Macrolides from Streptomyces sp. SN5452 and Their Antifungal Activity against Pyricularia oryzae

Pyricularia oryzae causes rice blast, the major destructive disease in nearly all rice fields. In order to obtain highly active compounds against P. oryzae, four new 20-membered macrolides named venturicidins G–J (1–4) were isolated from the culture broth of Streptomyces sp. SN5452 along with two known ones, venturicidins A (5) and B (6). Their structures were determined by the cumulative analyses of nuclear magnetic resonance (NMR) spectroscopy and high-resolution electrospray ionization mass spectrometry (HRESIMS) data. All isolated compounds were evaluated for their antifungal activity against P. oryzae. Interestingly, these compounds exhibited obvious inhibition to mycelial growth and conidial germination of P. oryzae. Remarkably, the EC50 values of venturicidins A (5), B (6), and I (3) against mycelial growth were 0.11, 0.15 and 0.35 µg/mL, and their EC50 values of conidial germination were 0.27, 0.39 and 1.14 µg/mL, respectively. The analysis of structure-activity relationships (SARs) revealed that the methylated positions might be involved in the antifungal activity of venturicidins. These results indicate that the venturicidins are prospective candidates for novel fungicides that can be applied in controlling rice blast.


Introduction
The need for food quality and quantity is urgent due to the increase in population and the improvement of people's living standards [1]. Rice is the staple food for most peoples in different countries around the world [2], including China, Bangladesh, and Malaysia. However, the yield of rice is threatened by a variety of plant diseases every year, among which rice blast can cause up to 30% yield loss in some regions [3,4]. Rice blast is caused by Pyricularia oryzae [5], a haploid filamentous ascomycete [6,7], which can infect any stage of plant growth, causing leaf blast, node blast and panicle blast [8][9][10][11].
At present, the control of rice blast mainly depends on applying synthetic chemical agents [12,13]. These have been repeatedly applied for decades, which not only develops resistance of P. oryzae to them [14,15], but also destroys the balance of ecosystem [16]. Moreover, after several years of planting disease-resistant varieties [17], the pathogens may develop new pathogenic races, resulting in the collapse of resistance [11,18]. In order to solve these difficulties, scientists expect to seek for environmentally friendly and safe microbial biocontrol agents.
In our ongoing efforts to seek for new and bioactive secondary metabolites produced by actinomycetes [28,29], the crude extract from the fermentation culture of Streptomyces sp. SN5452 at 50 µg/mL could completely inhibit the mycelial growth of P. oryzae ( Figure S1). In this study, we isolated and purified the active compounds from the crude extract. The structures of these active compounds were determined based on the analysis of mass spectrometry (MS) and nuclear magnetic resonance (NMR) data. Furthermore, their antifungal activity against P. oryzae was also evaluated with mycelial growth inhibition and conidial germination assays.

General Experimental Procedures
Optical rotations were obtained at the sodium D line with a polarimeter (Atago, Tokyo, Japan), maintained at room temperature. NMR spectra experiments were operated on an Avance-600 NMR spectrometer (Bruker, Karlsruhe, Germany). Chemical shifts were calibrated by carbon signals and the residual proton signals of DMSO-d 6 (δ C 39.5 and δ H 2.50). High-resolution mass spectra (HRESIMS) were recorded on an Agilent 1260/6520 Q-TOF mass spectrometer. The crude extract was chromatographed on silica column with silica gel of 100-200 and 200-300 mesh (Qingdao Ocean Chemical Co., Ltd., Qingdao, China) and Sephadex LH-20 (GE Healthcare, Uppsala, Sweden). High-performance liquid chromatography (HPLC) analysis was performed using the C18 column (Agilent ZORBAX Eclipse XDB, 4.60 × 250 mm, 5 µm) on an Agilent 1260 series system (Agilent, Santa Clara, CA, USA). Active compounds were collected using semi-preparative HPLC with a C18 column (Agilent ZORBAX Eclipse XDB, 9.4 × 250 mm, 5 µm). The germination number of conidia was observed by microscope (Nikon, Tokyo, Japan). All chemical agents were purchased from Sinopharm Chemical Reagent company (Shanghai, China).

Actinomycete Material
The stain was isolated from the gut of millipede (Kronopolites svenhedind Verhoeff) which was obtained from campus of Shenyang Agricultural University. The processing of the Diplopoda gut sample and the isolation of the strain were conducted according to Heo et al.'s method [30]. For taxonomic identification, the strain 16S rRNA sequence was compared and analyzed by EzTaxon database. A phylogenetic tree was constructed based on 16S rRNA sequence using Molecular Evolutionary Genetics (MEGA v 7.0) [31,32]. Colonies were deposited in 20% (v/v) glycerol solution at −80 • C.

Fermentation and Extraction
Streptomyces sp. SN5452 was cultured on Gause's synthetic agar no. 1 (GS) [33] plates at 28 • C in the dark for 10 days for subsequent fermentation. The mycelia were inoculated into test tubes with 5 mL liquid ISP 2 medium [34] and shaken (180 rpm) at 28 • C in the dark for 2 days, and then were transferred to 250 mL Erlenmeyer flask with 50 mL liquid ISP 2 medium with shaking for 2 days to prepare the seed culture. Finally, the seed culture was shifted into 2 L Erlenmeyer flasks, which contained 400 mL of the GS liquid medium and 16 g of Amberlite XAD-16 resin. A total of 48 L fermentation culture was obtained after cultivating for 7 days under identical conditions.
The resin was collected from the fermentation broth by repeated deionized water washing. After it was dried in an oven at 30 • C to remove moisture, the resin was extracted four times with CH 3 OH. The CH 3 OH fractions were pooled and concentrated to obtain the CH 3 OH extract, and then it was redissolved in 50% CH 3 OH in H 2 O (0.6 L). The solution was extracted four times with the same volume of CH 2 Cl 2 . Collecting CH 2 Cl 2 fractions were concentrated to produce 11 g of crude extract.

Effect of Compounds on the Mycelial Growth of P. oryzae
Mycelial growth inhibition assay was performed as described by Li et al. [35] to measure the inhibitory effect of compounds 1-6 on P. oryzae. The tested compounds were dissolved in sterile water containing 0.5% DMSO and 0.25% Tween-80 to produce a desired agent concentration of 0.078125-8 µg/mL. The agents were blended with 40-45 • C potato dextrose agar, in which 0 µg/mL represented the negative control. The carbendazim was considered as the positive control. The 5 mm plugs of P. oryzae were placed in the center of treated plates (d = 90 mm). After 15 days of dark culture at 25 • C, the colony diameter of each test group was measured. Experiments were repeated in three replicates. The antifungal activities were calculated using the following formula: among which, D C is colony diameter in the negative control plate and D T is colony diameter in the plate containing tested compounds.
2.6. Effect of Compounds on the Conidia Germination of P. oryzae P. oryzae was incubated on oatmeal tomato agar medium as described by Miao et al. [36]. When P. oryzae reached the edge of the plate, the aerial growth was scraped off with sterilized cotton swabs. They were then placed in an incubator with black light lamp at 25 • C. After 7 days, the conidia were harvested from conidial colonies. The concentrations of conidia in suspension were adjusted to 1 × 10 5 conidia per mL by a hemocytometer [37] and 30 µL of various concentrations (0.078125-100 µg/mL) of tested compounds were dropped onto the glass slides with 30 µL conidial suspension before incubating both of them for 6 h. The sterile water containing 0.5% DMSO and 0.25% Tween-80 was considered as the negative control. The carbendazim was considered as the positive control. The conidia were considered to have germinated when the length of the germ tube was greater than the short radius of the conidia. When the negative control conidial germination rate was greater than 80%, the number of conidial germinated at various concentration was observed and assessed by under the microscope. Experiments were repeated in three replicates.

Identification of Strain SN5452
The 16S rRNA gene sequence analysis and comparisons showed that strain SN5452 was affiliated with the genus Streptomyces and shared the greatest gene sequence similarity to Streptomyces setonii (99.79%). The phylogenetic analysis of the 16S rRNA gene sequences indicated the stain forms a cluster with Streptomyces clavifer NRRL B-2557 T , Streptomyces mutomycini NRRL B-65393 T , Streptomyces atroolivaceus NRRL ISP-5137 T and Streptomyces finlayi NRRL B-12114 T ( Figure S2). Therefore, the strain SN5452 belongs to the genus Streptomyces and was named Streptomyces sp. SN5452 (Genbank accession no. ON358333).

Extraction, Separation and Purifcation of Extract
The CH 2 Cl 2 extract from the fermentation culture of Streptomyces sp. SN5452 was chromatographed on silica gel column followed by further purification on Sephadex LH-20 columns and reversed-phase HPLC, to afford compounds 1-6. Analysis of ESI-MS data and 1 H and 13 C NMR spectra suggested 5 and 6 to be venturicidins A and B, respectively, whose identities were unambiguously confirmed by extensive 2D NMR (COSY, HSQC and gHMBC) spectroscopic analyses as well as comparison to previously reported spectroscopic data (Figure 1) [38].
The CH2Cl2 extract from the fermentation culture of Streptomyces sp. SN5452 was chromatographed on silica gel column followed by further purification on Sephadex LH-20 columns and reversed-phase HPLC, to afford compounds 1-6. Analysis of ESI-MS data and 1 H and 13 C NMR spectra suggested 5 and 6 to be venturicidins A and B, respectively, whose identities were unambiguously confirmed by extensive 2D NMR (COSY, HSQC and gHMBC) spectroscopic analyses as well as comparison to previously reported spectroscopic data (Figure 1) [38].   Table 1 and Figure S3). The 13 C NMR and HSQC spectra of 1 showed eight methyl groups, ten methylene groups, sixteen methine groups, and five quaternary carbons ( Figures S4  and S5). The spectroscopic data also revealed two olefinic, one ketone, one ketal and one lactone carbon signals. These results indicated that 1 possesses the same venturicidin   (Table 1 and Figure S3). The 13 C NMR and HSQC spectra of 1 showed eight methyl groups, ten methylene groups, sixteen methine groups, and five quaternary carbons ( Figures S4 and S5). The spectroscopic data also revealed two olefinic, one ketone, one ketal and one lactone carbon signals. These results indicated that 1 possesses the same venturicidin scaffold as that of 6 ( Figure 2). Analyses of the 1 H and 13 C NMR spectra of 1 and 6 showed that one methyl group (δ H 0.78; δ C 12.9) and one methine group (δ H 1.80; δ C 34.6) found in 6 were not present in 1 (Table S1). Instead, 1 contains an additional methylene group (δ H 1.29, 1.47; δ C 29.0). The HMBC cross peaks of H 2 -18 (δ H 1.29, 1.47) with C-17 (δ C 32.8) and C-19 (δ C 78.3) indicated the methylene is located at C-18 (Figures 2 and S6). Therefore, the structure of 1 differs from that of 6 in only one aspect: the methyl group at C-18 in 6 is replaced by a proton in 1. This was also confirmed by the HRESIMS data and molecular formulas of 1 and 6 which indicated 6 has one more carbon and two more hydrogens than 1. The structure of compound 1 was thus elucidated, and compound 1 was named as venturicidin G.

Structure Elucidation of Compounds
Microorganisms 2022, 10, x FOR PEER REVIEW 7 of 12 scaffold as that of 6 ( Figure 2). Analyses of the 1 H and 13 C NMR spectra of 1 and 6 showed that one methyl group (δH 0.78; δC 12.9) and one methine group (δH 1.80; δC 34.6) found in 6 were not present in 1 (Table S1) Figures 2 and S6). Therefore, the structure of 1 differs from that of 6 in only one aspect: the methyl group at C-18 in 6 is replaced by a proton in 1. This was also confirmed by the HRESIMS data and molecular formulas of 1 and 6 which indicated 6 has one more carbon and two more hydrogens than 1. The structure of compound 1 was thus elucidated, and compound 1 was named as venturicidin G. Compound 2 was purified as a white powder. The molecular formula of 2 was determined as C41H68O10 based on HRESIMS data ( Figure S15). Similarly, 2 was identified to have the same scaffold as 6 based on their spectroscopic data (Figures S9-S14). However, the 1 H and 13 C NMR data revealed that the signals of methylene (δH 2.73, 2.54; δC 43.7) in 6 were substituted by signals of a methyl group (δH 1.12; δC 13.0) and a methine group (δH 2.57; δC 47.6) in 2, respectively (Tables 1, 2 and S1). The HMBC correlations of H3-2CH3 (δH 1.12) with C-1 (δC 176.0), C-2 (δC 47.6) and C-3 (δC 95.7) and COSY correlation between H3- Compound 2 was purified as a white powder. The molecular formula of 2 was determined as C 41 H 68 O 10 based on HRESIMS data ( Figure S15). Similarly, 2 was identified to have the same scaffold as 6 based on their spectroscopic data (Figures S9-S14). However, the 1 H and 13 C NMR data revealed that the signals of methylene (δ H 2.73, 2.54; δ C 43.7) in 6 were substituted by signals of a methyl group (δ H 1.12; δ C 13.0) and a methine group (δ H 2.57; δ C 47.6) in 2, respectively ( Table 1, Table 2 and Table S1). The HMBC correlations of H 3 -2CH 3 (δ H 1.12) with C-1 (δ C 176.0), C-2 (δ C 47.6) and C-3 (δ C 95.7) and COSY correlation between H 3 -2CH 3 (δ H 1.12) and H-2 (δ H 2.57) illustrated the position of the methyl at C-2 ( Figure 2). From these results, the structure of compound 2 was elucidated, and compound 2 was named as venturicidin H.
Compound 3 was obtained as white powder. Its molecular formula, C 40 H 65 NO 11 , was inferred by HRESIMS data ( Figure S21). The NMR data exhibited structure of 3 was highly similar to that of 5 ( Figure S16-S20). The analysis of 1 H NMR data of 5 exhibited a methyl (H 3 -16CH 3 , δ H 0.89) that was absent in 3, while 3 had a methylene at H 2 -16 (δ H 2.04, 1.97) rather than a methine at H-16 (δ H 2.06) in 5 (Tables 1 and S1). The HMBC correlations of H 2 -16 with C-15 (δ C 137.4) and C-17 (δ C 34.3) also revealed the methyl (δ H 0.89) in 5 was missing in 3 (Figure 2), corresponding to a 14 Da mass decrease. Consequently, 3 was elucidated and named as venturicidin I.
Compound 4 was isolated as white powder, and revealed spectroscopic data remarkable similar with those of 5 ( Figure S22-S27). The molecular formula of 4 was determined as C 42 H 69 NO 11 compatibly with its HRESIMS data ( Figure S28), which was 14 amu higher than that of 5, suggesting the presence of an additional methyl group (δ H 1.13, δ C 13.0) in 4 (Tables 1 and 2). The HMBC correlations of H 3 -2CH 3 (δ H 1.13) with C-1 (δ C 176.0), C-2 (δ C 47.6) and C-3 (δ C 95.7) and COSY correlation between H 3 -2CH 3 (δ H 1.13) and H-2 (δ H 2.57) confirmed that the additional methyl was located at C-2 of 4 ( Figure 2). Therefore, the structure of 4 was determined as illustrated in Figure 1, and compound 4 was named as venturicidin J.

Antifungal Activity Assay
Compounds 1-6 were evaluated for their antifungal activities against P celial growth inhibition and conidial germination assays. The results showe of the compounds 1-6 against mycelial growth were approximately 1.78, 1 0.11, 0.15 µg/mL, respectively (Table 3). Notably, the compounds 3, 5 and limit the mycelial radial elongation of P. oryzae (Figure 3), which showed co tifungal activity to the positive control carbendazim (EC50 = 0.30 µg/mL). Co and 4 showed moderate antifungal activities. In addition to inhibiting mycelial growth, compounds 1-6 also inhibite germination of P. oryzae in different levels, with the EC50 values estimated to  In addition to inhibiting mycelial growth, compounds 1-6 also inhibited the conidial germination of P. oryzae in different levels, with the EC 50 values estimated to be 24.95, 5.55, 1.14, 4.49, 0.27 and 0.39 µg/mL, respectively (Table 3). These results indicated that compounds 5 and 6 exert higher antimicrobial activity against P. oryzae than the positive control carbendazim (EC 50 = 3.99 µg/mL). Compounds 2, 3 and 4 showed comparable antifungal activity to the positive control, and compound 1 exhibited weaker inhibitory activity than carbendazim.

Structure Activity Relationship Analysis
The relationship between the structures and antifungal activity was elucidated based on the compounds' molecular structures, and the discussion upon their antimicrobial activity results against different developmental stages of P. oryzae. Both compounds 5 and 6 showed similar antifungal activity against P. oryzae, indicating the presence of acylamino group at the C-3 position does not cause a prominent effect on its activity. Comparing the structures of 1 with 6, a methyl group at the C-18 (R 4 ) was replaced by a hydrogen atom in 1, which impairs the antifungal activity against P. oryzae, so they have large activity differences in conidial germination inhibition. In contrast, between 3 and 5 the replacement in 3 happened on C-16 (R 3 ), which indicates that they have no differences in antifungal activity. Thus, the methyl group in position R 4 seems to increase the activity. Both compounds 2 and 4 introduce a methyl group at the C-2 position, lower the antimicrobial activity in comparison with that of 5 and 6. These results suggested that the antifungal activities of these compounds are influenced by the methylated positions in venturicidins. The SARs of these compounds provided a scientific basis for discovery of potent fungicides.

Discussion
With the increasing devastating rice blast disease, P. oryzae has provoked severe loss in the world [39]. The control of rice blast remains a long-standing agricultural issue associated with highly variable nature of P. oryzae [37,40]. In addition, P. oryzae has developed resistance to most of commercially available fungicides [16]. Thus, novel fungicides are a constant and critical need. Recently, microbial pesticides are widely applied in the control of plant diseases due to friendliness to the environment [41]. Unfortunately, there are few reports on the microbial pesticides against P. oryzae. This study indicated that secondary metabolites of Streptomyces sp. SN5452 may be promising fungicides in the control of rice blast, because they strongly inhibit the mycelial growth and the conidial germination of P. oryzae. The isolation and screening of Streptomyces are the prerequisites for obtaining bioactive natural products [42]. In our study, Streptomyces sp. SN5452 was isolated and purified from the gut of a millipede. By the analysis of 16S rRNA gene sequence, the strain belonged to Streptomyces genus, and the crude extract made from the fermentation culture of the strain showed prominent inhibitory activity against P. oryzae.
Streptomyces produce a number of secondary metabolites used in fields. The commercial kasugamycin produced by Streptomyces kasugaensis showed potential preventive effects against rice blast [43]. The insecticide avermectin has been used extensively for controlling Plutella xylostella and Pieris rapae in fields, which was produced by Streptomyces avermitilis [44][45][46]. Furthermore, Streptomyces fermentation products are considered as a valuable resource for the development of novel pesticides. The dimethyl sulfide and trimethyl sulfide produced by Streptomyces sp. AN090126, showed broad-spectrum antimicrobial activity against various plant-pathogenic bacteria and fungi, including Ralstonia solanacearum, Xanthomonas euvesicatoria, Sclerotinia homoeocarpa [47]. The LC 50 value of endostemonine J, an ionophore antibiotic produced by Streptomyces sp. BS-1, against Aphis gossypii was 3.55 µg/mL at 72 h via leaf dipping assay [48]. Compound cinnoline-4carboxylic acid was isolated from the fermentation culture of Streptomyces sp. KRA17-580 by Kim's group, and completely inhibited the growth Digitaria ciliaris at a concentration of 50 µg/mL [49]. In this study, venturicidins A, B and venturicidins G-J were isolated from the fermentation culture of Streptomyces sp. SN5452. These compounds inhibited mycelial growth and conidial germination of P. oryzae, with EC 50 values ranging from 0.11 to 1.78 µg/mL and from 0.27 to 24.95 µg/mL, respectively.
As we all know, structures of compounds were determined by the cumulative analyses of NMR, MS and X-ray data. The current study showed that the absolute configuration of compounds 1, 3, 5 and 6 were elucidated based on NMR and MS analyses and previously reported X-ray data [46]. However, the absolute configuration of compounds 2 and 4 were uncertain by spectroscopic data. We also tried to cultivate single-crystals of the compounds 2 and 4, but unfortunately it was not successful due to their structural specificity. Named as venturicidins G-J, compounds 1-4 enriched the structural diversity of 20-membered macrolide compounds [50][51][52][53][54][55], and compounds 5-6 were identified as venturicidins A-B. Further, venturicidins A, B and I exhibited good inhibition to mycelial growth and conidial germination of P. oryzae, which was comparable or superior to the positive control carbendazim. However, the actual field efficacy of these compounds should be further studied.

Conclusions
In conclusion, the crude extract made from the fermentation culture of Streptomyces sp. SN5452 showed inhibitory activity against P. oryzae. Compounds 1, 2, 3 and 4 were isolated and identified for the first time from the fermentation culture of Streptomyces sp. SN5452. Compounds 3, 5 and 6 exhibited good antimicrobial activity against P. oryzae. Given that microbial secondary metabolites exhibit excellent fungicide activity, they have the potential to become lead molecules for agricultural fungicides.

Acknowledgments:
The authors thank Yongming Yan of Shenzhen University for technical assistance with NMR and MS spectra.

Conflicts of Interest:
The authors declare no conflict of interest.