Phenolic Metabolites from a Deep-Sea-Derived Fungus Aspergillus puniceus A2 and Their Nrf2-Dependent Anti-Inflammatory Effects

Four undescribed phenolic compounds, namely asperpropanols A–D (1–4), along with two known congeners 5 and 6, were isolated from Aspergillus puniceus A2, a deep-sea-derived fungus. The gross structures of the compounds were established by detailed analyses of the HRESIMS and NMR data, and their absolute configurations were resolved by modified Mosher’s method and calculations of ECD data. Compounds 1–6 were found to have excellent anti-inflammatory effect on lipopolysaccharide (LPS)-induced RAW264.7 cells at 20 μM, evidenced by the reduced nitric oxide (NO), tumor necrosis factor α, and interleukin 6 production. Among them, 5 and 6 showed inhibitory effects on NO production comparable with the positive control (BAY11-7083 at 10 μM). Additionally, the LPS-induced mRNA expressions of inducible nitric oxide synthase and cyclooxygenase-2 were also decreased. Interestingly, mRNA expression of nuclear factor erythroid 2–related factor 2 (Nrf2) was downregulated by LPS and recovered by 1–6, suggesting a vital role of Nrf2 in their effect. We further found that pharmacological inhibition of Nrf2 by ML385 largely abrogated the effects of 1–6 on RAW264.7 cells. Therefore, 1–6 may share a common anti-inflammatory mechanism via Nrf2 upregulation and activation.


Introduction
Aspergillus fungi are distributed in nearly every marine habitat explored, and their metabolites encompass a great diversity in both chemistry and bioactivity [1]. The metabolites of Aspergillus include polyketides, alkaloids, terpenoids and other structural types, which have biological activities such as antibacterial, antitumor, and metabolic regulation [2]. For example, the well-known cholesterol-lowering agent lovastatin was found in a fermentation broth of Aspergillus terreus [3]. As for the marine-derived Aspergillus, the firstly reported novel compounds were fumiquinazolines A-C, discovered by Numata et al. in 1992 [4]. Since then, the number of novel compounds from marine-derived Aspergillus has been increasing each year, and so far, more than 500 novel compounds have been discovered from marine-derived Aspergillus [5].
Previously, eight novel diketopiperazine-type alkaloids were isolated from a deep-seaderived fungus Aspergillus puniceus SCSIO z021, and they were found to be LXRα agonists with the EC50 values of 1.7−50 µM [6]. Later, the same team found nine new mycotoxins (xanthone-type and anthraquinone-type) and fourteen new isoquinoline alkaloids as inhibitors of protein tyrosine phosphatases in the extract of the strain [7,8]. Aspergillus Previously, eight novel diketopiperazine-type alkaloids were isolated from a deepsea-derived fungus Aspergillus puniceus SCSIO z021, and they were found to be LXRα agonists with the EC50 values of 1.7−50 μM [6]. Later, the same team found nine new mycotoxins (xanthone-type and anthraquinone-type) and fourteen new isoquinoline alkaloids as inhibitors of protein tyrosine phosphatases in the extract of the strain [7,8]. Aspergillus puniceus A2 was newly isolated from the deep-sea sediment, and its secondary metabolites were investigated in order to find novel and bioactive compounds. Under our continuous efforts, four novel phenolic metabolites and two known congeners were isolated from the fermentation broth of A. puniceus A2 (Figure 1). Inflammation is a basic response to pathogens, toxic compounds and damaged cells. It is triggered by the immune system, and is an essential part of self-healing [9]. However, chronic inflammation caused by lifestyle, genetic, environmental and other factors can be detrimental. Therefore, inflammation is associated with a wide range of diseases such as arteriosclerosis, type 2 diabetes, fatty liver disease, asthma, obesity, Alzheimer's and Parkinson's diseases, inflammatory bowel disease and cancers [10]. Herein, the anti-inflammatory effect of the isolated compounds 1-6 from A. puniceus A2 were investigated in a lipopolysaccharide (LPS)-stimulated Raw264.7 cell model. Additionally, the mechanism underlying their effect was explored and a common regulator was revealed. Our discoveries may provide a novel insight into marine-derived compounds with therapeutic implications for inflammatory-related diseases.

Chemistry
Asperpropanol A (1) was isolated as a brownish red powder, and its molecular formula was determined to be C10H14O5 on the basis of the sodium adduct ion peak at m/z 237.0727 [M + Na] + in the HRESIMS spectrum, requiring four degrees of unsaturation. The IR spectrum exhibited absorption bands for OH (3425 cm −1 ) and benzene (1604 cm −1 ) functional groups. The 1 H NMR spectrum of 1 displayed two aromatic protons at δH 6.20, two oxymethines (δH 3.60, 4.48), two methyls (δH 0.87, 3 Figure 2). In addition, the HMBC correlations from 3-OH (δH 9.07) to C-2 (δC 138.3), C-3 (δC 150.4) and C-4 (δC 103.0), from 5-OH (δH 8.86) to C-4, C-5 (δC 153.6) and C-6 (δC 104.7), from H-7 to C-1 (δC 137.2)/C-2/C-6 and from OCH3 (δH 3.62) to C-2 deduced two hydroxy groups, a methoxy group and Inflammation is a basic response to pathogens, toxic compounds and damaged cells. It is triggered by the immune system, and is an essential part of self-healing [9]. However, chronic inflammation caused by lifestyle, genetic, environmental and other factors can be detrimental. Therefore, inflammation is associated with a wide range of diseases such as arteriosclerosis, type 2 diabetes, fatty liver disease, asthma, obesity, Alzheimer's and Parkinson's diseases, inflammatory bowel disease and cancers [10]. Herein, the antiinflammatory effect of the isolated compounds 1-6 from A. puniceus A2 were investigated in a lipopolysaccharide (LPS)-stimulated Raw264.7 cell model. Additionally, the mechanism underlying their effect was explored and a common regulator was revealed. Our discoveries may provide a novel insight into marine-derived compounds with therapeutic implications for inflammatory-related diseases.

Biology
Raw264.7 cells were employed to evaluate the anti-inflammatory effect of compounds 1-6. We firstly investigated their cytotoxic effect, and found that none of them have significant inhibitory effect on the cell viability of Raw264.7 cells at the concentration of 20 µM ( Figure 5A). Next, we found that LPS-induced nitric oxide (NO) production was significantly lowered by 1-6 at 20 µM. Among them, 5 and 6 exhibited comparable effect to BAY11-7083 (BAY, the positive control) of 10 µM ( Figure 5B). Major inflammatory cytokines tumor necrosis factor-α (TNF-α) and interleukin 6 (IL-6) were also robustly elevated by LPS and significantly lowered by the treatment of compounds 1-6 ( Figure 5C,D).
The above results indicated that all the tested phenolic metabolites had anti-inflammatory effect. To explore the underlying mechanism, we determined the mRNA expression of several key inflammation-related genes. In consequence, it was found that inducible nitric oxide synthase (iNOS) and cyclooxygenase-2 (COX-2) expression were sharply elevated by LPS treatment and significantly lowered by 1-6 ( Figure 6A,B).
NO is a free radical with an unpaired electron, and iNOS produces log-fold higher amounts of NO as an immune defense mechanism [17]. COX-2 was reported to elevate IL-6 in a pristane-treated murine model of inflammation [18]. Inhibition of COX-2 reduced IL-6 expression in both cell and animal models [19,20]. The literatures suggest a synergistic pro-inflammatory effect of COX-2 and TNF-α [21,22]. Therefore, the result of the alteration of the expression of pro-inflammatory genes iNOS and COX-2 were consistent with the regulation of NO, TNF-α and IL-6 production ( Figure 5). significantly lowered by 1-6 at 20 μM. Among them, 5 and 6 exhibited comparable effect to BAY11-7083 (BAY, the positive control) of 10 μM ( Figure 5B). Major inflammatory cytokines tumor necrosis factor-α (TNF-α) and interleukin 6 (IL-6) were also robustly elevated by LPS and significantly lowered by the treatment of compounds 1-6 ( Figure 5C,D). The above results indicated that all the tested phenolic metabolites had anti-inflammatory effect. To explore the underlying mechanism, we determined the mRNA expression of several key inflammation-related genes. In consequence, it was found that inducible nitric oxide synthase (iNOS) and cyclooxygenase-2 (COX-2) expression were sharply elevated by LPS treatment and significantly lowered by 1-6 ( Figure 6A,B). NO is a free radical with an unpaired electron, and iNOS produces log-fold higher amounts of NO as an immune defense mechanism [17]. COX-2 was reported to elevate IL-6 in a pristane-treated murine model of inflammation [18]. Inhibition of COX-2 reduced IL-6 expression in both cell and animal models [19,20]. The literatures suggest a synergistic pro-inflammatory effect of COX-2 and TNF-α [21,22]. Therefore, the result of the alteration of the expression of pro-inflammatory genes iNOS and COX-2 were consistent with the regulation of NO, TNF-α and IL-6 production ( Figure 5).
Furthermore, we found that Nrf2 was inhibited by LPS and partly rescued by the phenolic metabolites' treatment ( Figure 6C). Since Nrf2 is a master regulator that protects against oxidative damage and inflammatory responses [23], we inferred from the above result that the effect of 1-6 may be Nrf2-dependent. Hence, we used a specific Nrf2 inhibitor ML385 to observe the corresponding alteration of the inflammatory phenotypes under the treatment of 1-6. Intriguingly, the inhibitory effects of 1-6 on NO, TNF-α and IL-6 production were largely abolished by the antagonism of Nrf2 by ML385 (Figure 7A-C). The upregulation of Nrf2 by the compounds also partly diminished except for compound 4 ( Figure 7D). Furthermore, the downregulation of COX-2 and iNOS by 1-6 treatment was Furthermore, we found that Nrf2 was inhibited by LPS and partly rescued by the phenolic metabolites' treatment ( Figure 6C). Since Nrf2 is a master regulator that protects against oxidative damage and inflammatory responses [23], we inferred from the above result that the effect of 1-6 may be Nrf2-dependent. Hence, we used a specific Nrf2 inhibitor ML385 to observe the corresponding alteration of the inflammatory phenotypes under the treatment of 1-6. Intriguingly, the inhibitory effects of 1-6 on NO, TNF-α and IL-6 production were largely abolished by the antagonism of Nrf2 by ML385 ( Figure 7A-C). The upregulation of Nrf2 by the compounds also partly diminished except for compound 4 ( Figure 7D). Furthermore, the downregulation of COX-2 and iNOS by 1-6 treatment was abrogated to a large extent ( Figure 7E,F). production were largely abolished by the antagonism of Nrf2 by ML385 ( Figure 7A-C). The upregulation of Nrf2 by the compounds also partly diminished except for compound 4 ( Figure 7D). Furthermore, the downregulation of COX-2 and iNOS by 1-6 treatment was abrogated to a large extent ( Figure 7E,F). Nrf2 is a transcript factor that binds to antioxidant response element (ARE) to activate the downstream genes involved in antioxidative and anti-inflammatory defenses [24]. ML385 interacts with Nrf2, and most likely blocks the binding of NRF2 to AREs [25]. Since ML385 largely abolished the anti-inflammatory effect of 1-6, we inferred that 1-6 may activate Nrf2 to exert their anti-inflammatory effect in LPS-stimulated RAW264.7 cells. It should be noted that 1-6 upregulated the expression of Nrf2 ( Figure 5C), which may also contribute to their anti-inflammatory effect.

General Experimental Procedures
The UV data were measured on a UV-1800 spectrophotometer (Shimadzu, Kyoto, Japan). Optical rotation data were acquired on an MCP 500 automatic polarimeter (Anton Paar, Graz, Austria). The ECD spectra were recorded on a Chirascan Circular Dichroism Spectrometer (Applied Photophysics, Leatherhead, UK). The IR spectra were recorded on a Tensor27 FT-IR spectrophotometer (Bruker Optics, Ettlingen, Germany). The 1 H, 13 C, HSQC, COSY and HMBC spectra were recorded on an AV-600 MHz spectrometer (Bruker, Karlsruhe, Germany), and the NMR chemical shifts (δ) were referenced to the solvent peaks of DMSO-d 6 at 2.50 and 39.5 ppm for proton and carbon, respectively. The HRESIMS spectra were acquired on the basis of the Xevo G2 Q-TOF mass spectrometer (Waters, Manchester, UK). Semipreparative HPLC was conducted on an Alltech LS class pumpequipped with UV/Vis detector (Alltech, Deerfield, IL, USA), and a packed column (ODS-A, 250 × 10 mm, 5 µm, Cosmosil, Kyoto, Japan) was used for the isolation and purification. Column chromatography (CC) was performed on the basis of Sephadex LH-20 (Pharmacia, Uppsala, Sweden), silica gel and ODS (Osaka Soda, Tokyo, Japan). Precoated silica gel plates GF-254 (Jiangyou Silicon Development, Yantai, China) were used for TLC analysis. All solvents used for CC were analytical grade.

Fungal Material and Fermentation
The producing fungus, which was obtained from the deep-sea sediment at the depth of 4841 m sampling from the Pacific Ocean, was identified to be Aspergillus puniceus on the basis of its ITS gene sequence, which exhibited 99.83% homology to those of the Aspergillus puniceus SRRC 2155 (accession no. AY373863). Thus, the fungus was named Aspergillus puniceus A2, and the ITS sequence information was deposited in the GenBank given the accession no. OM063154. The strain was deposited at the Technology Innovation Center for Exploitation of Marine Biological Resources, Third Institute of Oceanography, Ministry of Natural Resources, China. For the fermentation, the fungus was activated on PDA medium (28 • C, 4 d), and then fresh mycelia and spores were inoculated into PDB medium under rotary culture (100 rpm, 25 • C, 3 d) to obtain seed cultures. Large-scale fermentation was carried out on rice solid medium in 1-L Erlenmeyer flasks, each containing 130 g of rice, 5.6 g of sea salt, and 170 mL of H 2 O. After autoclaving, each flask was inoculated with seed cultures (3 mL) and then incubated at 25 • C in static conditions for 50 days.

Extraction and Isolation
The fermented solid mash was extracted with ethyl acetate (EtOAc) four times to get a dark oily residue. The EtOAc extract was chromatographed on silica gel column CC eluting with petroleum ether (PE) and EtOAc (1:0-3:1) to yield five fractions (Fr.

Cell Viability
Raw264.7 cells were seeded in a 96-well plate at the intensity of 1 × 10 4 cells per well and cultured overnight. Then, the compounds (10 mM stock solutions, dissolved in DMSO) were added at the working concentration of 20 µM and cultured for 24 h (n = 6). After that, the original DMEM were discarded, and 100 µL of new culture medium with 10% Cell Counting Kit-8 (CCK8, MCE, Monmouth Junction, NJ, USA) was added in each well and cultured for 2 h. The absorbance at 450 nm was measured with a microplate reader (Tecan Sunrise, TECAN Deutschland GmbH, Crailsheim, Germany), and cell viability was calculated according to the manufacturer's instructions.

Nitrite Determination
Raw264.7 cells were seeded in a 96-well plate at 1 × 10 4 cells/well and treated with vehicle (normal control, NC), 1 µg/mL LPS (L4391, Sigma-Aldrich, St. Louis, MO, USA), LPS combined with 10 µM of BAY11-7083 (BAY, Selleck, Huston, USA), LPS combined with the compounds (20 µM), or LPS combined with the compounds and 5 µM of ML385 (Selleck). The cells were treated for 24 h, and the nitrite in the culture medium was determined with Griess reagent (Thermo Fisher Scientific, Shanghai, China) according to the manufacturer's protocol.

Enzyme-Linked Immunosorbent Assays
Cytokine levels in the cell culture medium were determined with corresponding kits (EK282/4-96 for TNF-α and EK206/3-96 for IL-6, Multi Sciences, Hangzhou, China) according to the manufacturer's protocols.

Quantitative RT-PCR Analysis
Total RNA of Raw264.7 cells was extracted using Trizol reagent (Biouniquer Technology Co., Ltd., Nanjing, China) and converted to cDNA using a commercial kit (R333-01, Vazyme Biotech, Nanjing, China). Quantification of gene expression was performed on the LightCycler 96 instrument (Roche, Basel, Switzerland) using Taq Pro Universal SYBR qPCR Master Mix (Q712-02, Vazyme Biotech) according to the manufacturer's instructions. The 2 −∆∆Ct methods were used to measure relative transcript mRNA level of iNOS, COX-2, and Nrf2. β-actin was used as the invariant control. The primers used are listed in Table 3.

Conclusions
Asperpropanols A-D (1-4), four undescribed phenolic compounds, along with two known analogous 5 and 6, were isolated from a deep-sea-derived fungus Aspergillus puniceus A2. Their structures, including the absolute configurations, were established by detailed analyses of the HRESIMS and NMR data, modified Mosher's method and calculations of ECD spectra. Furthermore, 1-6 exerted their anti-inflammatory effects probably via Nrf2 upregulation and activation.