First report of a Citrobacter sp. for fungal plant disease control: phenethyl alcohol‐producing NJC30 suppresses target spot on soybean

Abstract

BACKGROUND

Target spot, caused by the fungal pathogen Corynespora cassiicola, is a worldwide devastating disease that can affect multiple soybean tissues, including leaves, petioles, stems, pods, roots, and seeds. The control of target spot has traditionally relied on the application of synthetic fungicides, such as carbendazim.

RESULTS

In this study, Citrobacter si. NJC30 (Enterobacteriaceae family), which was isolated as an endophyte from healthy common comfrey leaves (Symphytum officinale) and identified by 16S rRNA, gyrB, and rpoB gene sequencing, was screened as a biocontrol agent to control target spot. NJC30 inhibited C. cassiicola mycelial growth in vitro and exhibited multiple antifungal mechanisms, such as competition for space and nutrients, and the production of phenethyl alcohol, which was present at 66.2 ± 8.6 mg L⁻¹ concentration after 3 days of culture. Preventive application of 1 × 10⁹ NJC30 cells/mL reduced target spot symptoms by 83.9% on pods and 74.2% on leaves, showing higher efficacy than 0.25 mg mL⁻¹ carbendazim.

CONCLUSION

Collectively, a new efficient biocontrol method for the management of target spot has been developed in this study. Although Citrobacter strains have been previously used to promote plant growth and enhance plant tolerance to abiotic stresses, this is the first report on the use of Citrobacter for the management of fungal plant pathogens, revealing a novel and promising biocontrol function for this genus and expanding the toolkit for sustainable plant disease management. © 2026 Society of Chemical Industry.

Citrobacter antifungal mechanisms were explored for the first time. Citrobacter sp. NJC30 competed for space and nutrients, produced phenethyl alcohol, and inhibited Corynespora cassiicola symptoms on soybean pods and leaves.

1. INTRODUCTION

Soybean (Glycine max) is a nutritionally valuable crop rich in isoflavones, which show estrogenic activities and have been associated with numerous health benefits, including cancer prevention, alleviation of menopausal symptoms, reduced risk of cardiovascular diseases, and protection against osteoporosis. ¹ , ² Soybean is essential in the diet of Asian countries, where it is usually consumed as sprouts or fermented products. ³ , ⁴ , ⁵ Soybean is consumed in vegan diets as an alternative to meat proteins. ⁶ , ⁷ The global production of soybean in 2023 reached 371 million metric tons (https://www.fao.org/faostat/en/#data/QCL/visualize). Despite the enormous nutritional and economic importance of soybean, a wide range of plant pathogens can infect soybean plants, reducing soybean production and quality. ⁸ , ⁹ , ¹⁰ It has been estimated that soybean diseases account for yield losses of approximately 21.4% worldwide. ¹¹ , ¹²

Target spot, caused by Corynespora cassiicola, is a devastating soybean disease. ¹³ Target spot symptoms have been detected in major soybean‐producing areas, such as the United States, China, Argentina, and Brazil, where yield losses between 10% and 42% have been reported. ¹⁴ , ¹⁵ C. cassiicola can infect multiple soybean tissues, including leaves, petioles, stems, pods, roots, and seeds. ¹⁶ Target spot outbreaks are especially common during the summer months under warm temperatures, ranging from 26 to 30 °C. ¹⁷ Foliage infection is favored when relative humidity in the soybean canopy exceeds 80%. ¹³ The control of C. cassiicola is a challenging task because of its broad host range, which includes other major crops, such as cantaloupe, cotton, and cucumber. ¹⁸ , ¹⁹ , ²⁰ In addition, in recent years, several C. cassiicola strains have been reported that are resistant to the commonly used fungicides for target spot control, including quinone outside inhibitors (QoIs; such as azoxystrobin, picoxystrobin, pyraclostrobin, and trifloxystrobin), methyl benzimidazole carbamates (MBCs; such as carbendazim and thiophanate methyl), and succinate dehydrogenase inhibitors (such as fluopyram, boscalid, pydiflumetofen, isopyrazam, and fluxapyroxad). ²¹ , ²² , ²³ , ²⁴ , ²⁵ Conventional breeding has been used to identify resistant soybean germplasm, such as the varieties ‘Bedford’ and ‘Council’, to combat this pathogen. ²⁶ , ²⁷ To date, only one biocontrol method for target spot management has been reported in the literature. The corresponding biocontrol method involves the use of the yeast Meyerozyma caribbica NJC41. ²⁸ Although NJC41 showed various antifungal mechanisms, including space and nutrient competition, and the production of antifungal phenethyl alcohol, application of 1 × 10⁹ cells/mL NJC41 reduced C. cassiicola symptoms by approximately 47.1%, indicating that NJC41 has only weak efficacy.

Bacterial genera that are underutilized in biocontrol research, such as Citrobacter (Enterobacteriaceae family), represent a promising source of novel antifungal agents. In general, some members from the genus Citrobacter are emerging as a new source of plant growth‐promoting rhizobacteria. ²⁹ Citrobacter strains can be mainly detected in soil and plant tissues, and have been reported to promote nitrogen fixation. ³⁰ , ³¹ A Citrobacter freundii strain, which was isolated from the desert cactus Euphorbia trigonas Mill, was reported to promote tomato plant growth under drought stress conditions, ³² whereas Citrobacter freundii N52 and Citrobacter freundii ZT14 were reported to promote wheat and cabbage growth, respectively. ³³ , ³⁴ , ³⁵ As far as we know, the potential use of Citrobacter strains to control fungal plant pathogens remains largely unexplored.

In this study, a new strain, Citrobacter sp. NJC30, was isolated from common comfrey (Symphytum officinale) plants. The antifungal mechanisms of NJC30 against C. cassiicola were examined. This study reveals an efficient biocontrol method for target spot management.

2. MATERIALS AND METHODS

2.1. General information

2.1.1. Fungal and bacterial sources, media, and growth conditions

C. cassiicola NJC25, which was previously isolated from soybean pods in Nantong Municipality, China, ²⁸ was used in the experiments. Additional fungi used in mycelial inhibition assays included Alternaria alternata HN‐5, ³⁶ Aspergillus flavus NJC03, ³⁷ Penicillium oxalicum NJC80, ³⁸ and Sclerotinia sclerotiorum NJC09. ³⁹ The fungi were maintained on potato–dextrose agar [PDA; 200 g of potato, 20 g of dextrose, and 15 g of agar in 1 L of double‐distilled water (ddH₂O)] and subcultured onto fresh PDA every 4 weeks. For experiments, cultures were grown at 28 °C in darkness for 3 days. C. cassiicola was grown in yeast extract–peptone–dextrose (YEPD) medium (20 g of glucose, 20 g of tryptone, and 10 g of yeast extract in 1 L of ddH₂O) at 28 °C and 200 rpm.

NJC30 was maintained in Luria–Bertani (LB) agar plates (10 g of tryptone, 5 g of yeast extract, 10 g of sodium chloride, and 15 g of agar in 1 L of ddH₂O, pH adjusted to 7.0) and grown in LB broth (LB medium without agar). Fungal and bacterial strains were preserved at −80 °C in 30% (v/v) glycerol solutions.

2.1.2. Plant material

Soybean pods, collected from a local field (Nantong Municipality, China) from 3‐month‐old plants (during the harvest stage), and soybean plants from the ‘ZH30’ cultivar were used in the experiments. ⁴⁰ After harvest, soybean pods were surface‐sterilized with 5% (v/v) NaOCl for 10 min, washed twice with ddH₂O, and immediately used in the experiments. Plants were grown in autoclaved soil within a growth chamber (RDN‐260A, Yanghui Ningbo Yanghui Instruments, Ningbo, China) at 28 °C and 60% relative humidity, with 12 h of light per day.

2.1.3. Chemicals and instruments

All reagents used were purchased from Macklin (Shanghai, China). Polymerase chain reaction (PCR) amplifications were performed in a thermocycler (Hema 9600 PCR, Zhuhai, China). High‐performance liquid chromatography (HPLC) analysis was performed using an Agilent 1200 instrument (Hewlett‐Packard, Waldbronn, Germany) equipped with a C₁₈ column (250 × 3.0 mm, Phenomenex, Tianjin, China). Mass spectrometry (MS) analysis was performed using a Triple Quadrupole/AB SCIEX 5500 QTRAP system (Sciex, Framingham, Massachusetts, USA). Scanning electron microscopy (SEM) images were collected using a Gemini 300 Instrument (Gemini, Oberkochen, Germany).

2.2. Isolation of Citrobacter sp. NJC30

To isolate potential biocontrol bacteria from medicinal plants, comfrey (S. officinale) leaf tissue was processed as follows. Comfrey leaf tissue was collected in Mozota Municipality, Spain, in August 2023. Comfrey leaf tissue (0.5 cm × 0.5 cm) was cultured on LB agar plates, allowing the formation of yellow biofilms after 3 days of cultivation at 28 °C. A small aliquot of the biofilm was collected by briefly touching a sterilized loop to the biofilm. The collected biofilm was then resuspended in 1 mL of phosphate‐buffered saline medium (PBS; 137 mm NaCl, 2.7 mm KCl, 8 mm Na₂HPO₄ 12H₂O, and 2 mm KH₂PO₄) (Solarbio, Beijing, China). An aliquot (100 μL) was spread on LB agar medium. A representative strain, Citrobacter sp. NJC30, was isolated as a single colony.

2.3. Molecular identification of Citrobacter si. NJC30 and construction of the phylogenetic tree

Genomic DNA from the isolated bacterial strain was extracted using a commercial Bacterial Genome DNA Extraction Kit (cat. no. D1600, Solarbio). The procedure was performed according to the manufacturer's instructions. NJC30 was cultured in LB medium (5 mL) at 28 °C for 12 h to an optical density at 600 nm (OD₆₀₀) of 0.5. One milliliter of bacterial culture was pelleted by centrifugation at 4 °C and 8000 g for 10 min, resuspended in 250 μL of Solution A containing 4 μL of RNase A (provided), and incubated at room temperature for 5 min. After adding 20 μL of proteinase K (10 mg mL⁻¹, provided) and mixing, the lysate was incubated at 70 °C for 10 min until it was clear and viscous. Subsequently, 220 μL of Solution B was added, followed by incubation at 70 °C for 10 min. DNA was precipitated by adding 220 μL of absolute ethanol, and the mixture was loaded onto a silica membrane adsorption column. The bound DNA was washed twice with 600 μL of wash buffer (containing ethanol) and eluted with 200 μL of elution buffer (pre‐warmed to 65 °C). The purified genomic DNA was stored at −20 °C.

The primers used for the amplification of the 16S ribosomal RNA (16S rRNA), gyrase subunit B (gyrB), and RNA polymerase (rpoB) genes were synthesized by GenScript (Nanjing, China) and are indicated in Supporting Information, Table S1. PCR amplification, which was carried out using a Hema 9600 PCR thermocycler (Zhuhai, China), was performed with an initial denaturalization step of 2 min at 95 °C, followed by 32 cycles, which included denaturalization at 95 °C for 10 s, annealing at 60 °C (16S rRNA), 55 °C (gyrB), or 53 °C (rpoB) for 10 s, and extension at 72 °C for 90 s, with a final extension period at 72 °C for 2 min. The obtained DNA fragments were deposited in GenBank under accession numbers PV875337 (16S rRNA), PX561077 (gyrB), and PX561078 (rpoB). The DNA sequences are presented in Supporting Information, Table S2.

A phylogenetic tree based on the 16S rRNA, gyrB, and rpoB sequences was constructed using MEGA11, with reference Enterobacteriaceae strains retrieved from GenBank. The strains used for the construction of the phylogenetic tree are detailed in Supporting Information, Table S3. Sequences were aligned using MUSCLE (https://www.ebi.ac.uk/Tools/msa/muscle/). Ambiguous regions and autapomorphic insertions were removed manually. The evolutionary history was inferred using the maximum likelihood method based on the Hasegawa–Kishino–Yano model. Initial trees for the heuristic search were obtained automatically by applying Neighbor‐Join and BioNJ algorithms to a matrix of pairwise distances estimated using the Maximum Composite Likelihood approach. The number of bootstrap replications was 1000. A discrete gamma distribution was used to model evolutionary rate differences among sites [five categories (+G, parameter = 0.46)]. The tree with the greatest likelihood (−44 865.729) is shown, with branch lengths measured in the number of substitutions per site. There were a total of 7934 positions in the final dataset. Apart from NJC30, the DNA sequences from 13 Enterobacteriaceae strains were used for the construction of the tree.

2.4. Pathogenicity of NJC30 on soybean pods

‘ZH30’ soybean pods, visually healthy (free of lesions or damage), were collected from a local field in Nantong Municipality, China. Soybean pods were used because the previously reported biocontrol approach to control soybean target spot also used soybean pods, ²⁸ which facilitates comparison with NJC30. NJC30 was cultured in LB medium (50 mL) at 28 °C for 2 days until OD₆₀₀ = 1.5. After collecting NJC30 cells via centrifugation at 4 °C and 8000 g for 10 min, the cells were washed twice with 20 mL of sterilized ddH₂O and resuspended in ddH₂O to OD₆₀₀ = 1.0. The resulting NJC30 bacterial suspension was sprayed–inoculated evenly, as a fine mist, on the soybean pods (10 mL of bacterial suspension on 40 soybean pods) using a 300‐mL Poly Spray bottle (Vazyme, Nanjing, China; with hand pressure) at a distance of 30 cm from the pods. Sterilized ddH₂O was used in the control experiment. The inoculated soybean pods were placed on sterile filter paper in opened containers within an RDN‐260A plant growth chamber (Yanghui Ningbo Yanghui Instruments) and maintained at 28 °C and 80% relative humidity in darkness for 3 days. Then, the pods were examined visually. The lack of pathogenicity was assessed by the absence of necrotic spots or abnormal discoloration compared to the control. Two independent repetitions were performed, each with 20 soybean pods (n = 40).

2.5. Population dynamics of NJC30 on soybean pods

NJC30 was cultured in LB medium (50 mL) at 28 °C for 2 days to OD₆₀₀ = 1.5. After collecting the NJC30 cells via centrifugation at 4 °C and 8000 g for 10 min, the cells were washed twice with 20 mL of sterilized ddH₂O and resuspended in ddH₂O to OD₆₀₀ = 1.0. Ten milliliters of the resulting suspension were sprayed evenly, as a fine mist, on 40 soybean pods using a 300 mL Poly Spray bottle (Vazyme; with hand pressure) at a distance of 30 cm from the pods, which were then incubated in a growth chamber at 28 °C for 7 days. Population dynamics were monitored every 24 h for 7 days post‐inoculation. At each time point, five soybean pods were added to 50‐mL Falcon tubes containing 10 mL of sterilized ddH₂O (one soybean pod in each Falcon tube). The experiment was performed three times. For each repetition, five soybean pods were sampled at each time point. After shaking at 1000 rpm for 2 min, an aliquot (20 μL) of the aqueous solution was diluted 100 times, and 100 μL of the diluted solution was spread on LB agar medium. The plates were incubated at 28 °C for 48 h. NJC30 population dynamics were monitored according to the observed colony‐forming units (CFU). The data are presented as CFU/mL of the spread solution.

2.6. In vitro antifungal activity screening

The inhibitory activity of NJC30 was screened in co‐cultures against five fungal pathogens, including A. alternata, A. flavus, C. cassiicola, P. oxalicum, and S. sclerotiorum. The strains are indicated in Section 2.1.1. All fungal pathogens were grown on PDA at 28 °C for 3 days, and then the mycelia were divided into plugs 3 mm in diameter. NJC30 was grown in LB medium (50 mL) at 28 °C and 200 rpm for 2 days to OD₆₀₀ = 1.5. Then, 20 μL of NJC30 culture medium (cell suspension) was applied as a straight band on the Petri dish, and a plug containing the mycelia of each fungal pathogen was placed in the middle of the Petri dish. The distance between the plug and NJC30 was ~2 cm. Control treatments were carried out using plugs of the fungal pathogens in the absence of NJC30. Antifungal activity was assessed by measuring the colony diameter in the direction facing the NJC30 band after incubating the plates at 28 °C for 3 days. The following formula was used to calculate the mycelial growth inhibition: Percent inhibition of mycelial growth = [(Dc – Dt)/Dc] × 100, where Dc is the colony diameter in the control and Dt is the colony diameter in the treatment (in the direction toward the NJC30 band). The experiment was performed independently three times, with five replicates per treatment in each repetition.

The viability of C. cassiicola hyphae after co‐culturing with NJC30 for 3 days was visualized via staining with Evans Blue and Neutral Red. ³⁵ , ⁴¹ Evans Blue (product number: E718217) and Neutral Red (product number: N718288) staining solutions were purchased from Macklin. C. cassiicola hyphae were collected from different sites of the plate, near and far away from the NJC30 band, and mixed with 5 μL of the commercial staining solutions on a microscope slide. After staining for 3 min, the cells were observed using an Olympus BX43 Microscope (Tokyo, Japan), with ×40 magnifications. Hyphae were collected from five different plates, sampling both near and far from the NJC30 band.

2.7. Antifungal activity of NJC30 cell‐free supernatant

To determine whether antifungal activity was due to diffusible compounds, cell‐free supernatant was tested. NJC30 was cultured in LB medium (50 mL) at 28 °C for 2 days until OD₆₀₀ = 1.5. After centrifugation of the culture suspension at 4 °C and 8000 g for 5 min, the cell‐free supernatant was collected and filtered using a 0.22‐μm nylon commercial filter (JinTeng, Tianjin, China) for complete sterilization. The filtered supernatant was incorporated into molten PDA to achieve final concentrations of 1.25%, 3.5%, 6.25%, and 12.5% (v/v). C. cassiicola was grown on PDA at 28 °C for 3 days and then, the mycelia were divided into plugs 3 mm in diameter. A plug of C. cassiicola was inoculated in the center of each PDA plate. The plates were then incubated at 28 °C for 3 days. C. cassiicola mycelial diameter in the treatment groups was measured and compared with that of the control group. The control was carried out in the absence of cell‐free supernatant. The following formula was used to calculate the mycelial growth inhibition: Percent inhibition of mycelial growth = [(Dc – Dt)/Dc] × 100, where Dc is the colony diameter in the control and Dt is the colony diameter in the treatment. The experiment was performed independently three times, with five replicates per treatment in each repetition.

2.8. Detection of phenethyl alcohol in NJC30 culture medium

To identify the primary antifungal metabolite produced by NJC30, culture extracts were fractionated by HPLC and assayed against C. cassiicola. NJC30 was cultured in LB medium (50 mL) at 28 °C and 200 rpm for 3 days to OD₆₀₀ = 1.5. The metabolites were extracted twice using 50 mL of ethyl acetate. After evaporating the organic solvent in a rotary evaporator at 35 °C, the resulting residue was dissolved in 10 mL of methanol. The metabolites from the methanolic solution were analyzed by HPLC using a C₁₈ column (250 × 3.0 mm, Phenomenex, Tianjin, China). ²⁸ A linear gradient from 13% to 30% acetonitrile in 0.1% (v/v) formic acid/water for 70 min was established. Flow rate, absorbance, temperature, and injection volume were set at 1.0 mL min⁻¹, 260 nm, 35 °C, and 10 μL, respectively. Each peak was collected from the HPLC detector outlet, the solvent was evaporated using a freeze drier (Labconco 117 Instrument, Kansas City, Missouri, USA), and the residue was finally resuspended in 100 μL of ddH₂O. Plates containing PDA were prepared. A hole 3 mm in diameter was created in the PDA plates, and 30 μL of the collected peaks was poured into the hole. A plug containing C. cassiicola mycelia was located in the center of the PDA plates. The distance between the hole and plug was ~2 cm. The plates were then incubated at 28 °C for 3 days. Only one collected peak (retention time 8.0 min) inhibited mycelial growth, causing visible directional inhibition on the side of the colony nearest to the compound well. This compound was further studied by MS (ESI ionization source) using a Triple Quadrupole/AB SCIEX 5500 QTRAP system (Sciex, Framingham, Massachusetts, USA), providing a main peak at 123.2 Da in positive mode. This m/z peak was analyzed by tandem mass spectrometry (MS/MS), yielding m/z peaks at 107.8, 96.1, 80.2, and 78.0. The active metabolite was identified as phenethyl alcohol. ²⁸ The concentration of phenethyl alcohol was calculated based on the peak area. Five independent repetitions were carried out. A calibration curve (linear regression) was established using commercial phenethyl alcohol (Macklin) from 5 μg mL⁻¹ to 1 mg mL⁻¹.

2.9. Antifungal activity of phenethyl alcohol

C. cassiicola was grown on PDA at 28 °C for 3 days and the mycelia were then divided into plugs 3 mm in diameter. A mycelial plug of C. cassiicola was cultured in the center of PDA plates containing 0.1, 0.2, 0.5, 1, and 2 mg mL⁻¹ phenethyl alcohol at 28 °C for 3 days. Phenethyl alcohol was mixed with molten PDA and then poured on the plates. The control experiment was carried out on PDA in the absence of phenethyl alcohol. Antifungal activity was monitored according to the inhibition of C. cassiicola mycelial growth. The following formula was used to calculate the mycelial growth inhibition: Percent inhibition of mycelial growth = [(Dc – Dt)/Dc] × 100, where Dc is the colony diameter in the control and Dt is the colony diameter in the treatment. Five independent repetitions were carried out.

2.10. Competition for space assay (SEM)

The competition for space assay was carried out using SEM following previously reported procedures. ⁴² , ⁴³ Briefly, healthy ‘ZH30’ soybean pods, visually healthy (free of lesions or damage), were collected from a local field, sterilized with 5% (v/v) NaOCl for 10 min, and washed twice with ddH₂O. A wound 2 mm in diameter was made on the soybean pod surface using a sterilized knife. NJC30 was cultured in LB medium (200 mL) at 28 °C and 200 rpm for 2 days to OD₆₀₀ = 1.5. NJC30 cells were collected by centrifuging at 4 °C and 8000 g for 10 min. The NJC30 cells were then washed twice with 10 mL of sterilized ddH₂O and resuspended in ddH₂O to 1 × 10⁶ NJC30 cells/mL. C. cassiicola was cultured on PDA plates at 28 °C for 3 days under continuous light. Spores were collected from the plate using 1 mL of sterilized ddH₂O. After discarding the mycelia, the spore concentration in the solution was adjusted to 1 × 10⁶ spores/mL using sterilized ddH₂O. Cell concentrations were adjusted using an XBK‐25 hemocytometer (Qiujing, Shanghai, China).

Aqueous solutions containing NJC30 cells (5 μL, 1 × 10⁶ cells/mL) and C. cassiicola spores (5 μL, 1 × 10⁶ spores/mL) were poured into the soybean wounds and mixed. The control experiments were carried out using only NJC30 (5 μL, 1 × 10⁶ cells/mL) or C. cassiicola (5 μL, 1 × 10⁶ spores/mL). The inoculated pods were kept on sterile filter paper in open containers in an RDN‐260A plant growth chamber (Yanghui Ningbo Yanghui Instruments) at 28 °C and 80% relative humidity in darkness for 24 h. The experiment was performed independently three times, with five replicates per treatment in each repetition.

Wounds were examined by SEM using standard procedures. ⁴² , ⁴³ Briefly, the wound tissue was extracted and immersed in a 2.5% (v/v) glutaraldehyde aqueous solution in 0.1 m phosphate buffer (5 mL, pH 7) for 24 h. The fixed tissue samples were then collected and dehydrated by submerging the samples in 65%, 75%, 85%, 95%, and 100% (twice) (v/v) ethanol (in ddH₂O) at room temperature. Each dehydration condition was carried out for 5 min. The samples were then stored in a desiccator (EM COD 300; Leica, Wetzlar, Germany) for 6 h. Dried samples were placed on the sample holder and sputter‐coated with gold (5–20 nm particles) using an AGB7340 Manual Sputter Coater (Agar Scientific, Stansted, UK). Images were obtained by SEM (Gemini 300 Instrument, Oberkochen, Germany) with an acceleration voltage of 20 kV.

2.11. Competition for nutrients assay

The effects of NJC30‐mediated nutrient deprivation on C. cassiicola spore germination were evaluated following the procedure reported by Di Francesco et al. ⁴⁴ This method involved the use of two polystyrene cylinders that were separated by a hydrophilic polytetrafluoroethylene (PTFE) membrane (Shanghai Yaxin, Shanghai, China), with a pore size of 0.45 μm. The PTFE membrane allowed the exchange of nutrients but not cells between the polystyrene cylinders. NJC30 and C. cassiicola cell suspensions were prepared as indicated in Section 2.10.

Soybean juice was obtained by boiling 100 g of smashed soybean pods for 20 min in 500 mL of ddH₂O. The pods used were healthy and surface‐sterilized with 5% (v/v) NaOCl prior to homogenization. After boiling, the suspension was filtered through four layers of medical gauze dressing (pore size: 800–1000 μm; Jiang Huang Medical, Jinhua, China), and the aqueous solution was collected and sterilized using an autoclave (LDZX‐50 L‐I, Shenan, Shanghai, China) at 120 °C for 1 h. The polystyrene cylinders were filled with 25 mL of 0% (only ddH₂O), 0.5%, and 5% (v/v) soybean juice. Soybean juice was diluted using sterilized ddH₂O. In one cylinder, 1 mL of C. cassiicola spore suspension (1 × 10⁶ spores/mL) was added; 1 mL of NJC30 cell suspension (1 × 10⁶ cells/mL) was added in the other cylinder. The control experiments were carried out by only inoculating C. cassiicola (1 × 10⁶ spores/mL), in the absence of NJC30. The cylinders were shaken at 28 °C and 200 rpm for 12 h using an MQD‐S2R shaker (Minquan Instruments, Shanghai, China). Then, 20 μL of C. cassiicola spore suspension was collected and added to a microscope slide, and the ratio of germinated C. cassiicola spores was calculated using an Olympus BX43 Upright microscope (Tokyo, Japan). The inhibitory effects caused by NJC30‐promoted nutrient deprivation on C. cassiicola spore germination were calculated by comparing the ratio of germinated spores in the treatment and control groups. To calculate the ratio of germinated spores, the germination of ~500 spores was observed in each replicate. The spore germination ratio was calculated based on the following formula: Number of germinated spores/Number of total spores × 100. A spore was considered germinated when the germ tube length was at least 20% the spore length. Three independent repetitions were carried out for each treatment condition.

2.12. In vivo efficacy of NJC30 for target spot control on detached soybean pods

Curative and preventive efficacies of NJC30 were screened following previously reported procedures with minor modifications. ²⁸ , ⁴² , ⁴³ Healthy ‘ZH30’ soybean pods (800 pods in total), in the absence of lesions or damage, were harvested from a local field. The pods were surface‐sterilized with 5% (v/v) NaOCl for 30 min, and washed twice with ddH₂O. NJC30 and C. cassiicola cell suspensions were prepared as indicated in Section 2.10. NJC30 concentration was adjusted to 1 × 10⁷, 1 × 10⁸, and 1 × 10⁹ cells/mL, and the C. cassiicola concentration was adjusted to 1 × 10⁶ spores/mL. After inoculation, the pods were placed on sterile filter paper in opened containers within an RDN‐260A plant growth chamber (Yanghui Ningbo Yanghui Instruments).

For the curative assay (wound and C. cassiicola inoculation at 0 h; NJC30 application at 24 h; symptom observation at 72 h) wounds 2 mm in diameter were created on soybean pods using a sterilized knife. Then C. cassiicola (20 μL, 1 × 10⁶ spores/mL) was inoculated into the soybean wounds. The inoculated soybean pods were kept in a plant growth chamber at 28 °C and 80% relative humidity in darkness for 24 h; afterwards, brown lesions were observed around the inoculation site. Cell suspensions containing 1 × 10⁷, 1 × 10⁸, and 1 × 10⁹ NJC30 cells/mL were then sprayed on the soybean pods (10 mL of bacterial suspension on 40 soybean pods). Sterilized ddH₂O (10 mL) was used for the negative control experiment. The pods were then incubated in a plant growth chamber at 28 °C and 80% relative humidity in darkness for 48 h.

For the preventive assay (wound and NJC30 application at 0 h; C. cassiicola inoculation at 12 h; symptom observation at 60 h) wounds 2 mm in diameter were created on the soybean pods with a sterilized knife, and cell suspensions containing 1 × 10⁷, 1 × 10⁸, and 1 × 10⁹ NJC30 cells/mL were sprayed on the soybean pods (10 mL suspension on 40 soybean pods). Sterilized ddH₂O (10 mL) was used in the negative control experiment. The soybean pods were then incubated in a plant growth chamber at 28 °C in darkness for 12 h. C. cassiicola (20 μL, 1 × 10⁶ spores/mL) was inoculated on the wounds. Finally, the soybean pods were maintained in a plant growth chamber at 28 °C and 80% relative humidity in darkness for 48 h.

Target spot symptoms (visible round black necrotic spots) were detected in all inoculated soybean pods. The experiments were performed using 40 soybean pods for each treatment condition in each repetition, and the experiments were repeated twice (n = 80). The antifungal activities of NJC30 cell suspensions were calculated based on the lesion diameter, which was measured using a digital caliper. The lesion diameter observed after application of 1 × 10⁷, 1 × 10⁸, and 1 × 10⁹ NJC30 cells/mL was compared with that of the control experiment. The following formula was used to calculate the inhibitory activity: Inhibitory activity = [(Lc – Lt)/Lc] × 100, where Lc is the lesion diameter in the control and Lt is the lesion diameter in the treatment. The curative and preventive efficacies of commercial carbendazim (only active ingredient; product number: C888676, Macklin) were evaluated as a positive control. Carbendazim was applied in assay at 0.25 mg mL⁻¹, which is the application rate suggested by commercial suppliers (Tonu Agrotech Ltd, Shanghai, China). To prepare the carbendazim solution, carbendazim (12.5 mg) was first dissolved in 1 mL of dimethyl sulfoxide (DMSO) and then suspended in 49 mL of sterilized ddH₂O.

2.13. In vivo efficacy of NJC30 for target spot control on soybean plants

‘ZH30’ soybean plants were grown in soil pots 30 cm in diameter (one plant per pot) in a growth chamber (RDN‐260A, Yanghui Ningbo Yanghui Instruments). The soil was autoclaved before use. The plants were grown at 28 °C and 60% relative humidity, with 12 h of light per day, for 3 weeks. Each plant contained three trifoliate leaves. NJC30 and C. cassiicola cell suspensions were prepared as indicated in Section 2.10. NJC30 concentration was adjusted to 1 × 10⁷, 1 × 10⁸, and 1 × 10⁹ NJC30 cells/mL; the C. cassiicola concentration was adjusted to 1 × 10⁶ spores/mL. NJC30 efficacy was assessed using curative and preventive assays.

For the curative assay (wound and C. cassiicola inoculation at 0 h; NJC30 application at 24 h; symptom observation at 72 h), C. cassiicola was sprayed–inoculated (5 mL per plant, 1 × 10⁶ spores/mL) on the soybean leaves. The inoculated soybean plants were kept at 28 °C and 60% relative humidity (with 12 h of light irradiation per day) for 24 h; afterwards, weak yellow–brown lesions were observed. Cell suspensions containing 1 × 10⁷, 1 × 10⁸, and 1 × 10⁹ NJC30 cells/mL were then sprayed on the soybean leaves (5 mL of bacterial suspension per plant). Sterilized ddH₂O (5 mL) was used for the negative control experiment. The plants were then grown at 28 °C and 60% relative humidity (with 12 h of light irradiation per day) for 48 h.

For the preventive assay (wound and NJC30 application at 0 h; C. cassiicola inoculation at 12 h; symptom observation at 60 h), bacterial cell suspensions containing 1 × 10⁷, 1 × 10⁸, and 1 × 10⁹ NJC30 cells/mL were sprayed on the soybean leaves (5 mL of bacterial suspension per plant). Sterilized ddH₂O (5 mL) was used in the negative control experiment. The soybean plants were then incubated at 28 °C and 60% relative humidity in darkness for 12 h. C. cassiicola was then sprayed–inoculated (5 mL per plant, 1 × 10⁶ spores/mL) on the leaves. Finally, the soybean leaves were maintained at 28 °C and 60% relative humidity (with 12 h of light irradiation per day) for 48 h.

Target spot symptoms (irregular yellow–brown lesions) were detected on all inoculated soybean leaves. The experiments were performed using ten soybean plants for each treatment condition in each repetition, with each plant containing three trifoliate leaves (approximately nine leaflets in each plant), and the experiments were repeated three times (n = 270). The antifungal activities of NJC30 cell suspensions were measured by comparing the number of lesions per leaf in the treatment and the control groups. A lesion was defined as any visible spot, with >1 mm in diameter. The following formula was used to calculate the inhibitory activity: Inhibitory activity = [(Nc – Nt)/Nc] × 100, where Nc is the number of lesions in control and Nt is the number of lesions in the treatment. The curative and preventive efficacies of commercial carbendazim (only active ingredient; product number: C888676, Macklin,) were evaluated as a positive control. Carbendazim was applied at 0.25 mg mL⁻¹ (5 mL per plant), which is the application rate suggested by commercial suppliers (Tonu Agrotech Ltd). To prepare the carbendazim solution, carbendazim (12.5 mg) was first dissolved in 1 mL of DMSO and then suspended in 49 mL of sterilized ddH₂O.

2.14. Statistical analysis

All obtained results regarding the in vivo assays were pooled to calculate the mean value and standard deviation (SD) (no data were discarded). Statistical analyses were performed using SPSS (Statistical Package, Version 20.0; IBM, Armonk, New York, USA). The obtained data were processed using one‐way analysis of variance (via post‐hoc multiple comparisons followed by Tukey's test (P < 0.05). The error bars in the figures represent SD.

3. RESULTS

3.1. NJC30 was isolated from common comfrey leaves and identified

NJC30 was isolated from healthy common comfrey leaves in Mozota Municipality, Spain. NJC30 formed yellow biofilms on LB medium (Fig. 1(A)). The 16S rRNA sequence of NJC30 showed 95.62%, 95.62%, 95.70%, and 95.61% identity to those of Citrobacter farmeri Lzp2 (GenBank number: MW012678.1), Citrobacter farmeri A24352 (GenBank number: CP181748.1), Citrobacter farmeri AUSMDU00008141 (GenBank number: CP022695.1), and Citrobacter farmeri FDAARGOS 1423 (GenBank number: CP077291.1), respectively. The gyrB gene from NJC30 showed 95.46%, 95.46%, 95.34%, and 94.72% homology to the gyrB genes from Citrobacter freundii complex si. CFNIH2 (GenBank number: CP025757.1), Citrobacter si. AHL6 (GenBank number: AY462061.1), Citrobacter amalonaticus FDAARGOS_122 (GenBank number: CP014015.2), and Citrobacter telavivensis 6105 (GenBank number: CP045205.1), respectively, whereas the rpoB gene from NJC30 showed 96.75%, 97.47%, 97.47%, and 96.39% homology to the rpoB genes from Citrobacter amalonaticus 2024CK‐01695 (GenBank number: CP180152.1), Citrobacter farmeri AUSMDU00008141 (GenBank number: CP022695.1), Citrobacter farmeri FDAARGOS 1423 (GenBank number: CP077291.1), and Citrobacter amalonaticus 86 (GenBank number: LT556084.1), respectively. These results confirmed that NJC30 belongs to the genus Citrobacter. A phylogenetic tree was constructed with representative strains and indicated that NJC30 showed high similarity compared with Citrobacter farmeri CDC 2991‐81 (Fig. 1(B)). The lack of pathogenicity of NJC30 on soybean pods was confirmed (Fig. 1(C)).

Figure 1.

Isolation and identification of NJC30 from common comfrey plants (Symphytum officinale). (A) NJC30 biofilms on LB medium. (B) Phylogenetic tree of NJC30 with representative Enterobacteriaceae strains, based on 16S rRNA, gyrB, and rpoB genes. The phylogenetic tree was constructed using MEGA11. The maximum likelihood method with 1000 bootstrap replicates was used. The evolutionary history was inferred using the maximum likelihood method based on the Hasegawa–Kishino–Yano model. A discrete gamma distribution was used to model evolutionary rate differences among sites [five categories (+G, parameter = 0.46)]. The tree with the highest likelihood (−44 865.729) is shown, with branch lengths measured in the number of substitutions per site. (C) Images showing the lack of pathogenicity of NJC30 on soybean pods. (D) Population dynamics of NJC30 on soybean pods. The results are expressed as colony‐forming units (CFU) per mL of the spread solution. Differences between means were considered significant when P ≤ 0.05. Different letters indicate that the variables are significantly different. Error bars represent standard deviation.

Interestingly, NJC30 colonized soybean pods (Fig. 1(D)). The population of NJC30 (CFU/mL) increased over time, showing a maximum at day 4. The number of NJC30 colonies at day 4 was 4.65‐fold higher than the number of colonies at day 0. After day 4, the number of colonies decreased.

3.2. NJC30 inhibited C. cassiicola mycelial growth

NJC30 reduced the mycelial growth of C. cassiicola by 42.7% (Fig. 2). C. cassiicola hyphae near NJC30 stained blue with Evans Blue but showed no color with Neutral Red, indicating that these cells were dead (Fig. 2). This indicated C. cassiicola cell death in the zone of NJC30 inhibition. Instead, C. cassiicola hyphae far away from NJC30 showed no color with Evans Blue, and red color with Neutral Red, indicating that these cells were alive. In addition, NJC30 inhibited the mycelial growth of other fungal pathogens, including A. alternata, A. flavus, P. oxalicum, and S. sclerotiorum, exhibiting a broad spectrum of antifungal activity in vitro (Supporting Information Fig. S1).

Figure 2.

Co‐culture of NJC30 and Corynespora cassiicola. A control experiment was carried out by culturing C. cassiicola in the absence of NJC30. The viability of C. cassiicola hyphae was examined by staining with Evans Blue and Neutral Red. Evans Blue stains dead cells, whereas Neutral Red stains live cells. C. cassiicola hyphae near NJC30 showed blue color after staining with Evans Blue, indicating cell death in the zone of NJC30 inhibition.

3.3. NJC30 produced antifungal phenethyl alcohol

NJC30 cell‐free supernatant significantly inhibited C. cassiicola mycelial growth. NJC30 cell‐free supernatant at 1.25%, 3.5%, 6.25%, and 12.5% (v/v) reduced C. cassiicola mycelial growth by 7.8%, 13.5%, 17.8%, and 45.6%, respectively (Fig. 3(A)). These results indicated that NJC30 can biosynthesize metabolites with antifungal activity.

Figure 3.

Antifungal activity of NJC30 cell‐free supernatant against Corynespora cassiicola. (A) The antifungal effects of NJC30 cell‐free supernatant on C. cassiicola mycelial growth. (B) High‐performance liquid chromatography‐based analysis of phenethyl alcohol in NJC30 culture medium. A linear gradient from 13% to 30% acetonitrile for 70 min was used as the mobile phase. ²⁸ (C) Tandem mass spectrometry‐based analysis of phenethyl alcohol. A main m/z peak was detected at 123.2 Da in positive mode. (D) The antifungal effects of phenethyl alcohol on C. cassiicola mycelial growth.

To identify the metabolites from NJC30 with antifungal activity, NJC30 cell‐free supernatant was analyzed by HPLC. Only one peak in the HPLC chromatogram showed obvious antifungal activity, confirming the presence of a main antifungal metabolite in NJC30 secretions (Fig. 3(B)). MS analysis of this compound provided a m/z peak at 123.2 Da in positive mode, which is consistent with the expected exact mass of phenethyl alcohol (calculated for C₈H₁₁O, [M + H]⁺ = 123.0804) (Fig. 3(C)). MS/MS analysis of the peak at 123.2 Da provided m/z peaks at 107.8, 96.1, 80.2, and 78.0. This MS/MS fragmentation is consistent with the reported MS/MS fragmentation profile of phenethyl alcohol. ²⁸ Commercial phenethyl alcohol appeared at the same retention time in the HPLC chromatogram compared with the active metabolite. The concentration of phenethyl alcohol in the secretions of NJC30 after 3 days of culture was 66.2 ± 8.6 mg L⁻¹.

To demonstrate the antifungal properties of phenethyl alcohol against C. cassiicola, this fungal pathogen was cultured on PDA plates containing 0.1, 0.2, 0.5, 1, and 2 mg mL⁻¹ phenethyl alcohol (Fig. 3(D)). Phenethyl alcohol at 1 and 2 mg mL⁻¹ inhibited C. cassiicola mycelial growth by 87.9% and 100%, respectively, demonstrating the antifungal activity of this metabolite.

3.4. NJC30 competed for space with C. cassiicola

NJC30 and C. cassiicola showed different morphological characteristics and therefore could be easily differentiated on the soybean pod surface. NJC30 cells were 2.11 ± 0.38 μm in length (number of observations = 100) (Fig. 4). In several cases, NJC30 cytokinesis was observed (number of observations = 18), indicating that the soybean pod surface is a suitable environment that allows NJC30 cell division and proliferation. C. cassiicola also colonized the soybean pod surface, forming spider nest‐like mycelia (Fig. 4). C. cassiicola hyphae were between 1.5 and 2.2 μm in width. When NJC30 and C. cassiicola were applied together, NJC30 colonized C. cassiicola hyphae, causing substantial irregularities in the C. cassiicola cell wall (Fig. 4). NJC30 formed a dense biofilm on C. cassiicola hyphae. NJC30 cytokinesis was also observed on C. cassiicola hyphae, indicating that NJC30 can co‐inhabit with C. cassiicola and proliferate in its mycelium.

Figure 4.

Scanning electron microscopy (SEM) observations showing the competition for space between NJC30 and Corynespora cassiicola on soybean pods. NJC30 colonized C. cassiicola hyphae and caused irregularities in C. cassiicola cell wall.

3.5. NJC30 competed for nutrients with C. cassiicola

The germination rate of C. cassiicola increased with higher soybean juice concentration (Table 1). In the absence of NJC30, the ratio of C. cassiicola germinated spores was 31.1 ± 1.5%, 50.7 ± 5.6%, and 67.4 ± 2.1% with 0%, 0.5%, and 5% (v/v) soybean juice, respectively. This result can be explained by the high availability of nutrients in soybean juice, which may promote C. cassiicola spore germination compared with ddH₂O.

Table 1.

Percentage of germinated Corynespora cassiicola spores after 12 h of incubation in a competition for nutrients assay with NJC30†

Combination	Soybean juice (%)‡
	0	0.5	5
C. cassiicola	31.1 ± 1.46 a	50.7 ± 5.6 a	67.4 ± 2.1 a
C. cassiicola + NJC30	8.8 ± 0.8 b	23.7 ± 1.2 b	59.7 ± 1.8 b

^†The experiments were carried out using a PTFE membrane following the conditions reported by Di Francesco et al. ⁴⁴

^‡Differences between means were considered significant when P ≤ 0.05. Different letters in the same column indicate that the variables are significantly different.

NJC30 significantly inhibited C. cassiicola spore germination (Table 1). In the presence of 0%, 0.5%, and 5% (v/v) soybean juice, NJC30 reduced C. cassiicola spore germination by 71.7%, 53.3%, and 11.4%, respectively, compared with the respective control experiments. Thus, the inhibitory effects were higher in 0% and 0.5% (v/v) soybean juice than in 5% (v/v) soybean juice. This trend suggests that nutrient competition is more effective under low‐nutrient conditions.

3.6. NJC30 inhibited target spot symptoms on detached soybean pods

The inhibitory effects of NJC30 on C. cassiicola mycelial growth, as well as the ability of NJC30 to compete for space and nutrients with C. cassiicola, suggested that NJC30 could be a potential biocontrol agent for target spot management. To evaluate the efficacy of NJC30 for controlling target spot on detached soybean pods, the preventive and curative efficacies of 1 × 10⁷, 1 × 10⁸, and 1 × 10⁹ NJC30 cells/mL were examined (Fig. 5 and Table 2). The higher the NJC30 concentration, the higher the inhibitory efficacy. NJC30 showed higher preventive than curative efficacy. In curative application, 1 × 10⁷, 1 × 10⁸, and 1 × 10⁹ NJC30 cells/mL inhibited C. cassiicola‐caused lesion diameter by 7.4% (not significant), 29.8%, and 69.2%, respectively. In preventive application, 1 × 10⁷, 1 × 10⁸, and 1 × 10⁹ NJC30 cells/mL inhibited C. cassiicola‐caused lesion diameter by 2.5% (not significant), 48.2%, and 83.9%, respectively.

Figure 5.

The efficacy of NJC30 to control Corynespora cassiicola on soybean pods. Detached soybean pods were used in the assay. The curative and preventive efficacies of NJC30 were screened. NJC30 was applied at 1 × 10⁷, 1 × 10⁸, and 1 × 10⁹ cells/mL. The negative control experiment was carried out by spraying sterilized ddH₂O, whereas the positive control experiment was carried out by spraying 0.25 mg mL⁻¹ carbendazim.

Table 2.

Efficacy of NJC30 to control target spot on soybean pods

Application	Treatment	Lesion diameter (cm)* , †	Inhibitory activity (%)
Curative	Negative control ‡	0.94 ± 0.10 a	—
	1 × 107 NJC30 cells/mL	0.87 ± 0.11 ab	7.4
	1 × 108 NJC30 cells/mL	0.66 ± 0.13 b	29.8
	1 × 109 NJC30 cells/mL	0.29 ± 0.05 c	69.2
	0.25 mg mL−1 carbendazim	0.25 ± 0.04 c	73.4
Preventive	Negative control ‡	0.81 ± 0.10 a	—
	1 × 107 NJC30 cells/mL	0.79 ± 0.02 a	2.5
	1 × 108 NJC30 cells/mL	0.42 ± 0.03 b	48.2
	1 × 109 NJC30 cells/mL	0.13 ± 0.02 d	83.9
	0.25 mg mL−1 carbendazim	0.24 ± 0.06 c	70.4

^*Lesion diameter (each treatment mean represents the average of 80 pods) ± standard deviation (SD).

^† Differences between means were considered significant when P ≤ 0.05. Different letters in the same column indicate that the variables are significantly different. Curative and preventive applications were considered different statistical groups.

^‡ The negative control experiments were carried out by spraying ddH₂O (in the absence of NJC30 and carbendazim).

In curative application, 0.25 mg mL⁻¹ carbendazim (73.4% lesion diameter inhibition) reduced target spot symptoms to a similar extent compared with 1 × 10⁹ NJC30 cells/mL (Fig. 5 and Table 2). However, 1 × 10⁹ NJC30 cells/mL was more efficient than 0.25 mg mL⁻¹ carbendazim (70.7% lesion diameter inhibition) for controlling target spot symptoms in preventive application.

3.7. NJC30 inhibited target spot symptoms on soybean leaves

To confirm the ability of NJC30 to control target spot symptoms on soybean plants, the preventive and curative efficacies of 1 × 10⁷, 1 × 10⁸, and 1 × 10⁹ NJC30 cells/mL for controlling target spot on soybean leaves were examined (Fig. 6 and Table 3). As in the case of the detached soybean pods, the inhibitory efficacy increased proportionally with NJC30 concentration. Interestingly, 1 × 10⁹ NJC30 cells/mL allowed higher curative and preventive efficacies than carbendazim, demonstrating its practical applicability for target spot management. Curative and preventive applications of 1 × 10⁹ NJC30 cells/mL reduced the number of lesions caused by C. cassiicola on soybean leaves by 46.1% and 74.5%, respectively. Instead, curative and preventive applications of 0.25 mg mL⁻¹ carbendazim resulted in 12.9% and 40.0% inhibition, respectively.

Figure 6.

The efficacy of NJC30 to control Corynespora cassiicola on soybean leaves. Three‐week‐old soybean plants were used in the assay. The curative and preventive efficacies of NJC30 were screened. NJC30 was applied at 1 × 10⁷, 1 × 10⁸, and 1 × 10⁹ cells/mL. The negative control experiment was carried out by spraying sterilized ddH₂O, whereas the positive control experiment was carried out by spraying 0.25 mg mL⁻¹ carbendazim.

Table 3.

Efficacy of NJC30 to control target spot on soybean leaves

Application	Treatment	Number of lesions per leaf* , †	Inhibitory activity (%)
Curative	Negative control ‡	32.69 ± 3.50 a	—
	1 × 107 NJC30 cells/mL	31.36 ± 2.46 a	4.1
	1 × 108 NJC30 cells/mL	22.29 ± 1.37 c	31.8
	1 × 109 NJC30 cells/mL	17.62 ± 1.28 d	46.1
	0.25 mg mL−1 carbendazim	28.48 ± 1.87 b	12.9
Preventive	Negative control ‡	30.48 ± 1.96 a	—
	1 × 107 NJC30 cells/mL	20.66 ± 1.77 b	32.2
	1 × 108 NJC30 cells/mL	16.11 ± 2.18 d	47.1
	1 × 109 NJC30 cells/mL	7.77 ± 1.51 e	74.5
	0.25 mg mL−1 carbendazim	18.29 ± 1.74 c	40.0

^*Number of lesions per leaf (each treatment mean represents the average of 270 leaflets) ± standard deviation (SD). The experiments were repeated three times, and the number of lesions was measured in approximately 90 leaflets in each repetition.

^† Differences between means were considered significant when P ≤ 0.05. Different letters in the same column indicate that the variables are significantly different. Curative and preventive applications were considered different statistical groups.

^‡ The negative control experiments were carried out by spraying ddH₂O (in the absence of NJC30 and carbendazim).

4. DISCUSSION

Common comfrey has been reported to alter soil microbial communities, increasing Proteobacteria abundance. ⁴⁵ Citrobacter strains with the ability to promote plant growth have been mainly isolated from soils containing heavy metals and showed heavy metal biosorption capacity, stimulating plant growth under heavy metal stress. ²⁹ For example, Citrobacter enshiensis si. nov. was isolated from seleniferous soil, ⁴⁶ whereas Citrobacter si. JH 11‐2, which showed cadmium biosorption ability, was isolated from mining site soil. ⁴⁷ A few strains have been also isolated from plant tissues. For example, Citrobacter freundii AYS58 was isolated from maize roots. ⁴⁸ To the best of our knowledge, NJC30 is the first Citrobacter strain that has been isolated from plant leaves. The identification of Citrobacter strains, as well as other Enterobacteriaceae species, is commonly done via 16S rRNA and gyrB sequencing. ⁴⁹ For this reason, this approach was also used in this study. The sequencing analysis indicated that NJC30 shows high homology to some representative Citrobacter farmeri, Citrobacter telavivensis, and Citrobacter amalonaticus strains. Although Citrobacter freundii has been reported to cause rot lesions on ginger and trunk canker in mulberry, ⁵⁰ , ⁵¹ NJC30 lacked pathogenicity on soybean pods.

NJC30 inhibited the mycelial growth of diverse pathogenic fungi, suggesting that NJC30 could be a suitable agent for the management of a wide range of fungal plant diseases. This study represents the first report on the antifungal properties of a bacterium from the genus Citrobacter. This discovery has significant implications for the field of biological control. First, it substantially expands the known taxonomic range of bacteria with direct biocontrol potential against fungal plant pathogens, highlighting Citrobacter, which is a genus primarily studied for plant growth promotion and environmental remediation, ²⁹ , ³² , ³³ , ³⁴ , ³⁵ as a novel and previously overlooked reservoir of antifungal agents. Second, this finding encourages the re‐evaluation of other rhizospheric and endophytic bacteria within the Enterobacteriaceae family, particularly those already known for beneficial plant interactions, for undiscovered antimicrobial capabilities. Exploring these related strains could accelerate the discovery of new bioactive metabolites and biocontrol mechanisms. Supporting this notion, other Enterobacteriaceae species have been successfully deployed for plant pathogen management. For example, Enterobacter asburiae L95 was applied for the biocontrol of Dickeya oryzae in rice plants, showing a mode of action based on the modulation of the quorum sensing signal. ⁵² Kluyvera ascorbata YAFL9 and YCFR5 showed antifungal activity against Colletotrichum gloeosporioides and Fusarium oxysporum. ⁵³

Although this is the first report regarding the ability of Citrobacter to synthesize phenethyl alcohol, various reports have confirmed that phenethyl alcohol is a common metabolite produced by Enterobacteriaceae strains. For example, Enterobacter si. CGMCC 5087 and Escherichia coli have been reported to biosynthesize phenethyl alcohol. ⁵⁴ , ⁵⁵ The identification of phenethyl alcohol as a key antifungal compound from NJC30 provides a specific and tractable mode of action for this novel biocontrol agent and suggests a potential conserved metabolic pathway worthy of further investigation within the Enterobacteriaceae family. Notably, M. caribbica NJC41, which was previously screened to control C. cassiicola, was also reported to produce phenethyl alcohol as the main antifungal metabolite. ²⁸ The concentration of phenethyl alcohol in NJC30 culture medium after 3 days (66.2 ± 8.6 mg L⁻¹) was slightly higher than that observed when culturing M. caribbica NJC41 for 2 days (44.6 mg L⁻¹). ²⁸ Phenethyl alcohol is commercially available and inexpensive ($0.07/g), suggesting that the direct application of phenethyl alcohol on soybean plants could be a suitable alternative for target spot control.

As with NJC30, M. caribbica NJC41 also competed with C. cassiicola for space in soybean pods and for nutrients in a soybean juice‐based medium. ²⁸ Serratia marcescens, which also belongs to the Enterobacteriaceae family, has also been reported to compete for nutrients and space with the fungal pathogen Rhizoctonia. ⁵⁶

Although a few soybean varieties, such as ‘Bedford’ and ‘Council’, have been reported to exhibit moderate resistance to C. cassiicola, ²⁶ , ²⁷ none of the varieties allow a complete control of the disease symptoms. ⁵⁷ For this reason, C. cassiicola management mainly relies on fungicide application. Here, an alternative efficient method for C. cassiicola management has been developed. It was observed that the higher the NJC30 concentration, the higher the inhibitory efficacy. This can be explained by the fact that higher concentrations of NJC30 cells may lead to greater accumulation of phenethyl alcohol and in higher competition for space and nutrients. The higher preventive efficacy compared with curative efficacy can be explained by the fact that the preventive application may allow NJC30 to proliferate in the soybean pods forming biofilms and to produce phenethyl alcohol, preventing C. cassiicola colonization. Instead, in curative application, C. cassiicola is already colonizing the soybean pods when NJC30 is applied.

NJC30 showed one of the highest efficacies among reported agents for target spot management. The curative efficacy of M. caribbica NJC41 for target spot control in soybean pods was only 47.1%. ²⁸ Fluazinam (0.1 mg mL⁻¹) was reported to decrease target spot symptoms in cucumber leaves by 97.47% and 58.23% in preventive and curative applications, respectively. ⁵⁸ In C. cassiicola‐infected cucumber leaves, benzovindiflupyr (0.15 mg mL⁻¹) exhibited preventive and curative efficacies of 85% and 80%, respectively, whereas fluxapyroxad (0.15 mg mL⁻¹) reduced target spot symptoms by 60% and 58% in preventive and curative applications, respectively. ⁵⁹

Although strain NJC30 was originally isolated from common comfrey (S. officinale) leaves, its successful colonization and biocontrol efficacy on soybean pods represents a key finding of this study. NJC30 exhibited the ability to effectively colonize soybean pod surfaces and proliferate, confirming its adaptability to different niches. NJC30 exhibited high efficacy for controlling target spot on both soybean pods and leaves, suggesting that NJC30 can be used for the management of C. cassiicola on different soybean plant tissues. The fact that NJC30 was isolated from common comfrey but successfully colonized soybean tissues suggests that NJC30 may be able to colonize different plant species, which could be a relevant characteristic for its actual application. Further studies should assess NJC30 colonization ability across different plant species.

NJC30 can be easily grown in common culture media, indicating that its production may not incur significant costs. Although Citrobacter strains are not typically major human pathogens, some species can cause opportunistic infections. Some Citrobacter strains have been associated with urinary and bloodstream infections in hospitalized immunocompromised patients, with Citrobacter freundii being the main causative agent. ⁶⁰ For this reason, toxicity studies are necessary to confirm that NJC30 is non‐pathogenic to humans and to support its potential agricultural application.

Collectively, a new biocontrol method for the management of C. cassiicola on soybean pods was developed in this study. This method was based on the application of NJC30, which was isolated from common comfrey leaves. NJC30 inhibited C. cassiicola mycelial growth and spore germination, and showed various antifungal mechanisms, including production of antifungal phenethyl alcohol, and competition for space and nutrients with the pathogen. NJC30 inhibited C. cassiicola‐caused symptoms on soybean pods and leaves in curative and preventive applications, showing higher preventive efficacy than the commercial fungicide carbendazim. NJC30 exhibited one of the highest efficacies among the reported methods for target spot management. Citrobacter has been used to promote plant growth and enhance plant tolerance to abiotic stress. However, this is the first study on the use of Citrobacter for the management of fungal plant diseases. This study reveals that the Citrobacter genus is an unexplored potential source of efficient biocontrol agents.

CONFLICT OF INTEREST

The authors declare no conflict of interest.

Supporting information

Table S1. Primers used for the amplification of the 16S rRNA, gyrB, and rpoB genes in NJC30.

Table S2. Sequences of the 16S rRNA, gyrB, and rpoB genes from NJC30.

Table S3. Strains used for the construction of the phylogenetic tree.

Figure S1. Antifungal activity in vitro of NJC30 against various fungal pathogens. The images show the co‐cultures of NJC30 with Alternaria alternata, Aspergillus flavus, Penicillium oxalicum, and Sclerotinia sclerotiorum. NJC30 inhibited the mycelial growth of the fungal pathogens. Control experiments were carried out by culturing the fungal pathogens in the absence of NJC30.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.