Ralstonia solanacearum Secretions Induce Shifts in Macrolactin Composition and Reduction in Antimicrobial Activity of Bacillus amyloliquefaciens BNC5

Abstract

Ralstonia solanacearum, a pathogen causing bacterial wilt, threatens global crop production, necessitating sustainable biocontrol solutions. Bacillus amyloliquefaciens is a promising biocontrol agent, but the mechanisms of its interaction with R. solanacearum are not well understood. This study investigates the effects of R. solanacearum secretions on the growth, metabolite production, and antimicrobial properties of B. amyloliquefaciens BNC5. Untargeted metabolomics revealed increases in two known macrolactins (A and M), six novel macrolactins (BA1–6), and one novel polyketide (BAP1), along with decreases in 7-O-malonyl macrolactin A and a new macrolactin, BA7, following exposure to the secretions. Real-time polymerase chain reaction showed a downregulation of the tailoring enzyme gene bmmGT3. The reduced antimicrobial activity in the exposed extracts (78 vs 39–78 μg mL^–1 in unexposed extracts) is likely due to diminished macrolactin efficacy in macrolactin fractions from the exposed (512–1024 μg mL^–1) compared to those from the unexposed samples (4 μg mL^–1). Supported by tomato pot experiments, these results suggest that applying B. amyloliquefaciens extracts prior to introducing live bacteria may enhance the management of R. solanacearum in severely affected fields, providing a sustainable alternative to chemical pesticides.

Introduction

Ralstonia solanacearum is a soil-borne plant pathogen that causes bacterial wilt in over 310 plant species, including tomatoes, chilies, potatoes, and gingers, posing a significant threat to global agriculture. This Gram-negative bacterium thrives in warm and humid climates; spreads rapidly through infected soil, plant materials, and water; and persists in soil for many years. Its genetic diversity, adaptability, and production of virulence factorssuch as exopolysaccharides, cell wall-degrading enzymes, and effector proteins , complicate control efforts. The infection alters root development to enhance host colonization and disrupts water and nutrient transport, ultimately resulting in plant wilting and death. Despite its devastating impact on crops, there are currently no highly effective and universally applicable chemicals or measures to control this pathogen. Efforts to develop resistant cultivars have been hindered by the bacterium’s genetic variability and the frequent occurrence of undesirable agronomic traits.

One promising method to control plant diseases is the use of biocontrol agents, beneficial organisms that can suppress plant pathogens. Biocontrol has several advantages: it is safer for farm workers and communities than synthetic pesticides, requires no waiting period after harvest, causes no toxic damage to plants, and is inexpensive to mass-produce. Additionally, biocontrol has succeeded in many circumstances where pesticides have failed. However, its effectiveness can be limited by environmental factors, such as host species, local microbes, temperature, humidity, and nutrients, which affect the agents’ ability to colonize roots or survive. Introduced microbes might also disrupt local ecosystems. Therefore, using local biocontrol microbes and understanding their interactions with other organisms are crucial for long-term success and environmental sustainability.

Recently, we successfully isolated Bacillus amyloliquefaciens BNC5 from Nan province, Thailand, based on its exceptional ability to inhibit the growth of R. solanacearum as well as other fungal pathogens in agar diffusion assays. B. amyloliquefaciens is a Gram-positive, spore-forming bacterium, known for its diverse applications in agriculture, particularly as a biocontrol and plant growth-promoting agent. , The use of Bacillus species as biocontrol agents offers several advantages, including easy spore storage, long life expectancy, and ease of application into the soil. ,

The biocontrol capabilities of B. amyloliquefaciens are primarily attributed to its production of lipopeptides with strong antimicrobial activities, including surfactins, iturins, and fengycins. , Previous research on Bacillus velezensis, another Bacillus species in the operational group Bacillus amyloliquefaciens, indicated increased lipopeptides levels in the presence of R. solanacearum. However, despite these increases, B. velezensis could not completely eradicate wilt diseases in some tested tomato plants. A deeper understanding of the effects of plant pathogens and their metabolites on the biocontrol bacterium’s metabolite production and survival might lead to more efficient control measures for the wilt diseases caused by R. solanacearum.

Consequently, in this work, we first explored the growth of B. amyloliquefaciens BNC5 in the presence of substances released from R. solanacearum. We then investigated how B. amyloliquefaciens BNC5 metabolically responded to these secretions using a liquid chromatography (LC)–mass spectrometry (MS)-based untargeted metabolomics approach. This quantitative technique allows for the simultaneous and unbiased comparison of multiple classes of metabolite levels across different sample groups. Subsequently, the transcript levels of genes related to the R. solanacearum secretion-induced changes in metabolites were quantified using quantitative real-time polymerase chain reaction (qRT-PCR). Finally, the antimicrobial activities against R. solanacearum of the isolated altered metabolites, as well as the extracts of B. amyloliquefaciens grown in the presence and absence of R. solanacearum secretions, were evaluated to correlate their activities.

Materials and Methods

Chemicals

Peptone (CAS 91079-38-8) was sourced from Merck (Rahway, New Jersey, USA) and HiMedia (Maharashtra, India); glucose from Kemaus (Cherrybrook, NSW, Australia); and casamino acid from Bio Basic (Toronto, Canada). For metabolite analysis, CHCl₃ was obtained from RCI Labscan (Bangkok, Thailand); MeOH, water, and ammonium hydroxide (25%) from Merck; isopropanol from Honeywell (Charlotte, North Carolina, USA); and formic acid from Sigma-Aldrich (Milwaukee, Wisconsin, USA). All chemical standards, including fengycins, iturins, surfactins, phosphatidylethanolamine (PE), and phosphatidylglycerol (PG), were obtained from Sigma-Aldrich (Milwaukee, Wisconsin, USA). All chemicals were of reagent grade, LC grade, or higher, and were used without further purification. For large-scale extraction and purification, methanol and ethyl acetate were obtained from RCI Labscan and were distilled prior to use. Anhydrous magnesium sulfate was purchased from Daejung (Gyeonggi-do, Korea), and SiliaBond C18 (17%) from SiliCycle (Quebec, Canada).

Bacterial Strains and Cultivation Conditions

B. amyloliquefaciens BNC5 was isolated from sandy loam soil with a pH of 6.5 in the Sob-Sai watershed management area, Wang Pha district, Nan province, Thailand, and identified through 16S rRNA and genome sequencing. R. solanacearum race 1 was obtained from the Plant Protection Research and Development Office, Department of Agriculture, Ministry of Agriculture and Cooperatives, Thailand. Bacteria were grown aerobically at 150 or 200 rpm and 30 °C in casamino acid–peptone–glucose media (CPG; 1 g L^–1 casamino acid, 10 g L^–1 peptone, and 5 g L^–1 glucose). All bacterial cultures were incubated overnight and then diluted 50-fold into fresh CPG media, with or without R. solanacearum supernatant (RSsup), before further cultivation to the specified duration.

The RSsup was prepared by centrifuging the 24 h bacterial cultures at 8960g, 10 °C for 30 min, followed by sterile filtration through a 0.2 μm membrane filter, and stored at −20 °C prior to use. To construct growth curves, B. amyloliquefaciens was diluted in the specified media containing the indicated concentration of RSsup (N = 3 per group; 20 mL). Samples were collected at 3, 6, 9, 12, 15, and 24 h after incubation to measure the optical density at 600 nm (OD₆₀₀). For untargeted metabolomics and qRT-PCR analyses, B. amyloliquefaciens was cultivated in CPG media with or without 50% (v/v) RSsup (20 mL) for 12 h before harvesting. Larger cultures (150 mL in each cultured bottle) were cultivated similarly before being combined and harvested for metabolite extraction and isolation.

Metabolite Extraction, LC–MS, and LC–MS/MS Analyses

B. amyloliquefaciens was incubated for 12 h with 50% RSsup or without RSsup, along with no-bacteria RSsup controls (i.e., 50% (v/v) RSsup in CPG medium without B. amyloliquefaciens). Metabolites were then extracted and analyzed using LC–MS, and LC–MS/MS on an UltiMate DGP-3600SD LC coupled to a Bruker MicrOTOF Q-II MS instrument, as previously described with some modifications. The modifications for metabolite extraction included the following: 1 mL of cultures (instead of 700 μL) was collected for OD₆₀₀ measurement. Centrifugation conditions were adjusted to 8960g, 10 °C for 30 min (instead of 4472g, 4 °C for 15 min), to separate bacterial cells from the supernatant. Additionally, CPG medium replaced nutrient broth medium for cell resuspension and washing, and centrifugation was performed at 1500g, 10 °C for 5 min (instead of 1000g, 4 °C for 5 min), to separate the organic from the aqueous layers. Data were collected using mass ranges of 100–1600 Da for LC–MS and 50–1600 Da for LC–MS/MS.

LC–MS Untargeted Data Analysis

The total ion chromatograms were obtained in two independent sets of triplicates for each sample group. For intracellular pellet samples, the chromatograms for each ion mode were obtained for both RSsup-exposed and unexposed B. amyloliquefaciens sample groups, totaling 12 chromatograms. For extracellular supernatant samples, a no-bacteria RSsup control group was also included, resulting in a total of 18 chromatograms. These intracellular and extracellular data were then subjected to comparative analyses separately between the exposed and unexposed B. amyloliquefaciens groups, as previously described. However, to allow for a more thorough investigation of metabolite changes, some cutoff filters for depicting metabolite ions with varying levels following exposure to RSsup were revised. The minimum fold change was adjusted to ≥3 (instead of ≥4), and the minimum integrated mass ion intensity (MSII) in the increased sample groups was lowered to 10,000 (instead of 30,000). Additionally, to eliminate changes in extracellular samples resulting from existing metabolites in RSsup, additional filters were applied. Elevated metabolite ions in the RSsup-exposed B. amyloliquefaciens group had to meet the same filters when compared to the no-bacteria RSsup control group. Metabolite ions (defined by m/z and retention time) were included in the final list of altered metabolite ions only if they met the criteria in both sets of experiments.

Structural Characterization of Differential Metabolites

Metabolite ions in the final list were manually characterized based on accurate mass measurements, retention times, and tandem mass spectra. To further confirm metabolite structures, supernatants from 12 h cultures of B. amyloliquefaciens exposed to RSsup were extracted with ethyl acetate and fractionated as described in the Metabolite Extraction and Enrichment section below. The macrolactin A fraction was analyzed by ¹H nuclear magnetic resonance (NMR) spectroscopy using a Jeol JNM-ECZ500R/S1 500 MHz NMR spectrometer (Jeol, Tokyo, Japan). Fractions containing novel metabolites, macrolactins BA1 and BA5, were further examined by ¹H, ¹³C, correlation spectroscopy (COSY), and heteronuclear single quantum coherence (HSQC) NMR spectroscopy using a Bruker Avance Neo 600 MHz NMR spectrometer equipped with a cryoprobe (Bruker, Ettlingen, Germany).

Macrolactin A

¹H NMR (CD₃OD, 500 MHz): δ_H 7.18 (1H, dd, J = 15.2, 11.3 Hz, H3), 6.60 (1H, t, J = 11.4 Hz, H9), 6.53 (1H, dd, J = 15.2, 11.1 Hz, H17), 6.17–5.98 (4H, m, H2, H4, H10, and H18), 5.71 (1H, dd, J = 15.1, 5.8 Hz, H16), 5.61 (1H, dt, J = 14.5, 7.0 Hz, H19), 5.55–5.48 (3H, m, H5, H8, and H11), 5.01–4.93 (1H, m, H23), 4.60 (2H, br s, 7-OH and 15-OH), 4.30–4.18 (2H, m, H7 and H15), 3.85–3.77 (1H, m, H13), 3.62 (1H, br s, 13-OH), 2.50–2.24 (6H, m, H6, H12, and H14), 2.21–2.00 (2H, m, H20), 1.60–1.53 (2H, m, H22), 1.51–1.40 (2H, m, H21), 1.22 (3H, d, J = 6.3 Hz, H24). The obtained spectrum aligns with previously reported data.

Macrolactin BA1

¹H NMR (CD₃OD, 600 MHz): δ_H 7.26–7.20 (1H, m, H3), 6.64 (1H, t, J = 11.5 Hz, H9), 6.57 (1H, dd, J = 15.2, 11.0 Hz, H17), 6.22–5.96 (5H, m, H2, H4, H10, H11, H12, and H18), 5.76 (1H, dd, J = 15.1, 5.9 Hz, H16), 5.66 (1H, dt, J = 14.6, 7.0 Hz, H19), 5.60–5.51 (3H, m, H5, H8, and H13), 5.07–4.98 (1H, m, H23), 4.59 (2H, br s, 7-OH and 15-OH), 4.29 (2H, dq, J = 26.9, 6.3 Hz, H7 and H15), 2.54–2.38 (4H, m, H6 and H14), 2.23–2.17 (2H, m, H20), 1.62 (2H, ddd, J = 7.4, 5.5, 2.9 Hz, H22), 1.54–1.46 (2H, m, H21), 1.26 (3H, d, J = 6.2 Hz, H24). ¹³C NMR (CD₃OD, 151 MHz): δ_C 168.0 (C1 = O), 145.0 (C9), 142.2 (C10 and C11), 137.5 (C16), 135.2 (C19), 135.1 (C5), 131.7–131.2 (C2, C4, C12 and C18), 130.3 (C3), 128.4 (C13), 126.0 (C17), 118.0 (C8), 101.0 (C23), 72.3–72.2 (C7 and C15), 42.9 (C14), 36.0 (C6), 33.0 (C20), 27.2 (C22), 25.7 (C21), 14.4 (C24).

Macrolactin BA5

¹H NMR (CD₃OD, 600 MHz): δ_H 7.26 (1H, d, J = 8.3 Hz, H3), 6.68–6.59 (1H, m, H4), 6.24–6.03 (2H, m, H2 and H9), 5.78–5.65 (1H, m, H10), 5.56 (2H, td, J = 16.5, 15.9, 9.7 Hz, H5 and H11), 5.41–5.32 (1H, m, H8), 4.92 (1H, d, H1′), 4.59 (6H, br s, OH), 4.34–4.28 (1H, m, H13), 4.03–3.97 (2H, m, H15 and H23), 3.81–3.77 (1H, m, H7), 3.67 (2H, ddd, J = 7.2, 3.7, 2.2 Hz, H6′), 3.45–3.41 (2H, m, H2′ and H5′), 3.37–3.32 (2H, m, H3′ and H4′), 2.50–2.25 (4H, m, H6 and H12), 1.63–1.57 (8H, m, H14, H16, H22 and H24), 1.36–1.28 (12H, m, H17–H21 and H25), 0.91 (3H, t, H26). ¹³C NMR (CD₃OD, 151 MHz): δ_C 170.3 (C1 = O), 137.6 (C10), 131.2 (C2 and C9), 127.9 (C3), 103.4 (C13), 78.2 (C3′ and C4′), 75.4 (C2′ and C5′), 72.4 (C1′), 71.1 (C6′), 69.5 (C15 and C23), 69.1 (C7), 54.8 (C5 and C11), 34.8 (C6 and C12), 32.7 (C18–C20), 30.6 (C21), 30.5 (C14), 27.1 (C16 and C17), 26.0 (C22 and C24), 23.6 (C25), 14.4 (C26).

Assessment of Potential Macrolactin Enzymatic Conversion Activities in RSsup

Supernatant from 12 h cultures (20 mL each) of B. amyloliquefaciens without RSsup (BNC5sup) was collected by centrifugation at 8960g, 10 °C for 30 min, sterilized by filtration through a 0.2 μm membrane filter, and stored at −20 °C before use. Three sample groups (N = 3) were then set up as follows: (i) a mixture of BNC5sup (10 mL) and RSsup (10 mL), (ii) BNC5sup (20 mL), and (iii) RSsup (20 mL), and incubated at 200 rpm, 30 °C for 12 h. Subsequently, a control was prepared by combining BNC5sup (10 mL) from group (ii) and RSsup (10 mL) from group (iii). This control was immediately extracted and analyzed by LC–MS. Additionally, a mixture of BNC5sup (10 mL) and RSsup (10 mL) from group (i) was extracted and analyzed, according to the protocols outlined in the Metabolite Extraction, LC–MS, and LC–MS/MS Analyses section above. However, in this case, the extracts were reconstituted in 100 μL (instead of 200 μL) of chloroform prior to injection. The levels of each macrolactin ion across all samples were then quantitated using the raw MSII obtained from the XCMS program.

qRT-PCR Analysis

The cell pellets from 1.5 and 1 mL of each RSsup-exposed and unexposed B. amyloliquefaciens culture (N = 4) were collected by centrifugation at 16,000g for 1 min, respectively. The total RNA was immediately isolated using the Presto Mini RNA Bacteria Kit (Geneaid, New Taipei City, Taiwan) with in-column DNase I digestion, according to the manufacturer’s instructions. Subsequently, 800 ng of total RNA was converted to cDNA, and the resulting cDNA solution was diluted 16-fold for qRT-PCR analysis using a Bio-Rad CFX96 Touch Real-Time PCR detection system, as previously described. The list of primer sequences (Bionics, Seoul, South Korea) used for amplifying the target and internal control (gyrA and recA) genes , is provided in the Supporting Information (SI), Table S1.

Metabolite Extraction and Enrichment

Supernatant from 12 h cultures (4.5 L) of B. amyloliquefaciens with or without RSsup was collected by centrifugation at 9000g, 10 °C for 30 min. The resulting supernatant samples, along with the no-bacteria RSsup control sample (1.5 L), were then extracted twice with ethyl acetate (1:1 (v/v) each time). The organic layers were dried over anhydrous MgSO₄, filtered, and concentrated in vacuo at ≤55 °C to yield rust-orange oils (488 and 422 mg), and bright orange oil (76 mg) for B. amyloliquefaciens with RSsup (BNC5RSe), without RSsup (BNC5e), and no-bacteria RSsup (RSsup-e) samples, respectively.

To enrich the metabolites of interest, BNC5RSe and BNC5e extracts were fractionated using flash reversed-phase column chromatography with a MeOH-H₂O gradient. For structural characterization by NMR spectroscopy, metabolites were isolated from a 6 L culture using the following gradient: 40% MeOH (1 L), 45% (1 L), 50% (1.5 L), and then increasing in 2% increments to 60% (1.5 L), followed by 65% (1 L) and 70% (1 L). Each fraction was analyzed by electrospray ionization (ESI)-MS in positive ion mode using a Bruker MicrOTOF Q-II MS instrument. Fractions containing macrolactin A (m/z 403), BA1 (m/z 385), and BA5 (m/z 597) (all as [M + H]⁺ ions) were pooled and concentrated in vacuo at ≤40 °C, yielding yellow oils of macrolactin A (2.6 mg from a 6 L culture), and BA1 (1.7 mg) and BA5 (1.6 mg) from a 12 L culture.

For metabolite enrichment for antimicrobial assays, the gradient started at 40% MeOH (400 mL) and increased to 100% in 10% increments (800 mL each step). In the BNC5RSe extract, fractions containing macrolactins that increased in response to RSsup (m/z 425, 597, and 611) and surfactins (m/z 1008, 1022, 1036, and 1050) were pooled and concentrated in vacuo at ≤40 °C, yielding rust-orange oil (26.3 mg) and white solid (65.5 mg), respectively. A similar approach was used for BNC5e extract, isolating macrolactins that decreased in response to RSsup (m/z 511) and iturins (m/z 1043), yielding rust-orange oil (2.9 mg) and dark-orange solid (2.7 mg), respectively. All final fractions were analyzed by ESI-MS in positive ion mode before use.

Agar Well Diffusion Assay

The protocols for the agar well diffusion assay were adapted from those previously reported in the literature. Briefly, an R. solanacearum culture with an OD₆₀₀ of 0.2 (500 μL) was spread onto CPG solid agar plates (20 mL per 10 cm-diameter dish). A 9.5 mm-diameter cylindrical well was created in each quadrant of the plate, followed by the addition of a 10 mg mL^–1 solution (100 μL) of the indicated extract to each well. The plates were incubated at 30 °C for 24 h. The diameter of the clear zone was then measured in four different positions, averaged, and subtracted from the well diameter to obtain the inhibition zone value.

Determination of Minimum Inhibitory Concentration (MIC) and Minimum Bactericidal Concentration (MBC)

MIC and MBC values for the BNC5RSe (N = 5), BNC5e (N = 5), and RSsup-e (N = 3) extracts (independent extracts for each replicate), as well as the metabolitesmacrolactins from BNC5RSe (N = 8), macrolactins from BNC5e (N = 4), iturins (N = 8), and surfactins (N = 8)were determined by a broth dilution assay similar to previously described protocols. , Briefly, a 10% DMSO solution in CPG medium (50 μL) was added to each well in columns 2–12 of a deep 96-well plate. A 20 mg mL^–1 solution of extract or a 2048 μg mL^–1 solution of metabolite in a 10% DMSO solution in CPG medium (100 μL) was prepared in each well in column 1, and 50 μL was used for a twofold serial dilution across columns 1–11 (i.e., the final volume of 50 μL in each well). Thereafter, 50 μL of an R. solanacearum culture in CPG medium with an OD₆₀₀ of 2.5 × 10^–4 (i.e., the density close to a 0.5 McFarland standard typically employed) was added to each well and incubated with shaking for 26–30 h until OD₆₀₀ reached 0.6–1.0. The plate was then visually inspected, and the MIC value was reported as the lowest concentration of each extract or metabolite that resulted in a clear culture in the well.

To obtain the MBC values, each culture (50 μL) from the wells with a concentration of extract or metabolite double the MIC value and above was spread on a CPG solid agar plate and incubated at 30 °C for 42 h. The lowest concentration that showed no growth of R. solanacearum was then reported as the MBC value.

Preparation of Tomato Seedlings

A tomato variety susceptible to R. solanacearum infection (Sugarmoon, Thepmongkhol Seed Co., Ltd., Thailand) was used in this study. Tomato seeds were soaked in water at 60 °C for 12 h, washed with sterile water, and incubated in wet tissue paper for 2–3 days. The germinated seeds were then transferred to a transplant tray, placing one seed per cell, and watered every 2 days. Two-week-old seedlings were used for the biological control experiments.

For each treatment, 20 pots filled with sterile peat moss were prepared. To establish group 2–6 seedlings, a 3-day-old CPG liquid culture (10⁷ CFU mL^–1, 10 mL) of R. solanacearum was poured around the seedling’s stem base on the day of transplantation.

Evaluation of the Bacterial Wilt-Controlling Ability of B. amyloliquefaciens Extracts and Cultures in Tomato Pot Experiments

Six experimental groups were established to assess the bacterial wilt-controlling ability of B. amyloliquefaciens extracts and cultures. Group 1 (control) received neither R. solanacearum nor antagonistic bacteria, with water (50 mL) applied per seedling. Groups 2–6 were inoculated with R. solanacearum, and upon the appearance of early infection symptoms, treatment (10 mL) was applied around the stem base. Group 2 (infected control) received no treatment, while group 3 (vehicle control) received ethanol, the solvent for BNC5e. Group 4 (extract treatment) was treated with 780 μg mL^–1 BNC5e in ethanol, group 5 (culture treatment) received B. amyloliquefaciens BNC5 culture at 10⁷ CFU mL^–1, and group 6 (combined treatment) was first treated with 780 μg mL^–1 BNC5e in ethanol, followed by B. amyloliquefaciens BNC5 culture (10⁷ CFU mL^–1) 2 days later.

All potted tomato seedlings were watered regularly throughout the experiment. The number of plants exhibiting wilting symptoms was recorded daily for 30 days. To confirm R. solanacearum infection, the freshly cut stems of wilted plants were placed in distilled water to observe bacterial ooze.

Results

Effects of R. solanacearum Secretions on the Growth of B. amyloliquefaciens

As a biocontrol bacterium, B. amyloliquefaciens BNC5 is expected to inhibit the growth of the plant pathogen R. solanacearum. However, it remains unclear how R. solanacearum or its secreted substances might, in turn, affect the growth or behavior of B. amyloliquefaciens. To investigate this, we monitored the growth of B. amyloliquefaciens BNC5 (20 mL) in CPG medium containing varying concentrations of RSsup, a solution containing substances released from a 24 h culture of R. solanacearum. CPG medium was selected, because it is a standard medium for culturing R. solanacearum and supports the growth of B. amyloliquefaciens BNC5, allowing for the study of their interactions under similar conditions. Growth was measured over 24 h using OD₆₀₀.

B. amyloliquefaciens grew in all tested RSsup concentrations (25–100% v/v), although at lower cell densities compared to the control without RSsup (0%) (Figure a). At lower RSsup concentrations (25 and 50% v/v), growth was similar to the control up to 6 h. However, after 9 h, growth began to decrease with increasing RSsup concentrations. At higher RSsup concentrations (75 and 100% v/v), growth inhibition was noticeable even at 6 and 3 h, respectively, with particularly low cell density at 100% RSsup. In addition, the growth curves displayed a distinct biphasic trend between 6 and 15 h, characterized by an initial decrease followed by an increase at RSsup concentrations of 25–75%, suggesting a potential adaptive response to the presence of RSsup. Overall, B. amyloliquefaciens exhibited reduced growth with increasing RSsup concentrations, with evidence of possible adaptation to RSsup at intermediate levels.

1.

Growth curves of Bacillus amyloliquefaciens BNC5. (a) Growth curves of B. amyloliquefaciens in CPG with concentrations of R. solanacearum supernatant at 0% (blue diamonds with solid line), 25% (green squares with dotted line), 50% (purple triangles with dashed line), 75% (orange crosses with dot–dash line), and 100% (v/v; dark red asterisks with dash–dot–dot line). (b) Growth curves of B. amyloliquefaciens in CPG at 1.25× (green squares with dotted line), 1× (normal strength; blue diamonds with solid line), 0.75× (purple triangles with dashed line), 0.50× (orange crosses with dot–dash line), and 0.25× (dark red asterisks with dash–dot–dot line) strengths. The data represents the average values ± standard errors of the mean from three independent cultures for each concentration.

While the growth inhibition suggested that substances in RSsup might negatively affect B. amyloliquefaciens, we considered the possibility that reduced growth could also be due to decreased nutrient availability in the CPG medium containing RSsup. To explore this, we monitored the growth of B. amyloliquefaciens in CPG medium at different strengths (0.25–1.25× the standard strength) over 24 h. The results indicated that B. amyloliquefaciens grew better with increasing medium strength (Figure b). Notably, the growth differences were observable from as early as 3 h and became clear by 6 h. In contrast, the differences in growth at 25% and 50% RSsup concentrations compared to the unexposed control became apparent only after 9 h. Together, although these findings could not precisely determine the contribution of nutrient deficiency to growth, they suggest that the lower cell densities, particularly in 25% and 50% RSsup, were mainly due to inhibitory substances in RSsup rather than nutrient deficiency.

Untargeted Metabolomics of B. amyloliquefaciens BNC5 upon Exposure to R. solanacearum Secretions

To investigate whether R. solanacearum secretions affect other behaviors (e.g., metabolic profiles) of B. amyloliquefaciens, untargeted metabolomics analyses were conducted on both the extracellular (supernatant) and intracellular (cell pellets) components of B. amyloliquefaciens BNC5 cultures. The cultures were exposed to 50% RSsup for 12 h, along with unexposed controls and no-bacteria controls (consisting of 50% RSsup in CPG medium without B. amyloliquefaciens). These conditions were chosen, because the nutrient deficiency had minimal impact on the growth of B. amyloliquefaciens at this RSsup concentration. Nonetheless, the concentration was sufficient to induce changes in growth behavior, as evidenced by lower cell densities compared to the unexposed control, which became apparent at the 12 h mark when the cells were in the stationary phase (Figure a). The extracellular and intracellular lipophilic metabolites from each sample group were separately extracted using a 2:1 (v/v) ratio of chloroform to methanol. The extracts were then concentrated and analyzed using LC–MS-based platforms for untargeted metabolomics analysis.

To identify metabolite ions associated with B. amyloliquefaciens BNC5′s response to substances secreted by R. solanacearum, the relative levels of metabolite ions in the LC–MS chromatograms of RSsup-exposed, unexposed, and no-bacteria control samples were identified and quantified using the XCMS program. The data were normalized by the OD₆₀₀ value of each sample to account for differences in cell numbers, and fold changes were calculated. Given that the cell numbers in B. amyloliquefaciens cultures exposed to 50% RSsup could be approximately 50–70% of those in unexposed cultures at 12 h, the criteria for identifying significant changes in metabolite ion levels were set to a threefold increase or decrease with statistical significance (determined by Student’s t-test, p < 0.05). Furthermore, to ensure that the observed increases in metabolite levels in the RSsup-exposed samples were not merely due to the presence of metabolites in the RSsup itself, metabolites were considered upregulated in response to R. solanacearum secretions only if their levels also showed a threefold or greater increase compared to the no-bacteria RSsup control group.

Using these criteria, 48 extracellular and 37 intracellular metabolite ions were found to increase, while 14 extracellular and 24 intracellular metabolite ions were found to decrease in B. amyloliquefaciens exposed to RSsup in positive ion mode. In negative ion mode, one extracellular and seven intracellular metabolite ions increased, whereas one extracellular and six intracellular metabolite ions decreased, compared to the unexposed and no-bacteria controls (SI, Figure S1 and Tables S2–S9).

Characterization of Metabolite Ions Associated with Response to R. solanacearum Secretions

Following the identification of differential metabolites potentially associated with the response to R. solanacearum secretions, we undertook a manual structural characterization of some of these metabolites. This was achieved by integrating potential molecular formulas derived from accurate mass measurements, ion fragment data from tandem mass spectrometry, and lipophilicity matched with known metabolites of similar structures, inferred from retention time data. This approach enabled us to propose the structures for differential positively charged ions of three known compoundsmacrolactin A, macrolactin M, and 7-O-malonyl macrolactin Aas well as seven novel macrolactin derivatives (BA1–7) and one related polyketide derivative (BAP1), found in extracellular samples (Table , Figure , and SI, Figure S2).

1. Relative Levels of Metabolites Identified in the Untargeted Metabolomics Analyses of Bacillus amyloliquefaciens BNC5 Exposed to Ralstonia solanacearum Supernatant (RSsup) for 12 h and in the Macrolactin Enzymatic Conversion Activity Assay between B. amyloliquefaciens Supernatant (BNC5sup) and RSsup, Compared to Their Respective Controls .

metabolite class and derivative	ion	m/z	RT (min)	BNC5RSm/BNC5m ,	BNC5sRSs/ (BNC5s+RSs) ,
Elevated metabolites in RSsup-treated extracellular samples
Macrolactin
BA1	[M + Na]+	407.2207	25.4	4.7	0.9
A	[M + Na]+	425.2328	20.7	4.6	1.1
BA2	[M + Na]+	427.2463	22.3	3.3	ND
M	[M + Na]+	439.2485	22.7	5.5	1.3
BA3	[M + Na]+	451.2477	22.8	4.5	ND
BA4	[M + Na2 – H]+	463.2404	22.1	12.8	ND
BA5	[M + H]+	597.3580	23.3	3.8	1.1
BA6	[M + H]+	611.3451	20.2	3.7	0.7
Other polyketides
BAP1	[M + K]+	377.2432	24.9	3.6	ND
Decreased metabolites in RSsup-treated extracellular samples
Macrolactin
BA7	[M + Na]+	605.3458	26.0	3.2	1.7
A, 7-O-malonyl	[M + Na]+	511.2344	21.0	2.1	1.1
	[2 M + Na]+	999.4720	21.0	5.1	ND
Other extracellular metabolites (fold <3 or p > 0.05)
Macrolactin
A, 7-O-succinyl	[M + Na]+	525.2518	21.4	1.6
B	[M + Na]+	587.2856	18.9	2.5
Iturin
14:0	[M + H]+	1043.5543	23.4	2.0
15:0	[M + H]+	1057.5663	24.5	2.0
Fengycin
16:0	[M + 2H]2+	732.4103	27.7	2.8
17:0	[M + 2H]2+	739.4201	28.4	1.3
Surfactin
14:0	[M + H]+	1022.6823	37.8	1.4
15:0	[M + H]+	1036.6996	38.4	1.2
Other intracellular metabolites (fold <3 or p > 0.05)
Phosphatidylethanolamine (PE)
15:0/15:0	[M + H]+	664.4989	40.4	1.7
15:0/17:0	[M + H]+	692.5313	41.5	1.9
Phosphatidylglycerol (PG)
15:0/15:0	[M – H]−	693.4719	37.2	1.7
15:0/17:0	[M – H]−	721.5034	38.9	1.5

^aAbbreviations: empty cell, not investigated; m/z, mass-to-charge ratio; ND, not detected; RT, retention time.

^bBNC5RSm/BNC5m value represents the ratio of the average mass ion intensity of RSsup-exposed samples (BNC5RSm) to that of the unexposed control (BNC5m) from untargeted metabolomics analyses, indicated in regular font. The opposite ratio (BNC5m/BNC5RSm) is indicated in underlined italics.

^cStudent’s t test.

^d p < 0.05, N = 3.

^e p < 0.01, N = 3.

^f p < 0.005, N = 3.

^gBNC5sRSs/(BNC5s+RSs) value represents the ratio of the average mass ion intensity of the RSsup and BNC5sup mixture with 12 h incubation (BNC5sRSs) to that of the no-incubation control (BNC5s+RSs) from the macrolactin enzymatic conversion activities assay.

^hAlthough this ion did not pass through the cutoff filter in the untargeted metabolomics analyses, it is included here to verify the changes observed in the [2 M + Na]⁺ ion.

2.

Structural characterization of differential ions as macrolactins using tandem mass spectrometry. MS/MS spectra of ions with increased levels (m/z 407, 425, 427, 451, 463, 597, 611) and decreased levels (m/z 605, 999) following exposure to Ralstonia solanacearum supernatant were analyzed, along with accurate mass and retention times. This led to the identification of the ions as macrolactins: (a) BA1, (b) A, (c) BA2, (d) BA3, (e) BA4, (f) BA5, (g) BA6, (h) BA7, and (i) 7-O-malonyl macrolactin A, respectively. Precursor ions are indicated by diamonds in the MS/MS spectra. Proposed fragments are highlighted with green dashed lines, and their corresponding m/z values are shown on the chemical structures. For a comprehensive list of differential ions and their identification, refer to SI, Tables S2–S9, and Figure S2.

The proposed structures of macrolactins A, BA1, and BA5 were further validated by NMR spectroscopy of the corresponding isolated fractions (SI, Figures S3–S5). The ¹H NMR spectrum of macrolactin A was consistent with previously reported data. Although BA1 and BA5 were obtained in limited quantities and contained some impurities, their ¹H, ¹³C, COSY, and HSQC NMR spectra, when compared to that of macrolactin A, provided additional support for their proposed structures.

For other novel macrolactins that could not be isolated in sufficient quantity and purity for characterization by NMR spectroscopy, we proposed structural elements that could not be determined from tandem mass spectra based on the known positions of these groups in established macrolactins. These elements included the positions of hydroxyl groups, double bonds, and the O-linkage at C-7 on the 24-membered lactone ring.

Specifically, the proposed structure of BA2 was based on modifications of macrolactin A, with the reduction of a double bond at C-16, akin to macrolactins F and N. The structure of BA7 was analogous to 7-O-malonyl macrolactin A, although the precise positions of the double bonds and the hydroxyl group on the ester side chain were tentatively assigned. For BA3 and BA4, structural assignments were based on macrolactin M, with BA3 incorporating an n-propyl group at C-23, resembling macrolactin BA5, and a ketone group replacing a hydroxyl group at C-15, similar to macrolactin E. BA4 was characterized by the reduction of a double bond in the ring at C-16, akin to macrolactin BA2. Lastly, the structure of BA6 was based on macrolactins P, Y, and BA5, featuring an O-β-d-glucosyl linkage at C-7, the addition of water across the double bond at C-18, and an n-propyl group at C-23.

Untargeted metabolomics analyses revealed that the levels of macrolactins A, M, BA1–6, and polyketides BAP1 were elevated in extracellular B. amyloliquefaciens samples exposed to RSsup compared to unexposed samples, while the levels of 7-O-malonyl macrolactin A and macrolactin BA7 decreased (Table ). Using accurate mass and retention time data, we also screened our data sets for some known metabolites, including 7-O-succinyl macrolactin A, macrolactin B, iturins, fengycins, and surfactins in the extracellular fraction, as well as PE and PG in the intracellular fraction. The identification of iturins, fengycins, surfactins, PE, and PG was confirmed by comparing their retention times and tandem mass spectra with those of reference chemical standards. The levels of these metabolites exhibited changes of less than threefold or were statistically insignificant after exposure to R. solanacearum substances in RSsup, consistent with the findings from untargeted metabolomics analyses (Table and SI, Table S10). Overall, the data suggest that R. solanacearum secretions induce changes in the macrolactin composition of B. amyloliquefaciens.

Assessment of Potential Macrolactin Enzymatic Conversion Activities in RSsup

To understand the mechanisms underlying the changes in macrolactin composition observed in the untargeted metabolomics analyses, we examined the structures of the two macrolactins with decreased levels upon exposure to R. solanacearum substances in RSsup: 7-O-malonyl macrolactin A and macrolactin BA7. Both macrolactins contain an ester bond at the 7-hydroxyl group of macrolactin A (Figures and ). In contrast, the levels of macrolactin A increased in the presence of RSsup. This raised the possibility that RSsup might not alter gene regulation in B. amyloliquefaciens. Instead, it might contain substances, such as esterases, that cleave the ester bonds of macrolactin derivatives, converting them into macrolactin A post-transcriptionally.

3.

Known and potential biosynthetic pathways of macrolactin A and its derivatives in Bacillus amyloliquefaciens. − The biosynthesis of macrolactin A begins with malonyl-CoA and involves enzymes encoded by the macrolactin biosynthetic genes mln A-I. This compound can then be converted into various derivatives by enzymes produced by the bmmGT1/2/3 genes or potentially unknown tailoring enzymes. In the metabolomics and quantitative real-time polymerase chain reaction analyses, changes in macrolactin levels in response to exposure of B. amyloliquefaciens to Ralstonia solanacearum supernatant are indicated by red (increased), green (decreased), and blue (unchanged) letters. Abbreviations: ACP, acyl carrier protein; CoA, coenzyme A; NDP, nucleoside diphosphate.

To differentiate between an enzymatic conversion mechanism and gene regulation, we prepared sterilized-filtered B. amyloliquefaciens BNC5 supernatant (BNC5sup) from cultures grown under the same conditions as the unexposed controls. BNC5sup was expected to contain the same composition of macrolactins and other metabolites as in the unexposed controls. We then incubated BNC5sup with RSsup at a 1:1 (v/v) ratio for 12 h, mimicking a 50% RSsup exposure condition but without live B. amyloliquefaciens cells. The incubated mixture was extracted using a chloroform–methanol system and analyzed by LC–MS, similar to the untargeted metabolomics analyses, along with controls containing the same ratio of BNC5sup and RSsup but without incubation.

If enzymatic conversion in RSsup contributed to the observed changes, then the levels of 7-O-malonyl macrolactin A and macrolactin BA7 would decrease, while macrolactin A levels would increase in the RSsup-incubated samples compared to the unincubated samples. However, the LC–MS results showed only minimal, insignificant changes (fold change ≤ 1.7) in the levels of all macrolactins in the incubated samples compared to the unincubated samples (Table and SI, Table S11). These findings thus exclude enzymatic conversion by RSsup as a cause of the changes in macrolactin composition in B. amyloliquefaciens upon exposure to R. solanacearum secretions.

Expression Levels of Genes Involved in the Biosynthesis of Macrolactin A and Its Derivatives in B. amyloliquefaciens

To further explore the mechanisms behind alterations in macrolactin composition, we assessed the gene expression levels associated with macrolactin biosynthesis using qRT-PCR analysis. Previous studies have described only the biosynthetic genes for macrolactin A, specifically mln A-I, with mln B-H being part of the same gene cluster and regulated by the same operons (Figure ). It is hypothesized that macrolactin A is converted into various derivatives by other tailoring enzymes. However, only three such enzymes, bmmGT1–3, which are involved in the glycosylation of the hydroxyl group on the 24-membered lactone ring of macrolactin A, have been identified. −

Based on this information, we conducted qRT-PCR analysis on total RNA extracted from B. amyloliquefaciens BNC5 cultured under the same conditions as those in the untargeted metabolomics experiments. The target genes included mln A, mln B, mln I, and bmmGT1–3. Interestingly, among the macrolactin biosynthetic and known tailoring enzymes, only bmmGT3 showed a statistically significant change, with a modest 1.3-fold downregulation upon exposure to RSsup (Figure ). Although this change was opposite to what might be expected for an enzyme involved in glycosylation, the findings indicate altered regulation of macrolactin tailoring enzymes in B. amyloliquefaciens when exposed to R. solanacearum secretions.

4.

Relative transcript expression levels of genes associated with the biosynthesis of macrolactin A and its derivatives in Bacillus amyloliquefaciens. mRNA expression was quantified in B. amyloliquefaciens 12 h post-exposure to Ralstonia solanacearum supernatant (BNC5RSr; black bars) and displayed in the graph, normalized to the control group without exposure (BNC5r; white bars). Quantitative real-time polymerase chain reaction was carried out using the 2^–ΔΔCt method, with gyrA and recA transcripts serving as internal controls. Data from four independent experiments are presented as the mean ± standard error of the mean. Student’s t-test: *p < 0.05; **p < 0.01.

Antimicrobial Activity of B. amyloliquefaciens Extracts against R. solanacearum after Exposure to Its Secretions

Next, we assessed the antimicrobial activities of B. amyloliquefaciens BNC5 extracts against R. solanacearum following exposure to substances secreted by the pathogen. Specifically, we evaluated the ethyl acetate extracts of B. amyloliquefaciens BNC5 grown in the presence (BNC5RSe) and absence (BNC5e) of R. solanacearum secretions (RSsup) under the same conditions as the untargeted metabolomics analyses. These extracts, along with a no-bacteria control extract (RSsup-e) and a solvent control (DMSO), were tested for antimicrobial activity using an agar well diffusion assay. The results showed a slight decrease in the inhibition zone for BNC5RSe compared to BNC5e across three independent extract sets (Figure a,b). In contrast, RSsup-e and DMSO exhibited significantly smaller inhibition zones.

5.

Antimicrobial activities against Ralstonia solanacearum. (a, b) Agar well diffusion assay was performed using ethyl acetate extracts of Bacillus amyloliquefaciens exposed (BNC5RSe) and unexposed (BNC5e) to R. solanacearum supernatant (RSsup), alongside extracts of a no-bacteria RSsup control (RSsup-e) and a solvent control (DMSO). The results are presented in (a) a representative plate showing clear zones and (b) a table detailing the inhibition zones from three independent extract sets. (c) Minimum inhibitory concentration (MIC) and minimum bactericidal concentration (MBC) assays were conducted using similar extracts (N = 3), along with macrolactin (MLN) mixtures isolated from BNC5RSe (N = 8) and BNC5e (N = 5) (i.e., enriched in macrolactins with increased and decreased levels in the RSsup-exposed B. amyloliquefaciens samples compared to unexposed samples in the metabolomics experiment, respectively), as well as iturins (N = 8) and surfactins (N = 8).

To confirm these findings, we also determined the MIC and MBC of the extracts. The MIC and MBC values, ranked from lowest to highest, corresponded to BNC5e, BNC5RSe, and RSsup-e (Figure c). Some values are reported as ranges, because both values were detected at approximately the same frequencies from independent extracts. These results corroborated the agar well diffusion assay findings, indicating a slight reduction in antimicrobial activity of B. amyloliquefaciens BNC5 extracts after exposure to R. solanacearum secretions.

Antimicrobial Activities of Macrolactins with Altered Levels against R. solanacearum

We investigated why the antimicrobial activity of B. amyloliquefaciens BNC5 extracts decreased after exposure to R. solanacearum secretions (RSsup). To explore potential links with changes in macrolactin composition observed in metabolomics analyses, we evaluated the antimicrobial activities of macrolactins with altered levels, along with iturins and surfactins, by determining their MIC and MBC values.

Some macrolactins with increased and decreased levels upon exposure to RSsup were enriched from the BNC5RSe and BNC5e extracts, respectively, using flash reverse-phase column chromatography, along with iturins and surfactins. ESI-MS analysis confirmed that the macrolactins from the BNC5RSe fraction included macrolactins A, BA5, and BA6, whereas the BNC5e fraction contained 7-O-malonyl macrolactin A (SI, Figures S6 and S7). Both fractions, however, also contained some macrolactin fragments (m/z 329–399). Other macrolactins with varying levels were also detected in other chromatographic fractions but were excluded due to their relatively low intensities or because their intensities were lower than those of undesirable compounds. The iturins fraction contained Iturin 14:0, while the surfactins fraction comprised Surfactin 13:0, 14:0, 15:0, and 16:0 (SI, Figures S8 and S9).

The combined fractions were tested for their MIC and MBC values against R. solanacearum (Figure c). The MIC and MBC values indicate that macrolactins from BNC5e exhibited the highest antimicrobial activity, while macrolactins from BNC5RSe showed the lowest. The lipopeptides displayed intermediate activity, with surfactins being more effective than iturins. These findings suggest that the reduced antimicrobial activity of macrolactins observed when B. amyloliquefaciens was exposed to RSsup may partially explain the slight decline in the overall antimicrobial activity of B. amyloliquefaciens extract following exposure to R. solanacearum secretions.

Bacterial Wilt-Controlling Ability of B. amyloliquefaciens Extracts in Tomato Pot Experiments

The reduced anti-R. solanacearum activity of B. amyloliquefaciens after exposure to R. solanacearum secretions suggests that applying B. amyloliquefaciens extracts before live bacteria may enhance disease control. To test this hypothesis, we assessed the effectiveness of B. amyloliquefaciens BNC5 extract (BNC5e), live BNC5 culture, and a sequential treatment (BNC5e for 2 days followed by BNC5 culture) in mitigating tomato wilt caused by R. solanacearum in pot experiments. Infected tomato plants exhibited early bacterial wilt symptoms within 3 days, confirmed by bacterial ooze in freshly cut stems (SI, Figure S10). Without treatment, all plants in the infected control (group 2) and vehicle control (group 3) remained wilted. Treatment with BNC5e (group 4) reduced wilt incidence to 90%, while BNC5 culture (group 5) and the sequential treatment (group 6) further lowered wilt incidence to 75 and 70%, respectively (Figure ). These preliminary findings suggest that pretreating plants with B. amyloliquefaciens extract before introducing the live biocontrol agent may improve disease suppression.

6.

Biocontrol efficacy of Bacillus amyloliquefaciens extracts and cultures in suppressing tomato bacterial wilt caused by Ralstonia solanacearum in pot experiments. (a) Wilt incidence (%) in each treatment group. (b) Representative side-view images of tomato plants, highlighting healthy plants (groups 1 and 6) and those with typical wilt symptoms (groups 2–5). (c) Top-view images of tomato plants in each group (N = 20). Group 1: control; group 2: infected control; group 3: vehicle control; group 4: extract treatment; group 5: culture treatment; group 6: combined treatment.

Discussion

Biocontrol bacteria offer a promising solution to combat bacterial wilt, a devastating disease caused by R. solanacearum. We recently isolated B. amyloliquefaciens BNC5 for its potential efficacy against the pathogen. While B. amyloliquefaciens has shown promise in reducing bacterial wilt in pot experiments, , its effectiveness in heavily infected, real-world conditions remains uncertain. This study examines how B. amyloliquefaciens responds to R. solanacearum secretions and how these interactions may influence its biocontrol efficacy.

Untargeted metabolomics revealed shifts in macrolactin composition in B. amyloliquefaciens upon exposure to R. solanacearum secretions. Macrolactins, known for their antimicrobial properties against various bacteria, including R. solanacearum, ,,− can differ in antibacterial efficacy. For instance, agar well diffusion assays show that 7-O-malonyl macrolactin A is more active against R. solanacearum than macrolactin A. Consistently, macrolactins from the BNC5e fraction, rich in 7-O-malonyl macrolactin A, exhibited stronger antibacterial activity against R. solanacearum than those from the BNC5RSe fraction, which contained macrolactin A, BA5, and BA6. This implies that BA5 and BA6 may have lower activity relative to 7-O-malonyl macrolactin A.

Our metabolomics data revealed a downregulation of 7-O-malonyl macrolactin A in response to R. solanacearum secretions, potentially as a defensive strategy by the pathogen to counteract B. amyloliquefaciens. This decrease, likely due to reduced tailoring enzyme activity, may lead to macrolactin A accumulationa precursor for other macrolactins (see Figure ). Consequently, other macrolactins may be upregulated in the presence of R. solanacearum secretions. Alternatively, the upregulation of macrolactins A and BA1–6 could be compensatory mechanisms by B. amyloliquefaciens to counteract the reduced 7-O-malonyl macrolactin A.

Previous research described the stimulation of lipopeptides (fengycins, iturins, and surfactins) in Bacillus velezensis, a biocontrol bacterium in the operational group B. amyloliquefaciens, when cultured with R. solanacearum. However, our study detected no lipopeptide changes (Table ). This discrepancy may stem from differences in methods and culture conditions.

Earlier studies used B. velezensis cultured in wells in the center of CPG agar plates, with or without R. solanacearum spread on top, and sampled from the inhibition zone to measure lipopeptide. In contrast, we used liquid cultures, analyzing the entire culture instead of localized inhibition zones, where effects against R. solanacearum might be concentrated. Additionally, surfactins, which aid biofilm formation and motility, , are likely more active on solid agar than in liquid cultures, explaining their stimulation in prior studies.

Furthermore, earlier experiments involved direct interaction between both bacterial species on agar, while our study exposed Bacillus to R. solanacearum secretions without live cells. This suggests that R. solanacearum cells or specific secretions in their presence may be required to stimulate lipopeptides. Alternatively, the metabolic responses of these related biocontrol bacteria to R. solanacearum might differ.

The qRT-PCR results showed no significant changes in the transcript levels of macrolactin biosynthetic genes (mlnA, mlnB, and mlnI). While higher production of macrolactins would benefit B. amyloliquefaciens, this result is unsurprising, as macrolactin derivatives exhibit both increased and decreased levels upon exposure to R. solanacearum secretions.

The only gene with altered transcript levels was bmmGT3, which encodes a tailoring enzyme that glycosylates macrolactins. In vitro, bmmGT3 converts macrolactin A into derivatives, including 7-O-glucosyl macrolactin A (macrolactin B), as well as 13- and 15-O-glucosyl macrolactin A. In this study, two 7-O-glucosyl derivatives, macrolactins BA5 and BA6, increased following exposure to R. solanacearum secretions, despite bmmGT3 being downregulated. Since bmmGT3 has nearly 10-fold lower activity than bmmGT2 and reduced regioselectivity for 7-OH glycosylation compared to bmmGT1, , its downregulation may have increased macrolactin A availability. If macrolactin glycosylation is catalyzed by bmmGT1, bmmGT2, or other unidentified tailoring enzymes instead of bmmGT3, then it likely leads to more efficient production and higher levels of 7-O-glucosyl macrolactins, consistent with the increased abundance of macrolactins BA5 and BA6 observed in our metabolomics analysis.

The metabolite isolates’ MIC activities against R. solanacearum ranked as follows: macrolactins from BNC5e, surfactins, iturins, and macrolactins from BNC5RSe. While no prior MIC values for macrolactins against R. solanacearum exist, previous studies reported MIC values of 128 μg mL^–1 for surfactins and >256 μg mL^–1 for iturins. Our results align, with MIC values of 128–256 μg mL^–1 for surfactins and 256–512 μg mL^–1 for iturins.

To our knowledge, this study is the first to report macrolactin MIC values and compare them with lipopeptides against R. solanacearum. The 7-O-malonyl macrolactin fraction (BNC5e) exhibited higher antibacterial activity than lipopeptides, while BNC5RSe macrolactins were less active. Notably, macrolactins from BNC5e were the only isolates demonstrating bactericidal activity, suggesting that 7-O-malonyl macrolactin production plays a crucial role in suppressing R. solanacearum. The pathogen’s ability to reduce this production may enhance its resilience.

The MIC and MBC data revealed a more substantial reduction in antimicrobial activity for macrolactins from BNC5RSe compared to BNC5e, exceeding the overall difference between the extracts. This may be due to the presence of other antimicrobial agents, such as surfactins and iturins, which are relatively abundant and unaffected by R. solanacearum secretions, as indicated by metabolomics and antimicrobial activity data. The combined effect of these compounds likely mitigates the reduction in the BNC5RSe extract’s activity compared to BNC5e.

Overall, the lower cell densities and reduced antimicrobial activity of B. amyloliquefaciens after exposure to R. solanacearum secretions suggest that using live cells alone in heavily infected fields may be insufficient. A more effective strategy could involve first applying B. amyloliquefaciens extracts to reduce pathogen populations, followed by live cells to combat the pathogen and prevent reinfectionan approach preliminarily supported by the wilt-controlling ability of B. amyloliquefaciens in pot experiments.

In summary, this study examined how substances from the plant pathogen R. solanacearum impact the growth, metabolite production, and antimicrobial properties of the biocontrol bacterium B. amyloliquefaciens BNC5. Exposure to these substances compromised B. amyloliquefaciens growth, with decreased cell densities observed. Untargeted metabolomics analyses identified changes in 138 metabolite ions, including three known and seven novel macrolactins and one related polyketide. Levels of macrolactins A, M, BA1–6, and polyketide BAP1 increased, while 7-O-malonyl macrolactin A and macrolactin BA7 decreased. These shifts were linked to a 1.3-fold downregulation of bmmGT3, suggesting altered macrolactin tailoring enzyme activity. B. amyloliquefaciens extracts exposed to R. solanacearum secretions showed reduced antimicrobial activity in agar diffusion and MIC/MBC tests, partly due to changes in macrolactin efficacy. Future research will examine the effects of R. solanacearum cells on metabolite production in B. amyloliquefaciens, identify the substances driving macrolactin changes, pinpoint tailoring enzymes through transcriptomics, and evaluate how these metabolic shifts affect the structure of soil microbial communities. These findings underscore the potential of using B. amyloliquefaciens extracts in R. solanacearum-infected fields prior to applying live biocontrol bacteria for improved pathogen management.

Glossary

Abbreviations

BNC5

RSsup-unexposed Bacillus amyloliquefaciens BNC5 sample (ending: e, ethyl acetate extract; m, metabolomics; r, total RNA extract)

BNC5RS

RSsup-exposed Bacillus amyloliquefaciens BNC5 (ending: e, ethyl acetate extract m, metabolomics; r, total RNA extract)

BNC5sup

Bacillus amyloliquefaciens supernatant

COSY

correlation spectroscopy

CPG

casamino acid-peptone-glucose

ESI

electrospray ionization

HSQC

heteronuclear single quantum coherence

LC

liquid chromatography

m/z

mass-to-charge ratio

MBC

minimum bactericidal concentration

MIC

minimum inhibitory concentration

MS

mass spectrometry

MSII

integrated mass ion intensity

NMR

nuclear magnetic resonance

OD₆₀₀

optical density at 600 nm

PE

phosphatidylethanolamine

PG

phosphatidylglycerol

qRT-PCR

quantitative real-time polymerase chain reaction

RSsup

Ralstonia solanacearum supernatant (ending: e, ethyl acetate extract)

SI

Supporting Information

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.jafc.5c03489.

• List of oligonucleotides used for qRT-PCR experiments; volcano plots from metabolomics analyses; integrated mass ion intensities of ions with altered levels and other known metabolite ions from metabolomics analyses and macrolactin enzymatic conversion activity assay; MS/MS spectra of all identified macrolactins and other polyketides; NMR spectra of macrolactins; MS spectra of all metabolite isolates; and representative image of bacterial ooze (PDF)

Conceptualization: Nawaporn Vinayavekhin. Investigation: Augustine C. Onuh, Rutchaneekorn Manopaek, Pimsiri Tiyayon. Methodology: Augustine C. Onuh, Pimsiri Tiyayon, Nawaporn Vinayavekhin. Supervision: Pimsiri Tiyayon, Nawaporn Vinayavekhin. Visualization: Augustine C. Onuh, Pimsiri Tiyayon, Nawaporn Vinayavekhin. Writingoriginal draft: Augustine C. Onuh, Pimsiri Tiyayon, Nawaporn Vinayavekhin. Writingreview and editing: Alisa S. Vangnai, Pimsiri Tiyayon, Nawaporn Vinayavekhin.

This research was supported by the Thailand Science Research and Innovation Fund, Chulalongkorn University, and the National Research Council of Thailand (N42A640329). The opinions expressed in this paper are solely those of the authors and do not necessarily reflect the views of the funding agencies or Chulalongkorn University. The funders had no role in the study design, data collection and analysis, decision to publish, or manuscript preparation. All authors read and approved the final version of the manuscript.

The authors declare no competing financial interest.