The chromosome of S. globisporus SP6C4 was accessed on the NCBI genome database BVBRC (accession no. LWMQ00000000) and RAST server (version 2.0). Multiple sequence alignment at the gene level was performed using Clustal Omega (https://www.ebi.ac.uk/Tools/msa/clustalo/), and a phylogenetic tree was constructed using MEGA 11 with the maximum likelihood method. Domain and functional characterization at the protein level was carried out using the Conserved Domain Database (CDD) provided by NCBI. In secondary metabolite annotation at antiSMASH (version 6.0)

The S. globisporus SP6C4 genome was previously reported by Kim et al. (2019b), and a P450 gene was predicted using the RAST server (version 2.0) and antiSMASH (version 6.0). The gene was targeted for deletion using the CRISPR/Cas9 system with the pCRISPomyces-2 plasmid. Gene-specific primers were designed for the left arm (LA) and right arm (RA) of the gene. Two fragments were amplified by polymerase chain reaction (PCR) in step I, using bacterial chromosome DNA at a concentration of 1 μg/μl, under the following conditions: initial denaturation at 98°C for 10 min, followed by 30 cycles of denaturation at 98°C for 1 min, annealing at 52-58°C for 30 s (depending on the respective primer melting temperature), elongation at 72°C for 30 s, and final extension at 72°C for 5 min. In step II, the LA and RA fragments (300 ng/μl each) were linked by PCR, with initial denaturation at 98°C for 30 s, followed by 5 cycles of denaturation at 98°C for 8 s, annealing at 50°C for 30 s, elongation at 72°C for 30 s, and extension at 72°C for 5 min. In step III, the homologous regions were amplified using the p450LA-F primer and p450RA-R primer (the PCR conditions were similar to step I, with the annealing temperature, changed to 60°C). PCR reactions were performed using an MJ Research PTC-200 thermal cycler (Bio-Rad, Hercules, CA, USA), and the PCR products were loaded on a 1% agarose gel. The amplicons were purified using the Expin GelSV kit (Gene All, Seoul, Korea) and eluted in 30 μl EB buffer. The T-vector ligation also requires A-tailing using 100 mM of dATP (NEB, Ipswich, MA, USA) at 70°C for 30 min, followed by elution in 15 μl EB buffer. The purified solution was then ligated to a pGEM-T easy vector (Promega, Madison, WI, USA). After ligation, the pGEM-T easy vector was added to a homologous fragment selected by the blue/white colony method, and the mixture was transformed into Escherichia coli DH5α cells. The selective media used was LB (Luria broth; 30 g; agar, 20 g per liter) supplemented with ampicillin (100 μg/ml) and X-gal (40 μg/ml). A white colony was cultured in 5 ml of LB broth for 16 h at 37°C, and the plasmid was extracted using the Dokdo Mini-prep Kit (ELPIS-Biotech, Daejeon, Korea). Two different plasmids, the homologous ligated pGEM-T easy vector, and pCRISPomyces-gRNA were cut with XbaI (NEB) and ligated at a 1:3 backbone-to-insert ratio at 16°C overnight. The gRNA was cloned into pBHA (2,002 bp), and pCRISPomyces-2 was cut with BbsI (NEB). The linear pCRISPomyces-2 and gRNA fragments were then ligated using T4 ligase (NEB). The final construction, pCRISPomyces-gRNA-homologous, was transformed into E. coli ET12567 containing pUZ8002 for conjugation with Streptomyces. In this step, the selection was performed on LB agar containing kanamycin (30 μg/ml), chloramphenicol (25 μg/ml), and apramycin (50 μg/ml). Prior to conjugation, the E. coli ET12567 cells containing pUZ8002 and the final construction were cultured in 10 ml of LB broth with the three antibiotics at 37°C for 9 h with shaking at 150 rpm. The cells were then gently washed with 10 ml of LB broth, followed by centrifugation (1,914 ×g for 15 min), and resuspended in 500 μl of LB broth. The bacterial spore stock (OD600nm: 0.6, 10⁹ cfu/ml) was generated by mixing 500 μl of 2XYT medium with SP6C4 spore stock and E. coli cell and incubating at 50°C for 10 min, followed by cooling down and thorough mixing by pipetting. The mixture was then spread on MS agar (mannitol, 20 g; soya, 20 g; agar, 20 g per liter) and incubated for 16 h at 30°C. After 16 h, the plates were overlaid with antibiotics apramycin (50 μg/ml) and nalidixic acid (25 μg/ml) and air dried for 1 h. The conjugation colonies were transferred twice onto MS agar containing apramycin (50 μg/ml). All primers related to mutagenesis, gRNA sequences, and strain information are listed in Supplementary Tables 1 and 2.

All phenotypic assays were performed using the wild-type SP6C4 strain and its mutants. Spore stocks were cultured on MS agar plates for 10 days at 28°C, and sterilized cotton balls were added to 10-ml syringes to collect spores on media with 1 ml of ddH₂O. The filtered spore stock was detected at OD_600nm as 0.6 (10⁶ cfu/ml) and stored at −20°C with 20% glycerol stock.

For the attachment assay, bacteria were cultured in 10 mL of PDK broth (containing 10 g peptone and 10 g potato dextrose per liter) for 7 days at 28°C without shaking. Then, 5 μl of SP6C4 strain or its mutants’ spore stock was injected into a test tube (n = 7) containing the broth media. The tube was discarded from the broth media using a pipette, and 10 ml of 1% crystal violet was added to stain the tube for 15 min at room temperature. Subsequently, the tube was washed twice with 70% ethanol to remove the unstained dye.

The bacterial growth curve was determined by measuring the optical density at 600 nm over a 24-h period. For the sporulation assay, 10 μl of SP6C4 stain or its mutant spore stock was added to minimal media (L-asparagine 0.5 g, K₂HPO₄ 0.5 g, MgSO₄·7H₂O 0.2 g, FeSO₄·7H₂O 0.01 g, glucose 10 g, agar 15 g per liter, adjusted the pH 7.2) in a 96-well plate. The plate was then incubated at 28°C with shaking at 150 rpm for 5 days, with a sample size of n = 7.

The sporulation assay was performed using four different types of media: MS (mannitol 20 g, soya flour 20 g, agar 20 g per liter), PDK (peptone 10 g, potato dextrose broth 20 g, agar 20 g per liter), TSA (tryptic soya broth 25 g, agar 20 g per liter), and minimal media (L-asparagine 0.5 g, K₂HPO₄ 0.5 g, MgSO₄·7H₂O 0.2 g, FeSO₄·7H₂O 0.01 g, glucose 10 g, agar 15 g per liter, adjusted the pH 7.2). The media were dried at room temperature for 1 h, and then 2 μl of SP6C4 and mutant spore stock was spotted onto the media. The plates were incubated at 28°C for 10 days, with a sample size of n = 5. Sporulation was assessed at 24 h, 48 h, and 72 h.

The antifungal activity was evaluated using PDK (peptone 10 g, potato dextrose broth 20 g, agar 20 g per liter) agar plates. Five microliters of spore stock were dotted at two sites and treated in a line 3 cm long from the center. After 3 days, Fusarium oxysporum f. sp. lycopersici, a cultured plant pathogen, was inoculated at the center of the plant using a 4-mm diameter cork bore and needle (n = 7). Seven days later, the antifungal activity was assessed based on the size of the inhibition zone, following the methods described by Cha et al. (2016) and Kim et al. (2019b).

Tomato seeds (cv. Heinz) were surface sterilized by washing with 1% NaOCl and 70% ethanol for 30 s each, followed by two washes with distilled water, and dried in a biosafety cabinet for 30 min. The surface-sterilized seeds were germinated on wetted cotton within a 9-cm diameter petri dish for 3 days. The germinated seeds were then transferred to 9-cm plastic pots and cultured for up to 2 weeks under conditions of 60% relative humidity, 16-h light, and 8-h dark, with a temperature maintained at 28-32°C. Two weeks later, SP6C4 bacterial cultured cells (10⁶ cfu/ml) and their mutants were drenched on each plant, with the culture containing a final concentration of 0.1% v/v of CMC (carboxyl-methylcellulose). Starting from 7 days after the addition of bacterial suspension, F. oxysporum f. sp. lycospersici was inoculated on the tomato plants, and the pathogen was cultured as chlamydospores in 1% oatmeal and soil inoculum. The chlamydospore density of the pathogen was calculated using a hemocytometer, and it was found to be 10⁵ spores/g. The disease index was recorded during the growth stage, and each treatment had 15 plant replications.

The quantitative real time polymerase chain reaction (qRT-PCR) primers used in this study were obtained from Kim and Kwak (2021). Total RNA was extracted twice: first at the stage of aerial hyphal formation for the expression of tsrD gene (48 h), and second after aerial hyphae formation for the expression of lanM gene (72 h). Bacterial cells were ground with liquid nitrogen, and the resulting pellet was stored at −80°C. RNA extraction was carried out using the TRIzol method and RNeasy Mini kit (Qiagen, Hilden, Germany) for column purification. The extracted RNA was then synthesized into cDNA using the ReverTraAce-α-® cDNA Synthesis Kit (TOYOBO, Osaka, Japan), and qRT-PCR was performed using the CFX96 Touch Real-Time PCR Detection System (Bio-Rad). Five replicates were performed for each mutant cDNA library in the qRT-PCR analysis.

The statistical analyses of all graphs were conducted using ANOVA, t-test. If the data did not meet the assumptions of normality and equal variance, they were analyzed using ANOVA post-hoc tests were performed using Tukey’s honestly significant difference (Konietschke et al., 2008; Munzel and Hothorn, 2001). The graphs were visualized using ggplot2 from the R software package.

The functions and structures of Streptomyces species cytochrome P450 enzymes have been reported (Cho et al., 2019). These enzymes, also known as P450d CYP, are involved in various catalytic, metabolic, and biosynthetic activities, as well as in toxicity and drug metabolism (Guengerich, 2001). The compared region for the P450 cluster in S. globisporus SP6C4 was analyzed using Seed Viewer version 2.0. The cutoff for similarity in this cluster was set at more than 40% identity. Multiple P450 biosynthesis genes of Streptomyces spp. were aligned using MEGA11 and a phylogenetic tree was generated using maximum likelihood (Supplementary Fig. 1). The results revealed that the P450 gene of SP6C4 belonged to a distinct group compared to other species, such as S. albus, S. baarnensis, and S. anulatus, and showed similarity only to S. griseus. The strain SP6C4 showed similarity to Streptomyces groups, specifically S. griseus NBRC13350 (E-value: 0). Additionally, it showed similarity to the class of Actinomycetia, including Streptosporangium (E-value: 3.00E⁻¹²⁴), Thermomonospora (E-value: 8.00E⁻⁹²), and Mycobacterium (E-value: 1.00E⁻⁷³), as determined by PSI-BLAST (Supplementary Table 3).

The SP6C4 stain was found to have melC1 and melC2 genes upstream of the P450 biosynthesis gene, and three different genes were aligned with DNA and protein levels using the NCBI database, showing more than 99% similarity (Table 1, Fig. 1A). Domain search revealed that SP6C4-P450 belonged to the cytochrome P450 (CYP) superfamily (cd00302) and CYP142-like (cd11033) (Fig 1B). These CYPs were found to have heme-binding sites associated with various oxidation reactions and exhibited diverse structural characteristics. Among the Streptomyces species, S. coelicolor, S. avermitalis, S. velenzuelae, S. rapamycinicus, S. hygroscopicus, and S. peucetius have the highest number of groups, including CYP105N1, CYP171A1, CYP107L1, CYP112A2, and CYP129A2, respectively (Cho et al., 2019; Lim et al., 2012). These groups belong to 144 P450 families and 377 P450 subfamilies (Moshoeshoe, 2019). Although the reactions catalyzed by these CYP enzymes have been reported, their participation in biological processes such as electron transfer, anti-functional compound synthesis, and oxidation, which may uncover additional functions, is still not fully understood (Han et al., 2015; Lee et al., 2016).

We annotated a number of P450 genes and conducted a comparative analysis of P450s at the SP6C4 genome level to validate their key role as secondary metabolite regulators. Twenty-three cytochrome genes were detected in the SP6C4 genome, including a putative melanin-related P450 gene, which showed low similarity to Streptomyces species at the protein level but high alignment with other Actinomycete classes at the gene level, with unassigned functions. The gene cluster of P450 was disrupted by a double homologous recombination process resulting in two mutants, ΔP450ΔtsrD and ΔP450ΔlanM, where an antifungal biosynthesis gene (tsrD and lanM) was deleted in a two-step process, followed by the loss of the P450 enzyme. Phenotypic assays were performed for attachment, bacterial growth, and sporulation. In the attachment assay, crystal violet staining revealed that the wild type (WT) and ΔP450 mutants showed a colored line in the glass tube, while the ΔP450ΔtsrD mutant exhibited a darker line (Fig. 2A). Other mutant lines ΔtsrD, ΔlanM, ΔP450, and ΔP450ΔlanM showed no attachment ability in glass tubes (Fig. 2A). Assays were conducted in 96-well plates for bacterial growth, and OD values were analyzed using the Gompertz model and rolling regression analysis. The mutant ΔP450ΔtsrD exhibited an exponential growth rate with a maximum slope (μ = 0.4/1 h) and 95% confidence intervals predicted (Fig. 2B). The growth rates of the WT and other mutant lines ΔtsrD, ΔlanM, ΔP450, and ΔP450ΔlanM were within the 95% confidence intervals, and the growth rate of ΔP450ΔtsrD was higher (0.49-0.57/1 h) than that of ΔP450ΔtsrD (Supplementary Fig. 2). Consequently, the exponential growth rate was determined by employing a rolling regression approach, and the slope of the linear regression between the optical density (OD_600nm) and time was utilized as the estimate for exponential growth. The sporulation test was performed using various media, ranging from high nutrient richness (MS, PDK, TSA) to low (minimal media) to mimic different Streptomyces growth conditions. The wild-type strain exhibited high sporulation across all media at 24 to 72 h (Fig. 2C). Among the mutant lines, ΔP450, ΔlanM, and ΔP450ΔlanM showed good sporulation in multiple types of media at different time points, while ΔtsrD did not show any sporulation at 24 h in any medium (Fig. 2C). However, the double knockout mutant ΔP450ΔtsrD showed sporulation even at 24 h in rich media such as MS, PDK, and TSA (Fig. 2C). Therefore, the sporulation ability of ΔP450ΔtsrD was found to be higher than that of ΔtsrD. These changes in phenotypic characters are likely to affect secondary metabolite production, as Streptomyces undergoes a complex developmental life cycle involving mycelial growth and sporulation, and morphological differentiation is required for new activities (Bobek et al., 2014; McCormick and Flärdh, 2012).

The strain SP6C4 exhibits antifungal activity against the pathogen Fusarium oxysporum in strawberries (Kim et al., 2019b, 2021). This finding is consistent with the antifungal activity observed in the tomato pathogen Fusarium oxysporum f. sp. lycopersici (Kim et al., 2021). Mutagenesis lines showed differential activity against the pathogen, with ΔP450 showing no significant difference compared to the WT, while signal mutants ΔtsrD and ΔlanM lost their antifungal ability (Fig. 3). Interestingly, one of the double deletion mutants, ΔP450ΔtsrD, showed regained inhibition of fungal growth against F. oxysporum f. sp. lycopersici (Fig. 3). Volatile tests using four fungal pathogens, including Botrytis cinerea, F. oxysporum f. sp. fragariae, F. oxysporum f. sp. lycopersici, and Phytophthora fragariae, showed no difference between WT and mutant lines (Supplementary Fig. 3).

In biocontrol assays using wild-type and mutant lines, a similar pattern was observed in the in vitro test results on agar plates. The disease index values ranged from 0 to 5, with WT and untreated showing a value of 0, ΔP450 and ΔP450ΔtsrD showing a value of 1.2, which was not significantly different in Tukey’s test (Fig. 4A, Supplementary Table 4). However, values higher than 3 were observed in ΔlanM, ΔtsrD, ΔP450ΔlanM, and all values were counted after 4 weeks (Fig. 4A and B). The data was generated using a time serial model, which revealed significant differences in week intervals (P = 6.030E⁻¹⁴) among the treatments (P < 2.2E⁻¹⁶), and interactions between the two factors with a P-value of 3.956E⁻¹¹ (Fig. 4A, Supplementary Fig. 4). This result indicates that the recovered antifungal activity in ΔP450ΔtsrD can also be demonstrated in plants.

We elucidated the expression patterns of eight regions of biosynthetic genes, including lanM (grisin biosynthesis) and tsrD (conprimycin biosynthesis), as well as three types of NRPS and PKS genes, which were previously reported by Kim and Kwak (2021) as secondary metabolite-related genes in S. globisprus SP6C4. Among these genes, lanM showed a significant increase in expression with a fold change greater than 3 Log2 fold in the ΔP450ΔtsrD mutant, indicating its key role in antifungal effects. In contrast, tsrD was not expressed in mutant lines, including ΔP450, ΔP450ΔlanM, and ΔP450ΔtsrD (Fig. 5C). When examining other secondary metabolite-related genes using Log2 fold change as a metric, the NRPS and PKS genes showed expression levels lower than 0 at 48 h in ΔP450 and ΔP450ΔlanM mutants (Supplementary Fig. 5A and B). The ΔP450ΔtsrD mutant exhibited slightly increased expression at 48 h, but no significant upregulation, whereas at 72 h, NRPS and PKS genes showed 3-5 Log2 fold change in each mutant (Supplementary Fig. 5C). The key finding of this study was the antagonism assay, which unexpectedly showed that the ΔP450ΔtsrD mutant regained antifungal activity against the pathogen. Furthermore, although the mechanism is unclear, the expression of lanM antifungal-related genes was upregulated in the ΔP450ΔtsrD mutant.

Our results suggest that the enzyme P450-CYP142 is involved in the morphological processes of S. globisporus SP6C4, including antifungal activity. Specifically, we propose that P450-CYP142 acts as a repressor at the lanM gene under normal conditions, without specific signaling to favor. These findings highlight the significant role of the P450-CYP142 enzyme in modulating the morphological response of S. globisporus SP6C4, as well as its contribution to the strain’s antifungal ability. These results imply the potential involvement of P450-CYP142-mediated metabolic pathways in shaping the growth and secondary metabolism of S. globisporus SP6C4, thus underscoring its relevance as a putative target for the development of novel antifungal agents.