Neuroprotective Metabolites from Vietnamese Marine Derived Fungi of Aspergillus and Penicillium Genera

Low molecular weight secondary metabolites of marine fungi Aspergillus flocculosus, Aspergillus terreus and Penicillium sp. from Van Phong and Nha Trang Bays (Vietnam) were studied and a number of polyketides, bis-indole quinones and terpenoids were isolated. The structures of the isolated compounds were determined by 1D and 2D NMR and HR-ESI-MS techniques. Stereochemistry of some compounds was established based on ECD data. A chemical structure of asterriquinone F (6) was thoroughly described for the first time. Anthraquinone (13) was firstly obtained from a natural source. Neuroprotective influences of the isolated compounds against 6-OHDA, paraquat and rotenone toxicity were investigated. 4-Hydroxyscytalone (1), 4-hydroxy-6-dehydroxyscytalone (2) and demethylcitreoviranol (3) have shown significant increasing of paraquat- and rotenone-treated Neuro-2a cell viability and anti-ROS activity.


Introduction
Marine fungi have to adapt to difficult environments and therefore they have a unique secondary metabolism that differs from terrestrial strains [1,2]. For instance, entomopathogenic fungus Beauveria (Isaria) felina usually produces a big variety of insecticide depsipeptides [3] but its marine strain was a source of rare oxygenated isochromenes and unique cytotoxic α-unsubstituted pyran polyketides [4]. Moreover, marine areas differ in a number of characteristics (salinity, depth, content of nutrients and organic compounds in the water), and therefore, fungi isolated from different seas can also produce different substances [5].
Van Phong and Nha Trang Bays are located in the central part of the Vietnamese coastline of the South China Sea [6]. Water in these sea areas is warm, and the temperature ranges from 23 • C in January up to 28 • C in May-June due to being closely connected with a monsoon climate. Salinity of coastal waters is close to the ocean normal one, ranging from 32% to 34% . The living conditions of determined as 3S,4S by comparison of its experimental CD data (Supplementary Materials, Figure S7) with literature data [27] and experimental CD data for 1 (Supplementary Materials, Figure S3). Thus, the structure of 2 was established as (3S,4S)-4-hydroxy-6-dehydroxyscytalone.
The fungus Aspergillus terreus LM.1.5 was cultivated for 21 days on modified rice medium. The dry EtOAc extract of the culture was separated successively over a column of silica gel and by normal-phase and reversed-phase HPLC to afford pure compounds 6-16 ( Figure 1).
Pseudo-molecular peak [M+Na] + at m/z 439.1257 in the (+)-HR-ESI-MS spectrum of 6 suggested the molecular formula as C24H20N2O5 (calculate for C24H20N2O5Na, 439.1264), which was confirmed by NMR data. A thorough analysis of 1 H and 13 C NMR spectra together with DEPT and HSQC data ( Table 1) indicated the presence of two methoxy groups (δC 61.5, 60.4; δH 4.12, 3.58), one oxygen-bearing sp 3 -metine carbon (δC 104.6; δH 6.24), one oxygenated quaternary sp 3   A chemical composition of sediment-derived fungus Aspergillus flocculosus was recently studied by us and reported earlier [18]. Dihydroaspyrone (5) was the main metabolite of the extract of this fungus.
Unfortunately, compounds 13-16 were obtained in insufficient amounts and were not studied for biological activities.

Biological Activities of the Studied Compounds
The neuroprotective activity of the compounds in Parkinson's disease (PD) in vitro models was investigated in murine malignant Neuro-2a cells, which are widely used for this purpose [41,42]. Firstly, all investigated compounds were tested on cytotoxicity against neuroblastoma Neuro-2a cells (Table 2).  A direct comparison of the NMR data of 6 with literature data for known varioloid D [29,30] revealed the close similarities, except the signals of C-1, C-2, C-4, C-6 and C-3" in 6. These data proved the presence of a tetracyclic system formed via a reduction of C-2' and C-3' in indole moiety followed by cyclization at C-4 and C-2'. Thus, compound 6 has a planar structure 3-(1H-indol-3-yl)-1, 4-dimethoxy-5a,6-dihydro-10bH-benzofuro [2,3-b]indole-2,10b-diol. Absolute configurations of stereocenters at C-2' and C-3' in 6 were determined based on a comparison of experimental ECD data of 6 with those for closely related known varioloid C [29,30]. Compound 6 was named asterriquinone F. It should be noted that a compound with this planar structure was once reported by Arai and Yamamoto [31] as the unnamed derivative of asterriquinone D. Nevertheless, these authors have not provided reliable structure elucidation evidence. The similarity of optical rotation values of asterriquinone F (6) and the unnamed compound from Arai and Yamamoto's report (+47 and +39, respectively) may indicate an identity of these compounds. In addition, another Arai and Yamamoto's paper proved an origin of asterriquinone F (6) via oxidation of asterriquinone D (11) [32].
Unfortunately, compounds 13-16 were obtained in insufficient amounts and were not studied for biological activities.
Questin (12) showed moderate cytoprotective activity in the rotenone-induced PD cell model and increased the viability of cells by only 23.7% (Figure 5e). At the same time, 12 was inactive against 6-OHDA and PQ neurotoxicity while the significant in vitro anti-ROS activity of 12 was found (Figure 5b,d,f). Earlier DPPH-radical scavenging activity has been reported for questin (12) however no exact data have been published [44]. In our investigation 12 did not show any significant DPPH radical scavenging up to 100 µM (Table 2). Questin (12) showed moderate cytoprotective activity in the rotenone-induced PD cell model and increased the viability of cells by only 23.7% (Figure 5e). At the same time, 12 was inactive against 6-OHDA and PQ neurotoxicity while the significant in vitro anti-ROS activity of 12 was found (Figure 5b,d,f). Earlier DPPH-radical scavenging activity has been reported for questin (12) however no exact data have been published [44]. In our investigation 12 did not show any significant DPPH radical scavenging up to 100 µM ( Table 2). (1) is a well-known metabolite intermediate of pentaketide pathway of melanin biosynthesis [45]. Melanin occurs in the cell walls of many fungi. Melanized fungal cells  [45]. Melanin occurs in the cell walls of many fungi. Melanized fungal cells survive desiccation and ultraviolet irradiation notably better than their hyaline counterparts. The ability of certain fungi to produce melanin also appears to be an important determinant of pathogenicity.  [45].
Demethylcitreoviranol (3) have been reported only twice [25,54]. It was investigated in the ARE luciferase assay and showed a weak inducing of the Nrf2-ARE pathway, which may suppress of oxidative genes and oxidative stress induced neurodegenerative diseases and carcinogenesis [54]. This has been confirmed in PQ-and rotenone-induced PD cell models.
The structural differences between 1 and 3 did not affect their neuroprotective activity. Probably, 1-naphtalenone or geometrically similar isocoumarine core is of great importance for the neuroprotective effect of these compounds against PQ and rotenone toxicity.
The lack of protective effect of 1-3 against the 6-OHDA-induced neuronal damage is apparently associated with a different mechanism of 6-OHDA, PQ and rotenone actions. All these neurotoxins cause oxidative stress and enhance ROS accumulation in cells. 6-OHDA induces oxidative stress both during its autoxidation to p-quinone and, also, during one-electron reduction of p-quinone to p-semiquinone, catalyzed by flavoenzymes that transfer one electron. PQ causes generation of intracellular free radicals via reducing of the divalent paraquat ion (PQ2+) to monovalent paraquat ion (PQ+) by NADPH-oxidase of mitochondrial complex I and reestablishing a new redox reaction by PQ+. In turn, rotenone was reported as a direct inhibitor of mitochondrial complex I. Thus, neuroprotective activity of compounds 1-3 is due not so much to the ROS scavenging (since it decreased intracellular ROS level in all three PD models), but also to the influence on some other aspects of PQ and rotenone neurotoxicity. 4-Hydroxy-3,6-dimethyl-2-pyrone (4) earlier was tested for antibacterial activities (methicillinresistant Staphylococcus aureus (MRSA), S. aureus, Enterococcus faecalis and Acinetobacter baumannii, E. coli and Klebsiella pneumonia) and cytotoxic activities against K562, BEL-7042, SGC-7901, A-549 and Hela cell lines and was ineffective [55]. In addition, it demonstrated a weak cytotoxicity against murine lymphoma L5178Y cell line [56]. As it was reported by Abe at al. 4 shows DPPH radical scavenging activity via donation of a hydrogen atom to DPPH radical and forming adduct with DPPH radical [57]. We suggest this mechanism may be realized in PQ-treated Neuro-2a cells. Differences between the mechanism of ROS generation by 6-OHDA, PQ and rotenone resulted in this fact that 4 reduced the ROS level in PQ-treated cells only. Dihydroaspyrone (5) earlier showed a weak cytotoxicity against HeLa [58], while it did not show cytotoxicity against a number of other cell lines [59,60] and antimicrobial activity against several microbial pathogens [59,61]. In present study 5 reduced ROS level in all neurotoxin-treated Neuro-2a cells but it statistically increased viability of PQ-treated cells only. Additional hydroxylation of 5 compared with 4 may play a key role for realizing of anti-ROS activity of 5 in neurotoxin-induced PD cell models.
It was earlier published that synthetic analogues of the asterriquinones, 1H5 and 5E5, have activated TrkA tropomyosin receptor kinase A directly in the cells and protected differentiated PC12 cells (rat adrenal pheochromocytomas) or contributed to the differentiation of neurons [62,63].
Thus, neuroprotective effects of asterriquinone B4 (8) and questin (12) from A. terreus in toxin-induced PD cell models were found at the first time. Earlier some butenolides and butyrolactone I from tropical strains A. terreus were reported as antineuroinflammatory compounds [64,65]. Moreover, butyrolactones I and VII from A. terreus exhibited protective activity against the glutamate-induced excitotoxicity [66]. Butyrolactone aspernolide F from endophytic strain A. terreus shown protective activity against doxorubicin-induced cardiotoxicity [67].
Metabolism of marine fungi is focused on survival in difficult highly competitive environments. For this reason, antimicrobial and cytotoxic fungal metabolites were the main goal of researchers during many years [68]. At the same time, the presence of cytoprotective compounds also should help fungi to successfully compete with bacteria in microbial communities. However, the cytoprotective potential of marine fungi is unappreciated today and these research are quite promising.

Fungal Strains
The strain Penicillium sp. KMM 4672 was isolated from a brown alga Padina sp. (Van Phong Bay, South China Sea, Vietnam) on malt extract agar, and identified on the basis of morphological and molecular features, as described earlier [69].
The strain Aspergillus flocculosus was isolated from a sediment sample (Nha Trang Bay, South China Sea, Vietnam) by inoculating on the modified Sabouraud medium and on the basis of morphological and molecular features, as described earlier [17].
The strain of Aspergillus terreus was isolated from leaves of the unidentified mangrove tree collected in Khanh Hoa province (Vietnam, South China Sea) by inoculating on the modified Sabouraud medium (peptone 10 g, glucose 20 g, agar 18 g, natural sea water 1000 mL, penicillin 1.5 g and streptomycin 1.5 g, pH 6.0-7.0). The fungus was identified according to a molecular biological protocol by DNA amplification and sequencing of the ITS region (GenBank accession number MN788658.1). BLAST search results indicated that the sequence was 98.06% identical (858/875 bp) with the sequence of Aspergillus terreus strain DTO 403-C9 (GenBank accession number MT316343.1). The strain is stored at the collection of microorganisms of the Nha Trang Institute of Technology and Research Application VAST (Nha Trang, Vietnam) under the code LM.5.2

Cultivation of the Fungi
All the fungal strains were cultured at room temperature for three weeks in 60 × 500 mL Erlenmeyer flasks each containing rice (20.0 g), yeast extract (20.0 mg), KH 2 PO 4 (10 mg) and natural seawater (40 mL).

Extraction and Isolation
The main part of the isolation procedures of compounds from Penicillium sp. KMM 4672 was described in a previous paper [22]. The n-hexane-EtOAc . The isolation of compound 4 was described in a previous paper [22].
The isolation of compound 5 from Aspergillus flocculosus was described in a previous paper [18]. The fungal mycelia of Aspergillus terreus with the medium were extracted for 24 h with 12.0 L of EtOAc. Evaporation of the solvent under reduced pressure gave a dark brown oil (2.7 g). To this residue was added 150 mL of H 2 O-EtOH (4:1), and the mixture was thoroughly stirred to yield a suspension. The suspension was sequentially extracted with hexane (100 mL × 3), EtOAc (150 mL × 3) and n-BuOH (150 mL × 2). The EtOAc fraction was concentrated under reduced pressure to give a dry residue (1.6 g), which was separated on a silica gel column (20.

DPPH Radical Scavenger Assay
The DPPH radical scavenging activities of the compounds were tested as described [70]. The compounds were dissolved in MeOH, and the solutions (120 µL) were dispensed into wells of a 96-well microplate. In all, 30 µL of the DPPH (Sigma-Aldrich, Steinheim, Germany) solution in MeOH (7.5 × 10 −3 M) was added to each well. The concentrations of the test compounds in the mixtures were 10 and 100 µM. The mixtures were shaken and left to stand for 30 min, and the absorbance of the resulting solutions was measured at 520 nm with a microplate reader MultiscanFC (ThermoScientific, Waltham, MA, USA). The concentration scavenging 50% of the DPPH radical (EC 50 ) was calculated for each investigated compound.

Cytotoxicity Assay
The in vitro cytotoxicity of the individual substances was evaluated using an MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay, which was performed according to the manufacturer's instructions (Sigma-Aldrich, USA). Absorbance of the converted formazan was measured using a Multiskan FC plate photometer (Thermo Scientific, Waltham, MA, USA) at λ = 570 nm. The results were presented as percent of control data, and concentration of cell viability inhibition on 50% (IC 50 ) was calculated [71].

Neurotoxin-Induced Cell Models of Parkinson's Disease
The neuroblastoma Neuro-2a line cells (1 × 10 4 cells/well) were treated with the test compounds at concentrations of 1 and/or 10 µM for 1 h, and then the neurotoxins were added to the neuroblastoma cell suspensions [17]. Rotenone (Sigma-Aldrich, USA) was used at the concentration of 10 µM. Paraquat (Sigma-Aldrich, USA) was used at 500 µM. 6-Hydroxydopamine (Sigma-Aldrich, USA) was used at 50 µM. Cells incubated without neurotoxins and the test compounds and cells incubated with neurotoxins only were used as positive and negative controls, respectively. After 24 h of incubation, the cell viabilities were measured using the MTT method. The results are presented as the percent of positive control data.

ROS Level Studying in Neurotoxin-Treated Neuro-2a Cells
The cells (1 × 10 4 cells/well of a 96-well plate) were incubated with compound solutions (10 µM) during 1 h. Then, 6-OHDA/PQ/rotenone were added to cell suspension to resulting concentration of 50 µM, 500 µM and 10 µM respectively for incubation during 1 h. Cells incubated without neurotoxins and compounds and with neurotoxins alone were used as positive and negative controls, respectively. The 20 µL of 2,7-dichlorodihydrofluorescein diacetate solution (H 2 DCFDA, Molecular Probes, Eugene, OR, USA) was added to each well (10 µM, final concentration) and the plate was incubated for an additional 10 min at 37 • C. The intensity of dichlorofluorescein fluorescence was measured with PHERAstar FS plate reader (BMG Labtech, Ortenberg, Germany) at λ ex = 485 nm and λ em = 518 nm. The data were processed by MARS Data Analysis v. 3.01R2 (BMG Labtech, Germany). The results were presented as the percent of positive control data [71].