Actinofuranones D-I from a Lichen-Associated Actinomycetes, Streptomyces gramineus, and Their Anti-Inflammatory Effects

Six new metabolites, actinofuranones D-I (compounds 1–6), were isolated together with three known compounds—JBIR-108 (7), E-975 (8), and E-492 (9)—from a fermentation broth of Streptomyces gramineus derived from the lichen Leptogium trichophorum. The structures of the new compounds 1–6 were established using comprehensive NMR spectroscopic data analysis, as well as UV, IR, and MS data. The anti-inflammatory activity of these isolated compounds were evaluated by examining their ability to inhibit nitric oxide (NO) production in LPS-stimulated RAW 264.7 macrophage cells. Compounds 4, 5, 8, and 9 attenuated the production of NO due to the suppression of the expression of nitric oxide synthase (iNOS) in LPS-induced RAW 264.7 cells. Moreover, 4, 5, 8, and 9 also inhibited LPS-induced release of proinflammatory cytokines interleukin-6 (IL-6) and tumor necrosis factor α (TNF-α).


Introduction
Lichens are symbiotic organisms, composed of fungi (mycobiont) and algae or cyanobacteria (photobiont or phycobiont). Like mycorrhizal fungi and fungal endophytes, they frequently host diverse bacterial communities as symbiotic niches [1]. The lichen-forming fungi and endolichenic fungi inhabiting the lichen thalli had been widely investigated as a source for discovering new bioactive natural products [2,3]. Recently, with discovery of some new secondary metabolites from them, lichen-associated actinomycetes have attracted extensive attention [4][5][6][7][8].
In our previous work, the diversity of cultivable actinomycetes associated with lichen symbiosis in samples collected from Yunnan Province, P. R. China were investigated. A total of 213 actinomycetes strains were isolated from 35 lichen samples and 16S rRNA gene sequence analysis of the isolates exhibited a high level of diversity among these strains [9]. Antimicrobical activities and biosynthetic potential studies of the isolated actinomycetes were also conducted in the research and the results showed the actinomycetes associated with lichen could be considered as a potential microbial resource for discovering novel bioactive natural products [9]. As a subsequent work, Streptomyces gramineus (YIM 130461), occurring in the lichen Leptogium trichophorum, collected from an evergreen broad-leaf forest in Benzilan, Diqing (Yunnan Province, China) was chosen to investigated its secondary metabolites. A small scale (100 mL) fermentation broth of S. gramineus showed diverse investigated its secondary metabolites. A small scale (100 mL) fermentation broth of S. gramineus showed diverse chemical constituents by HPTLC and HPLC-MS analysis. Chromatographic separation of an EtOAc extract of a large scale fermentation broth of S. gramineus led to the isolation of six new 2-hydroxy-2-(1-hydroxyethyl)-2,3-dihydro-3(2H)-furanones, named actinofuranones D-I (compounds 1-6), as well as three known compounds: JBIR-108 (7) [10], E-975 (8) [11], and E-492 (9) [11].
The structures of compounds 1-9 were elucidated using comprehensive NMR spectroscopic data analysis. Meanwhile, as part of our continuous search for the new anti-inflammatory compounds from lichen-associated actinomycetes, we investigated the anti-inflammatory activity of the isolates by examining their ability to inhibit production of nitric oxide (NO), interleukin-6 (IL-6) and tumor necrosis factor α (TNF-α), and the expression of inducible nitric oxide synthase (iNOS) in LPS-stimulated RAW 264.7 macrophage cells. Herein, details of the isolation, structure elucidation, anti-inflammatory effects of these derivatives are reported.

Structural Elucidation of Actinofuranones D-I
A large-scale (70 L) fermentation broth of S. gramineus was centrifuged and the supernatant was extracted with ethyl acetate. Then, the dried extract was fractionated by sequential chromatography over Sephadex LH-20, silica gel, and octadecylsilyl-bonded silica (ODS) to yield nine actinofuranones 1-9 (Figure 1), including the six new actinofuranone analogues 1-6. All of them were isolated as a mixture of C-1 and C-2 diastereomers. Here the major components were employed for the structural establishment.  Compound 1 was isolated as a colorless amorphous substance. Its molecular formula of C 24 13 C NMR data of 1 showed typical characteristic of actinofuranone analogues and were very similar to those of 1S*-JBIR-108 (7) [10], which was also obtained from the same strain S. gramineus as an isomer mixture (major 1S*-isomer). Herein, we identified the structure of 1 through comparing NMR data in DMSO-d 6 and methanol-d 4 with those in the literature (methanol-d 4 ). Compared with NMR data between major isomers of 1 and 1S*-7, the only difference was an extra hydroxyl located at the unsaturated alkyl chain in 1. COSY  The identical 13 C NMR data (both in DMSO-d 6 and MeOH-d 4 ) of C-1 to C-15 and C-21 to C-24 of the major isomer in 1 with 1S*-7 revealed that it possessed the same relative configuration at C-1, C-7, C-14, and C-15 as 1S*-7. The coupling constant 3 J H-14/H-15 = 8.5 Hz (in MeOH-d 4 ) further confirmed the configuration of H-14 and H-15. Therefore, the structure of the major isomer in 1 was elucidated as 1S*-actinofuranone D as shown in Figure 1. By the same deduction above, the minor isomer in 1 was identified as 1R*-actinofuranone D.  13 C NMR data of 1 showed typical characteristic of actinofuranone analogues and were very similar to those of 1S*-JBIR-108 (7) [10], which was also obtained from the same strain S. gramineus as an isomer mixture (major 1S*-isomer). Herein, we identified the structure of 1 through comparing NMR data in DMSO-d6 and methanol-d4 with those in the literature (methanol-d4). Compared with NMR data between major isomers of 1 and 1S*-7, the only difference was an extra hydroxyl located at the unsaturated alkyl chain in 1.  JBIR-108 was firstly isolated from S. gramineus IR087Pi-4 as a mixture of two diastereomers at C-1 and its planar structure and absolute configuration were determined by spectral analysis and total synthesis (1S*-/1R*-isomer, 5:2) [10]. The NMR data of JBIR-108 was measured in MeOH-d4 without presenting any active protons signals in the literature. They could not get the relative content of all isomers from the characteristic protons since most 1 H-NMR signals of isomers were same. However, hydroxyl groups of isomers, especially the hydroxyl of hemiacetal group, presented individual 1 H-NMR signals in the DMSO-d6 ( Figure 3). Since the hemiacetal formation was mainly influenced by the steric hindrance of C-1 substitute, the ratio of epimers should be same (1Sα/1Sβ = 1Rβ/1Rα). Thus, we could deduce that 7 is a 1.4:1 mixture of diastereomers at C-1, actually. Even if it exhibited a 3:1 mixture of C-2 epimers for each C-1 diastereomer (R-or S-isomers), we could not give an absolute content for those 1Sβ, 1Sα, 1Rα or 1Rβ isomers ( Figure 3). Approximately the same ratio of isomer was also deduced from the individual 1 H-NMR signals of C-1 and C-7 hydroxyls in Figure 1. JBIR-108 was firstly isolated from S. gramineus IR087Pi-4 as a mixture of two diastereomers at C-1 and its planar structure and absolute configuration were determined by spectral analysis and total synthesis (1S*-/1R*-isomer, 5:2) [10]. The NMR data of JBIR-108 was measured in MeOH-d 4 without presenting any active protons signals in the literature. They could not get the relative content of all isomers from the characteristic protons since most 1 H-NMR signals of isomers were same. However, hydroxyl groups of isomers, especially the hydroxyl of hemiacetal group, presented individual 1 H-NMR signals in the DMSO-d 6 ( Figure 3). Since the hemiacetal formation was mainly influenced by the steric hindrance of C-1 substitute, the ratio of epimers should be same (1Sα/1Sβ = 1Rβ/1Rα). Thus, we could deduce that 7 is a 1.4:1 mixture of diastereomers at C-1, actually. Even if it exhibited a 3:1 mixture of C-2 epimers for each C-1 diastereomer (R-or S-isomers), we could not give an absolute content for those 1Sβ, 1Sα, 1Rα or 1Rβ isomers ( Figure 3). Approximately the same ratio of isomer was also deduced from the individual 1 H-NMR signals of C-1 and C-7 hydroxyls in Figure 1.  Therefore, the structure of 2 was identified as that in Figure 1 and named as actinofuranone E.
Compound 3 was obtained as a colorless amorphous powder. Its 13 C-NMR data and HRESIMS data (m/z 461.2654 [M + Na] + ) suggested that 3 has the same molecular formula as that of 1 and 2 (C24H38O7). The structure of 3 was almost identical to 1 by comparison with 1 H-and 13 C-NMR data ( Table 2 and Supplementary Material) of 1. However, 1 H-1 H COSY correlation between the terminal methyl protons H3-20 (δH 1.00) and oxygenated methine proton H-19 (δH 3.58) indicated that the hydroxyl group was located at C-19 in 3 instead of C-18 in 1 (Figure 2). The assignment of the hydroxyl was further confirmed by the clear HMBC correlations between H3-20 (δH 1.00) and C-19 (δC 66.6), C-18 (δC 37.8), between H-18 (δH 1.98) and C-16 (δC 138.2), C-17 (δC 123.1) ( Figure 2). The identical 13 C-NMR data of C-1 to C-15 and C-21 to C-24 in 3 with those of 1 deduced the relative configurations of chiral carbons in 3 except for C-19. Thus, the structure of 3 was elucidated and named as actinofuranone F.
As displayed by 13 C-NMR data and [M + Na] + ion at m/z 447.2363 in HRESIMS, compound 4 had the molecular formula of C23H36O7. The difference in mass compared to 1 owed to a missing CH2 group. Furthermore, the 1 H-NMR and 13 C-NMR data of 4 ( Table 2   Therefore, the structure of 2 was identified as that in Figure 1 and named as actinofuranone E. Compound 3 was obtained as a colorless amorphous powder. Its 13 C-NMR data and HRESIMS data (m/z 461.2654 [M + Na] + ) suggested that 3 has the same molecular formula as that of 1 and 2 (C 24 H 38 O 7 ). The structure of 3 was almost identical to 1 by comparison with 1 H-and 13 C-NMR data ( Table 2 and Supplementary Material) of 1. However, 1 H-1 H COSY correlation between the terminal methyl protons H 3 -20 (δ H 1.00) and oxygenated methine proton H-19 (δ H 3.58) indicated that the hydroxyl group was located at C-19 in 3 instead of C-18 in 1 (Figure 2). The assignment of the hydroxyl was further confirmed by the clear HMBC correlations between H 3 -20 (δ H 1.00) and C-19 (δ C 66.6), C-18 (δ C 37.8), between H-18 (δ H 1.98) and C-16 (δ C 138.2), C-17 (δ C 123.1) (Figure 2). The identical 13 C-NMR data of C-1 to C-15 and C-21 to C-24 in 3 with those of 1 deduced the relative configurations of chiral carbons in 3 except for C-19. Thus, the structure of 3 was elucidated and named as actinofuranone F.
As displayed by 13  (δ H 6.10), H-10 (δ H 5.54) and C-9 (δ C 69.2) indicated the migrated hydroxyl was located at C-9 ( Figure 2 13 C-NMR data with the total synthesis JBIR-108 (7) ( Table 3 and Supplementary Material) combining with the biosynthesis procedure [11], the configuration of 5 was determined except for C-9. The UV, IR, 1 H-NMR, and 13 C-NMR data indicated 6 was closely related to 4 and has the molecular formula of C 23 H 36 O 7 based on the ion peak at m/z 447.2337 [M+Na] + in its HRESIMS. Comparison of NMR data of 6 ( Table 3 and Supplementary Material) with those of 4 showed that the substituted position of C-18 hydroxyl in 4 changed. Moreover, one oxygenated quaternary carbon and a methylene appeared in 6 instead of two methines of 4. The singlet methyl at C-14 in 6 indicated the migrated hydroxyl located at C-14, which was confirmed by the HMBC correlations from H-22 to C-13, C-14, C-15, from H-13 to C-12, C-14 ( Figure 2). Thus, the structure of 6 was determined, and named actinofuranone I. The ratios of the major isomers in 1-4, and 6 could be approximately determined through analysis of 1 H-NMR data of 2-OH and 7-OH (Figure 1), even though the 1-OH of those isomers did not present an ideally individual 1 H-NMR signal.

Cell Viability and Effects of Compounds on the Production of NO in LPS-Induced RAW 264.7 Cells
Nitric oxide (NO) produced by activated RAW 264.7 macrophages plays an important role in inflammation diseases [22]. Excess production of NO have been reported to be involved in inflammatory disorders [23]. In the present study, the anti-inflammatory activity of isolated compounds were investigated by evaluating their effects on production of NO in LPS-induced RAW 264.7 cells. In order to exclude influence of the cytotoxic of compounds 1-9 on the anti-inflammatory evaluation, the viability of RAW 264.7 cells with test compounds treatment was carried out by an MTT assay. The result showed that cell growth inhibitory rate of compounds 1, 2, 6, and 7 were more than 40% at 60 µM and others had no obvious effect on cell viability at the test concentration ( Figure 4A). Hence, the compounds 3-5, 8, and 9 were selected for the anti-inflammatory activity evaluation process. As displayed in Figure 4B, compounds 4, 5, 8, and 9 significantly inhibited the production of LPS-induced NO in a dose-dependent manner at the varying concentration (0, 15, 30, and 60 µM). However, compound 3 have no inhibition effect on NO production in LPS-stimulated RAW 264.7 cells which were treated at the low concentration. HMBC interactions between H-11 (δH 6.10), H-10 (δH 5.54) and C-9 (δC 69.2) indicated the migrated hydroxyl was located at C-9 ( Figure 2 13 C-NMR data with the total synthesis JBIR-108 (7) ( Table 3 and Supplementary Material) combining with the biosynthesis procedure [11], the configuration of 5 was determined except for C-9. The UV, IR, 1 H-NMR, and 13 C-NMR data indicated 6 was closely related to 4 and has the molecular formula of C23H36O7 based on the ion peak at m/z 447.2337 [M+Na] + in its HRESIMS. Comparison of NMR data of 6 ( Table 3 and Supplementary Material) with those of 4 showed that the substituted position of C-18 hydroxyl in 4 changed. Moreover, one oxygenated quaternary carbon and a methylene appeared in 6 instead of two methines of 4. The singlet methyl at C-14 in 6 indicated the migrated hydroxyl located at C-14, which was confirmed by the HMBC correlations from H-22 to C-13, C-14, C-15, from H-13 to C-12, C-14 ( Figure 2). Thus, the structure of 6 was determined, and named actinofuranone I. The ratios of the major isomers in 1-4, and 6 could be approximately determined through analysis of 1 H-NMR data of 2-OH and 7-OH (Figure 1), even though the 1-OH of those isomers did not present an ideally individual 1 H-NMR signal.

Cell Viability and Effects of Compounds on the Production of NO in LPS-Induced RAW 264.7 Cells
Nitric oxide (NO) produced by activated RAW 264.7 macrophages plays an important role in inflammation diseases [22]. Excess production of NO have been reported to be involved in inflammatory disorders [23]. In the present study, the anti-inflammatory activity of isolated compounds were investigated by evaluating their effects on production of NO in LPS-induced RAW 264.7 cells. In order to exclude influence of the cytotoxic of compounds 1-9 on the anti-inflammatory evaluation, the viability of RAW 264.7 cells with test compounds treatment was carried out by an MTT assay. The result showed that cell growth inhibitory rate of compounds 1, 2, 6, and 7 were more than 40% at 60 μM and others had no obvious effect on cell viability at the test concentration ( Figure  4A). Hence, the compounds 3-5, 8, and 9 were selected for the anti-inflammatory activity evaluation process. As displayed in Figure 4B, compounds 4, 5, 8, and 9 significantly inhibited the production of LPS-induced NO in a dose-dependent manner at the varying concentration (0, 15, 30, and 60 μM). However, compound 3 have no inhibition effect on NO production in LPS-stimulated RAW 264.7 cells which were treated at the low concentration. . Cells were incubated with compounds or minocycline (MINO, 25 μM) for 2 h then stimulated with or without LPS (1 μg/mL) for 24 h. Cell viability was evaluated by MTT assay. The NO production in the medium was measured using Griess agent. All conditions were run in triplicate, and data show mean ± SD values. # , p < 0.05 was compared to control. *, p < 0.05 compared to LPS. . Cells were incubated with compounds or minocycline (MINO, 25 µM) for 2 h then stimulated with or without LPS (1 µg/mL) for 24 h. Cell viability was evaluated by MTT assay. The NO production in the medium was measured using Griess agent. All conditions were run in triplicate, and data show mean ± SD values. # , p < 0.05 was compared to control. *, p < 0.05 compared to LPS.

Compounds 4, 5, 8, and 9 Attenuated LPS-Induced iNOS Expression in RAW 264.7 Cells
Inhibition of NO overproduction through blocking inducible nitric synthase (iNOS) expression have been proved to be potential target of anti-inflammatory drug [24]. In the following, the inhibitory effect of compounds 4, 5, 8, and 9 on iNOS expression in LPS-stimulated RAW 264.7 cells were investigated. Western blotting analysis revealed that compounds 4, 5, 8, and 9 markedly suppressed the iNOS protein expression in LPS-induced RAW 264.7 cells in a concentration-dependent manner ( Figure 5). The downregulation of the expression of iNOS corresponded to the reduction of the production of NO. Inhibition of NO overproduction through blocking inducible nitric synthase (iNOS) expression have been proved to be potential target of anti-inflammatory drug [24]. In the following, the inhibitory effect of compounds 4, 5, 8, and 9 on iNOS expression in LPS-stimulated RAW 264.7 cells were investigated. Western blotting analysis revealed that compounds 4, 5, 8, and 9 markedly suppressed the iNOS protein expression in LPS-induced RAW 264.7 cells in a concentrationdependent manner ( Figure 5). The downregulation of the expression of iNOS corresponded to the reduction of the production of NO.

Compounds 4, 5, 8, and 9 Suppressed Release of IL-6 and TNF-α in LPS-Induced RAW 264.7 Cells
Proinflammatory cytokines, such as IL-6 and TNF-α produced by activated macrophages contribute to the inflammatory responses in inflammation-related diseases [25,26]. Therefore, they are regarded as targets for inhibiting the inflammatory process. To evaluate the effects of compounds 4, 5, 8, and 9 on the production of proinflammatory mediators, the levels of IL-6 and TNF-α in the culture medium were measured by ELISA. In accordance with the NO results, compounds 4, 5, 8, and 9 also significantly inhibited LPS-induced IL-6 and TNF-α released by RAW 264.7 cells in a concentration-dependent manner ( Figure 6).
As illustrated in Figure 4B, anti-inflammatory effect of actinofuranones was dramatically influenced by the locality of hydroxyl substitute of C-5 unsaturated alkyl chains (3 vs . 4, 5). Meanwhile, the more hydroxyl located at alkyl chain, the weaker anti-inflammation effect exhibited (3-5 vs . 8, 9). The additional hydroxyl possibly adjust the physico-chemical properties of drugs and change their biological activities. Proinflammatory cytokines, such as IL-6 and TNF-α produced by activated macrophages contribute to the inflammatory responses in inflammation-related diseases [25,26]. Therefore, they are regarded as targets for inhibiting the inflammatory process. To evaluate the effects of compounds 4, 5, 8, and 9 on the production of proinflammatory mediators, the levels of IL-6 and TNF-α in the culture medium were measured by ELISA. In accordance with the NO results, compounds 4, 5, 8, and 9 also significantly inhibited LPS-induced IL-6 and TNF-α released by RAW 264.7 cells in a concentration-dependent manner ( Figure 6).

Microbial Material
The strain (No. YIM 130461) was isolated from Leptogium trichophorum collected from an evergreen broad-leaf forest at an elevation of 2500 m in Benzilan, Diqing (Yunnan Province, China). On the basis of NCBI BLAST analysis of 16S rRNA gene sequences, this strain was identified as Streptomyces gramineus because it had 99.90% sequence identity with previously reported S. gramineus (GenBank accession no. HM748598). The strain (No. YIM 130461) was deposited at the Yunnan Institute of Microbiology, Yunnan University, China.

Microbial Material
The strain (No. YIM 130461) was isolated from Leptogium trichophorum collected from an evergreen broad-leaf forest at an elevation of 2500 m in Benzilan, Diqing (Yunnan Province, China). On the basis of NCBI BLAST analysis of 16S rRNA gene sequences, this strain was identified as Streptomyces gramineus because it had 99.90% sequence identity with previously reported S. gramineus (GenBank accession no. HM748598). The strain (No. YIM 130461) was deposited at the Yunnan Institute of Microbiology, Yunnan University, China.

Fermentation, Extraction and Isolation
The strain, grown on agar plate, was prepared to inoculate 500 mL Erlenmeyer flasks each containing 100 mL of sterile seed medium composed of glucose 0.4%, yeast extract 0.4%, malt extract 0.5%, multiple vitamin (thiamine 0.5 mg, riboflavin 0.5 mg, niacin 0.5 mg, pyridoxine 0.5 mg, inositol 0.5 mg, calcium pentothenate 0.5 mg, p-aminobenzoic acid 0.5 mg, biotin 0.25 mg) 3.75 mg per liter, and trace element solution (2g L −1 FeSO 4 ·7H 2 O, 1 g L −1 MnCl 2 ·4H 2 O, 1 g L −1 ZnSO 4 ·7H 2 O) 1.0 mL per liter at a pH of 7.2 with no adjustment. These flasks cultures were incubated at 28 • C for 2 days on a rotary shaker set at 180 rpm. For large-scale fermentation, 20 mL of seed medium was used to transferred into 1-L Erlenmeyer flasks containing 200 mL of sterile fermentation medium composed of 10 g L −1 soybean meal, 2 g L −1 peptone, 20 g L −1 glucose, 5 g L −1 soluble starch, 2 g L −1 yeast extract, 4 g L −1 NaCl, 0.5 g L −1 K 2 HPO 4 , 0.5 g L −1 MgSO 4 ·7H 2 O, and 2 g L −1 CaCO 3 at a pH of 7.8 with no adjustment. The fermentation batches were cultured at 28 • C for 7 days on a rotary shaker set at 180 rpm. The mycelium and broth filtrate (70 L) were separated by centrifugation (4000 rpm, 5 min). The resultant aqueous phase filtrate was extracted with EtOAc, then the organic partition layer was collected. Meanwhile, the mycelium cake was steeped in MeOH for 24 h to produce cell extracts. The methanol solution was centrifuged, concentrated, diluted with water and extracted with EtOAc. The combined extracts were evaporated in vacuo to yield 13 g of dried crude extract and further separated by chromatography on a silica gel column with a gradient of CH 2 Cl 2 -MeOH (v/v 50:1-1:1) to obtain nine fractions. Fraction 4 was subjected to Sephadex LH-20 chromatography (MeOH) to produce five subfractions.