Rhizosphere soils were collected from Kaijiang, Sichuan province (31°08′N, 107°87′E). Rhizosphere soils were diluted with sterile water and dilutions of 10^–3 to 10^–9 were selected for bacterial isolation. Bacteria were separated on nutrient agar (NA) and potato dextrose agar (PDA) at 28°C. All morphologically distinct colonies were collected for confrontation culture on PDA with the indicator strains M. oryzae and B. cinerea. Moreover, 16S was amplified by the specific primer (27F: AGA GTT TGA TCM TGG CTC AG; 1,492R: CGG TTA CCT TGT TAC GAC TT) (Lane, 1985) and the phylogenies were reconstructed using the neighbor-joining (NJ) method using MEGA 7.0 software (Kumar et al., 2016).

The antifungal activity of 183 diverse bacterial species was tested in a crossing-culture assay against B. cinerea and M. oryzae: 18 species showed a clear zone of growth inhibition. For further antagonistic tests, KJ-34 was inoculated on one side of a PDA plate and the fungi (Ustilaginoidea virens, Alternaria solani, Fusarium oxysporum, Phytophthora capsica, Corynespora cassiicola, Magnaporthe oryzae, Rhizoctonia solani, Bipolaris maydis, Botrytis cinerea, Colletotrichum lagenarium, Epicoccum nigrum, Fusarium proliferatum, Alternaria brassicae, and Phomopsis vexans, which are maintained by the Laboratory of Integrated Plant Disease Control, Northwest Agriculture and Forestry University) mycelial plugs were inoculated on the other side after KJ-34 growth at 28°C for 1 day. At least three plates were used for each phytopathogenic fungus.

The culture media used in this study included Luria-Bertani (LB), Potato Dextrose Agar (PDA), Potato-Saccharose-Agar (PSA), Nutrient Agar (NA), Complete medium (CM), Ashby nitrogen-free medium and Tryptic Soytone Agar (TSA), Ashby nitrogen-free medium, and TSA. The optimum culture medium for measuring the antifungal activity of KJ-34 was determined by testing various combinations of different carbon components (sucrose, fructose, maltose, xylose, galactose, raffinose, mannose, sorbitol, corn flour, etc.) and nitrogen components (NH₄Cl, (NH₄)₂SO₄, KNO₃, NH₄NO₃, beef paste, peptone, yeast extract, potato powder, etc.) Different culture conditions, such as fermentation time, fermentation temperature, and shaker rotation speed, were also tested to determine the optimum conditions

The M. oryzae strain Guy11 was incubated on CM at 28°C for 5 days. For M. oryzae, the spores were collected and sprayed on 14-day-old rice leaves (LTH) with a final concentration of 1 ×10⁵ (Sakulkoo et al., 2018). Lesions produced in each treatment were counted. B. cinerea was obtained from a local greenhouse and was incubated on PDA medium at 28°C for 5 days. The clumps of B. cinerea were placed on the center of the leaves of tomato landraces, and lesion areas were photographed and measured using ImageJ. Spores and mycelia were observed under an optical microscope (Olympus, Japan).

For the identification of specialized metabolites produced by KJ-34, KJ-34 was shaken (5 days, 28°C, 200 rpm) in PDB medium under optimum flask culture conditions; samples were collected and stored at − 80°C prior to ultra-performance liquid chromatography (UPLC)–tandem mass spectrometry (MS/MS) analysis using equipment manufactured by Metware Biotechnology Co., Ltd ., Wuhan, China.

Antagonism between microorganisms has attained a broad consensus: biocontrol strains from local areas of agricultural farms are eco-friendly and important for sustainable agriculture. To screen the antifungal activities of bacteria, rhizosphere soils were collected from local rice paddies in Kaijiang, Sichuan province (31°08′N, 107°87′E). Bacteria were separated on NA and PDA at 28°C. A total of 183 morphologically distinct bacterial species were tested for their antifungal activity in crossing-culture assay against B. cinerea and M. oryzae: 18 of them showed a clear zone of growth inhibition (Figure S1). Among these strains, KJ-34 exhibited the highest antifungal activity on the PDA plate; we then confirmed its antagonistic activities against B. cinerea, as shown in Figures 1A, B. Strain morphology was also observed using a scanning electron microscope (Figure 1C): 16S rRNA was amplified and sequenced to construct the phylogenetic tree, which revealed a high degree of similarity with B. gladioli (Figures 1D, E).

In order to determine the most efficient culture medium for the isolation of antifungal compounds, we tested the growth velocity of KJ-34 on different culture media (LB, PDA, PSA, NA, and TSA) by measuring the bacterial concentration at 6, 12, 24, 48, 72, 96, and 120 hours. As shown in Figures S2A, B, in the first 48 hours of shaking flask fermentation, KJ-34 grew most rapidly in TSB, whereas after 72 hours, growth of KJ-34 was most rapid on PDB medium and the OD600 reached 1.94 (TSB and PDB are fluid medium of TSA and PDA without agar). The morphology of KJ-34 grown on solid medium was markedly different from that of KJ-34 subjected to shaking flask fermentation, although the growth ratio was similar. As shown in Table 1, the optimum flask culture conditions were selected through orthogonal tests; the optimum flask culture conditions were 2% cornmeal, 1.5% beef extract, and 0.01% FeSO₄ (5 days, 28°C, 200 rpm). For the antifungal activity against M. oryzae. As shown in Figure S3, the antifungal activity of KJ-34 against M. oryzae was maximized by using CM as the fermentation broth and by shaking flask fermentation at 28°C in PDB for 6 days; the antifungal activity against M. oryzae was significantly increased by shaking for 5 days. Growth of To test the inhibition of KJ-34 shake flask fermentation against plant pathogen (6 days, 28°C, 200 rpm).

To test the inhibition of KJ-34 shake flask fermentation against plant pathogen, we used B. cinerea as the representative strain. After 5 days of fermentation under optimum flask culture conditions, the fermentation broth was filtered and the filtrate diluted (concentrations: 100%, 10%, and 1%). As shown in Figure 2A, B and Table S1, the disease inhibition rate was 98.2% after treatment with 100% fermentation broth had a control efficiency of 75.6%, but the 1% fermentation broth treatment was also quite effective, with a control efficiency of 45.8%. Furthermore, the protective effect of KJ-34 fermentation filtrate was also analyzed, as shown in Table 2: control efficiency of complete fermentation broth treatment was 75.6%, and even the centesimal fermentation broth treatment was 45.8%. KJ-34 is thought to have a great developing potential in the control of plant fungicidal disease. To understand the effects of the KJ-34 strain on the growth of B. cinerea, the bacterial broth-treated spore germination was observed, as shown in Figure 2C, D. Spore germination in B. cinerea was inhibited by 24 hours’ treatment with KJ-34 metabolites and the results are consistent with the effective inhibition of tomato gray mold disease. The B. cinerea hypha were also swollen and deformed after co-culture. Scanning electron microscopy of mycelia also showed the same results. These results indicate that KJ-34 produces various antifungal metabolites, for use in the development of bactericides and fungicides.

To test the ability of the KJ-34 shaking flask fermentation filtrate to inhibit rice blast fungus, the fermentation broth was filtered after 5 days of fermentation with the optimum flask culture conditions added in CM medium. As shown in Figure 3A, the growth of M. oryzae were significantly inhibited and the bacterial broth treated spore germination and mycelial growth were observed, as shown in Figure 3B, M. oryzae spore produced more appressorium thus was failure to complete infection. The germinal tubes were also malformed. The M. oryzae hypha were swollen and deformed by co-culture with KJ-34 on plates containing solid PDA medium. Moreover, the fermentation broth was filtered after 5 days of fermentation under optimum flask culture conditions and the fermentation filtrate was then diluted (concentrations: 100%, 10%, and 1%). Spores of M. oryzae were mixed with the diluted fermentation filtrate and sprayed on to the rice leaves (Figure 3C, D): lesion numbers reduced by 34.6% after treatment with 1% fermentation filtrate, and by 77.5% after treatment with 100% fermentation filtrate. The results are consistent with the effective inhibition of tomato gray mold disease. These results indicate that KJ-34 has great potential in plant protection.

Fungal growth inhibition was observed in the co-culture experiments. As shown in Figure 4 and Table 2, several well-known fungi (Ustilaginoidea virens, Alternaria solani, Fusarium oxysporum, Phytophthora capsica, Corynespora cassiicola, Magnaporthe oryzae, Rhizoctonia solani, Bipolaris maydis, Botrytis cinerea, Colletotrichum lagenarium, Epicoccum nigrum, Fusarium proliferatum, Alternaria brassicae, and Phomopsis vexans) that are pathogenic to crops, vegetables, or fruits were grown along with KJ-34 on solid PDA medium. KJ-34 strongly inhibited the filamentous growth of the fungal pathogens tested. KJ-34 co-culture with C. lagenarium significantly inhibited growth of the fungus, and the diameter of the inhibitory zone was 41.2 mm. The growth of other fungi co-cultured with KJ-34 was also inhibited, but the magnitude of the inhibitory effect did not depend on the concentration of the fermentation filtrate. The flat growth restraint of the other fungi co-cultures with KJ-34 also showed an obviously inhibitory effect. The rice-pathogenic fungi U. virens and R. solani were only weakly inhibited, largely because their growth rate was either too fast or too slow, and another reason could be because of the nutrient competition and the accompanying of antifungal compounds change.

KJ-34 appeared to produce antimicrobial metabolites effective against multiple phytopathogenic fungi. To explore the antifungal activity compounds produced by KJ-34, we first dialysis the medium optimized with 10 kDa filter attenuates and then 1 kDa (Li et al., 2019). The metabolites with a molecular weight below 1,000 Da produced by KJ-34 also exhibited high inhibitory ability and a broad antibiotic spectrum (data not shown). With these results, we performed the untargeted metabolomics to explore the pronounced inhibitory effects of KJ-34 against M. oryzae as compared with the medium before optimized. As shown in Figure S4 and Table S2 and S3, 2,352 metabolites were included in the positive ionization model and 663 metabolites were included in the negative ionization model. The detected metabolites were subjected to qualitative and quantitative analyses and, based on the results, divided into groups. Fold changes in the quantitative information for the metabolites in each group were then compared. Differential metabolites with a fold change of ≥ 6 are listed in Tables S3, S4, and metabolites annotated in accordance with the KEGG (Kyoto Encyclopedia of Genes and Genomes) pathway database are displayed in Figure S4. Differential metabolites are involved in 2-oxocarboxylic acid metabolism, ovarian steroidogenesis, protein digestion and absorption, serotonergic synapse pathway, steroid hormone biosynthesis, bile secretion, porphyrin and chlorophyll metabolism, and arachidonic acid metabolism. After optimizing the flask culture conditions, metabolites significantly increased in both type and quantity (Figure 5A, B, Figure S4).

We also analyzed the differential metabolites from multiple perspectives according to the characteristics of publicly available databases, and finally accurately analyzed the main differential antifungal compounds that may be responsible for the broad antibiotic spectrum (Figure 5C). Metabolites marked with a yellow background in Table S3 and S4 (Ac-Yvad-cho, benzoylstaurosporine, TAXOL C, morellin, jubanine B, trichostatin A, thapsigargin, kabiramide B, scopolamine, enniatin B, latrunculin A, rifaximin, rigin, and garcinone C) appear to be promising bioactive compounds. These results are consistent with antifungal activity against a wide range of phytopathogenic fungi. Moreover, oligopeptides were identified in the metabolome; whether or not the oligopeptides are functional needs further validation (Tables S4 and S5). KJ-34 is not surprising to be a library of antifungal compounds.