Microbial Biopesticides against Bacterial, Fungal and Oomycete Pathogens of Tomato, Cabbage and Chickpea

: Biological control is an environmentally friendly approach that holds promise to complement or replace chemicals to effectively protect crop plants against pests and pathogens. Environmental samples with highly diverse and competitive microbiomes that harbor antagonistic microbes with diverse modes-of-action can provide a rich source of microbial biopesticides. In the current study, bacteria isolated from rhizosphere soil and food spoilage samples were subsequently screened against various plant fungal and oomycete pathogens in growth inhibition assays. These included the new potential biocontrol bacteria Corynebacterium ﬂavescens , Sporosarcina aquimarina and Sporosarcina saromensis with anti-fungal and antioomycete activities. Potential candidates selected by preliminary screening in plant assays were then applied to tomato, cabbage and chickpea plants to control bacterial ( Pseudomonas syringae pv. tomato ), fungal ( Alternaria brassicicola ) and oomycete ( Phytophtora medicaginis ) phytopathogens. Ten potential microbial biopesticides were demonstrated to be effective against these diseases, and led to signiﬁcant ( p < 0.05) reductions in symptoms and/or pathogen DNA compared to mock-treated diseased plants. We conclude that new and effective microbial biopesticides to control crop pathogens can be rapidly isolated from biodiverse microbiomes, where bacteria may employ these features to effectively compete against each other. in Student’s t -test) to the S. saprophyticus control. This bacterium was chosen for comparisons, as it showed low in vitro bioactivity against plant pathogens.


Introduction
Pests and pathogens use plants as sources of energy and pose significant threats to agricultural production and food security [1]. Food production must increase by 60% by 2050 to support human demands; however, 10 to 16% of crop production is estimated to be destroyed by pests and pathogens [2]. These include the five most important cropswheat, rice, maize, potato and soybean-with estimated global crop losses caused by pests and pathogens of 21.5, 30.0, 22.6, 17.2 and 21.4%, respectively [3]. Agrochemicals are frequently applied to control pests and pathogens, but they often harm beneficial microbes and their residues remain in almost 40% of foods, according to FAO-WHO reports [4]. Furthermore, pesticides pose significant health risks to humans and their environment, and provoked or induced the development of resistance both in plant pathogens and pests [5,6]. Many beneficial microorganisms isolated from plants or other environments can promote plant growth and/or suppress plant pathogens [7]. These microbes can be applied as biocontrol agents called biopesticides [8]. These biopesticides are considered much more environmentally friendly alternatives to protect crops against pathogens in the future.
There are currently only about 25 microbial products that are in regular use [9,10]. This contrasts with the thousands of chemical products that are in the market, and there is thus a need to develop more biopesticides that have diverse modes-of-action [11]. The present study focused on bespoke treatments of three important vegetable crops (tomato, cabbage and chickpea) and three types of economically significant pathogens with different nutrient

Anti-Fungal and Antioomycete Assays
A total of 68 rhizosphere soil and food spoilage bacteria were previously found to be active against bacterial pathogens Pseudomonas syringae pv. tomato DC3000 (Pst) and Clavibacter michiganensis [23]. These came from clay-rich soil collected in Tennyson, Queensland, Australia (27 • 31 37.0 S 152 • 59 51.7 E) and food spoilage samples (mixed vegetable and fruit scraps from a compost bin), and were identified by 16S rRNA sequencing. In the present study, these bacteria were further evaluated for their potential as biocontrol agents against various fungal and oomycete plant pathogens. Bacterial isolates were grown on various liquid media including Yeast Extract Peptone (YEP), Liquid Medium (LM), Luria Bertani (LB) and De Man, Rogosa and Sharpe (MRS) at 28 • C in a shaker incubator at 120 rpm (Supplementary Table S1). All growth inhibition assays were then performed on Petri dishes using PDA (Supplementary Table S1) [24]. The potential antagonizing effects of bacterial isolates were examined against fungal and oomycete plant pathogens, including Fusarium oxysporum, Alternaria brassicicola, Alternaria solani, Phytophthora capsici, Phytophthora medicaginis and Phytophthora cinnamomi. Potential biocontrol bacteria were cultured overnight at 28 • C in a shaker incubator. The pure fungal inoculum was placed in the center of a new PDA plate with two lines of streaked biocontrol bacteria, including Staphylococcus saprophyticus for comparison in three replicate plates for each isolate (Supplementary Figure S1). Staphylococcus saprophyticus was chosen as a reference microbe because its antagonistic activity against the pathogens tested was low, it came from the same soil samples and can be incubated under identical conditions. Plates were incubated at 28 • C until the control plate for each pathogen (without biocontrol bacteria) reached~80% fungal coverage of the plate (1-2 weeks). Anti-fungal/oomycete effects were quantified by the following formula [24]: (Average control plate colony diameter − Average bacteria treated plate colony diameter)/Average control plate colony diameter × 100.
A minimum of 50% in vitro pathogen inhibition was applied as a threshold for further experimentation with these microbial isolates. These were then used in preliminary plant assays where they were sprayed onto tomato and cabbage or watering to chickpea seedlings (one plant per isolate) to select four isolates for subsequent in plant biocontrol assays.

Biopesticide and Plant Bacterial Disease Suppression Assays
Solanum lycopersicum (tomato) plants (cv Moneymaker) were grown in 55 mm pots in UQ23 potting mix for 4 weeks in growth cabinets with 16 h light at 26 • C and 8 h night at 21 • C, with a light intensity of 100 µmol photons m −2 s −1 (white fluorescent lamps). Eight-week-old tomato plants, in three biological replicates containing 10 pooled plants per replicate, were sprayed with potential beneficial biocontrol bacteria (33YE, 46YE, 14th, 28MC and a mixture of all isolates) that also previously showed antibacterial activity against Pst [23], 1 day before and 1 day after bacterial pathogen treatment. Briefly, the four bacterial isolates were grown overnight in 10 mL cultures that were upscaled to five 100 mL cultures each in 500 mL flasks. The cultures were centrifuged at 4170× g for 10 min, the supernatants were discarded and bacterial pellets were resuspended in 10 mM MgCl 2 containing 0.02% Silwet L-77. The OD 600 for all bacterial cultures was adjusted to 1.57. The same procedure was repeated for Pst, and the OD 600 was adjusted to 0.2 based on previous studies [25]. Bacterial suspensions were then sprayed onto plant leaves by applying both sides of each leaflet of all leaves for each plant, and mock-treatments comprised application with MgCl 2 solution only. Six days after pathogen inoculation, all plant leaves were collected. Disease symptom scores were based on the percentage of leaflet area displaying chlorosis, brown spots or necrosis (1 = 0-20%; 2 = 20-40%; 3 = 40-60%; 4 = 60-80%; 5 = 80-100%). Total DNA from the same leaflets was extracted with the CTAB method [26]. For pathogen quantification using real-time quantitative PCR (qPCR), 40 ng of genomic DNA was mixed with 6 µL of FastStart Essential DNA Green Master mix (Bioline, UK) containing 3 µM each of the corresponding primers (Supplementary Table S2). The copy number of a specific Pst gene (gyrA) was determined for each sample using qPCR (Roche LightCycler ® 96, Basel, Switzerland) and normalized to tomato ACTIN [27,28].

Biopesticide and Plant Fungal Disease Suppression Assays
Golden Acre cabbage plants (Brassica oleracea var. capitata) were grown in a growth cabinet under 16 h light at 26 • C and 8 h night at 21 • C with a light intensity of 100 µmol photons m −2 s −1 (white fluorescent lamps). 4-week-old plants in three biological replicates containing 10 plants per replicate were inoculated with potential biocontrol isolates 46YE, 28M, 4YE, 44LGS and a mixture of all isolates 1 day before and 1 day after fungal pathogen treatment. Cultivations of bacterial strains and preparations for mock treatments are explained in Section 2.2. Cabbage plants were inoculated with Alternaria brassicicola 24 h after initial treatment, with biocontrol bacteria based on the established protocol for A. solani [29]. Briefly, spores from 2-week-old A. brassicicola cultures on PDA plates were harvested using sterile water containing 0.005% (v/v) Tween 40. The spore concentration was then measured using a hemocytometer, and adjusted to make up 500 mL of a 1.8 × 105 spores/mL suspension, which was sprayed on cabbage plants (application on both sides of the leaves). Five days after the initial inoculation, plants were symptom-scored based on the percentage of leaf area displaying necrosis (1 = 0-20%; 2 = 20-40%; 3 = 40-60%; 4 = 60-80%; 5 = 80-100%). The same leaves were harvested for qPCR (Roche LightCycler ® 96, Basel, Switzerland) using primers listed in Table S2, Supplementary File. Pathogen amplicons were normalized to cabbage DLH [30,31].

Biopesticide and Plant Oomycete Disease Suppression Assays
Chickpea (Cicer arietinum var genesis 090) plants (14-17 plants per treatment) were grown in UQ23 potting mix in growth cabinets with 16 h light at 26 • C and 8 h night at 21 • C and a light intensity of 100 µmol photons m −2 s −1 (white fluorescent lamps) for 3 weeks. Cultivations of bacterial strains and preparations for mock treatments are explained in Section 2.2. Each pot with a 3-week-old plant in the treatment was watered with 10 mL biocontrol bacteria (44LGS, 43M, 30LM, 14TH and a mix of all four) 1 day before oomycete inoculation. The P. medicaginis infection assay was conducted based on the method described by Ozgonen et al. [32]. Briefly, P. medicaginis was grown on PDA plates at 28 • C in a dark incubator for 2 weeks. Then, 1 cm 2 agar plugs from plates were transferred into a 1 L flask containing 300 g of autoclaved wheat seeds, and the oomycete was grown for 2 weeks at 28 • C under continuous light. Eight infected wheat seeds were placed on the soil surface for each pot, which contained a 3-week-old chickpea plant after the biocontrol isolate treatments. An additional 20 infected wheat seeds were randomly distributed in the chickpea water tray where pots were kept. Biocontrol treatments continued every week until the end of the experiment. At 4 weeks post-inoculation with P. medicaginis, plants were uprooted, dried and weighed.

Statistical Analyses
Results of fungal plate assays were assessed for significant differences (p < 0.05) using Microsoft Office Excel Student's t-test compared to S. saprophyticus assays. Student's t-test was performed on biopesticide raw data to determine significant differences (p < 0.05) between mock-treated control and biocontrol treatment plant samples.

Use of Biocontrol Bacteria as Biopesticides to Control Bacterial Plant Pathogens in Tomato
The four best biocontrol bacteria, which inhibited Pst on pathogen growth inhibition plate assays, were applied on tomato plants as potential biopesticides agents against Pst (Figure 2A

Use of Biocontrol Bacteria as Biopesticides to Control Bacterial Plant Pathogens in Tomato
The four best biocontrol bacteria, which inhibited Pst on pathogen growth inhibition plate assays, were applied on tomato plants as potential biopesticides agents against Pst (Figure 2A,B). After symptom analysis and qPCR quantification, disease symptoms and the number of pathogens per plant cell significantly (p < 0.05) declined for plants treated with isolates Paenibacillus peoriae (14TH), Comamonas jiangduensis (28MC), B. methylotrophicus (46YE), Bacillus amyloliquefaciens (33YE) and the mixture of all four compared to mock (MgCl 2 ) treatments ( Figure 2C,D).

Use of Biocontrol Bacteria as Biopesticides to Control Fungal Plant Pathogens in Cabbage
Based on preliminary screening in plant assays, the four most promising A. brassicicola antagonists, B. methylotrophicus (46YE), Brevibacillus laterosporus (4YE), B. licheniformis (28M), Bacillus megaterium (44LGS) and a mixture of these four bacterial isolates were examined for their potential as biopesticides to control dark leaf spot in cabbage ( Figure 3A,B). The symptom scoring suggests that individual isolates significantly (p < 0.05) reduced disease symptoms on cabbage plants, but not the mixture of isolates ( Figure 3C). However, qPCR data show that all treatments, including the mixture of isolates, significantly decreased the amount of pathogen DNA (p < 0.05; Figure 3D).

Use of Biocontrol Bacteria as Biopesticides to Control Oomycete Plant Pathogens in Chickpea
Based on preliminary screening in plant assays, the four most promising biocontrol isolates B. megaterium (44LGS), C. flavescens (43M), Bacillus mojavensis (30LM), Paenibacillus peoriae (14TH) and a mixture of all four were used as pretreatment to the soil of chickpea plants before they were challenged with P. medicaginis. No leaf symptoms were observed at 4 weeks post-infection, but upon inspection of the roots, it was evident that all bacterial pretreatments significantly (p < 0.05) prevented root biomass decline compared to mock-treated (MgCl 2 ) P. medicaginis-infected control plants (Figure 4). This suggests that in the presence of biocontrol bacteria, P. medicaginis was not able to infect plants effectively and significantly prevented root damage. The efforts to quantify P. medicaginis using molecular techniques were unsuccessful due to the lack of specific primers for P. medicaginis quantification.
Based on preliminary screening in plant assays, the four most promising biocontrol isolates B. megaterium (44LGS), C. flavescens (43M), Bacillus mojavensis (30LM), Paenibacillus peoriae (14TH) and a mixture of all four were used as pretreatment to the soil of chickpea plants before they were challenged with P. medicaginis. No leaf symptoms were observed at 4 weeks post-infection, but upon inspection of the roots, it was evident that all bacterial pretreatments significantly (p < 0.05) prevented root biomass decline compared to mocktreated (MgCl2) P. medicaginis-infected control plants (Figure 4). This suggests that in the presence of biocontrol bacteria, P. medicaginis was not able to infect plants effectively and significantly prevented root damage. The efforts to quantify P. medicaginis using molecular techniques were unsuccessful due to the lack of specific primers for P. medicaginis quantification.

Discussion
In this study, previously identified bacteria active against bacterial phytopathogens and food pathogens [23] were tested in the current study against fungal and oomycete pathogens (Figure 1 and Supplementary Figure S1). Several reviews summarised reports on anti-fungal activities of Bacillus, Brevibacillus, Lactobacillus and Paenibacillus species, as they produce a wide range of anti-microbial compounds [33][34][35][36]. There are a few reports

Discussion
In this study, previously identified bacteria active against bacterial phytopathogens and food pathogens [23] were tested in the current study against fungal and oomycete pathogens (Figure 1 and Supplementary Figure S1). Several reviews summarised reports on anti-fungal activities of Bacillus, Brevibacillus, Lactobacillus and Paenibacillus species, as they produce a wide range of anti-microbial compounds [33][34][35][36]. There are a few reports that show anti-fungal activity of Ochrobactrum and Klebsiella [37][38][39], and a recent study has shown the potential anti-fungal activity of Pseudochrobactrum kiredjianiae [40], which was also confirmed in our study. To our knowledge, however, there is currently no scientific report on C. flavescens, S. aquimarina and Sporosarcina saromensis anti-fungal or oomycete activities. To evaluate the potential application of biocontrol bacteria as biopesticides, we have developed three different assays against bacterial, fungal and oomycete pathogens.
Bacterial speck caused by Pst is an important economical bacterial phytopathogen in tomato plants [41]. In our biopesticide assay, we have applied four potential biocontrol bacteria as biopesticides, and B. amyloliquefaciens (33YE) was the best biopesticide among other treatments to suppress Pst symptoms (Figure 2). B. amyloliquefaciens is a well-known plant growth-promoting bacterium; for example, it can alleviate the symptoms of Pst in sugar beet plants [42].
Alternaria leaf spot causes severe infection and significant loss to brassica plants, such as canola [43]. Effects of Bacillus sp. on Alternaria alternata and Paenibacillus sp. on A. solani have been investigated before on tomato plants [44,45]; however, there is no evidence to demonstrate the effectiveness of Bacillus sp. and Brevibacillus sp. as biocontrol agents for A. brassicicola in cabbage. In our in vivo study, we found that B. laterosporus (4YE) supplied the best biocontrol performance in reducing the symptom and number of pathogens compared to other treatments (Figure 3). Although B. laterosporus has insecticidal effects, it has not been used as a biopesticide to control fungal pathogens or insects on plants [46]. The treatment with bacterial mixture also reduced the number of pathogens in cabbage leaf cells (as shown by qPCR), but the disease symptoms were exacerbated. This may have occurred because pathogen DNA in necrotic tissue could be degraded. A mixture of several biocontrol bacteria may also increase the likelihood for the activation of defense signaling pathways and plant hypersensitive response [47,48].
Phytophthora root rot is an important soil-borne disease of chickpea in Australia caused by P. medicaginis with up to $8.2 million annual yield losses [49]. Anti-fungal activity of Mesorhizobium ciceri and its growth promotion on chickpeas against P. medicaginis has been evaluated, and it has been shown that this bacterium can increase the biomass, improve nodulation and enhance disease resistance [50]. We chose four potential biocontrol bacteria and tested them in vivo against this phytopathogen. The mixture of the four bacterial isolates was by far the most effective treatment to prevent chickpea root biomass loss from P. medicaginis infection, although each individual bacterial treatment decreased root infection (Figure 4). A possible explanation could be a synergistic effect [51] from the mixture of bacteria that complement each other, and further studies may focus on the mechanisms on how this occurs. This could be done by direct anti-oomycete activity from all four isolates used. In addition, some of these isolates may also prime the plants to trigger defense responses, and induce systemic resistance or provide direct root growthpromoting effects.
B. amyloliquefaciens is one of the best-studied biocontrol bacteria, which produces a wide range of anti-microbial compounds, including amylocyclicin [55][56][57][58]. Our B. amyloliquefaciens isolate 33YE possesses AMM activity against the tomato pathogen C. michiganensis and AMP activity against the human pathogen Listeria monocytogenes [23]. Potential AMPs were derived from thioredoxin, septation protein SpoVG, 50S ribosomal protein L7/L12, phosphocarrier protein HPr, non-specific DNA-binding protein Hbs, a chaperonin and other hypothetical and cold shock proteins.
Further anti-microbial activity has been reported for B. mojavensis (30LM) against L. monocytogenes (AMP activity) [23]. B. mojavensis anti-microbial agents include mojavensin and the heat-stable toxin amylosin [62][63][64]. C. jiangduensis, a biosurfactant-producing bacterium [65], also produces AMPs and AMMs against P. syringae for the isolate (28MC) tested in the current study [23], which may explain the observed resistance against this pathogen in tomato (Figure 2). Apart from the observed activity against P. medicaginis (Figure 2), C. flavescens (43M) also had AMM activity against C. michiganensis and P. syringae [23]. C. flavescens is a beneficial bacterium applied in food industries to make cheese pigmentation, and its application as a biocontrol agent needs to be investigated further [66]. Apart from the observed activity against A. brassicicola (Figure 3), B. licheniformis (28M) also possessed AMP against L. monocytogenes [23]. B. licheniformis produces bacteriocin-like molecule lichenin [67,68]. B. megaterium (44LGS) was active against all pathogens tested in the present study, and previous studies found that B. megaterium produces anti-microbial metabolites and a cyclic polypeptide [69,70].
Symptom scoring and molecular analysis have shown that biopesticide treatments significantly lowered disease symptoms and the number of pathogens in tomato, cabbage and chickpea. This could happen due to direct interactions between biocontrol bacteria and their anti-microbial compounds and pathogens. In the current study ten potential biopesticides have been developed from Bacillus, Brevibacillus, Paenibacillus, Comamonas and Corynebacterium species that should be further tested under field conditions. The application of Bacillus species as biopesticides has been widely explored [71], but Brevibacillus, Paenibacillus and Comamonas species, while known for their anti-microbial activity, have not yet been tested as potential biopesticides. Several reports have suggested the potential application of Paenibacillus species as biopesticide agents. However, this is the first report that shows the potential application of P. peoriae as a biopesticide [45,72,73]. In addition, this study has shown potential uses of Brevibacillus, or Comamonas and Corynebacterium species as biopesticides against bacterial, fungal and oomycete pathogens.
Regulatory and technical barriers have limited biopesticide applications in agricultural industries [74]. Compared to conventional pesticides, biopesticides typically have a lower efficacy and stability [75]. Further studies will be required to determine if these bacteria when applied to plants would remain dominant in the soil and phyllosphere microbiomes or require formulation (e.g., preferred carbon source) to improve their stability. Biopesticides may have potential side effects on the environment, which need to be well studied. It has been shown that a fungus-based bioinsecticide could impact insect pollinators' functioning [76]. Therefore, possible environmental tests are needed before industrial applications of biopesticides are implemented.

Conclusions
Biopesticide application is an environmentally-friendly alternative for chemical pesticides. However, the current lack of diversity of available microbial biopesticides and their limited adaptability to different climate and farming conditions need to be investigated and addressed [11,77,78]. The current study took a biodiscovery approach by accessing diverse microbial taxa from environmental samples with the aim to rapidly develop microbial biopesticides to control the most devastating plant pathogens. The technology to rapidly develop new biopesticides is relevant for both agriculture and natural ecosystems where fungal threats to plant and ecosystem health increased 5-fold within 15 years [79]. The current study reports the application of soil bacteria against bacterial, fungal and oomycete plant pathogens as biocontrol agents. Our work suggests that environmental microbiomes, where bacteria compete against each other, such as rhizosphere soil, are a plentiful source for biocontrol agents against a wide range of diseases. Among these, three new potential biocontrol bacteria, Corynebacterium flavescens, Sporosarcina aquimarina and Sporosarcina saromensis, have been isolated that displayed anti-fungal and antioomycete activities against vegetable crop diseases that have not been reported previously. In total, we developed ten potential microbial biopesticides that significantly reduced the number of pathogens and disease symptoms in tomato, cabbage and chickpea plants.
Supplementary Materials: The following supplemental information is available online https:// www.mdpi.com/article/10.3390/applmicrobiol2010021/s1. Table S1: Media used in this study, Table S2: Primers and qPCR conditions, Table S3: Media used in this study, Figure S1: Screening microbial isolates against fungal and oomycete pathogens.