Detection of Antagonistic Compounds Synthesized by Bacillus velezensis against Xanthomonas citri subsp. citri by Metabolome and RNA Sequencing

Abstract

Biological control of plant diseases has gained attraction for controlling various bacterial diseases at a field trial stage. An isolated endophytic bacterium, Bacillus velezensis 25 (Bv-25), from Citrus species had strong antagonistic activity against Xanthomonas citri subsp. citri (Xcc), which causes citrus canker disease. When Bv-25 was incubated in Landy broth or yeast nutrient broth (YNB), the ethyl acetate extract of Landy broth exhibited higher levels of antagonistic activity against Xcc compared to that of YNB. Therefore, the antimicrobial compounds in the two ethyl acetate extracts were detected by high performance liquid chromatography–mass spectrometry. This comparison revealed an increase in production of several antimicrobial compounds, including difficidin, surfactin, fengycin, and Iturin-A or bacillomycin-D by incubation in Landy broth. RNA sequencing for the Bv-25 grown in Landy broth were performed, and the differential expressions were detected for the genes encoding the enzymes for the synthesis of antimicrobial compounds, such as bacilysin, plipastatin or fengycin, surfactin, and mycosubtilin. Combination of metabolomics analysis and RNA sequencing strongly suggests that several antagonistic compounds, especially bacilysin produced by B. velezensis, exhibit an antagonistic effect against Xcc.

1. Introduction

Phytopathogenic bacteria cause numerous plant diseases worldwide, resulting in considerable losses in crop production and storage [1]. The control strategies for these diseases pose one of the greatest challenges to sustainable agriculture. Due to the difficulty of managing bacterial diseases naturally, growers heavily rely on chemical agents, especially pesticides [2]. Extensive and repeated applications of these chemicals have led to the development of pesticide resistance in the target phytopathogens and even non-target organisms, including human beings [2]. Besides chemical pesticides, the public is becoming more concerned about using antibiotics in the environment due to their adverse effects on human health and are suspicious of the massive application of antibiotics due to the advent of antibacterial resistance [3]. Therefore, in recent years, biological control agents (i.e., specialized microbes or its metabolite to control plant diseases) are gaining attention from scientists and farmers as an alternative mean of plant disease management. Until now, numerous biocontrol agents have been commercially available to control plant diseases [4].

Citrus canker is a devastating disease caused by the bacterial pathogen Xanthomonas citri subsp. citri (Xcc), which affects most commercial citrus varieties. Infection of Xcc causes necrotic lesions on citrus leaves, stems, and fruits, resulting in premature fruit loss, defoliation, twig dieback, and tree decline [5]. The disease impacts fruit production detrimentally by both decreasing fruit yields in most citrus-producing countries and making the infected fruits less valuable and unmarketable [6]. The most common approach to control citrus canker is the usage of copper-based bactericidal sprays such as copper hydroxide, cuprous oxide, and copper oxychloride [7]. Copper-resistant strains, however, have been detected on several phytopathogenic bacteria, including Xcc, X. axonopodis pv. vesicatoria, X. alfalfae subsp. citrumelonis, and Pseudomonas syringae pv. tomato [7,8]. A copper-resistant strain of Xcc was first reported in 1994 in the province of Corrientes, Argentina, where the citrus nurseries were exposed to regular application of copper-based bactericides [8]. The antibiotic streptomycin is extensively used to control Xcc; however, streptomycin resistance emerges several- or thousand-folds due to chromosomal or gene mutations [9]. Streptomycin resistance exhibits an increasing pattern due to the repeated usage of the chemical. For example, in the case of pathogenic X. perforans, resistance increased from 5% in 2006 to 14% in 2011 during field applications of streptomycin [10].

Plant pathologists and microbiologists, therefore, have developed novel strategies for developing biocontrol agents to control phytopathogenic bacteria in more cost-effective and environmentally friendly ways [11]. Numerous new microbes containing biologically active agents have been identified in an effort to reduce the excessive use of chemical pesticides in agriculture [12]. Among the commercially developed biocontrol agents, the most used organisms are based on the isolation of Bacillus spp. For example, Botrybel (B. velezensis), RhizoVital^® (B. velezensis FZB42^T), Serenade (B. velezensis QST713), Kodiak ^TM (B. velezensis GB03), and Taegro^® (B. velezensis FZB24) [13]. Pseudomonas- and Strepotomyces-based biocontrol agents are also available in the market [12]. These biocontrol agents exhibit various modes of action, such as antimicrobial compound secretion, systemic resistance induction, and nutrient competition with the pathogenic microbes [11].

Next-generation genome sequencing has revealed that many biosynthetic gene clusters encode bioactive secondary metabolites in the genomes of several Bacillus species, e.g., B. velezensis, B. subtilis, B. amyloliquefaciens, and B. siamensis [14,15,16,17]. B. velezensis is an endospore-forming and free-living soil bacterium, first described by Ruiz-Garcia in 2005 [18]. The enzymes encoded by these biosynthetic genes synthesize antimicrobial compounds that could serve as a reservoir of novel scaffolds for drug screening to treat numerous plant diseases [19,20]. Following complete sequencing and annotation of the B. velezensis genome, considerable parts of the genomic DNA were found harboring various antimicrobial gene clusters. The clusters of srf, bmy, fen, mln, dfn, bae, dhb, and bac encode the active compounds surfactins, bacillomycin D, fengycins, macrolactin, difficidin, bacillaene, bacillibactin, and bacilysin, respectively [21]. These bioactive secondary metabolites have numerous applications in plant-microbe interactions with antimicrobial and nematocidal activities [21]. Among the compounds, surfactins, bacillomycin D, and fengycins are grouped as lipopeptides (LPs), and the classification of these compounds is based on the length variation of fatty acid chain and amino acid sequences of these compounds [22]. Macrolactin, difficidin, and bacillaene are classified as polyketide molecules (PKs) that are synthesized by giant modular polyketide synthases with multifunctional domains [23].

LPs are mainly associated with antifungal action, especially against filamentous fungi. For example, bacillomycin D synthesized by B. velezensis FZB42 exhibited strong antifungal action against Fusarium graminearum that causes Fusarium head blight [24]. Lipopeptide fengycins produced by B. velezensis Bs006 suppressed mycelial growth of F. oxysporum f. sp. physali that is responsible for golden berry wilt disease [25]. Surfactin B and surfactin C produced by B. velezensis 9D-6 exhibits antifungal activities against Monilinia fructicola that causes brown rot of nectarine and some other stone fruits [26]. Similarly, the antifungal activity of B. velezensis LM2303 was mainly associated with fengycin B, iturin A, and surfactin A [25]. Under field conditions, LM2303 exhibited strong antagonist activity against F. graminearum by significantly reducing disease severity of Fusarium head blight, with a control efficiency of 72.3% [27].

The antibacterial activity of B. velezensis is mainly associated with PKs including macrolactin, difficidin, and bacillaene [28]. For example, difficidin and bacilysin produced by B. velezensis FZB42 exhibited antibacterial activity against Xanthomonas oryzae pv. oryzae and X. oryzae pv. oryzicola [29]. B. velezensis LM2303 encodes several antibacterial compounds such as surfactin A, butirosin, plantazolicin, kijanimicin, bacilysin, difficidin, bacillaene, and macrolactin. In addition, other metabolites including iron–siderophore bacillibactin, molybdenum cofactor, and teichuronic acid were involved with nutrient uptake during plant–microbe interaction [27]. Moreover, the antifungal action of dipeptide antibiotic bacilysin was reported; for example, bacilysin produced by FZB42 damaged the hyphal structure of Phytophthora sojae that causes root rot disease in soybean [30]. Two other ribosomally synthesized bacteriocin, known as plantazolicin and amylocyclicin in B. velezensis, displayed high antibacterial activities against closely related Gram-positive bacteria [31,32]. Bacillibactin secreted by B. velezensis can bind with environmental iron to form siderophore–iron complex and can be internalized into the bacterium, where iron is released for cellular processes or nutritional requirements [33].

Previously, we isolated B. velezensis-25 (Bv-25), which exhibited antibacterial activity against wild-type and streptomycin-resistant Xcc strains [34,35]. Therefore, we conducted this research to further understand the antibacterial compounds produced by B. velezensis that can control citrus canker disease effectively and efficiently. Through the application of various state-of-the-art techniques, such as transcriptome analysis, high performance liquid chromatography–mass spectrometry (HPLC-MS) analysis, and antimicrobial assays, we identified the differentially expressed genes and antibacterial compounds, which may be involved in exerting the antagonistic effects on Xcc.

2. Materials and Methods

2.1. Strains and Culture Conditions

Bv-25 and B. amyloliquefaciens strain 2 (Ba-2) were incubated in yeast nutrient broth (YNB: yeast extract, 5 g; and nutrient broth, 8 g in 1 L distilled water) and in Landy broth (L-glutamic acid 5 g, glucose 20 g, yeast extract 1 g, phenylalanine 2 mg, MgSO₄·7H₂O 0.5 g, KCl 0.5 g, MnSO₄ 5 mg, CuSO₄·5H₂O 0.16 mg, FeSO₄·7H₂O 0.15 mg, KH₂PO₄ 1 g in 1 L H₂O, pH 7.0). Bv-25 was isolated from citrus species in our laboratory, and Ba-2 was provided by the Korean Agricultural Culture Collection (KACC, RDA, Wanju-gun, Republic of Korea). The pathogenic Xcc strains used for bioassay were composed of two Xcc strains, wild-type XccW1 and streptomycin-resistant mutant XccM4 [9]. The pathogenic Xcc strains were kind gifts from Dr. Hyun (Citrus Research Station, NIHH, RDA, Jeju, Republic of Korea). The wild-type and streptomycin-resistant strains of Xcc were routinely grown on yeast nutrient agar (YNA, yeast extract, 5 g; nutrient broth, 8 g; and agar, 1.5 g in distilled water, 1 L) or YNA supplemented with 50 ppm streptomycin, respectively. All the strains, such as Bv-25, Ba-2, and Xcc, were maintained in YNB containing 50% glycerol at −80 °C until use.

2.2. Antibacterial Activity of Ethyl Acetate Extracts of Bv-25 and Ba-2 in Landy Broth and YNB

To compare the production of active compounds in different media, 50 μL of B. velezensis Bv-25 and Ba-2 cells from the stock were added to 2 mL of different media including Landy broth and YNB. The cultures were then incubated for 24 h at 28 °C. The seed cultures (0.1 mL) of B. velezensis Bv-25 and Ba-2 were then transferred into 100 mL of Landy and YNB broth and cultured for 48 h at 28 °C in a shaking incubator with 180 rpm. The growth of bacterial cells in different media was monitored using OD600_nm by a spectrophotometer (ASP-3700, ACTGene, Piscataway, NJ, USA). An equal volume of ethyl acetate (100 mL) was added to the bacterial cultures, sonicated for 5 min, and maintained overnight with vigorous shaking. The culture mixture was centrifuged at 3000 rpm for 10 min at 4 °C, the supernatants collected, and dried in a rotary evaporator (A-1000S; Eyela, Tokyo, Japan) at 50 °C.

The residues were dissolved In 1.5 mL HPLC-grade methanol and air-dried under a chemical hood. After drying, metabolites were weighed and the metabolite concentrations were prepared (10.0 mg mL⁻¹, 1.0 mg mL⁻¹ and 0.1 mg mL⁻¹) using HPLC-grade methanol.

Disk-diffusion assays were used to test the antibacterial activity of ethyl acetate extracts against Xcc strains. XccW1 and XccM4 strain grown overnight was mixed with 5 mL of 0.7% YNA soft agar and directly poured onto 1.5% agar YNA plates. After gel solidification, 30 µL of the metabolites in three different concentrations, such as 10.0 mg mL⁻¹, 1.0 mg mL⁻¹, and 0.1 mg mL⁻¹, were placed on a sterile paper disk. The antibacterial activity of the ethyl acetate extract was compared with the positive control composed of two streptomycin concentrations (S1: streptomycin 1.0 mg mL⁻¹; S2: streptomycin 0.1 mg mL⁻¹) and a negative control composed of 100% methanol. The inhibition zones after 24 h incubation at 28 °C around the paper disks were measured using an electronic digital caliper, and the antibacterial activities were compared. The experiment was conducted twice with three replicates.

2.3. Minimum Inhibitory Concentration (MIC) and Minimum Bactericidal Concentration (MBC) of Bv-25 and Ba-2 Extracts Cultured in Landy Broth and YNB

After selecting Landy medium as the highest antibacterial activity-inducing medium, MIC and MBC of ethyl acetate extracts of Bv-25, Ba-2, and streptomycin were determined using Landy broth and YNB by the broth microdilution method [36]. The metabolite was collected from the bacterial culture, weighed and dissolved in methanol at a concentration of 2.0 gm mL⁻¹ (stock solution). For MIC analysis, the metabolites of Bv-25 and Ba-2 were two-fold serially diluted using methanol in order to make concentrations ranging from 15.625–1000 μg mL⁻¹. The study of MIC was conducted in 96-well plates. Each well contained a total 200 µL, composed of 100 µL metabolites of Bv-25 and Ba-2, 90 µL broth, and 10 µL indicator bacterial inoculum with an optical density of 2.0 at 600 nm. One well served as a positive control (YNB plus inoculum), and one served as a negative control (only YNB). The plates were incubated at 28 °C for 24 h. MIC was defined as the lowest concentration of metabolite extract that prevented the visible growth of the indicator bacterial strain. Additionally, MBC was determined by plating 10 µL of the cultures from each well in YNA agar plates. MBC was defined as the lowest concentration of metabolite extract that did not show any bacterial growth in YNA plates after an incubation period of 24 h at 28 °C. Both the MIC and MBC were expressed in µg mL⁻¹.

2.4. HPLC-MS Analysis of Ethyl Acetate Extracts from Bv-25 and Ba-2

The ethyl acetate extracts of Bacillus spp. were filtered by using 0.22 µm nylon filter that was then analyzed by using liquid chromatograph–mass spectrometry (HPLC-MS). The HPLC-MS system was composed of an HPLC apparatus (model 2695; Waters, Milford, MA, USA) equipped with a pentafluorophenyl column (Luna C18 reversed phase column, 4.6 × 150 mm, 5μm; Phenomenex, Torrance, CA, USA) and a mass spectrometer (Waters model 3100, Milford, MA, USA). The HPLC conditions were as follows: injection volume = 5 µL, solvent A = 20 mM Ammonium acetate buffer (pH adjusted to 7.2), solvent B = acetonitrile, and solvent program = 50: 50 (A/B) to 50: 50 (A/B) over 10 min at a flow rate of 0.5 mL/min. MS data was obtained under the following conditions: desolvation gas (N₂) flow rate = 4 L/h, desolvation temperature = 350 °C, capillary voltage = 4 kV, cone voltage = 30 V, ionization mode = electrospray positive, and single ion recording m/z = 163.

2.5. Differential Gene Expression Analysis of Bv-25 Grown in Landy Broth and YNB

RNA-sequence analysis was employed to disclose the transcriptomic features of Bv-25 genes that were expressed in two different media (Landy and YNB). In this process, genetic information from a gene is translated in the synthesis of a functional gene product, typically proteins. Total RNA was purified from Bv-25 and propagated for 24 h using Landy broth and YNB. Bacteria were collected by centrifugation at 8000 rpm at 4 °C for 10 min, and the cell pellets (bacterial cells) were then immediately frozen in liquid nitrogen and stored at −80 °C until use. RNA extraction (performed with the RNeasy Midi kit; Qiagen, Hilden, Germany), RNA quantification, and cDNA synthesis and labeling were performed as described previously [37]. Log2 fold change (log2 FC) was calculated through dividing FPKM (Fragments per kilobase of exon model per million mapped reads) value of Bv-25 grown in Landy by Bv-25 grown on YNB to obtain fold change (FC) and then putting the equation in excel as Log(FC, 2).

2.6. Antibacterial Activity Assay of Pure Surfactin and Fengycin against Xcc

Several antibacterial compounds were detected by metabolome and transcriptome analysis. Several compounds such as fengycin and surfactin (purchased from Sigma-Aldrich, St. Louis, MO, USA) and macrolactin isolated from Bv-25 (purified and identified by Dr. Sanghee Shim from Seoul National University, Seoul, Republic of Korea) were tested for antagonistic effects against Xcc based on the procedure described in Section 2.2.

3. Results

3.1. Antibacterial Activity of Ethyl Acetate Extracts of Bv-25 in Landy and YNB

In the disk diffusion assays, ethyl acetate extracts of Bv-25 were screened for their antibacterial potentiality against the wild-type XccW1 and streptomycin resistant XccM4 strains (Figure 1). The metabolite extracted from Bv-25 using Landy broth showed higher antibacterial activity against both Xanthomonas strains than the one extracted using YNB. At a concentration of 10.0 mg mL⁻¹ (extracted from Landy broth), the zones of inhibition for Bv-25 metabolites against XccW1 and XccM4 were 26.59 ± 1.29 mm and 30.97 ± 2.3 mm, respectively. At a concentration of 1.0 mg mL⁻¹, these metabolites inhibited XccW1 and XccM4 with zones of inhibition of 17.75 ± 0.91 mm and 20.81 ± 0.78 mm, respectively. On the other hand, the metabolites that was isolated from Bv-25 using YNB broth displayed a substantially reduced zone of inhibition in comparison to the Landy broth extracted metabolites. The zones of inhibition against XccW1 and XccM4 at a dose of 10 mg mL⁻¹ were 22.83 ± 0.53 and 22.43 ± 1.53 mm, respectively. For the same metabolites at a dose of 1 mg mL⁻¹, the zones of inhibition against XccW1 and XccM4 were 12.91 ± 0.37 and 12.84 ± 0.55 mm, respectively. The metabolites extracted from Ba-2 did not exhibit any antibacterial effect against Xanthomonas pathogens.

Streptomycin concentration S1 and S2 had antagonistic effects on the wild-type XccW1, with the zone of inhibition 30.68 ± 2.12 mm and 19.26 ± 0.46 mm, respectively. However, S1 and S2 did not exhibit any antibacterial effect on the streptomycin-resistant XccM4 strain (Figure 1).

3.2. Determination of MIC and MBC of the Ethyl Acetate Extracts of Bv-25 and Ba-2 Incubated in Landy Broth and YNB

MIC and MBC of crude ethyl acetate extract of strain Bv-25 were determined by testing against two Xanthomonas pathogens, XccW1 and XccM4, using the broth micro-dilution method. For Bv-25 extract using Landy broth culture, the MIC and MBC were 31.25 µg mL⁻¹ and 62.5 µg mL⁻¹, respectively against two Xcc strains. On the other hand, MIC and MBC of Bv-25 metabolites using YNB culture was 62.5 µg mL⁻¹ and 125.0 µg mL⁻¹, respectively. MIC and MBC of streptomycin (control antibiotic) against XccW1 ranged from 1.25–2.5 µg mL⁻¹, but MIC and MBC of streptomycin against the streptomycin-resistant XccM4 strains were significantly high and ranged from >500–1000 µg mL⁻¹ (

Table 1. Determination of the MIC and MBC of the metabolic extract of B. velezensis Bv-25 against wild type XccW1 and streptomycin resistant XccM4 using the culture broth of Landy and YNB.

Xcc Strains	Extract of Bv-25 Using Landy Broth		Extract of Bv-25 Using YNB		Streptomycin
	MIC (µg mL−1)	MBC (µg mL−1)	MIC (µg mL−1)	MBC (µg mL−1)	MIC (µg mL−1)	MBC (µg mL−1)
XccW1	31.25	62.5	62.5	125.0	1.25	2.5
XccM4	31.25	62.5	62.5	125.0	>500	>1000

). Metabolite extracted from Ba-2 had no effect on any Xanthomonas strains.

3.3. Detection of Antimicrobial Compounds by HPLC-MS Analysis

HPLC-MS analysis was performed to identify the differentially produced molecules, based on the molecular weight from the MS database (Mass Spectrometry Data Center, NIST, Gaithersburg, MD, USA). These compounds mostly belonged to the class of LPs (iturin, fengycin, bacillomycin-D, and surfactin) and PKs (oxydifficidin, macrolactin, and bacillaene) (

Table 2. Cyclic lipopeptide and polyketide molecules produced by the B. velezensis Bv-25 strain as detected by HPLC-MS analysis. The values of Ba-2 and Bv-25 were the means of the data from three independent replications.

Compound Name	M/Z	RT (min)	Ba-2	Bv-25	Fold Change (Bv-25/Ba-2)	p-Value
Oxydifficidin	559.0	43.3	216,380.3	30,417,069.1	140.6	0.0027
Bacillaene	582.2	31.3	291,622.7	2,774,552.5	9.5	0.0001
Macrolactin	524.2	38.6	308,031.1	2,211,408.9	7.2	0.0046
Surfactin	1036.6	27.0	93,166.2	1,420,585.5	15.2	0.0000
Fengycin A	1475.9	30.5	128,474.2	1,092,443.5	8.5	0.0006
Fengycin B	1489.3	31.3	158,266.9	983,475.3	6.2	0.0005
Bacillomycin D	1118.6	40.5	92,804.4	19,132,037.6	206.2	0.0031
Iturin	1083.7	29.3	272,286.6	2,513,178.3	9.2	0.0036

). HPLC-MS analyses indicated the presence of several active substances with protonated molecular ions at m/z 1036.6, corresponding to surfactin isoforms. The molecular mass of bacillomycin-D in the range m/z 1008–1036 matched with the molecular database. Notably, these results corresponded to the MS data obtained from the previous reports [29].

3.4. Differential Gene Expression Analysis of Bv-25 Grown in Landy and YNB Media

To reveal the transcriptomic correlates of overproduction, we performed transcriptome profiling via RNA-sequencing of Bv-25 in Landy broth and YNB with an incubation period of 3 days at 28 °C. Although most genes were medially expressed in both Landy and YNB throughout the cultivation period, genes were expressed at a higher level in Landy broth. A total of 3930 genes were expressed during the mRNA expression analysis. Based on the expression data, genes encoding bacilysin (bac), plipastatin (pps) or fengycin (fen), surfactin (srf), and mycosubtilin (myc) were increased based on fragments per kilobase of exon model per million mapped reads (FPKM) value (Figure 2,

Table 3. Genes associated with the antibacterial compounds of B. velezensis Bv-25 during mRNA expression analysis.

Active Compound	Gene	UniProt ID	Product
Bacilysin	bacA	Q8KWT6	Prephenate decarboxylase
	bacB	Q8KWT5	3-[(4R)-4-hydroxycyclohexa-1,5-dien-1-yl]-2-oxopropanoate isomerase
	bacC	Q8KWT4	Dihydroanticapsin dehydrogenase
	bacD	Q8KWT3	L-alanine-anticapsin ligase
	bacE	Q8KWT2	Putative bacilysin exporter
	ywfG	P39643	Probable aspartate aminotransferase
	ywfH	P39644	Bacilysin biosynthesis oxidoreductase
Plipastatin or Fengycin	ppsA	P39845	Plipastatin synthase subunit A
	ppsB	P39846	Plipastatin synthase subunit B
	ppsC	P39847	Plipastatin synthase subunit C
	ppsD	P94459	Plipastatin synthase subunit D
	ppsE	O31827	Plipastatin synthase subunit E
Mycosubtilin	fenF	Q9R9J2	Malonyl CoA-acyl carrier protein transacylase
	mycA	Q9R9J1	Mycosubtilin synthase subunit A
	mycB	Q9R9J0	Mycosubtilin synthase subunit B
	mycC	Q9R9I9	Mycosubtilin synthase subunit C
Surfactin	srfAD	Q08788	Surfactin synthase thioesterase subunit
	srfAC	Q08787	Surfactin synthase subunit 3
	srfAB	Q04747	Surfactin synthase subunit 2
	srfAA	P27206	Surfactin synthase subunit 1

). Besides a substantial number of genes or gene clusters involved in rhizosphere adaptation in plant beneficial traits through the formation of biofilm or by assimilating nitrogen, iron (through secretion of siderophore bacillibactin), potassium, manganese (Mn²⁺), and magnesium (Mg²⁺). Numerous genes related to plant growth promotion and induction of systemic resistance in plants were found to be expressed during mRNA expression study and are summarized in

Table 4. Genes associated with the bacterium-plant interactions during mRNA expression analysis of B. velezensis Bv-25.

Bioactivity	Gene	Product	Log2 FC
Biofilm formation (Exopolysaccharide component)	epsC	Probable polysaccharide biosynthesis protein	−3.17
	epsD	Putative glycosyltransferase	−2.08
	epsE	Putative glycosyltransferase	−4.00
	epsF	Putative glycosyltransferase	−0.45
	epsG	Transmembrane protein	0.04
	epsH	Putative glycosyltransferase	−1.25
	epsI	Putative pyruvyl transferase	−0.85
	epsJ	Uncharacterized glycosyltransferase	0.51
	epsK	Uncharacterized membrane protein	0.20
	epsL	Uncharacterized sugar transferase	−0.88
	epsM	Putative acetyltransferase EpsM	−1.37
	epsN	Putative pyridoxal phosphate-dependent aminotransferase	−0.55
Plant growth promotion and ISR induction (3- hydroxy-2-butanone)	alsR	HTH-type transcriptional regulator	2.77
	alsS	Acetolactate synthase	−3.58
	alsD	Alpha-acetolactate decarboxylase	−7.67
Plant growth promotion (Indole acetic acid)	ysnE	Uncharacterized N-acetyltransferase	inf
	yhcX	Hydrolase	−0.48
Plant growth promotion (Trehalose)	treR	Trehalose operon transcriptional repressor	1.91
	treA	Trehalose-6-phosphate hydrolase	1.32
	treP	PTS system trehalose-specific EIIBC component	0.72

and

Table 5. Expression of key genes of B. velezensis Bv-25 related to ion assimilation during mRNA expression analysis.

Bioactivity	Gene	Product	Log2 FC
Bacillibactin for iron (Fe2+) assimilation	dhbA	2,3-dihydro-2,3-dihydroxybenzoate dehydrogenase	0.80
	dhbC	Isochorismate synthase	−1.52
	dhbE	2,3-dihydroxybenzoate-AMP ligase	−2.45
	dhbB	Isochorismatase	−3.08
	dhbF	Dimodular nonribosomal peptide synthase	−1.32
Manganese (Mn2+) assimilation	mntH	Divalent metal cation transporter	3.71
	mntP	Putative manganese efflux pump	0.69
	mntR	Transcriptional regulator	−0.75
Magnesium (Mg2+) assimilation	mgtE	Magnesium transporter	2.19
	corA	Magnesium transport protein	−2.66
Potassium (K) assimilation	ktrD	Ktr system potassium uptake protein D	1.06
	ktrC	Ktr system potassium uptake protein C	−2.75
Nitrogen (N) assimilation	moaA	Cyclic pyranopterin monophosphate synthase	−1.03
	moaB	Molybdenum cofactor biosynthesis protein B	1.59
	moaC	Cyclic pyranopterin monophosphate synthase	0.79
	moaD	Molybdopterin synthase sulfur carrier subunit	0.00
	nasB	Assimilatory nitrate reductase electron transfer subunit	1.34
	nasD	Nitrite reductase [NAD(P)H]	2.82
	nasF	Uroporphyrinogen-III C-methyltransferase	0.41
	nrgB	Nitrogen regulatory PII-like protein	−3.03
	narK	Nitrite extrusion protein	3.04

.

3.5. Antibacterial Activity Assay of Pure Surfactin and Fengycin against Xcc

By analyzing the metabolome of Bv-25, several antimicrobial substances, including surfactin, fengycin or plipastatin, and macrolactin, were detected. Therefore, we checked the antibacterial activity of fengycin, surfactin, and macrolactin; however, we were unable to detect any antibacterial impact of these compounds against the Xanthomonas spp. (XccW1 and XccM4).

4. Discussion

Biological control of diseases by Bacillus spp. have been extensively used to combat plant pathogenic microbes due to their specific characteristics of heat and desiccation resistance that help in the formulation of stable and dry powder with longer shelf lives. Furthermore, Bacillus spp. are less toxic to the environment, are capable of producing antibiotics, spores, and biofilm, and induce systemic resistance in plants [38]. Metabolite produced by a particular microorganism is crucial in determining its efficacy to be a potential biocontrol agent to combat plant diseases. In this study, we showed that the ethyl acetate extracts of Bv-25 exhibited strong antibacterial activity against both wild type and streptomycin resistant Xcc. Furthermore, metabolite extracted from Landy broth incubation exerted stronger antibacterial activity than that extracted from YNB incubation.

HPLC-MS data obtained during this study showed that the antagonistic isolates were able to produce a variety of LPs (i.e., surfactins, bacilliomycin-D, and fengycin) and PKs type molecules (i.e., difficidin, bacillaene, and macrolactin). The result of this study also agree with previous investigations, which reported the presence of several non-ribosomally synthesized LPs- and PKs-type molecules in B. velezensis that showed antibacterial activity against phytopathogenic microbes [39]. These compounds have been investigated for their antagonistic effects against a wide variety of phytopathogenic microorganisms including Xanthomonas oryzae pv. oryzae and X. oryzae pv. oryzicola (bacterial blight and bacterial leaf streak disease of rice, respectively) [29], Rhizoctonia solani (bottom rot of lettuce) [40], and X. campestris pv. cucurbitae (bacterial leaf spot of cucurbits) [41]. Co-culture of B. velezensis FZB42 and pathogenic strain of X. campestris pv. campestris had antagonistic effect on the pathogen that causes black rot disease on cruciferous vegetables [42].

Transcriptome analysis of Bv-25 revealed the differentially increased expression of several gene clusters encoding bacilysin (bac), plipastatin (pps) or fengycin (fen), surfactin (srf), and mycosubtilin (myc) in the Landy broth incubation in comparison to YNB incubation. Bacilysin produced by B. velezensis is a dipeptide antibiotic, which relies on peptide transporters for uptake into the target cells and causes changes in the cell wall structure and efflux of intracellular components of X. oryzae pv. oryzae and X. oryzae pv. oryzicola after 12 h of exposure [29]. This dipeptide antibiotic also reported to show antifungal activity against Candida albicans. Upon the transportation of bacilysin into the C. albicans cells, intracellular proteinase hydrolyzes bacilysin into anticapsin and L-alanine. Anticapsin interacts with glucosamne-6-phosphate synthase of C. albicans, which is essential for cell wall formation [13,43].

Plipastatin or fengycin, a well-studied lipopeptide at a genetic level, is known to develop antifungal activity against filamentous fungi. Plipastatin A, secreted by B. amyloliquefaciens S76-3, displayed strong fungicidal activity against Fusarium graminearum by completely killing the conidial spores at the MIC range of 100 µg mL⁻¹. Microscopy analyses revealed severe morphological changes in conidia and substantial distortions in F. graminearum hyphae and caused vacuolation [44]. Mycosubtilin is a potent LP-type antifungal peptide compound, mainly detected in B. subtilis. This compound is characterized by a β-amino fatty acid moiety linked to the circular heptapeptide [45]. Surfactins produced by B. subtilis 9407, which includes C13- to C16-surfactin A, were the primary antibacterial compound against Acidovorax citrulli (syn. Acidovorax avenae subsp. citrulli), which cause bacterial fruit blotch of cucurbit crops. A srfAB deletion mutant strain of B. subtilis 9407 ∆srfAB, was unable to synthesize surfactin as well as showed reduced ability to form biofilms, swarming motility and root colonization during in vivo experiments [46].

Based on the metabolome and transcriptome analysis, we prepared the commercially available pure fengycin and surfactin and an isolated compound macrolactin from Bv-25. We found the antagonistic effect against Xcc to be non-existent, therefore, there are only few B. velezensis-producing compounds left, including bacilysin, difficidin, bacillaene, bacillomycin D, and iturin. We assumed that the most promising candidate could be bacilysin because the activity of ethyl acetate decreased after only one week, indicating the most active compound is unstable due to presence of epoxy group on bacilysin. The unstable nature of this compound can be the most difficult challenge for agricultural application. Collectively, this study indicates that B. velezensis could be a promising candidate to control citrus canker caused by X. citri subsp. citri and may provide an effective strategy to combat plant pathogens.

5. Conclusions

In this study, we investigated the antibacterial compounds synthesized by B. velezensis Bv-25 against the phytopathogenic Xcc. Our combined investigation by HPLC-MS and differential gene expression analysis identified several LPs and PKs type antimicrobial compounds, as well as a dipeptide antibiotic, bacilysin. We assumed that bacilysin produced by B. velezensis could be the controlling agent against Xcc. Though bacilysin has been identified as a strong antimicrobial agent, the epoxy group in the compound makes it unstable. In addition to the investigation of freshly prepared bacilysin for the antagonistic effect against Xcc, future work will focus on the stabilization of bacilysin activity by modifying the molecular structure to interfere with the spontaneous degradation of bacilysin.