The microbial natural products derived from terrestrial and marine environment have played an important role in the development of various pharmaceuticals, including antibiotics and anticancer agents, recognized as a crucial source for the discovery of new drug candidates.^1–3 Especially, marine ecosystems, occupying 70% of the Earth's surface, are rich in biological diversity and are considered unexplored reservoirs of microorganisms.⁴ Marine microorganisms, adapting to extreme environmental conditions such as high-salinity, low temperature, high pressure, various nutrient availability, have been revolutionized their metabolic pathways and enzymatic systems compared to terrestrial counterparts, possessing the potential to produce structurally and biologically novel secondary metabolites.^5,6

As part of our ongoing efforts to discover novel bioactive compounds from marine microorganisms, we mainly focused on actinomycetes collected from unexplored marine environments such as volcanic islands and mudflats. However, conventional screening for bioactive secondary metabolites of actinomycetes often leads to the frequent rediscovery of known compounds, limiting the identification of novel compounds.⁷ Genomic analyses of actinomycetes have shown that each strain's biosynthetic potential far exceeds the number of biosynthetic pathways expressed under laboratory cultivation conditions.⁸ Accordingly, the One Strain Many Compounds (OSMAC) strategy, developed to broaden microbial chemical diversity, proposes the systematic variation of cultivation conditions can activate silent biosynthetic pathways.⁹ In our previous reports, Streptomyces sp. SNM55, isolated from an intertidal mudflats in Mohang, Korea, was reported to produce the structurally-novel pseudodimeric peptides mohangamides A and B.¹⁰ Further examination of its secondary metabolites under varied culture conditions led to the discovery of several additional novel bioactive compounds.^11–13

To extend the application of OSMAC strategy on Streptomyces sp. SNJ210, which was previously reported as producer for separacenes A–D,¹⁴ the secondary metabolites of strain SNJ210 were further profiled by changing the culture medium and time. During a prolonged cultivation in different culture media for 6 days, which is three times longer than the cultivation time required for separacenes A–D production, LC/MS-based chemical analysis indicated that the strain SNJ210 produced only a polyunsaturated compound with a terminal carboxyl group. In this study, we report the isolation, structural determination, and biological activity of a new separacenes derivative, separacene E (1).

General experimental procedures – UV spectra were measured with a Perkin Elmer Lambda 35 UV/VIS spectrometer (Perkin Elmer, Waltham, MA, USA). Optical rotation was obtained with a Jasco P-1020 polarimeter with a 1-cm cell (JASCO, Easton, PA, USA). IR spectra were acquired using a Thermo N1COLET iS10 spectrometer (Thermo, Madison, CT, USA). Electrospray ionization (ESI) low-resolution LC/MS data were recorded on an Agilent Technologies 6130 Quadrupole mass spectrometer (Agilent Technologies, Santa Clara, CA, USA) coupled with an Agilent Technologies 1200 series HPLC using a reversed-phase C₁₈ column (Phenomenex Luna, 100 × 4.6 mm). High-resolution fast-atom bombardment (HR-FAB) mass spectra were obtained with a Jeol JMS-600W high resolution mass spectrometer (Jeol, München, Germany), at NCIRF (National Center for Inter-University Research Facilities). Purified separacene E was acquired by using Gilson Semi Prep HPLC (Gilson, Middleton, WI, USA) coupled with a reversed-phase column (Kromasil C₁₈ (2): 250 × 10 mm, 5 μm). ¹H and 2D NMR spectra were collected with a Bruker Avance 600 MHz spectrometer (Bruker, Billerica, MA, USA) at NCIRF. All solvents required for these experiments were obtained from Honeywell International Inc. (Honeywell, Charlotte, North Carolina, USA).

Bacterial isolation, cultivation, and extraction – The procedure for bacterial isolation of Streptomyces sp. SNJ210 was previously reported.¹⁴ Based on OSMAC strategy, the strain was cultivated in 50 mL YEME media (4 g yeast extract, 10 g malt extract, and 4 g glucose in 1 L artificial seawater) in a 125 mL Erlenmeyer flask. After cultivation for 3 days on a rotary shaker at 200 rpm at 30 °C, 10 mL samples of the culture were inoculated in 1 L A1+C media (10 g starch, 4 g yeast extract, 2 g peptone and 1 g calcium carbonate in 1 L artificial seawater) in 2.8 L Fernbach flasks (12 ea × 1 L, total volume 12 L). The large culture was incubated under the same conditions used for the seed culture. After 6 days, the whole culture (12 L) was extracted twice with 18 L ethyl acetate. The ethyl acetate layer was separated and dried over anhydrous sodium sulfate. The ethyl acetate extract was concentrated in vacuo to yield 1.1 g of dried material. This procedure was repeated 4 times (48 L culture, the total amount of extract: 4.4 g) to obtain sufficient quantities for structure elucidation and bioassays. All reagents used for microbial cultivation were obtained from BD Difco (BD Difco, Franklin Lakes, New Jersey, USA).

Isolation of separacene E – The crude extract (4.4 g) was absorbed onto celite and then loaded onto a 2 g Sep-Pak C₁₈ cartridge. Celite mixture was subsequently fractionated using 20 mL each of 20%, 40%, 60%, 80%, and 100% MeOH/H₂O and 1:1 MeOH/dichloromethane. Separacene E (1) was identified in the 60% and 80% MeOH/H₂O fractions. Several chromatography steps were utilized to further purify compounds 1. Initially, the fraction containing compounds 1 were purified using reversed-phase HPLC [Kromasil C₁₈ (2): 250 × 10 mm, 5 μm] under isocratic conditions with a 6:2 acetonitrile/H₂O solvent system (UV 280 nm detection, flow rate: 2 mL/min). A major peak was collected at a retention time of 28 min. The collected fraction was further purified using a reversed-phase HPLC [Kromasil C₁₈ (2): 250 × 10 mm, 5 μm] with a gradient solvent system, starting from 65% MeOH/H₂O to 90% MeOH/H₂O over 40 min (UV 280 nm detection, flow rate: 2 mL/min). As a result, pure compound was isolated: Separacenes E (1) (4.2 mg) at 33 min.

Separacene E (1) – White powder; αD20 = –4 (c 0.05, MeOH); UV (MeOH) λ_max (log ε) 302 (4.27) nm; ¹H- and ¹³C-NMR, see Table 1; HR-ESI-MS m/z 223.1342 [M+H]⁺, calcd for C₁₃H₁₉O₃ at m/z 223.1340.

Determination of the absolute configuration of 10S- and 10R-separacene E by specific rotation calculations – Possible enantiomers of separacene E were generated based on NMR data using Avogadro 1.2.0, and their geometries were optimized in the ground-state based on density functional theory (DFT) calculation using Turbomole X 4.3.2. The B3LYP/DFT level and def-SVP basis set were used for the specific rotation calculations. The ground-state geometries were optimized with density functional theory (DFT) calculations, using Turbomole X 4.3.2. The B3LYP/DFT level and def-SVP basis set were used for the calculations.

Antibacterial activity assay – Pathogenic bacterial culture and sample preparation were performed as previously described.¹⁴ The minimum inhibitory concentration (MIC) values were determined as the lowest concentration of test compound that inhibited bacterial growth. Ampicillin was used as a reference compound.

Antifungal activity assay – Pathogenic fungal culture and sample preparation were performed as previously described.¹⁴ The minimum inhibitory concentration (MIC) values were determined as the lowest concentration of test compound that inhibited bacterial growth. Ampicillin was used as a reference compound.

Anti-proliferative activity assay – Cancer cell lines and cell proliferation were performed as previously described.¹⁴ The effects of separacene E on cell viability were calculated as percentages relative to the solvent-treated control. The IC₅₀ values were calculated using nonlinear regression analysis (percent survival versus concentration).

Separacene E (1) was isolated as a white powder, with its molecular formula identified as C₁₃H₁₈O₃ based on HR-ESI-MS data (obsd [M+H]⁺ at m/z 223.1342, calcd 223.1340). Analysis of the ¹H and HSQC NMR data (Table 1) of 1 in pyridine-d₅ revealed seven olefinic signals (δ_C/δ_H: 123.1/6.87, 129.0/6.44, 140.1/6.41, 140.1/6.30, 133.3/6.18, 133.5/5.80, 131.5/5.76), indicative of a highly unsaturated polyene system. Further analysis identified one carboxyl carbon (δ_C: 180.0), one non-protonated double-bond carbon (δ_C: 130.8), one methylene group (δ_C/δ_H: 43.9/3.13 and 2.90), and two methyl groups (δ_C/δ_H: 18.8/1.73, 12.6/1.78). The presence of eight olefinic carbons and seven olefinic protons suggests that separacene E (1) features a conjugated tetraene moiety as a new derivative of previously reported separacenes A–D (Fig. 1).

Fig. 1. 
The structures of 1 and separacenes A–D isolated from Streptomyces sp. SNJ210.

Analysis of the HSQC spectrum assigned all one-bond-proton-carbon correlations, followed by ¹H-¹H COSY and HMBC NMR to define the planar structure. In the COSY spectrum, doublet methyl proton H₃-1 (δ_H: 1.73) showed correlation with H-2 (δ_H: 5.76). COSY correlations between H-2/H-3 (δ_H: 6.44) and H-3/H-4 (δ_H: 6.18) confirm the C-2–C-3 bond and C-3–C-4 double bond. In the HMBC spectrum, singlet methyl proton H₃-13 (δ_H: 1.78) exhibited correlations to the olefinic proton H-4 and H-6 (δ_H: 6.41), as well as to the quaternary carbon C-5 (δ_C: 130.8), thereby unambiguously confirming that the methyl substituent (C-13) is attached at C-5. Further COSY analysis revealed sequential correlations between H-6–H-7 (δ_H: 6.87), H-7–H-8 (δ_H: 6.30), and H-8–H-9 (δ_H: 5.80), which defines the olefinic linkage from C-6 through C-9. A COSY correlation between H-9 and oxygen-bound methine proton H-10 (δ_H: 5.62) established the C-9–C-10 bond, with C-10 bearing a hydroxyl substituent. Additionally, COSY correlations of H-10 with H₂-11 confirmed the bond between C-10 and C-11. Finally, HMBC correlation between H₂-11 and C-12 (δ_C 180.0), along with the five degrees of unsaturation of separacene E (1), indicates the presence of a terminal carboxyl group. Consequently, the complete planar structure of separacene E (1) was determined to be a conjugated tetraene polyketide bearing a terminal carboxylic acid group. The double bond geometries of separacene E (1) were assigned by analysis of vicinal coupling constants. Only the large coupling constant observed for H-6/H-7 (J = 15.0 Hz) indicates an E configuration at C-6–C-7, resulting in the assignment of 2Z, 4Z, 6E, and 8Z (Fig. 2).

Fig. 2. 
Key 2D NMR correlations of 1.

To determine the absolute configuration of 1, we applied the specific rotation calculations. Energy-minimized conformers of the 10S and 10R enantiomers of 1 were generated (Table S3 and S4), and their specific rotations were computed by time-dependent density functional theory (TD-DFT) at the B3LYP/def-SVP level. The specific rotation values of 10S- and 10R-separacene E were calculated as -3.126 and 3.126, respectively. The experimental specific rotation value of 1 at -4 is consistent with the calculated value of 10S conformer, thereby determining the absolute configuration at C-10 as S (Fig. 3 and Table S1 and Table S2).

Fig. 3. 
Determination of the absolute configuration of 1 via specific rotation calculations.

We evaluated the biological activities of 1 followed by the previous report. The antimicrobial activity of the 1 was evaluated against diverse pathogenic bacterial strains such as Staphylococcus aureus ATCC 6538p, Bacillus subtilis ATCC 6633, Kocuria rhizophila NBRC 12708, Salmonella enterica ATCC 14028, Proteus hauseri NBRC 3851, and Escherichia coli ATCC 35270. Ampicillin was used as a positive control. Separacene E (1) did not display significant inhibitory activity against the tested bacteria whereas separacene A was reported to have weak antibacterial activity against B. subtilis ATCC 6633 and P. hauseri NBRC 3851, with MIC values of 50 µg/mL and 100 µg/mL, respectively. Furthermore, separacene E did not exhibit any remarkable activity in antifungal assays against Candida albicans ATCC 10231, Aspergillus fumigatus HIC 6094, Trichophyton rubrum NBRC 9185, and T. mentagrophytes IFM 40996. For the last, we evaluated 1 for anti-proliferative activity assay and separacene E (1) displayed weaker cytotoxicity against the colon cell line HCT116 with IC₅₀ values of 31.5 µg/mL rather than previously reported value of separacene A (IC₅₀ = 14.0 µg/mL). These results suggest that the absence of terminal methylene group coupled with hydroxyl groups in 1 is responsible for its attenuated antimicrobial and cytotoxic activities. Further structure-activity relationship study (SAR) is required to demonstrate biological potency of separacenes A–E.

To our best knowledge, type I polyketide synthase (PKS) is responsible for biosynthesis of separacenes A–E. Separacene E (1) contains a 13-membered carbon chain, whereas separacenes A–D bear 15-membered carbon backbones. The number of carbon chain differences suggests that the missing of one acetyl-CoA building block during polyketide biosynthesis resulted in the loss of two carbons. In our hypothesis, the terminal carboxylic acid of 1 is further reduced by a ketoreductase and enoyl-reductase to install a methylene moiety at C-15 found in separacenes A–D. To prove our hypothetical biosynthesis of separacenes A–E, a further genome mining study combined with a full genome sequence will be performed.

In summary, OSMAC strategy applied to marine-derived Streptomyces sp. SNJ210 led to the discovery of a new tetraene mono-ol compound, separacene E, a new derivative of separacenes A–D. Our findings demonstrate that the OSMAC approach is effective not only for yielding known compounds but also for uncovering novel derivatives beyond those previously reported. Moving forward, we plan to extend the OSMAC strategy combined with cutting-edge technologies such as genome mining, co-cultivation, and Hi-TES to overcome the bottleneck for the discovery of novel microbial secondary metabolites as drug candidates.