To achieve sterilization of 3 g of soybean seeds (Glycine max cv. Daewon), the seeds were aseptically transferred into 50-ml tubes. Subsequently, they underwent a 30-s exposure to 70% ethanol. After, the ethanol supernatant was decanted, and a 1% NaOCl solution was added to the seeds. The seed-NaOCl mixture was vigorously vortexed for 1 min. Following this, the seeds underwent thorough washing with ddH₂O for 1 min, and this rinsing process was repeated three or more times. Once the rinsing sequence was completed, the seeds were carefully placed within sterilized cotton paper in a Petri dish. The dishes were then incubated for 3 days at 28°C under dark conditions. After germination, the sprouted seeds were transferred to individual plastic square pots (18 cm × 18 cm), ensuring a distance of more than 20 cm between each pot to prevent cross-contamination. The plants were cultivated in a controlled glass greenhouse with a light exposure of 14 h at a temperature of 28 ± 3°C, followed by a dark period of 10 hours at a temperature of 15 ± 3°C. This growth schedule was maintained for a duration of 28 days.

The genome of S. bacillaris S8 was previously characterized by Kim and Kwak (2023), underscoring its importance in the synthesis of antifungal metabolites effective against large patch disease (Jeon et al., 2021). To identify potential secondary metabolite gene clusters within this strain, we utilized the antiSMASH 7.0 version tool (https://docs.antismash.secondarymetabolites.org) (Blin et al., 2023). The left flanking and right flanking regions for each deleted gene were amplified using 1 μg of strain S8 genomic DNA and gene-specific primers. The PCR amplification for each gene’s left flanking and right flanking regions followed a protocol comprising an initial denaturation at 98°C for 10 min, 30 cycles of denaturation at 98°C for 1 min, annealing at 52-58°C for 30 s (according to the respective primer melting temperature) (Supplementary Table 1), elongation at 72°C for 30 s, and a final extension at 72°C for 5 min. Subsequently, the left flanking and right flanking regions were fused through PCR, involving an initial denaturation at 98°C for 30 s, followed by five cycles of denaturation at 98°C for 8 s, annealing at 50°C for 30 s, elongation at 72°C for 30 s, and a final extension at 72°C for 5 min. Finally, the target homologous regions, including left flanking and right flanking, were amplified, and the last PCR step adhered to conditions similar to the earlier left flanking and right flanking region PCR, with the annealing temperature set at 60°C. The PCR procedures were executed using the MJ Research RTC-200 Thermal Cycler (Bio-Rad, Hercules, CA, USA).

The purified molecules were ligated to the pGEM-T easy vector (Promega, Madison, WI, USA) using T4 ligase (Promega) and incubated at 16°C for over 20 h. The ligated constructs were transformed into Escherichia coli DH5α via bacterial transformation. Subsequently, the vector and pCRISPomyces-gRNA were digested using XbaI (NEB, Ipswich, MA, USA), followed by elution into 30 μl of EB buffer. A 3:1 molar ratio ligation was performed using 400,000 units (400 units/μl) of T4 ligase. The gRNA was cloned into pBHA (2,002 bp), and pCRISPomyces-2 (Cobb et al., 2015) was digested using BbsI (NEB). The resulting fragments were ligated using T4 ligase, connecting the two fragments and subsequently linking them with the gRNA (guide RNA). The resultant plasmids were then transformed into E. coli ET12567 containing pUZ802 (Bierman et al., 1992).

The transformed cells cultured and underwent centrifugation (1,914 ×g, 15 min), followed by two gentle washes with 10 ml of Luria-Bertani (LB) solution. Each wash was accompanied by centrifugation. The resulting pellet was resuspended in 500 μl of LB broth medium. In a separate E-tube, 500 μl of 2× YT broth medium (10 g yeast extract, 16 g tryptone, 5 g NaCl per liter) was heated at 50°C for 10 min. Subsequently, 10 μl of strain S8 spore stock (10⁹ cfu/ml) containing 20% glycerol was mixed with the heated 2× YT broth. The washed E. coli (ET12567/pUZ8002) cells were gently mixed with the strain S8 spore stock using a pipette and spread onto mannitol soya agar (SFM; 20 g mannitol, 20 g soya flour per 1 liter) medium (Kieser et al., 2000). After 12 h, the agar plates were supplemented with apramycin (50 μg/ml) and nalidixic acid (25 μg/ml). After 7 days, colonies were transferred twice onto SFM agar containing apramycin (50 μg/ml).

C. truncatum and Fusarium sp. were grown onto potato dextrose agar (PDA; 24 g of potato dextrose, 20 g of agar per liter). C. truncatum underwent cultivation at 25°C for 10 days, while Fusarium sp. was cultured at 27°C for 7 days. The production of the Fusarium sp. inoculum as mixture of 99 g of sand and 1 g of oatmeal within a 1-liter plastic bottle, subjected to three cycles of sterilization. Subsequently, this sterilized mixture was inoculated with Fusarium sp., previously cultured on PDA medium for 7 days. The inoculated mixture was cultivated at a temperature of 28°C for 3 weeks to facilitate chlamydospore preparation. To assess the biological control efficacy of wild-type strain S8 and mutant strains, each was streaked on mannitol soya agar (SFM; 20 g mannitol, 20 g soya flour per liter), and cultured at 28°C for 5 days. A single colony from each strain was sub-cultured in 5 mL PDK (10 g potato dextrose broth, 10 g peptone per liter) broth at 28°C for 3 days. The cultured cells were then transferred to 200 ml of PDK broth and incubated at 28°C for 7 days, after which cell densities were adjusted to 10⁶ cfu/ml.

A greenhouse measuring 5 m in length was utilized to establish distinct zones for both infected and non-infected control groups. Subsequently, a suspension containing the wild-type strain S8 and mutant strains was prepared with a final concentration of 0.1% carboxymethyl cellulose. This suspension was applied to the proximity of soybean roots and leaves in both groups. Five days post-inoculation, C. truncatum was administered to each plant at a rate of 10 ml per plant using a spray method. Conversely, for Fusarium sp., a drenching method was employed, administering 10 ml (10⁵ spores) of the suspension per plant. Disease severity assessments were conducted at weekly intervals following pathogen inoculation, utilizing a 5-level index. For C. truncatum, the severity scale encompassed the following categories: 0 = no significant damage, 1 = spots covering <1-10% of the plant, 2 = spots covering 10.1-15% of the plant, 3 = spots covering 15.1-25% of the plant, 4 = complete plant death (Agam et al., 2019). In the case of Fusarium sp., the severity scale was defined as follows: 0 = no symptoms, 1 = mild symptoms (discoloration without visible lesions), 2 = obvious lesions (severe discoloration with lateral root reduction), 3 = severe lesions on the primary and lateral roots and diminished plant vigor, and 4 = stem rotten, plant dead (Chang et al., 2020).

Streptomyces bacillaris S8 has previously been acknowledged for its potent biocontrol capabilities against large patch disease in turfgrass, with the identification of the valinomycin biosynthetic gene cluster (Jeon et al., 2019). Building upon this, our investigation aimed to explore its potential as a biocontrol agent in soybeans, motivated by its relevance to plant health (Kim and Kwak, 2023). The identified gene clusters within the bacteria were found to include genes associated with the production of nonribosomal peptide synthetase (NRPS) or polyketide synthase (PKS) compounds, as detailed in Table 1. Various biosynthesis knockout mutants targeting these clusters were generated using the CRISPR/Cas9 system, to assess their antifungal activities against two soybean phytopathogenic fungi (Table 2).

In the present study, we generated three mutant strains in which nrps/pks2, nrps3, or nrps4 genes were abolished from the wild-type S8 strain (Fig. 1). Notably, the mutants of strain S8 in the nrps3 and nrps/pks2 clusters exhibited reduced efficacy against root rot disease, as evidenced by diminished activity in biocontrol assays (Fig. 2, Supplementary Figs. 1 and 2). Conversely, two distinct mutants showed no significant suppression of the two pathogens, with a statistical significance of P = 1.180e-10 (*** on C. truncatum) and P = 7.770e-13 (*** on Fusarium sp.) (Fig. 3). These observations lend support to the hypothesis that the NRPS and PKS serves as of antibiotics crucial for protecting the plant.

The biosynthesis genes were identified in Streptomyces spp., specifically in Streptomyces albus strain BK 3-25, which is known for salinomycin production, and in the deep-sea-derived S. atratus SCSIO ZH16NS-80S known to produce atratumycin (Almeida-Paes et al., 2021; Yang et al., 2019). Salinomycin has been reported for its anticancer and antifungal properties against Paracoccidioides brasiliensis (Zhou et al., 2013). While information regarding the antifungal effect of atratumycin is limited, its expression level was observed to increase in a co-culture system the fungus Nigrospora oryzae roseF7 and Streptomyces sp. MW-W600-10 (Siupka et al., 2021). This suggests that two distinct NRPS or PKS related compounds play a role in the antifungal effect against plant pathogens. Furthermore, atratumycin exhibited specific activity against C. tuncatum and Fusarium sp., suggesting its promise as a strain-specific biocontrol agent for the S8 strain.

The diminished biocontrol activity of specific mutants underscores the importance of these antibiotic clusters in combatting plant diseases. This study contributes valuable insights into the potential use of Streptomyces strains for sustainable crop disease management, with far-reaching implications for agricultural practices. Additionally, the findings provide a foundation for basic research in environmental biological control, offering potential strategies for controlling soybean diseases and enhancing overall plant health.