were collected from 5 to 7 day-old-colonies growing on PDA in darkness at 25uC by flooding each plate with 10 mL sterile distilled water containing 0.1% (v/v) tween 20 and rubbing the agar surface gently with a bent glass rod. The resulting spore suspension was filtered through four layers of Miracloth (Calbiochem, Merck Pty. Ltd., Australia). The spore concentrations were determined using a hematocytometer and adjusted to 1 3 10 8 spores/mL in sterile distilled water for each pathogen. The methanol extract was dissolved in sterile distilled water, and diluted to obtain the concentrations of 12.5, 25, 50, 100, 200, 400 and 800 mg/mL. Methanol extract solution (200 mL) from each concentration were inoculated with spore suspension (200 mL) and then mixed thoroughly. The final concentrations of the methanol extract in the mixtures were 6.25, 12.5, 25, 50, 100, 200 and 400 mg/mL. The mixture (10 mL) from each concentration was placed on a glass slide. The control for comparison was spore suspension without methanol extract. There were three replicates (one slide per replicate) for each treatment. The slides were incubated in a moisture chamber at 25uC in darkness for 24 h. The number of spores germinated at each concentration was counted under the microscope. The percentage of spore germination was calculated, and the inhibitory rate was determined as described 33 . This experiment was repeated once under the same conditions. Effect of methanol extract and cell-free filtrate on detached tomato fruits infected with B. cinerea. Fresh green tomato fruits (cv. L402) with similar size (about 6-cmdiameter), picked from a glasshouse in Yangling (Shannxi Province, China), were used. Fruits were wounded on the middle sides using a sterilised needle (2-mmdiameter) before inoculation. To evaluate the therapeutic effect of methanol extract and cell-free filtrate of TB culture, mycelial agar discs (5-mm-diameter) from the edges of 3 to 5-day-old B. cinerea colony growing on PDA in darkness at 25uC were placed onto the pre-wounded sites of tomato fruits with the mycelial side facing the fruit (one mycelial disc per fruit). There were three replicates (eight fruits per replication) for each treatment. After inoculation, fruits were placed in plastic containers with moistened filter papers at the bottom to maintain high humidity. The containers were kept in a climate chamber at 25uC in darkness. After 24 h, the fruits were sprayed with 50 mL methanol extract at three concentrations (250, 500 and 1000 mg/mL) in water and the original cell-free filtrate. The controls were sprayed with water as negative control and with 50% Sumilex (10003, Sumitomo Chemical Co. Ltd.) as positive control. To determine the protective effect, tomato fruits were sprayed with each solution prior to inoculation and kept under the same conditions as above. After 24 h, the fruits were inoculated with B. cinerea as above. The lesion diameter on each fruit was measured seven days post inoculation, and the inhibitory rate was determined as described 33 . This experiment was repeated once under the same conditions. Effect of methanol extract and cell-free filtrate on tomato plants infected with B. cinerea and pepper plants infected with P. capsici. Tomato seeds (cv. L402) were sown in plastic pots and kept in the controlled environmental room at 28/20uC (day/ night) with about 70%-humidity and a 12 h-photoperiod (light intensity 350 mE m 22 s 21 ). Pepper Seeds (cv. Shijihong) were sown and kept in the controlled environmental room at 18/15uC (day/night) with about 70%-humidity and a 12 hphotoperiod (light intensity 250 mE m 22 s 21 ). Tomato plants with three true leaves and pepper plants with six true leaves were used. To determine the therapeutic effect, leaves of tomato plants and pepper plants were inoculated with spore suspensions (1 3 10 8 spores/mL) of B. cinerea and P. capsici, respectively. Plants were covered with transparent polyethylene bags for 24 h to maintain high humidity, and plants were then sprayed with methanol extract at three concentrations (250, 500 and 1000 mg/mL) and the cell-free filtrate. For controls, plants were sprayed with water and commercial chemicals, respectively. Fifty percent Sumilex (10003, Sumitomo Chemical Co. Ltd.) is commonly used for effective control of grey mold disease, while 25% Metalaxyl (5003, Sumitomo Chemical Co. Ltd.) is commonly used for effective control of leaf scorch disease on plants in China. There were three replicates (eight plants per replicate) for each treatment. To determine the protective effect, plants were sprayed with each solution as above prior to inoculation. After 24 h, plants were inoculated as above. After 15 days, the disease severity of plants were evaluated as described 33,34 . This experiment was repeated once under the same conditions. Data analyses. Data analyses were conducted using the SPSS statistical package (Version 11 for windows). Regression analyses were conducted to determine the relationship between the concentrations of methanol extract of the cell-free filtrate of X. nematophila TB culture and their inhibitory rate on the mycelial growth of B. cinerea and P. capsici, and also, the relationship between the concentrations of methanol extract and their inhibitory rate on spore germination of B. cinerea and P. capsici, and the associated regression equation, R 2 and 50% effective concentration (EC 50 ) of methanol extract were calculated. The concentrations of methanol extract were log10-tranformed and the inhibitory rates were probit-tranformed. Analyses of variance were conducted to determine the effects of treatments (viz. methanol extract at different concentrations, cell-free filtrate and chemical control) on the disease caused by B. cinerea in detached tomato fruits, the disease caused by B. cinerea in tomato plants and the disease caused by P. capsici in pepper plants. Subsequent multiple comparisons among treatments of each experiment were conducted based on Fisher's protected least significant differences at P 5 0.05 and P 5 0.01. Standard errors (SE) of means were also computed. For all analyses, data from the two repeat experiments (i.e., the original and one repeat experiment) of each study were not significantly different (P . 0.05); therefore, data from the two repeat experiments of each study were combined and analysed together.