Screening, Identification, and Fermentation of a Biocontrol Strain against Peony Southern Blight and Extraction of Secondary Metabolites

Abstract

A bacterial strain (WM-37) was isolated from soil and identified as Streptomyces rectiviolaceus on the basis of morphological, physiological, biochemical, and 16S rRNA characteristics. The strain was screened regarding its potential use for controlling the pathogen causing peony southern blight. To enhance the secondary metabolite yield, submerged fermentation was conducted according to a single-factor trial and response surface method. Metabolite production peaked under the following conditions: 250.00 mL flask containing 100.00 mL culture medium consisting of 20.00 g·L⁻¹ soluble starch, 3.86 g·L⁻¹ ammonium sulfate, 0.50 g·L⁻¹ sodium chloride, 0.50 g·L⁻¹ dipotassium phosphate, 0.50 g·L⁻¹ magnesium sulfate, and 0.01 g·L⁻¹ ferrous sulfate; inoculation amount, 7.74%; temperature, 30 °C; initial pH, 7.00; incubation time, 7 days; and rotational speed, 160 rpm. The fermentation broth was absorbed by D101 macroporous resin and eluted with an ethanol-water gradient, after which the eluate fractions with antifungal compounds were collected, evaporated, and concentrated to obtain a crude extract. This extract was dissolved in methanol and separated by high-performance liquid chromatography. The active compound was identified as azelaic acid (C₉H₁₆O₄) on the basis of ultra-high-resolution mass spectrometry and analyses of publicly available data. These findings suggest that Streptomyces rectiviolaceus WM-37 may be a viable biocontrol agent effective against the pathogen responsible for peony southern blight.

1. Introduction

Peony (Paeonia suffruticosa) is a candidate national floral species and a unique aromatic plant in China, where it is known as the “king of flowers” because of its large flowers, bright colors, strong fragrance, and high ornamental, medicinal, and nutritive value [1,2]. Peony is also widely cultivated in urban parks abroad, such as Seattle in the United States, Tours in France, Okayama in Japan, and so on. Increases in the number of peony varieties and their planting area have been accompanied by an increase in disease-related damages, which has significantly affected the ornamental and medicinal value of peonies. Peony southern blight is considered to be one of the most serious threats to peony cultivation, and its incidence rate is around 5% [3]. The pathogen responsible for peony southern blight is Sclerotium rolfsii (subphylum: Ascomycota; class: Aspergillus; and genus: Sclerotium). This pathogen overwinters as sclerotia in infected roots, soil, and weeds, primarily spreading through seedlings or water flow. In the subsequent year, the spores and mycelium of S. rolfsii invade the roots of weak peony plants and the stems near the soil. Because damage is not easily detected in soil, it is often not until white mycelium is discovered that peonies are close to withering, and there is almost no medicine to treat them, with a mortality rate close to 100% [4]. Fungal diseases of plants are typically prevented and managed by the application of synthetic fungicides [5], but their excessive and improper use has led to environmental damage, human and animal health issues, and increased fungicide resistance among plant pathogens [6,7]. Therefore, alternative management practices are needed to maintain crop quality and productivity without damaging the environment [8]. Biological control (biocontrol) is an environmentally friendly strategy for minimizing crop damages caused by plant pathogens [9]. Currently, bacteria, fungi, and actinomycetes are widely used biocontrol agents that protect plants from diseases [10]. Actinomycetes are an extremely diverse group of microorganisms, with at least 180 genera and over 20,000 species identified to date [11]. Streptomyces, a genus comprising actinomycetes, is primarily used for the biocontrol of plant diseases. Several Streptomyces species have antagonistic effects on fungal plant pathogens [12].

Microbial fermentation is influenced by the composition of the culture medium and fermentation conditions [13,14]. These factors affect the production of secondary metabolites by Streptomyces species, ultimately determining the yield of these compounds [15]. Even minor changes in the fermentation medium composition can alter the yield and metabolic profile of microorganisms. Additionally, variations in fermentation conditions (e.g., time, temperature, salinity, pH, shaking conditions, and oxygen supply) can modulate gene expression in microorganisms [16], which can activate “silent metabolic pathways” in strains, thereby increasing the diversity of secondary metabolites and improving the utility of microbial resources [17,18]. Although traditional optimization methods for strains typically involve linear regression analyses and orthogonal testing, these methods have limitations in terms of accuracy and consistency. In contrast, the response surface method can effectively address these limitations, offering advantages over traditional methods, including shorter cycles and higher precision. This method is widely applied in various fields (e.g., food, biology, and chemical industries) [19,20].

Biologically active compounds derived from natural sources, such as plants, animals, and microorganisms, are defined as natural products [21]. Natural products with diverse biological activities target a wide range of receptors and have relatively low toxicity and few side effects. Their advantageous chemical structures make certain active compounds valuable resources for the development of new drugs [22]. Actinomycete-derived active compounds and modified bioactive molecules possess antibacterial, anti-inflammatory, antitumor, insecticidal, anti-aging, and plant stress resistance-enhancing activities [23]. Several compounds, such as vancomycin, jinggangmycin, avermectin, streptomycin, amphotericin, and erythromycin, have diverse uses (e.g., in food, green biopesticides, and healthcare products) [24]. Among the natural products that are commonly used, 70% are derived from actinomycetes [25]. Therefore, analyzing natural products from actinomycetes may have positive implications for life sciences research, preventing and controlling pathogenic microbes, and restricting drug resistance development.

In this study, microorganisms were isolated from the surface soil of Henan University of Science and Technology farm and screened, with S. rolfsii used as the target organism. Strains were identified on the basis of morphological characteristics, physiological and biochemical activities, and 16S rRNA sequences. The single-factor trial and response surface method were used to determine optimal conditions for large-scale fermentation. Fermentation products were separated and purified using various methods, including macroporous resin adsorption, ethanol-water gradient elution, and high-performance liquid chromatography (HPLC). The purified compounds were structurally identified using an ultra-high-resolution mass spectrometry system and analyses of publicly available data. The objectives of this study were to address the limitations of traditional control methods, decrease the reliance on chemical pesticides, prevent chemical pollution, and effectively manage peony southern blight. The study findings contribute to the cataloging of microbial biocontrol species resources, setting the stage for the identification of natural products with novel pharmacological activities.

2. Materials and Methods

2.1. Isolation, Screening, and Antifungal Activity Analysis of Bacterial Strains

In the farm of Henan University of Science and Technology, a five-point sampling method was adopted to collect soil samples 15 cm from the surface, put the samples into sterile sealed bags, and bring them back to the laboratory for natural air drying at room temperature. The soil samples were wrapped in gauze, crushed and screened, stored in a refrigerator at 4 °C for later use. Strains were isolated and purified using the spread plate method [26]. Actinomycetes were isolated and purified using Gao’s No. 1 medium (soluble starch 20.0 g, potassium nitrate 1.0 g, sodium chloride 0.5 g, dipotassium phosphate 0.5 g, magnesium sulfate 0.50 g, ferrous sulfate 0.01 g, distilled water 1000.0 mL, pH 7.0 ± 0.2, solids require 15.0–20.0 g of agar), and common bacteria were isolated and purified using NA medium (peptone 10.0 g, beef extract 3.0 g, sodium chloride 5.0 g, distilled water 1000.0 mL, pH 7.0 ± 0.2, solids require 15.0–20.0 g of agar). The test microorganisms included peony southern blight pathogen (S. rolfsii), peony root rot pathogen (Fusarium solani), peony yellow spot pathogen (Phyllosticta commonsii), tomato leaf spot pathogen (Fusarium proliferatum), rice sheath blight pathogen (Rhizoctonia solani), wheat sheath blight pathogen (Rhizoctonia cerealis), wheat crown rot pathogen (Fusarium pseudograminearum), tobacco root black rot pathogen (Thielaviopsis basicola), and maize ear rot pathogen (Fusarium verticillioides). Isolated strains were preserved on NA medium and Gao’s No. 1 medium. Primary screening involved the agar block method. Briefly, agar-solidified media were inoculated with the isolated strains. After mature colonies formed, a hole-puncher was used to collect agar blocks (8–10 mm diameter) containing mycelium. These agar blocks were then placed on the surface of culture plates with medium coated with S. rolfsii. Three replicates were produced, and culture plates were incubated at 30 °C for 36–48 h. The region surrounding the agar blocks of the tested strains was examined for the presence of inhibition zones. Secondary screening involved an antagonistic assay completed using the Oxford cup method [27] to determine the antifungal activity of the biocontrol agent. Finally, eight pathogenic fungi were used as the target, including F. solani, P. commonsii, F. proliferatum, R. solani, R. cerealis, F. pseudograminearum, T. basicola, and F. verticillioides, and the antibacterial activity and inhibitory rate of the biocontrol strain finally screened by the above method were determined by the Oxford Cup confrontation test [28].

2.2. Identification of Bacterial Strains

Bacterial strains were initially identified according to their growth characteristics and Gram staining using an optical electron microscope (EX30; Honglin Scientific Instruments Co., Ltd., Changsha, China). Morphological features were also examined using a scanning electron microscope (JSM-IT200; Scientific Instruments Beijing Co., Ltd., Beijing, China). The following physiological and biochemical tests were conducted: Voges-Proskauer (V-P) reaction, gelatin liquefaction, starch hydrolysis, nitrate reduction, utilization of carbon sources, cellulose hydrolysis, milk coagulation and curdling, catalase test, hydrogen sulfide production, temperature tolerance tests at 40, 45, 50, and 55 °C, pH tolerance tests at 2, 4, 6, 8, 10, and 12, and salt tolerance tests at concentrations of 5%, 7%, and 10% (w/v) [29,30]. A modified version of the CTAB (cetyltriethylammonium bromide) method was used to extract genomic DNA from endophyte isolates [31] for the PCR amplification involving universal primers 27 F (5′-AGAGTTTGATCCTGGCTCAG-3′) and 1492 R (5′-CTACGGCTACCTTGTTACGA-3′). PCR reaction procedure: pre-denaturation at 94 °C for 5 min; denatured 45 s at 94 °C, annealed 45 s at 55 °C, extended for 1 min at 72 °C (30 cycles); finally, 72 °C extension for 10 min. PCR products were purified and sent to Beijing Novogene Biotechnology Co., Ltd. (Beijing, China) for 16S rRNA sequencing. The obtained sequences were used as queries for a BLAST search of the NCBI database to determine the possible species of the examined isolates. BLAST results were downloaded, and phylogenetic trees were constructed using MEGA software (Version 7.0.26; Mega Limited, Auckland, New Zealand).

2.3. Seed Broth Preparation

The fermentation process was divided into two stages: seed growth and production of antifungal active substances. Gao’s No. 1 medium in plates was inoculated with the preserved strain WM-37 and incubated at 30 °C for 3 days. Subsequently, 100 mL Gao’s No. 1 liquid medium in a 250 mL erlenmeyer flask was inoculated with WM-37 (inoculation volume of 8%). The flask was then placed in a shaking incubator set at 30 °C and 160 rpm for 3 days.

2.4. Determination of Antifungal Activities in the Fermentation Broth

Centrifugation of the fermentation broth was performed using a tabletop high-speed refrigerated centrifuge (KH22R; Hunan Kaida Scientific Instrument Co., Ltd., Changsha, China) at 500 rpm for 3 min at 4 °C to obtain the supernatant, after which the antifungal activity was determined by measuring the diameter of the inhibition zone according to the Oxford cup method, with S. rolfsii used as the target fungus [32].

2.5. Trial Design and Optimization

(1)

Single-factor trial

Using Gao’s No. 1 medium as the base medium, the basic fermentation conditions were a 250 mL erlenmeyer flask with a liquid volume of 100 mL medium, inoculation amount of 8%, temperature of 30 °C, initial pH of 7, incubation time of 7 days, and rotation speed of 160 rpm. The antifungal activity of Streptomyces rectiviolaceus WM-37 against S. rolfsii was determined by measuring the inhibition zone diameter. Single-factor trials were conducted to optimize the following factors: carbon sources (addition of 2% soluble starch, glucose, sucrose, maltose, and galactose); nitrogen sources (addition of 0.1% potassium nitrate, ammonium nitrate, ammonium sulfate, peptone, and yeast extract); inoculation amount (4%, 6%, 8%, 10%, and 12%); temperature (26, 28, 30, 32, and 34 °C); initial pH (6, 6.5, 7, 7.5, and 8); incubation time (5, 6, 7, 8, and 9 days); and rotational speed (120, 140, 160, 180, and 200 rpm). After determining the optimal carbon and nitrogen sources, additional trials were conducted to determine the optimal amounts of carbon (1–5%) and nitrogen (0.1–0.5%) sources.

(2)

Response surface optimization

Design-Expert 13 software was used for the Plackett–Burman trial design. The Plackett–Burman test is frequently used to simultaneously screen multiple factors [33]. Gao’s No. 1 medium after single-factor optimization was used as the base medium, and factors selected from single-factor trials were assigned two levels each (i.e., −1 and 1). The low level reflects the initial culture conditions, whereas the high level was set at approximately 1.25-times the low level. A Plackett–Burman test involving seven main factors and a total of 12 trials (N = 12) was conducted. Factors with significant effects were identified on the basis of the Plackett–Burman test results. The gradient direction for changing the trial values was determined as the climbing direction, whereas the climbing direction for the other factors was decided according to the positive or negative coefficients of their impact coefficients. This enabled the determination of the central points of the key factors affecting antifungal activities. Building on the results of the single-factor trials and the Plackett–Burman test, factors that significantly affected the overall fermentation model were selected for the response surface analysis. Design-Expert 13 software was used for the Box–Behnken trial design [34], with three levels for each factor. The response surface analysis used significant factors as independent variables and the inhibition zone diameter (Y value) as the response value to determine the influence of each factor on the inhibitory effects against S. rolfsii. Design-Expert 13 software was used to analyze the experimental data, and the analysis of variance and regression equation were tested by F-test. p < 0.05 meant that the difference was significant. R² was used to represent the fitting of multiple regression models, and R² > 0.9 was judged to be optimal. p > 0.05 indicates that lack of fit was not significant, indicating that the regression model fits the data well. To validate the predicted values obtained using the response surface method, fermentation experiments were carried out under optimized conditions. Each trial was repeated three times, and the average value was calculated. By comparing the trial results with the predicted values, the reliability, accuracy, and practicality of the fermentation conditions optimized using the response surface method were determined.

2.6. Preparation of Crude Extract

The fermentation liquid of 20.0 L was prepared using the fermenter (MC-JSF-20L; Beijing Mancang Technology Co., Ltd., Beijing, China) under the above optimal fermentation conditions. Fermentation broth (20.0 L) was centrifuged at 8000 rpm for 15 min at 4 °C, after which the pelleted bacterial cells were removed, and the supernatant was retained. The supernatant was passed through a D101 macroporous resin column (D101; Anhui Sanxing Resin Technology Co., Ltd., Bengbu, China), with ethanol-water serving as the eluent (elution at room temperature until no liquid drips). The following volume ratios were used for the gradient elution 10:90, 30:70, 50:50, 70:30, and 90:10. The antifungal activity of each fraction (S. rolfsii as the indicator organism) was detected using the Oxford cup method. The fractions with antifungal activity were pooled to obtain the crude extract.

2.7. Purification and Identification of Compounds

After the compounds in the crude extract were dissolved in methanol and filtered through a 0.22 μm membrane, they were separated using a preparative chromatography column, whose specification was a C18 reversed-phase column (19 mm × 100 mm, 5 μm) (Symmetry C18 column; Waters Corporation, Milford, CT, USA). For the gradient elution, water (0.01% formic acid) and methanol were used as eluents. The gradient elution conditions were as follows: 40% methanol (volume fraction) at 0–20 min; gradual increase to 90% methanol at 45 min; further increase to 100% methanol in the aqueous solution at 60 min; hold at 100% methanol until 70 min; decrease to 40% methanol at 71 min; and hold at 40% methanol until 81 min. The operating conditions were as follows: temperature, 25 °C; flow rate, 1.0 mL·min⁻¹; and UV detection, 280 nm (Agilent 1260; Agilent Technologies Co., Ltd., Santa Clara, CA, USA). The separated fractions were collected and evaluated in terms of antifungal activities according to the Oxford cup method, with S. rolfsii as the indicator organism. Fractions with detectable antifungal activities were further dissolved in methanol, passed through a 0.22 μm membrane, and purified via HPLC (Agilent 1260; Agilent Technologies Co., Ltd., Santa Clara, CA, USA). After an individual component was confirmed via liquid chromatography, the purified fraction was evaporated using a rotary evaporator and lyophilized at −20 °C for 24 h to obtain a pure compound.

An appropriate amount of the purified compound was dissolved in methanol in a centrifuge tube prior to an ultrasonic treatment to ensure the compound was completely dissolved. The solution was filtered through a 0.22 µm membrane. The molecular structure of the active substance was determined by ultra-high-resolution mass spectrometry (LTQ Orbitrap Elite; Thermo Fisher Technologies Co., Ltd., Waltham, MA, USA) and the C18 reversed-phase chromatography column. Sample (25 μL) was injected after equilibration by 5% (v/v) acetonitrile containing 0.1% (v/v) trifluoroacetic acid (TFA). The elution program was as follows: 5% (v/v) acetonitrile containing 0.1% (v/v) TFA for 10 min; linear gradient elution from 5% to 35% (v/v) acetonitrile containing 0.1% (v/v) TFA for 5 min; 35% to 70% acetonitrile for 45 min. The elution was at a flow rate of 0.3 mL·min⁻¹. The conditions of ultra-high-resolution mass spectrometry were as follows: ionization mode was positive; capillary temperature was 320.0 °C; source voltage was 4.50 kV. The total ion chromatograms were taken in a mass range from 100 to 2000 m/z. The resulting spectra were compared with Global Natural Products Social Molecular Networking (GNPS) and publicly available literature data to identify the antifungal active substances and elucidate their structures.

2.8. Statistical Analyses

All experimental data underwent an analysis of variance (ANOVA) using SPSS Statistics software (SPSS 22.0; International Business Machines Corporation, Armonk, NY, USA). The mean values for each treatment were compared using Duncan’s multiple range test to determine significant differences (at the 0.05 level). Data are presented herein as the mean and their standard error. Design-Expert (version 13; Stat-Ease, Minneapolis, MN, USA) was used for response surface analysis, and Origin (Origin 2021; OriginLab Corporation, Northampton, MA, USA) was used for mapping.

3. Results

3.1. Isolation, Screening, and Assessment of Inhibitory Activities

In this study, 46 bacterial strains were isolated from soil samples (40 actinobacteria and six other bacteria). Culture plate-based analyses of the inhibitory effects of these strains on S. rolfsii (pathogen responsible for peony southern blight) resulted in the identification of three antagonistic strains (WM-16, WM-32, and WM-37) (Figure 1). According to the secondary screening results, WM-37 had the strongest inhibitory effect, with an inhibition zone diameter of 16.70 ± 0.85 mm (Figure 2). Thus, it was selected for subsequent analyses. The inhibition zone diameter reflects the strength of the antifungal effect. The Oxford cup confrontation test results indicated that WM-37 had a relatively broad-spectrum antifungal effect. More specifically, it effectively inhibited the growth of two pathogens that cause peony root rot (F. solani) and peony yellow spot (P. commonsii), while also inhibiting the growth of pathogens responsible for other plant diseases, including tomato leaf spot (F. proliferatum), rice sheath blight (R. solani), wheat sheath blight (R. cerealis), wheat crown rot (F. pseudograminearum), tobacco root black rot (T. basicola), and maize ear rot (F. verticillioides) (Figure 3), with antifungal rates of 21.95%, 42.22%, 43.81%, 41.56%, 46.60%, 35.97%, 14.09%, and 31.95%, respectively. Notably, WM-37 had the strongest inhibitory effect against the wheat sheath blight pathogen, with an antifungal rate of 46.60%.

3.2. Strain Identification

When cultured on Gao’s No. 1 agar medium, WM-37 colonies had morphological characteristics that were similar to those of Streptomyces colonies. Initially (2- to 3-day-old culture), colonies were white, but the colony center gradually changed from dark blue to purple (6- to 7-day-old culture), whereas the surrounding areas were light blue. The colony had a convex surface with distinct folds. Gram staining confirmed WM-37 as a Gram-positive bacterial strain. Scanning electron microscopy images indicated that WM-37 hyphae formed a complex, branched network structure with an irregular morphology (e.g., intertwined, bifurcated, and coiled branches) (Figure 4). The physiological and biochemical characteristics of WM-37 are summarized in

Table 1. Physiological and biochemical characteristics of strain WM-37. The results of physiological and biochemical tests such as Voges–Proskauer (V-P) reaction, gelatin liquefaction, and starch hydrolysis of strain WM-37.

Indicator		Strain WM-37
Starch hydrolysis		+
Nitrate reduction		+
Cellulose decomposition		−
Gelatin liquefaction		+
V-P reaction		−
H2S test reaction		−
Catalase test		+
Milk coagulation reaction		+
Utilization of carbon sources	Fructose	+
	Galactose	+
	Maltose	+
	Sucrose	+
pH	2	−
	4	−
	6	+
	8	+
	10	+
	12	−
Growth temperature	40 °C	+
	45 °C	+
	50 °C	+
	55 °C	−
Salt tolerance	5%	+
	7%	+
	10%	−

Note: “+” and “−” indicate positive and negative reactions, respectively.

. On the basis of the morphological features and “Actinomycetes Systematics”, WM-37 was classified as a member of the genus Streptomyces.

WM-37 genomic DNA was used as a template for a PCR amplification. The PCR product, which was analyzed by 1% agarose gel electrophoresis, was detected as a band at approximately 1400 bp. Sequencing of the target region revealed a 16S rRNA sequence comprising 1382 bp (NCBI accession number: OR342243). This sequence was used to construct a phylogenetic tree (Figure 5), which indicated that WM-37 was most closely related to the known S. rectiviolaceus strain NRRL B-16374T (DQ026660), with a homology of 99.78%. Considered together, the morphological, physiological, and biochemical characteristics and phylogenetic relationships suggested WM-37 was an S. rectiviolaceus strain. This strain was preserved at the General Microbiological Center of the China Committee for Culture Collection of Microorganisms (CGMCC number: 27937).

3.3. Fermentation Conditions

(1)

Effect of carbon sources on the inhibitory activity of S. rectiviolaceus

Trials involving the addition of soluble starch, maltose, sucrose, galactose, and glucose to the fermentation medium showed that supernatant samples of fermentation broth based on all these carbon sources inhibited pathogenic fungal growth. The rank-order of inhibition zone diameters for the tested carbon sources (largest to smallest) was as follows: soluble starch > sucrose > galactose > maltose > glucose. Notably, soluble starch had the strongest effect, which was consistent with the findings of an earlier study that determined soluble starch was the optimal carbon source for the fermentation of Streptomyces rochei [35]. There was a statistically significant difference (p < 0.05) between the effect of soluble starch (inhibition zone diameter of 16.68 ± 0.27 mm) and the effects of the other four carbon sources (Figure 6). Thus, soluble starch was selected as the optimal carbon source. The inhibition zone diameter increased as the soluble starch concentration increased from 10.0 to 20.0 g·L⁻¹, but further increases in the soluble starch concentration (up to 30.0 g·L⁻¹) decreased the inhibition zone diameter (Figure 7). Accordingly, the optimal soluble starch concentration was determined to be 20.0 g·L⁻¹.

(2)

Effect of nitrogen sources on the inhibitory activity of S. rectiviolaceus

When potassium nitrate, ammonium sulfate, yeast extract, peptone, and ammonium nitrate were added to the fermentation medium, the addition of all the nitrogen sources resulted in inhibitory effects on pathogenic fungal growth. The rank-order of inhibition zone diameters for the examined nitrogen sources (largest to smallest) was as follows: ammonium sulfate > potassium nitrate > ammonium nitrate > peptone > yeast extract. Notably, among these nitrogen sources, ammonium sulfate had the strongest effect, which was in accordance with the previously reported optimal nitrogen source for the fermentation of Streptomyces JD211 [36]. There were significant differences (p < 0.05) between the effect of ammonium sulfate (inhibition zone diameter of 18.97 ± 0.18 mm) and the effects of three other nitrogen sources (Figure 8). Hence, ammonium sulfate was selected as the optimal nitrogen source. Moreover, the inhibition zone diameter increased as the ammonium sulfate concentration increased from 1.0 to 4.0 g·L⁻¹, but further increases in the ammonium sulfate concentration (up to 5.0 g·L⁻¹) decreased the inhibition zone diameter (Figure 9). Ultimately, the optimal ammonium sulfate concentration was determined to be 4.0 g·L⁻¹.

(3)

Effect of inoculation amount on the inhibitory activity of S. rectiviolaceus

Of the tested inoculation amount (4%, 6%, 8%, 10%, and 12%), 8% inoculum resulted in the fermentation broth with the strongest inhibitory effect. The effect of this inoculation amount (inhibition zone diameter of 16.86 ± 0.28 mm) differed significantly (p < 0.05) from the effects of two other inoculation amounts (Figure 10). Thus, an 8% inoculum was considered to be the optimal inoculation amount.

(4)

Effect of temperature on the inhibitory activity of S. rectiviolaceus

Analyses of the effects of various temperatures (26, 28, 30, 32, and 34 °C) indicated that the inhibition zone diameter initially increased and then decreased as the temperature increased. The antifungal activity of the fermentation broth peaked at 30 °C, with an inhibition zone diameter (16.81 ± 0.16 mm) that was significantly larger (p < 0.05) than the inhibition zone diameters for three other temperatures (Figure 11). Hence, 30 °C was selected as the optimal temperature.

(5)

Effect of initial pH on the inhibitory activity of S. rectiviolaceus

Fermentation broth activity initially increased and then decreased as the initial pH increased (6, 6.5, 7, 7.5, and 8), with peak activity at initial pH 7.0 (inhibition zone diameter of 16.81 ± 0.52 mm). The effect of initial pH 7.0 was significantly different (p < 0.05) from the effects of three other initial pH levels (Figure 12). Therefore, initial pH 7.0 was selected as the optimal initial pH.

(6)

Effect of incubation time on the inhibitory activity of S. rectiviolaceus

Fermentation broth activity initially increased significantly (p < 0.05) as the incubation time increased, reaching peak levels at 7 days (inhibition zone diameter of 16.50 ± 0.40 mm). The observed activity was not significantly affected by further increases in the incubation time (8 and 9 days) (Figure 13). Hence, 7 days was selected as the optimal incubation time.

(7)

Effect of rotational speed on the inhibitory activity of S. rectiviolaceus

A trial was conducted to assess the effects of different rotational speeds (120, 140, 160, 180, and 200 rpm). As the rotational speed increased, the fermentation broth activity initially increased and then decreased (Figure 14). The antifungal activity was highest at a rotational speed of 160 rpm, with an inhibition zone diameter of 16.89 ± 0.15 mm. The effect of 160 rpm differed significantly (p < 0.05) from the effects of three other rotational speeds. Accordingly, 160 rpm was selected as the optimal rotational speed.

(8)

Placket–Burman test results

After considering the fundamental nutrient components required for microbial growth and the results of the single-factor trials, soluble starch (X₁), ammonium sulfate (X₂), inoculation amount (X₃), temperature (X₄), initial pH (X₅), incubation time (X₆), and rotational speed (X₇) were selected for error analyses. Each variable had two levels (−1 and +1). There were a total of seven factors and 12 trial groups, and three replicates were performed in each group. The factors and their levels are summarized in

Table 2. Plackett–Burman test factors and levels. The various factors in the Plackett–Burman test and their corresponding actual values encoded.

Source	Fermentation Conditions	Level
		−1	+1
X1	Amount of soluble starch added (g·L−1)	15	25
X2	Amount of ammonium sulfate added (g·L−1)	3	5
X3	Inoculation amount (%)	6	10
X4	Temperature (°C)	28	32
X5	Initial pH	6.5	7.5
X6	Incubation time (d)	6	8
X7	Rotational speed (r·min−1)	140	180

.

The Placket–Burman trial data are presented in

Table 3. Plackett–Burman test design and results.

Assay	Variable Levels							Inhibitory Diameter (mm)
	X1	X2	X3	X4	X5	X6	X7
1	−1	−1	1	1	−1	1	1	18.14
2	−1	1	1	−1	1	1	1	16.39
3	1	1	−1	−1	1	−1	1	17.28
4	1	1	−1	1	−1	1	1	17.49
5	1	−1	1	−1	−1	−1	1	17.45
6	−1	1	−1	−1	−1	1	−1	17.54
7	1	−1	1	−1	1	1	−1	18.11
8	−1	1	1	1	1	−1	−1	16.47
9	1	−1	−1	1	1	1	−1	18.87
10	−1	−1	−1	−1	−1	−1	−1	18.23
11	−1	−1	−1	1	1	−1	1	18.21
12	1	1	1	1	−1	−1	−1	16.73

. By analyzing the main effects and the significance of each factor (Appendix A,

Table A1. Significance analysis of the Plackett–Burman test.

Source	Sum of Squares	df	Mean Square	F	p
Model	6.40	7	0.91	21.96	0.0049 **
X1-Soluble starch	0.08	1	0.08	1.81	0.2501
X2-Ammonium sulfate	4.21	1	4.21	101.21	0.0005 **
X3-Inoculation amount	1.56	1	1.56	37.54	0.0036 **
X4-Temperature	0.07	1	0.07	1.66	0.2673
X5-Initial pH	0.01	1	0.01	0.13	0.7414
X6-Incubation time	0.39	1	0.39	9.43	0.0373 *
X7-Rotational speed	0.08	1	0.08	1.96	0.2339

Note: “*” indicates significant difference (p < 0.05); “**” indicates extremely significant difference (p < 0.01).

), ammonium sulfate concentration, inoculation amount, and incubation time were revealed as the three most significant factors affecting liquid fermentation. All three factors significantly affected fermentation (p < 0.05). Other factors have little effect on fermentation (p > 0.05). The regression model was statistically meaningful.

(9)

Results of the steepest ascent trial

The Plackett–Burman test indicated that among the factors influencing the conditions for the fermentation of WM-37, the ammonium sulfate concentration, inoculation amount, and incubation time significantly influenced the inhibition zone diameter (Y value). All three of these factors had significant positive effects. The direction and step size of these three factors should be determined according to their proportional impact. Trial condition 3 was optimal (

Table 4. Climbing test results.

Assay	X2 (g·L−1)	X3 (%)	X6 (d)	Inhibitory Diameter (mm)
1	5	10	6	16.52
2	4.5	9	6.5	18.67
3	4	8	7	21.85
4	3.5	7	7.5	20.15
5	3	6	8	18.91

), resulting in an inhibition zone diameter of 21.85 ± 0.49 mm. Therefore, it was selected for the trials involving individual factors.

(10)

Box–Behnken trial results

To determine the interactions among X₂, X₃, and X₆, a Box–Behnken design was employed to analyze 17 trial groups. The factors and their levels are summarized in

Table 5. Box–Behnken test factors and levels.

Source	Fermentation Conditions	Level
		−1	0	+1
X2	Amount of ammonium sulfate added (g·L−1)	3.5	4	4.5
X3	Inoculation amount (%)	7	8	9
X6	Incubation time (d)	6.5	7	7.5

. The trial design and results are presented in

Table 6. Box–Behnken trials and results.

Assay	Variable Levels			Inhibitory Diameter (mm)
	X2	X3	X6
1	1	0	1	18.73
2	0	0	0	21.36
3	1	1	0	18.99
4	0	0	0	22.05
5	−1	−1	0	21.02
6	0	0	0	21.96
7	0	−1	−1	18.56
8	−1	0	1	17.82
9	0	0	0	21.72
10	1	−1	0	17.69
11	0	−1	1	19.37
12	1	0	−1	16.36
13	0	1	1	18.93
14	0	1	−1	17.87
15	−1	0	−1	19.49
16	0	0	0	21.41
17	−1	1	0	18.63

.

On the basis of the trial data, a quadratic polynomial regression analysis was performed to predict the key factors associated with the maximum inhibition zone diameter. As shown in Appendix A,

Table A2. Response surface ANOVA.

Source	Sum of Squares	df	Mean Square	F	p	Significant
model	47.71	9	5.30	58.73	<0.0001 **	significant
X2	3.37	1	3.37	37.30	0.0005 **
X3	0.6160	1	0.6160	6.83	0.0348 *
X6	0.8256	1	0.8256	9.15	0.0193 *
X2X3	3.40	1	3.40	37.71	0.0005 **
X2X6	4.08	1	4.08	45.21	0.0003 **
X3X6	0.0156	1	0.0156	0.1731	0.6898
X22	10.78	1	10.78	119.42	< 0.0001 **
X32	4.36	1	4.36	48.30	0.0002 **
X62	16.84	1	16.84	186.59	< 0.0001 **
Residua	0.6318	7	0.0903
Lack of Fit	0.2416	3	0.0805	0.8256	0.5444	Not significant
Pure Error	0.3902	4	0.0976
Cor Total	48.34	16

Note: “*” indicates significant difference (p < 0.05); “**” indicates extremely significant difference (p < 0.01).

, the model had a highly significant value (p < 0.0001), whereas the lack of fit was 0.5444 (insignificant), indicating that the model was reliable and could be used for subsequent optimization and prediction work. The R² value of the regression equation was 0.9869, whereas the adjusted R² value was 0.9701 (i.e., both exceeded 0.8). The difference between R²_Adj (0.9701) and R²_Pre (0.9074) was less than 0.2, indicating the model was highly significant. Thus, the Box–Behnken model was reliable and accurate. Design-Expert 13 software was used to process the data, which resulted in the following regression equation: 21.70 − 0.6487X₂ − 0.2775X₃ + 0.3213X₆ + 0.9225X₂X₃ + 1.01X₂X₆ + 0.0625X₃X₆ − 1.60X₂² − 1.02X₃² − 2.00X₆². As shown in Figure 15 and Figure 16, when the concentration of ammonium sulfate was constant, the inoculation amount increased from 7% to 9%, and the incubation time increased from 6.5 days to 7.5 days, the inhibitory diameter showed a trend of first increasing and then decreasing. As shown in Figure 15 and Figure 17, when the inoculation amount was constant, the concentration of ammonium sulfate increased from 3.5 g·L⁻¹ to 4.5 g·L⁻¹, and the incubation time increased from 6.5 days to 7.5 days, the inhibitory diameter showed a trend of first increasing and then decreasing. As shown in Figure 16 and Figure 17, when the incubation time was fixed, the concentration of ammonium sulfate increased from 3.5 g·L⁻¹ to 4.5 g·L⁻¹, and the inoculated amount increased from 7% to 9%, the inhibitory diameter showed a trend of first increasing and then decreasing. The response surface analysis diagram and surface diagram of the above results were convex, indicating that the inhibitory diameter has a maximum value. After Design-Expert 13 software was used to process the regression equation, the predicted optimal conditions (3.86 g·L⁻¹ ammonium sulfate; 7.74% inoculation amount; and 7-day incubation time) were expected to result in an inhibition zone diameter of 21.83 mm. In summary, the basic fermentation medium composition and conditions were as follows: erlenmeyer flask of 250 mL, liquid volume of 100mL, soluble starch of 20.0 g·L⁻¹, potassium nitrate of 1.0 g·L⁻¹, sodium chloride of 0.5 g·L⁻¹, dipotassium phosphate of 0.5 g·L⁻¹, magnesium sulfate of 0.50 g·L⁻¹, ferrous sulfate of 0.01 g·L⁻¹, inoculation amount of 8%, temperature of 30 °C, initial pH of 7.0, incubation time of 7-day, and rotation speed of 160 rpm. The optimized fermentation medium composition and conditions were as follows: erlenmeyer flask of 250 mL, liquid volume of 100 mL, soluble starch of 20.0 g·L⁻¹, ammonium sulfate of 3.86 g·L⁻¹, sodium chloride of 0.5 g·L⁻¹, dipotassium phosphate of 0.5 g·L⁻¹, magnesium sulfate of 0.50 g·L⁻¹, ferrous sulfate of 0.01 g·L⁻¹, inoculation amount of 7.74%, temperature of 30 °C, initial pH of 7.0, incubation time of 7-day, and rotation speed of 160 rpm.

(11)

Model validation trial

The optimal fermentation conditions predicted using the response surface method (i.e., 3.86 g·L⁻¹ ammonium sulfate; 7.74% inoculation amount; and 7-day incubation time) resulted in an inhibition zone diameter of 21.56 ± 0.32 mm (i.e., 98.76% of the predicted diameter). Thus, the model accurately predicted the optimal fermentation conditions.

3.4. Preparation of Crude Extract

After a pretreatment step, the 20.0 L fermentation product for WM-37 was adsorbed by D101 macroporous resin and eluted with an ethanol-water gradient. The antifungal effect of each fraction against S. rolfsii was detected using the Oxford cup method. Antifungal activity was detected in the fractions eluted using ethanol-water volume ratios of 50:50, 70:30, and 90:10. These fractions were collected and concentrated using a rotary evaporator to obtain 7.36 g crude extract, which was dissolved in methanol.

3.5. Extraction and Purification

The crude extract was dissolved in methanol and passed through a 0.22 μm filter membrane prior to the separation via Waters preparative chromatography, with water (0.01% formic acid) and methanol used as eluents. Figure 18 presents the chromatogram of the crude extract. The collected chromatographic fractions yielded seven components, which were designated as I–VII. Each component was concentrated using a rotary evaporator to obtain extracts (12.2, 8.31, 8.9, 8.4, 7.5, 7.2, and 16.7 mg for components I, II, III, IV, V, VI, and VII, respectively).

Each component was dissolved in methanol, after which the antifungal activity of each component was determined using the Oxford cup method, with S. rolfsii as the indicator organism. Component VII had a significant antifungal activity, resulting in an inhibition zone diameter of 23.82 ± 0.54 mm (Figure 19). Figure 20 presents the HPLC chromatogram of component VII, with a peak appearing between 48 and 49 min. This component was analyzed further to identify its chemical constituents.

3.6. Identification of Active Substances with Inhibitory Effects on Fungal Growth

The lyophilized antifungal active substance was a slightly yellow powder. The total ion chromatogram of the antifungal compound included an ion peak at 0.809 min (Figure 21). The ultra-high-resolution MS spectrum (Figure 22) had a [M + H⁺] mass of 189.85783. Using this spectrum to screen the GNPS and the available literature enabled [37,38,39,40] us to determine that the antifungal compound was azelaic acid (C₉H₁₆O₄). The molecular structure of azelaic acid is illustrated in Figure 23. Initially discovered in plant tissues (e.g., Agave americana cuticle), azelaic acid has also been detected among the metabolites in various microorganisms. The current findings suggest that azelaic acid was the primary antifungal compound among the secondary metabolites of S. rectiviolaceus. This discovery sheds light on the biosynthetic process and potential biological activity of azelaic acid in microorganisms. Future studies should explore the biosynthesis and pharmacological effects of azelaic acid as well as its potential utility for developing new antifungal drugs and fungicides.

4. Discussion

Actinomycetes are a diverse group of microorganisms that are rich in valuable chemical compounds. For example, the antibiotics they produce are critical for the protection of plants, animals, and humans from harmful microorganisms [41,42,43,44]. Streptomyces species are the main microbes used for the biocontrol of plant diseases because of their wide host range. Moreover, they are currently the most studied actinomycetes. Three actinomycetes (S3, S12, and S40) isolated from the surface soil of chickpea fields by Mahboobeh et al. reportedly have significant inhibitory effects against the causative agent of chickpea wilt; S3 and S12 were most similar to Streptomyces antibioticus, whereas S40 was most similar to Streptomyces peruviensis [45]. Strain SM3-7, isolated from maize rhizosphere soil by Hou et al., has significant inhibitory effects against F. verticillioides, which causes maize stalk rot; this strain was identified as Streptomyces sclerogranulatus [46]. In the current study, we isolated S. rectiviolaceus WM-37, which may be useful for controlling peony southern blight, from the surface soil of the Henan University of Science and Technology farm. According to a detailed analysis of biological characteristics and antifungal activities, this strain can effectively suppress the pathogen responsible for peony southern blight, but it also has promising inhibitory effects against various other pathogens, including those responsible for peony root rot disease, peony yellow spot disease, tomato leaf spot disease, rice sheath blight disease, wheat sheath blight disease, wheat crown rot disease, tobacco root black rot disease, and maize ear rot disease. These findings suggest S. rectiviolaceus WM-37 may be a versatile biocontrol agent. Thus, the utility of WM-37 for developing biopesticides useful for preventing and controlling multiple diseases in peonies and other crops must be investigated. In recent years, there has been increasing interest in microbial antagonists, such as Streptomyces sp., in terms of their potential use as biocontrol agents to protect plants from disease [47]. Compared with traditional chemical pesticides, biopesticides are generally safer, more environmentally friendly, and more sustainable, making them ideal for modern agricultural practices.

Optimizing the fermentation process is crucial for the industrial application of microbial products. Refining fermentation conditions can shorten the fermentation cycle, decrease costs, and improve the input–output ratio, ultimately leading to optimized fermentation at an industrial scale. Fermentation conditions can significantly influence the antifungal activities of antagonistic microorganisms, which are reflected by the diameter and area of inhibition zones [48]. Although traditional single-factor trials and orthogonal trials are commonly conducted to optimize fermentation conditions, they have limitations (e.g., an inability to accurately assess the interactions between various factors and potential subjectivity and omissions in trial design). Alternatively, the response surface method is better for comprehensively and accurately analyzing interactions among multiple factors, thereby providing more reliable data for improving fermentation conditions. In addition, the response surface method can generate three-dimensional response surfaces and contour plots of the interactions between important factors via regression equation models, which can directly reflect the trends in the effects of certain factors on selected variables [49]. Gao et al. increased the antifungal effect of Streptomyces lavendulae by 20% by optimizing nutritional parameters [50]. Wang et al. showed that the production of ingredients with Strain Hhs.015^T (Saccharothrix yanglingensis sp. nov.) antimicrobial activity had been improved by 20% when compared with the basal fermentation conditions (14.4 ± 0.15 mm) [27]. In this study, the fermentation conditions were successfully optimized by combining single-factor tests and response surface methodology to maximize the antifungal effect and verify the reliability of the obtained model. The antifungal effect increased by 28.79%, which was an improvement of 8.79% higher than the optimization by Gao et al. [50] and also an increase of 8.79% compared to the optimization by Wang et al. [27]. This approach can enhance the antifungal effect of biocontrol agents, while also providing the scientific basis for the industrial production and application of biocontrol agents. Future studies will need to identify and analyze additional influencing factors to further improve the fermentation process and increase antifungal effects.

Actinomycetes are an important class of microorganisms partly because of the diversity in their secondary metabolites with significant biological activities. Accordingly, they have broad industrial uses (e.g., in agriculture and medicine). Their metabolites exhibit antimicrobial, insecticidal, herbicidal, and antitumor effects [11]. Guo et al. reported that Streptomyces sp. HMH1 and Streptomyces sp. HML1 have strong antagonistic effects against agricultural pathogens. Additionally, they isolated and purified four compounds, which were identified as actinomycin D, actinomycin X2, actinomycin XO-β, and K-252d, of which actinomycin D has broad-spectrum antimicrobial effects against several plant pathogenic fungi and bacteria [51]. Kong et al. determined that an endophytic actinomycete (Streptomyces californicus XY3-1) isolated from the leaves of a Chinese mosquito tree has strong inhibitory activities against plant pathogens. Nine compounds were isolated and identified from the ethyl acetate extract derived from the fermentation broth. The structures of these compounds were determined on the basis of nuclear magnetic resonance and mass spectrometry data as well as a literature search. Notably, benzodipyran, okicenone, and (6-methyl-8-oxo-5,8-dihydronaphthalen-1-yl)-acetic acid were identified as the main antibacterial compounds [52]. In the current study, we isolated and identified azelaic acid as the main antifungal compound produced by WM-37. This finding provides new insights into the mechanism underlying the antifungal activities of this bacterial strain, with potential implications for the development of novel azelaic acid-based antimicrobial agents. As an organic acid, it is not surprising that azelaic acid exhibits antimicrobial activity. Many organic acids have confirmed antibacterial effects, which can disrupt bacterial cell membranes, interfere with metabolic processes, or inhibit key enzyme activities. Organic acids, such as lactic acid, acetic acid, and propionic acid, are the main metabolites related to the antifungal activity of lactic acid bacteria [53]. Yun and Lee demonstrated that the inhibitory effect of propionic acid is associated with mitochondria-mediated apoptosis, leading to fungal cell death [54]. Although the specific mechanism underlying the antifungal effect of azelaic acid remains to be thoroughly investigated, its molecular structure and chemical properties suggest it may disrupt physiological functions, similar to other organic acids. The antifungal activity of azelaic acid may be modulated by various factors, including the concentration, pH, temperature, and interactions with other compounds. Furthermore, azelaic acid and other antimicrobial compounds may have synergistic effects, suggestive of a complex antimicrobial system that will need to be precisely characterized. In terms of its potential utility, the natural antifungal properties of azelaic acid may be exploited to develop new antimicrobial agents. Compared with conventional antibiotics, azelaic acid-based antimicrobial agents may be less toxic and have broader activities, making them potentially useful for addressing relevant issues (e.g., antibiotic misuse and resistance). Future studies should focus on isolating and identifying secondary active substances and clarifying their interactions and synergistic effects with azelaic acid. This approach may enhance our understanding of the antimicrobial mechanism of bacterial strains, which may lead to the development of effective and non-toxic antimicrobial agents. The current study revealed the importance of azelaic acid as the primary antifungal compound in WM-37. Further analyses of its antimicrobial mechanism and potential applications will supplement the available information regarding antimicrobial substances, while also leading to the development of innovative strategies for mitigating antibiotic resistance.

5. Conclusion

In this study, 46 strains were initially isolated from the surface soil of the Henan University of Science and Technology farm planting area (40 actinomycete strains and six bacterial strains). Using S. rolfsii, which can infect peony plants, as the target pathogen, three strains with inhibitory effects were selected. According to the Oxford cup method, WM-37 was the most effective strain for inhibiting the growth of fungal pathogens responsible for diseases of peony and other plants, indicative of its high research value. Morphological, physiological, and biochemical analyses as well as an examination of the 16S rRNA sequence identified WM-37 as an S. rectiviolaceus strain.

Optimizing fermentation conditions can accelerate the fermentation cycle, minimize costs, and improve efficiency to achieve fermentation objectives and enhance industrial production. This study applied a single-factor trial and a response surface method to optimize carbon and nitrogen sources, inorganic salts, and fermentation conditions for S. rectiviolaceus. The inhibition zone diameter on medium inoculated with S. rolfsii was used as the response value. Initially, individual factors were optimized in a single-factor trial. Subsequently, a Plackett–Burman test was conducted to evaluate several factors influencing fermentation, which led to the identification of three significant factors affecting fermentation (ammonium sulfate concentration, inoculation amount, and incubation time). On the basis of the steepest ascent and Box–Behnken trials, the optimal values for these three significant influencing factors were determined (3.86 g·L⁻¹ ammonium sulfate; 7.74% inoculation amount; and 7-day incubation time), which resulted in an inhibition zone diameter of 21.83 mm. The optimal conditions obtained using the response surface method were applied for a liquid fermentation test involving S. rectiviolaceus WM-37. The resulting inhibition zone diameter was 21.56 ± 0.32 mm, which was 98.76% of the predicted value, indicating that the model was effective and accurate. The optimized inhibition zone diameter was 28.79% larger than the pre-optimized inhibition zone diameter.

Following the fermentation under optimized conditions, the fermentation broth was added to D101 macroporous resin and eluted with an ethanol-water gradient. The collected fractions were evaporated and concentrated. The resulting crude extract was dissolved in methanol and separated by HPLC, revealing seven main components (I–VII). Among these components, VII had significant inhibitory effects. The antifungal compound was identified as azoic acid (C₉H₁₆O₄) by a comparison of mass spectrometry data, the GNPS database, and the existing literature.

In this study, a microbial strain with potential resistance to multiple plant diseases was selected, and its antibacterial effect was significantly improved through optimized fermentation conditions. The main antibacterial active substances were identified using advanced chromatography and mass spectrometry techniques. The research results provide strain resources and technical support for the biological control of plant fungal diseases of azelaic acid.