The P. frutescens samples used in this experiment were collected from Zisu Garden of Huanan Forestry Bureau (Heilongjiang, China). The plant material was divided into root, stem, and leaf sections, and subjected to seven-step surface sterilization procedure as previously described (Liu et al., 2019). After thorough drying under sterile conditions, the samples were ground using a mortar and pestle. Subsequently, 200 μL of the suspension was plated onto five different actinobacteria-selective media types, including Gause’s synthetic (GS) agar no. 1 (Atlas, 1993), sodium succinate-asparagine agar (SSA) (Li et al., 2012), cellulose-proline agar (CPA) (Qin et al., 2009), dulcitol-proline agar (DPA) (Guan et al., 2015), and amino acid agar (AAG) (Liu et al., 2019) and incubated at 28°C for 2–3 weeks. All media were supplemented with nalidixic acid (20 mg/L) and cycloheximide (50 mg/L) to inhibit the growth of Gram-negative bacteria and fungi. Then, the grown colonies were subsequently transferred onto oatmeal agar (ISP3) for further analysis (Shirling and Gottlieb, 1966).

Antimicrobial screening was conducted against 17 phytopathogenic fungi and 1 phytopathogenic bacterium, including Bipolaris sorokiniana NEAU-W4, Rhizoctonia solani CAAS-F11, Fusarium moniliforme NEAU-M1, Fusarium graminearum PH-1, Fusarium oxysporum f. sp. cucumerinum CAAS-F13, Fusarium oxysporum f. sp. melonis CAAS-F14, Fusarium solani CAAS-F15, Fusarium verticillioides NEAU-C3, Cochiioboius heterostrophus CAAS-F27, Setosphaeria turcica CAAS-F35, Corynespora cassiicola CAAS-F22, Alternaria tenuissima NEAU-5, Alternaria cucumerina NEAU-7, Alternaria alternata NEAU-10, Verticillium dahlia CAAS-F43, Botrytis cinerea CAAS-F48, Stagonosporopsis cucurbitacearum CAAS-F48 and Ralstonia solanacearum CAAS-B1. B. sorokiniana, F. moniliforme, F. verticillioides, A. cucumerina, A. tenuissima, and A. alternata were maintained at the Key Laboratory of Agricultural Microbiology of Heilongjiang Province (Harbin, China). R. solani, F. oxysporum f. sp. cucumerinum, F. oxysporum f. sp. melonis, F. solani, C. heterostrophus, S. turcica, C. cassiicola, S. cucurbitacearum, B. cinerea, and R. solanacearum were obtained from Chinese Academy of Agricultural Sciences (Beijing, China). F. graminearum was kindly provided by Zhejiang University (Hangzhou, China). For R. solanacearum, the bacterium was cultured on nutrient agar (NA) (Waksman, 1961) at 37°C for 24 h. The antagonistic activity of the isolates against R. solanacearum was assessed using the agar cylinders method. A bacterial suspension of 100 μL (OD₆₀₀ = 0.3) was plated onto NA agar. A fresh mycelial agar plug (5 mm diameter) from isolates cultured on ISP3 medium for 7 days was placed in the center of NA plate and incubated at 37°C. The diameter of the inhibition zone was measured after 24 h. Regarding the fungi, they were cultured on potato dextrose agar (PDA) (Cao et al., 2017) at 28°C. The antifungal activity of the isolates were evaluated using the dual culture plate assay (Hamzah et al., 2018). The isolates were point-inoculated at the margin of PDA plates and incubated for 3 days at 28°C, after which a fresh mycelial PDA agar plug of the fungus (5 mm diameter) was transferred to the opposite margin of the corresponding plate. After additional days of incubation at 28°C for 3–7 days, the diameter of the inhibition zones was measured, and the inhibition rates were calculated according to the following formula: (the width of inhibition/the width between pathogen and antagonistic bacteria) × 100%. All experiments were performed in triplicate.

Genomic DNA extraction and PCR amplification of the 16S rRNA gene sequence were performed according to previous study (Wang et al., 2020). The phylogenetic tree was constructed using the neighbor-joining algorithm implemented in Molecular Evolutionary Genetics Analysis (MEGA) software version 11 (Tamura et al., 2021). The stability of the phylogenetic tree topology was assessed using the bootstrap method with 1,000 repetitions (Felsenstein, 1985). An evolutionary distance matrix was calculated according to Kimura’s two-parameter model (Kimura, 1980). All positions containing gaps and missing data were completely eliminated from the dataset. The obtained 16S rRNA gene sequences were aligned using the EzBioCloud server (Yoon et al., 2017).

The biofertilizer of NEAU-ZSY13 was mixed in the soil (alkaline chernozem) at the final concentrations of 10⁶ CFU/g, 10⁷ CFU/g, and 10⁸ CFU/g of soil. The soil used in the experiment was steam-sterilized at 121°C for 30 min. Tomato seeds (Aoguan 1, moderate susceptible to R. solanacearum) were subjected to surface sterilization using 70% ethanol for 1 min, followed by a 2-min soak in 2% NaClO solution. After treatment, the seeds were pregerminated for 2 days at 25°C and then placed in seedling trays under natural conditions. Once they reached the four-leaf stage, the seedlings were transplanted individually into pots containing the biofertilizer, one seedling per pot. After a period of 7 days, a 10 mL suspension of R. solanacearum (OD₆₀₀ = 0.3) was applied to the soil surrounding the seedlings. Each treatment consisted of 10 pots, and the entire experiment was repeated three times, independently. These experiments were conducted in a growth chamber set to simulate a 12-h photoperiod. The chamber maintained a temperature of 37°C and humidity level ranging between 75 and 90%. The duration of the experiment was 14 days. The extent of the disease was assessed using a five-class scale, where 0 indicated no wilt, 1 represented 1–25% leaf wilt, 2 denoted 25–50% leaf wilt, 3 indicated 51–75% leaf wilt, and 4 indicated over 75% leaf wilt. To quantify the disease severity, the Disease Index (DI) was computed using the following formula: (Σ(plant numbers with the same rating of disease severity × disease rating)/(maximum rating value × total number of plants)) × 100. Furthermore, the biological control efficiency was calculated using the formula: ((DI of the negative control - DI of the treatment)/DI of the negative control) × 100%.

The biofertilizer was combined with the alkaline chernozem to achieve spore concentrations of 10⁶ CFU/g, 10⁷ CFU/g, and 10⁸ CFU/g of soil. The soil used in the experiment was steam-sterilized at 121°C for 30 min. The pathogenic fungus B. sorokiniana was cultured on wheat grains at 28°C for 10 days to obtain the disease inoculum, then mixed with the soil at a ratio of 1% (w/w). The variety Longmai 33, which is moderately susceptible to B. sorokiniana, was selected for the experiment. To prepare the seeds, the seeds were soaked in water for 24 h and subsequently planted in pots containing a mixture of disease inoculum and biofertilizer with different spore concentrations. As a negative control, a treatment without biofertilizer was included. Each treatment consisted of 10 pots, with one seedling per pot. The experiment was replicated three times. The potted plants were maintained at 25°C and a humidity level of 70% for a duration of 14 days. The severity of the disease was evaluated using a five-point scale, corresponding to the percentage of roots and above-ground parts displaying discoloration symptoms: 0 denoted no symptoms, 1 represented 1–25% affected, 2 indicated 26–50% affected, 3 denoted 51–75% affected, and 4 indicated over 75% affected. The DI and biological control efficiency were computed based on the formulas previously mentioned.

NEAU-ZSY13 was cultivated on ISP3 medium at a temperature of 28°C for 7 days and then transferred to 250 mL baffled Erlenmeyer flasks containing 50 mL of sterile seed medium (GY). The cultivation was carried out for 3 days at 28°C with shaking at 250 rpm. After that, 18 mL aliquots of the culture were aseptically transferred into 1 L baffled Erlenmeyer flasks containing 300 mL of the production medium. The production medium consisted of malt extract (1%), glucose (0.4%), yeast extract (0.4%), soluble starch (2%), and CaCO₃ (0.2%) with a pH of 7.2–7.4. The flasks were incubated at 28°C for 1 week with continuous shaking at 250 rpm. Following the fermentation period, the resulting broth (18 L) was centrifuged at 4,000 rpm for 10 min. The supernatant was collected and extracted three times with ethyl acetate. The extract was then evaporated under vacuum, resulting in 8.68 g of crude oily extract. The mycelia were extracted with methanol (1 L), and the resulting extract was concentrated under vacuum to remove the methanol. This process led to the production of 25.13 g of concentrated extract. Both of these extracts exhibited identical sets of metabolites, as confirmed by High-Performance Liquid Chromatography (HPLC) analysis (Figures 1A,B). Due to the similarity in metabolite profiles, the two extracts were combined for subsequent purification steps. The crude extract was subjected to column chromatography using a silica gel column (100–200 mesh). Elution was performed using a series of solvents, starting with petroleum ether/ethyl acetate mixtures (5/1, 1/1, 1/5, 1/10, and 0/1, v/v), followed by ethyl acetate/methanol mixtures (10/1, 5/1, 1/1, 1/5, and 1/10, v/v). This process resulted in the separation of 10 fractions labeled as fractions 1 through 10. Notably, fraction 10 demonstrated bioactivity. Subsequently, fraction 10 was further purified using a Sephadex LH-20 column, eluting with methanol. This step resulted in the collection of three sub-fractions denoted as F10A, F10B, and F10C. Among these, the fraction displaying bioactivity was F10B. Subsequently, F10B was subjected to semipreparative HPLC using a YMC-Triart C18 column. The mobile phase was a mixture of methanol and water (78/22 v/v) at a flow rate of 3 mL/min. This yielded compounds 1 (retention time, t_R = 11.5 min, 8 mg) and 2 (t_R = 17.5 min, 8.8 mg) (Figure 1C). The structures of compounds 1 and 2 were determined through spectroscopic analysis. NMR spectra were acquired using a Bruker Avance III-400 spectrometer in CD₃OD, with TMS as the internal standard. ESI-MS data were obtained using an Agilent G6230 Q-TOF mass spectrometer.

To assess the antibacterial activity of compounds 1 and 2 against R. solanacearum, a bacterial suspension (30 μL) was inoculated into 5 mL of NA broth containing different concentrations of the compounds. The tubes were then incubated at 37°C for 24 h, and the OD₆₀₀ was measured. The tube containing the same amount methanol was used as a negative control. The assay was conducted in triplicate to ensure accuracy and reliability of the results. The percentage of inhibition was calculated using the formula [1 - (mean OD₆₀₀ value of the treatment/mean OD₆₀₀ value of the negative control)] × 100%. Linear regression analysis was conducted on the data to determine the effective concentrations required for 50% inhibition (EC₅₀).

For the antifungal activity assay against B. sorokiniana, compounds 1 and 2 were dissolved in methanol and then diluted to various concentrations with water. These solutions were added to PDA medium. A fungal plug measuring 5 mm in diameter, taken from B. sorokiniana, was placed at the center of each PDA plate, which was then incubated at 28°C for 7 days. A control plate containing the same amount methanol was used as a negative control. The experiments were performed in triplicate to validate the findings. The percentage of inhibition was calculated using the formula [1 - (mean colony diameter of the treatment/mean colony diameter of the negative control)] × 100%. Linear regression analysis was conducted on the data to determine the EC₅₀ value.

All experiments were conducted in triplicate, and all data were presented as the mean values ± standard deviation. Data processing and statistical analysis were performed using SPSS statistical software package (SPSS Inc., Cary, NC, United States, v.26). Multiple comparisons were performed on the data using a one-way ANOVA program, followed by a Waller-Duncan multiple range test. Differences were considered significant when the probability was less than 0.05.

A total of 76 isolates were obtained from P. frutescens, comprising 24 from leaf, 17 from stem, and 35 from root. Among these, one particular isolate, NEAU-ZSY13, retrieved from leaf using GS no. 1 medium, displayed robust antagonistic activity against all tested fungi, showcasing inhibition rates spanning 53.3–84.8% (Figure 2). Notably, it exhibited remarkable antifungal potency against B. sorokiniana. Additionally, NEAU-ZSY13 suggested potent antibacterial capability against R. solanacearum, marked by a 24.3 mm inhibition diameter (Figure 2A).

The morphology of NEAU-ZSY13 exhibited typical characteristics of the genus Streptomyces, with white-colored mycelia when cultured on ISP3 medium at 28°C for 7 days (Figure 3). Elucidation of the 16S rRNA gene sequence provided unequivocal substantiation of its taxonomic classification within the genus Streptomyces, revealing a maximal sequence identity of 100% with Streptomyces melanosporofaciens DSM 40318. In the 16S rRNA-based phylogenetic tree (Figure 4), NEAU-ZSY13 formed a subclade with S. melanosporofaciens DSM 40318. Hence, drawing from these findings, NEAU-ZSY13 has been classified as a species within the genus Streptomyces.

In this study, the growth of NEAU-ZSY13 was investigated using wheat bran and vermicompost as solid fermentation substrates. Through systematic analysis, the optimal substrate ratio was identified as a 1:2 (vermicompost to wheat bran by mass), yielding a spore production of 2.27 × 10⁹ CFU/g (Supplementary Figure S1A).

Results from the single-factor tests revealed significant influences of water content, inoculum amount, and temperature on spore production, with pH exerting minor effect (Supplementary Figures S1B–E). Based on these findings, water content, inoculum amount, and temperature were selected as fixed variables for the subsequent BBD. The BBD encompassed 17 runs, systematically investigating the impact of each factor on spore production (Supplementary Table S1). The biomass yield was measured as the response variable, and the obtained data were fitted to a polynomial equation as follows:

Y=9.49+0.205X1+0.1163X2+0.0662X3−0.0025X1X2+0.0575X1X3+0.0050X2X3−0.9875X1 2−0.3500X2 2−0.4200X3 2

The statistical significance of the experimental model was rigorously evaluated through the application of ANOVA, as elucidated in Supplementary Table S2. Factors with a p-values beneath the 0.05 threshold were designated as statistically significant. Evidently, the model exhibited a pronounced degree of congruence, as evidenced by the coefficient of determination (R²) value of 0.9888. This value substantiates a robust correlation between the empirical and predicted outcomes. Furthermore, the lack of fit test rendered an insignificantly elevated p-value of 0.7878 (>0.05), thereby attesting to the model’s commendable accuracy. Elucidating the intricate interrelationships among variables, the 3D response surface plots and contour lines presented in Figure 5 provided a graphical medium for the visualization of the optimal fermentation conditions. Through a meticulous statistical analysis, the anticipated optimal parameters for fermentation were pinpointed: 61.1% water content, 20.8% inoculum amount, and 28.3°C. This specific combination was projected to yield the zenith of spore production, quantified at 3.18 × 10⁹ CFU/g. Subsequently, the projected parameters derived from the application of the RSM were subjected to empirical validation. To align with practical operational considerations, the fermentation conditions were judiciously calibrated to encompass 60% water content, 20% inoculum amount, and 28°C. The ensuing spore production quantified under these optimized parameters amounted to 3.29 ± 0.41 × 10⁹ CFU/g, demonstrating a close concurrence with the forecasted value. Furthermore, an optimal fermentation period of 7 days was established (Supplementary Figure S1F).

The application of the NEAU-ZSY13-based biofertilizer yielded a significant reduction in the severity of tomato bacterial wilt as compared to the negative control. Notably, the degree of disease severity exhibited a proportional decrease with the increasing spore concentration of the biofertilizer in the soil. Further elucidation of this relationship can be found in Table 1; Figure 6. Specifically, the achievement of a spore concentration of 10⁸ CFU/g within the soil corresponded to a noteworthy biocontrol efficacy of 72.15%. This compelling observation underscores the substantial potential of the biofertilizer at this particular concentration to effectively manage the growth and dissemination of tomato bacterial wilt.

After a growth period of 14 days within the experimental pots, we conducted a comprehensive evaluation of wheat root rot incidence and concurrently assessed the efficacy of the implemented biological control interventions. As visually depicted in Figure 7, the cohort designated as the negative control, comprising wheat seedlings deliberately inoculated with B. sorokiniana, exhibited distinct and conspicuous black necrotic lesions both on the subterranean root system and the above-ground portions, leading to manifest limitations in growth and development. Conversely, the application of NEAU-ZSY13 exhibited a notable mitigation of disease severity, as discerned from the visibly reduced necrotic manifestations. The investigation further revealed that the employment of a spore concentration of 10⁸ CFU/g within the soil resulted in the attainment of the highest level of biocontrol efficacy, reaching an impressive value of 78.28% (Table 2). These empirical findings underscore the substantial potential of NEAU-ZSY13 as an efficacious biological control agent for curtailing wheat root rot, substantiating its prospective utility in managing the disease in agrarian contexts.

This study employed an activity-guided approach to isolate biologically active compounds from NEAU-ZSY13, leading to the identification of two distinct compounds (Figure 8). The determination of EC₅₀ values revealed that compounds 1 and 2 exhibited significant activity against R. solanacearum, with values of 3.9 and 3.4 μg/mL, and against B. sorokiniana, with values of 3.6 and 2.4 μg/mL, respectively.

Compound 1 was obtained as a white crystal. Its ESIMS spectrum showed a molecular ion peak at m/z 1,142 [M + H]⁺ (Supplementary Figure S2). ¹H NMR (600 MHz, CD₃OD) and ¹³C NMR (150 MHz, CD₃OD) NMR data were shown in Supplementary Table S3; Supplementary Figures S3, S4. Through meticulous comparison with previously study (Hu et al., 2018), the identity of compound 1 was confirmed as niphimycin C.

Compound 2 was obtained as a white crystal. Its ESIMS spectrum showed a molecular ion peak at m/z 1,142 [M + H]⁺ (Supplementary Figure S5). ¹H NMR (600 MHz, CD₃OD) and ¹³C NMR (150 MHz, CD₃OD) NMR data were shown in Supplementary Table S3; Supplementary Figures S6, S7, substantiated its characterization as niphimycin A (Hu et al., 2018).