Rice blast disease, caused by the ascomycete fungus Pyricularia oryzae (P. oryzae), is one of the most damaging fungal diseases that leads to the reduction of rice production worldwide [1]. Several fungicides have been developed to control the disease; however, pathogenic fungi often acquire resistance to fungicides [2]. Thus, new antifungal agents need to be developed continuously to counter this disease.

Isono et al. at RIKEN previously carried out a screening for antifungal agents and identified polyoxins, nucleoside antibiotics, from the culture broth of Streptomyces cacaoi [3]. Polyoxins were then successfully developed as an agrochemical [4]. Polyoxins inhibit cell wall synthesis [5], leading to swelling of fungal cells; a characteristic morphological change caused by polyoxins [6]. Since then, we have continued the screening based on the changes in fungal cell morphology. Recently, we have constructed a database of the morphological changes in P. oryzae induced by the well-characterized antifungal agents and used it to screen for new fungicides [7].

In the course of screening, we found a new antifungal compound, YO-001A (Fig. 1a), from the culture broth of Streptomyces sp. YO15-A001. This strain was isolated from a soil sample collected at Tateyama, Toyama Prefecture, Japan. A BLAST search based on its 16S rRNA sequence revealed that YO15-A001 strain was designated as Streptomyces sp. A loopful of slant culture of YO15-A001 was inoculated into a 500-mL cylindrical flask containing 70 mL of SY medium (0.1% yeast extract (BD), 1% soluble starch (nacalai tesque), and 0.1% N-Z amine type A (Wako)). The seed culture was incubated at 28 °C for 3 days on a rotary shaker at 180 rpm. One milliliter of the seed culture was inoculated into five hundred milliliters cylindrical flasks containing seventy milliliters of oatmeal medium (6% oatmeal (Quaker Oats)), and then incubated at 28 °C for 4 days on a rotary shaker at 180 rpm. The culture broth (1.4 L) was adjusted to pH 9.0 with NaOH and extracted with EtOAc. The extracted oil sample was partitioned with Hexane/MeOH/H2O (10:9:1). The MeOH layer was collected and evaporated in vacuo to yield 297 mg of the extract, which was then fractionated via centrifugal partition chromatography (CPC) with CHCl3/MeOH/H2O (5:6:4). The active fractions were refractionated via CPC with Hexane/EtOAc/MeCN (5:1:4) and further purified via reverse-phase high-performance liquid chromatography using a C18 column with a linear gradient of acetonitrile/H2O to yield YO-001A (7.2 mg), having the following characteristics: white solid; UV (MeOH) λmax (log ε), 230 nm (4.67); [α]58924, +14.9 (c 0.28, CHCl3); IR (ATR) νmax, 3421, 2960, 2935, 2873, 1697, 1452, 1386, 1265, 1070, 985, and 752 cm−1; HR-ESI-MS m/z, 759.5022 [M + Na]+ (calculated for C42H72O10Na, 759.5018). IR spectra suggested a hydroxyl group (3421 cm−1), alkenes (2960, 2935, and 2873 cm−1), and a conjugated carbonyl moiety (1697 cm−1). The 1H NMR spectra showed six methyl signals containing one singlet signal, two doublet signals, and three triplet signals (Fig. S1 and Table 1). The 13C NMR showed 42 signals containing six methyls, fifteen methylenes, eighteen methines, and three quaternary carbon moieties (Fig. S2 and Table 1). These signals were confirmed by 13C DEPT spectra (Fig. S3). One characteristic quaternary carbon (δC 164.7) indicated the presence of an ester. Another quaternary carbon (δC 97.4) suggested the attachment of two oxygen substituents and two aliphatic carbon moieties, consistent with the spiroketal system. These partial structures suggested a potential similarity to maclafungin [8] and oligomycins [9, 10]. On comparing the 13C NMR spectra of maclafungin with those of YO-001A, a significant upfield shift at C-9 from δC 81.0 to δC 74.6, a downfield shift at C-12 from δC 45.3 to δC 74.9, an upfield shift at C-13 from δC 75.2 to δC 38.3, an upfield shift at C-24 from δC 40.0 to δC 37.4, and an upfield shift at C-34 from δC 71.6 to δC 31.2 were observed (Table 1). These differences indicated the following changes in the attached functional groups: for C-9, from a methyl ether to a hydroxyl group; for C-12, from a methyl group to hydroxyl and methyl groups; for C-13, from a hydroxyl group to a proton; for C-24, from an alkyl chain to a proton; for C-34, from a hydroxyl group to a proton. The planar structure was identified through the interpretation of 2D NMR spectra (Fig. 1b and Figs. S4, S5, S6, S7, and S8). DQF-COSY and HSQC-TOCSY spectra proved three partial structures from C-2 to C-11, from C-13 and C-26, and from C-28 to C-35. The linkage of partial structures was established by HMBC spectrum. The correlations from H-26 and H-28 to C-27 constructed the spiroketal system. A 26-membered macrocyclic lactone ring was suggested in accordance with HMBC correlations from H-39 to C-11, C-12 and C-13, and from H-25 and H-3 to the ester moiety at C-1 (δC 164.7). The presence of this macrocyclic lactone was confirmed by the degree of unsaturation of YO-001A (seven degrees of unsaturation: three double bonds, one carbonyl, and three ring systems). Four of the six hydroxyl groups (5-OH, 7-OH, 9-OH, and 11-OH) were assigned via 2D NMR correlation spectroscopy; the remaining two were inferred via 13C DEPT spectroscopy and chemical shift assignment. Finally, YO-001A was identified as a new 26-membered macrolide.

Fig. 1
figure 1

a Structure of YO-001A. b Key 2D NMR correlations of YO-001A. c Structure of maclafungin and oligomycin A

Full size image
Table 1 1H and 13C NMR spectra data of YO-001A in CDCl3
Full size table

The relative configuration of 6,6-spiroacetal core structure (C-23 to C-32) was analyzed on the basis of ROESY correlations (Figs. S9 and S10). ROESY correlations between H-23/H-25, H-24/H-26, H-28/30-Me, H-29/H-31, and H-31/H-23 were observed. These ROE correlation patterns were similar to those of other oligomycins and neomaclafungins [11, 12], indicating the relative configuration of spiroacetal moiety in YO-001A as shown in Fig. S10.

Besides, coupling constants (14.5–15.5 Hz) of olefinic protons, the ROE correlations between H-2/H-4, H-16/H-18, and H-18/H-20, and strong IR absorption at 985 cm−1 suggested that all three double bonds were in the trans configuration; subsequent stereochemical investigations are currently underway.

Antimicrobial and cytotoxic effects of YO-001A were evaluated (Table 2). YO-001A inhibited the growth of P. oryzae (IC50 = 0.012 μM), A. fumigatus (IC50 = 0.42 μM), and C. albicans (IC50 = 1.6 µM), displaying slight cytotoxicity to HeLa cells (IC50 = 8.2 μM) and HL-60 cells (IC50 = 5.8 µM). YO-001A was especially effective against filamentous fungi and the morphology induced by YO-001A was similar to that induced by oligomycin A. Based on the structural and functional similarities, YO-001A was considered to have FOF1-ATPase inhibitory activity. To test whether YO-001A interfered with mitochondrial respiration, a cell-based assay monitoring the Oxygen Consumption Rate (OCR) was performed using the Seahorse XFe96 Analyzer [13]. Our data showed that YO-001A markedly decreased the OCR in HeLa cells, indicating its inhibitory effects on mitochondrial respiration (Fig. S11, IC50 = 0.0025 µM). Furthermore, on comparing the pattern of changes in OCR values via the Mito stress test, wherein oligomycin A, FCCP, and rotenone/antimycin A were added in cell culture media in a stepwise manner, we confirmed that its pattern of OCR variation is identical to that of oligomycin A (Fig. S11). Consequently, we examined its effect in vitro on FOF1-ATPase activity, using isolated bovine heart mitochondria and observed potent and concentration-dependent inhibition of FOF1-ATPase (IC50 = 1 μM), similar to oligomycin A (Fig. S12). These results indicated that YO-001A inhibits FOF1-ATPase in cell-based and cell-free systems.

Table 2 Biological activity (IC50: μM) of YO-001A
Full size table

In this study, we report the new antifungal agent, YO-001A, as a new congener of 26-membered macrocyclic polyketides. Although YO-001A is similar to oligomycins and maclafungins, it is potentially interesting since there are no functionalities on C-13 and C-24 in its structure. Namely, C-13 in all reported 26-membered macrolides are hydroxylated, while that of YO-001A is not. OlmA5, a modular PKS synthesizing oligomycin, generates the C11–C13 carbon skeleton, and consists of acyltransferase, ketosynthase, β-ketoreductase, and inactive dehydratase (DH) domains [14], resulting in the hydroxyl group at C-13. Hence, we speculated that the corresponding module in Streptomyces sp. YO15-A001 may differ from that of oligomycin, containing active DH and additional enoylreductase in its biosynthetic cluster. We intend to further identify and analyze the functions of YO-001A biosynthetic genes.

Regarding the structure–activity relationships of oligomycins, Omelchuk et al. have semisynthesized 7,11-tetrahydrooligomycin A and revealed the importance of ketones in the 26-memberd ring [15]. So far, C7-C13 polyol region has been studied poorly in light of SAR partly due to the difficulty in site-selective chemical modification. Thus, when comparing with semisynthetic oligomycins and neomaclafungins, YO-001A might be a useful tool to study the SAR of this class of compounds.