Saprolegnia parasitica is an oomycete pathogen, a virulent water mold infecting fishes such as salmon and trout, the resultant saprolegniasis causing tremendous damage and losses in commercial aquaculture. Previously, malachite green was used to control saprolegniasis but its use has been widely banned as eating fish contaminated with malachite green poses a significant risk due to its reported carcinogenicity. Therefore, novel anti-saprolegniasis compounds are urgently needed [1, 2]. We focused on deep-sea fungi, which were known as producers of new source of novel useful compounds such as spiromastols A–K [3], emerixanthones A–D [4] and aspiketolactonol [5]. In this paper, we attempt to isolate anti-saprolegniasis compounds from deep-sea fungi.

Our screening for anti-saprolegniasis agents from 546 cultured broths of 91 deep-sea fungal strains has already resulted in the discovery of cladomarine produced by Penicillium coralligerum YK-247 [1]. In this paper, we additionally report a new compound, named quellenin (1) (Fig. 1a), from the cultured broth of a deep-sea fungus, Aspergillus sp. YK-76, which we isolated together with three known compounds, diorcinol (2) [6], violaceol-I (3) [7] and violaceol-II (4) [7] (Fig. 1c).

Fig. 1
figure 1

a Structure of quellenin (1). b COSY and HMBC correlation of quellenin (1). Bold lines show 1H–1H correlation, arrows show 1H–13C correlation. c Structure of diorcinol (2), violaceol-I (3) and violaceol-II (4)

Full size image

Strain YK-76 was isolated from an Osedax sp. annelid, commonly called bone-eating worm, collected at the São Paulo Ridge in off Brazil (water depth: 4203 m, sampling site: 28°31.1′S and 41°39.4′W), using the human-occupied vehicle Shinkai 6500 of the Japan Agency for Marine-Earth Science and Technology (JAMSTEC), as part of a Japan-Brazil cooperative deep-sea investigation named Iatá-Piúna cruise [8].

The internal transcribed spacer of ribosomal RNA gene (ITS-rDNA) sequence of YK-76 was compared to sequences in the GenBank database by BLASTN 2.7.1 analysis [9]. The sequence of YK-76 was 99.6% similar to the sequence of NRRL 58600 (ex-type of Aspergillus jensenii, GenBank accession number NR_135444). The producing strain YK-76 was assigned to the genus Aspergillus based on its morphology and sequence analysis.

A loop of spores from a colony growing on a Miura’s medium (LcA) slant was inoculated into 10 ml of GP seed medium (2.0% glucose, 0.2% yeast extract, 0.5% MgSO4 7H2O, 0.5% Polypeptone, 0.1% KH2PO4 and 0.1% agar, pH 6.0) in a test tube (total 20 ml) and incubated at 27 °C for 3 days on a shaker at 300 rpm. The seed-culture (20 ml) was inoculated into 500 g of rice medium in a single culture bag (Ulpak 47, Hokken Co. Ltd., Tochigi, Japan) and static fermentation was carried out at room temperature (about 25 °C) for 13 days. The cultured rice medium (500 g) was subsequently added to 500 ml of ethanol and then filtered. The filtrate was concentrated under reduced pressure to remove ethanol and extracted with EtOAc (500 ml) three times. The EtOAc extract (1.91 g) was applied to an octadecylsilane (ODS) gel column (25φ × 100 mm, YMC Co. Ltd., Kyoto, Japan), eluted stepwise with MeOH-H2O (0, 30, 50, 60, 70, 80, 90 and 100%) and fractionated. The 50% MeOH aq. fraction (343.5 mg) was separated by MPLC (silica gel column, 25φ × 100 mm, Shoko Scientific Co., Ltd., Kanagawa, Japan) with CHCl3-MeOH (100/0, 100/1, 100/3, 100/5, 9/1, 1/1 and 0/100). Fractions 31-42 eluted with CHCl3-MeOH (100/3-100/5) were combined and purified by HPLC (YMC Actus Triart C18 (20φ × 250 mm) YMC Co. Ltd., Kyoto, Japan) with 25% acetonitrile aq. with a flow rate of 14 ml/min, using detection by UV 210 nm to obtain 1 (4.7 mg, Rt = 15 min), 3 (20.4 mg, Rt = 33 min) and 4 (27.4 mg, Rt = 40 min). The 60% MeOH aq. fraction (321.6 mg) was separated by silica gel column (30φ × 100 mm, Merck KGaA, Darmstadt, Germany) with CHCl3–MeOH (100/0, 100/1, 100/3, 100/5, 9/1, 1/1 and 0/100). The 100/5 fraction (21.2 mg) was purified by HPLC (Develosil C30-UG-5 (20φ × 250 mm) Nomura Chemical Co. Ltd., Aichi, Japan) with 30% acetonitrile aq. with a flow rate of 14 ml/min, using detection by UV 210 nm to obtain 2 (4.7 mg, Rt = 80 min).

The molecular formulae of 2, 3 and 4 were elucidated by HR-ESI-MS to be C14H14O3, C14H14O5 and C14H14O5, respectively. The comparison and analysis of 1H, 13C and 2D NMR spectra data with the literature led to the identification of 2, 3 and 4 as diorcinol [6], violaceol-I [7] and violaceol-II [7], respectively.

Physico-Chemical properties of 1 are shown in Table S2. Compound 1 was obtained as a white solid (UV (MeOH) λmax nm (ε): 205 (30,129), 221 (38,550), 280 (6751) and 290 (5,190)). The molecular formula of 1 was elucidated as C21H21O3N3 by high-resolution electron spray ionization mass spectrometry (m/z 362.1501 [M–H]− (calcd. for C21H20O3N3, 362.1505)) with thirteen degrees of unsaturation. The IR characteristic absorptions of 1 at 3432, 1666 and 1639 cm−1 suggested the presence of hydroxyl and carbonyl groups. The 1H and 13C NMR spectra of 1 in CD3OD are summarized in Table 1. Analyses of 1H, 13C NMR and HMQC spectra indicated the presence of one nitrogenated methyl group, two sp3 methylene groups, two nitrogenated sp3 methine carbons and 14 sp2 aromatic carbons and two carbonyl carbons. The 1H-1H COSY analysis revealed two partial structures, from H-2′ to H2-3′ and from H-2 to H2-3 (Fig. 1b). The HMBC correlations from H-9 to C-7 and C-5a, from H-8 to C-6, from H-7 to C-9 and C-5a, from H-6 to C-9a, C-8 and C-5, from H-4 to C-9a, C-5a and C-5 and from H2-3 to C-5a, C-5, C-2 and C-1 suggested the presence of a tryptophan (Trp) residue, including one partial structure. Moreover, the HMBC correlations from H-6′ to C-4′, from H-5′ to C-7′, C-5′ and C-3′, from H2-3′ to C-5′, C-4′, C-2′ and C-1′ and H-2′ to C-8 revealed 1 has a N-methyl-tyrosine (N-methyl-Tyr) moiety. Finally, the HMBC correlations from H3-8′ to C-1 suggested 1 should be cyclo (Trp-N-methyl-Tyr). So far, cyclo (D-Trp-D-N-methyl-Tyr) has only been reported as a synthetic compound with possible four stereoisomers [10]. Because 1H and 13C NMR data in acetonitrle-d3 for 1 was identical with those of the synthetic one described in the literature, the relative configuration of 1 was the same as the synthetic compound. However, the specific rotation ([α]D25(c = 0.1, acetonitrle) −99.62) of 1 was opposite to that of the synthetic compound ([α]D25 + 157.1), suggesting that 1 is cyclo (L-Trp-L-N-methyl-Tyr), an enantiomer of the synthetic product.

Table 1 1H and 13C NMR data for (1) (recorded at 400MHz 1H NMR and 100MHz 13C NMR in CD3OD; ∂ in ppm)
Full size table

The absolute configuration of 1 was confirmed by the Advanced Marfey’s method after acid hydrolysis [11]. Compound 1 was hydrolyzed and derivatized with L-FDLA and analyzed by an UPLC-ESI-MS. As a result of the comparison of retention time with L-FDLA derivatives of standard, Trp was assigned to be the l configuration (Table S3, Fig. S3). The results described above confirmed that 1 is a new analog of diketopiperazine, which we named quellenin.

Antimicrobial activity against bacteria, filamentous fungus and yeast, as well as oomycetes, were measured using the Paper disc method, as described previously [1, 12]. Test organisms were as follows; two oomycetes (Saprolegnia parasitica kassi1, Pythium sp. sakari1), two bacteria (Bacillus subtilis ATCC 6633, Pseudomonas aeruginosa NBRC 12582), one filamentous fungus (Aspergillus niger ATCC 6275) and one yeast (Candida albicans ATCC 64548). Compounds 1, 2, 3 and 4 were evaluated at five doses (100, 30, 10, 3 and 1 µg). Compound 1 was also evaluated at a higher dosage of 200 ug.

Compounds 2, 3 and 4 showed anti-Saprolegnia activity, with inhibition zones of 13.7–17.5 mm at 30 µg/disc (Table 2). Although they also showed anti-Bacillus activity, they showed anti-Saprolegnia activity with better selectivity than malachite green. Compound 1 displayed weaker activity against S. parasitica, with an inhibition zone of 19.9 mm at 200 µg/disc, but showed better selectivity against S. parasitica (without anti-Bacillus activity) compared to 2, 3 and 4. We determined the minimum inhibitory concentrations (MICs) of 2, 3 and 4, according to a previously reported method [1], finding MICs of 8, 32 and 16, respectively.

Table 2 Antimicrobial activity of quellenin (1), diorcinol (2), violaceol-I (3) and violaceol-II (4)
Full size table

Compounds 3 and 4 have been reported to be actin inhibitors, inducing cell shape elongation in fibroblast cells and impairment of mitochondrial function causing uncoupling of oxdative phosphorylation [13, 14]. Therefore, the bioactivity of 3 and 4 may be associated with these functions. Moreover, 3 and 4 have been identified as anti-malarial agents [15], cladosporin, previously reported as an anti-saprolegniasis lead compound by our group [1], also displays anti-malarial properties. Thus, other anti-malarial agents might be a good source for anti-Saprolegnia screening. With regards to 1, other compounds containing a diketopiperazine moiety, such as cyclo(L-Pro-L-Leu) and cyclo-(D-Pro-L-Tyr) [16], have been reported to exert many biological activities. The enatiomer of 1 was known as a synthetic intermediate, but hasn’t been reported in any bioassays. Further structural modifications of 1 or isolation of other structural analogs may be valuable in the development of new anti-parasitic agents.

In conclusion, three known 2–4 and one new compound 1 were isolated from a cultured broth of deep-sea fungus Aspergillus sp. YK-76. Compounds 2, 3 and 4 showed inhibition of the growth of S. parasitica, together with several other microbes. Our results suggest that a diorcinol analog could be a good lead candidate in the development of novel agents to help prevent infection with Saprolegnia. Moreover, fungi from deep-sea locations represent a highly-promising new source for the possible discovery of novel bioactive lead compounds.