Phomopsichin A–D; Four New Chromone Derivatives from Mangrove Endophytic Fungus Phomopsis sp. 33#

Four new chromone derivatives, phomopsichins A–D (1–4), along with a known compound, phomoxanthone A (5), were isolated from the fermentation products of mangrove endophytic fungus Phomopsis sp. 33#. Their structures were elucidated based on comprehensive spectroscopic analysis coupled with single-crystal X-ray diffraction or theoretical calculations of electronic circular dichroism (ECD). They feature a tricyclic framework, in which a dihydropyran ring is fused with the chromone ring. Compounds 1–5 showed weak inhibitory activities on acetylcholinesterase as well as α-glucosidase, weak radical scavenging effects on 1,1-diphenyl-2-picrylhydrazyl (DPPH) as well as OH, and weak antimicrobial activities. Compounds 1–4 showed no cytotoxic activity against MDA-MB-435 breast cancer cells. Their other bioactivities are worthy of further study, considering their unique molecular structures.


Introduction
The chromone family of natural products exhibit a range of biological activities including anticancer, anti-inflammatory, antibacterial, antiviral, atypical antipsychotic, and anti-platelet properties [1][2][3][4][5][6][7][8][9]. In our continuous investigation of new bioactive secondary metabolites from the mangrove endophytic fungi in the South China Sea, four new chromone derivatives, phomopsichin A-D (1)(2)(3)(4), along with a known compound, phomoxanthone A (5), were isolated from the metabolic products of endophytic fungus Phomopsis sp. 33# from the bark of the mangrove plant Rhizophora stylosa. Compounds 1-3 ( Figure 1) featured a tricyclic framework in which a dihydropyran ring is fused at C-3 and C-4 of the chromone ring. To our knowledge, the compounds with this type skeleton number approximately 10, which were reported to exhibit the activity attenuating resistin-induced adhesion of HCT-116 colorectal cancer cells to endothelial cells [10], the activity interrupting the dimer formation of Aβ17-42 peptide associated to Alzheimer's disease [11], inhibitory activity against metallo-β-lactamases [12], moderate antibacterial activity and weak cytotoxic activity [13][14][15]. In this study, we report the isolation, structural elucidation, and exploration on the biological activities of compounds 1-5.
In this study, we report the isolation, structural elucidation, and exploration on the biological activities of compounds 1-5.

Structure Elucidation
Phomopsichin A (1, Figure 1) was obtained as a white solid and had a molecular formula of C16H16O7 as determined by its datum of high resolution electrospray ionization mass spectroscopy (HRESIMS) (observed m/z 319.08184 M − , calculated 319.08233), requiring nine degrees of unsaturation. The 13 C-NMR and distortionless enhancement by polarization transfer (DEPT) spectra (Table 1) indicated the presence of two carbonyl groups (δc 169.5 and 173.2), eight olefinic carbons, two sp 3 CH groups, one sp 3 CH2 group, two methoxy groups, and one methyl group. The 1 H-NMR and 1 H-1 H correlation spectroscopy (COSY) ( Table 1 and  The remaining two degrees of unsaturation supported a tricyclic carbon framework of dihydropyrano [4,3-b]chromen-10(1H)-one in 1, which was confirmed by the correlations between H-13 and C-2/C-11 in heteronuclear multiple-bond correlation (HMBC) spectroscopy. In the HMBC spectrum ( Figure 2), rich correlation data allowed us to unambiguously establish the locations of substituents on the carbon skeleton. The HMBC correlation between H-8 and C-14 revealed that the carbonyl group was located at the C-9 position; the correlation between H3-1 and C-3 demonstrated that the CH3-1 was located at the C-2 position; and the correlations between H3-15 and C-14 as well as between H3-16 and C-13 indicated that the two methoxy groups were located at the C-14 and C-13 positions, respectively. One hydroxyl group was identified at the C-7 position based on the lower field chemical shift (δc 162.6, C-7).

Structure Elucidation
Phomopsichin A (1, Figure 1) was obtained as a white solid and had a molecular formula of C 16 H 16 O 7 as determined by its datum of high resolution electrospray ionization mass spectroscopy (HRESIMS) (observed m/z 319.08184 M − , calculated 319.08233), requiring nine degrees of unsaturation. The 13 C-NMR and distortionless enhancement by polarization transfer (DEPT) spectra (Table 1) indicated the presence of two carbonyl groups (δ c 169.5 and 173.2), eight olefinic carbons, two sp 3 CH groups, one sp 3 CH 2 group, two methoxy groups, and one methyl group. The 1 H-NMR and 1 H-1 H correlation spectroscopy (COSY) ( Table 1 and Figure 2) showed the signals of two m-hydrogens of phenol (δ H 6.93 d J = 2.4 Hz; 6.85 d J = 2.4 Hz), two methoxy groups (δ H 3.85/3.42), and one 2-oxo-propyl group (δ H 1.34 d J = 6.0 Hz; 4.34 m; 2.67 dd J = 18.0, 4.0 Hz; 2.58 dd J = 18.0, 4.0 Hz). The remaining two degrees of unsaturation supported a tricyclic carbon framework of dihydropyrano [4,3-b]chromen-10(1H)-one in 1, which was confirmed by the correlations between H-13 and C-2/C-11 in heteronuclear multiple-bond correlation (HMBC) spectroscopy. In the HMBC spectrum ( Figure 2), rich correlation data allowed us to unambiguously establish the locations of substituents on the carbon skeleton. The HMBC correlation between H-8 and C-14 revealed that the carbonyl group was located at the C-9 position; the correlation between H 3 -1 and C-3 demonstrated that the CH 3 -1 was located at the C-2 position; and the correlations between H 3 -15 and C-14 as well as between H 3 -16 and C-13 indicated that the two methoxy groups were located at the C-14 and C-13 positions, respectively. One hydroxyl group was identified at the C-7 position based on the lower field chemical shift (δ c 162.6, C-7).  13 C-NMR data of compounds 1-4 (400/100 MHz, J in Hz).    The relative stereochemistry of 1 was established by its nuclear Overhauser effect spectroscopy (NOESY). The NOE correlation between H-2 and H 3 -16 indicated the relative stereochemistry of 1 as shown in Figure 3.

(in C 3 D 6 O) 2 (in CDCl 3 ) 3 (in CDCl 3 ) 4 (in CD 3 OD)
The complete structure and stereochemistry of 1 were further confirmed by X-ray diffraction analysis ( Figure 4). The final refinement of the Cu Kα data resulted in a small Flack parameter of 0.02(3), allowing an unambiguous assignment of the absolute configuration of 1 as 2S, 13R ( Figure 1).
Phomopsichin B (2, Figure 1) was obtained as a white solid and had a molecular formula of C 17 H 18 O 8 based on HRESIMS data (observed m/z 349.09241 M − , calculated 349.09289), with one more CH 3 O group than compound 1. The 1 H-NMR, 13 C-NMR, and HMBC spectra of 2 were very similar to those of 1 (Table 1), except for the absence of the H-6 signal, and an added CH 3 O-17 signals (δ H/C 3.98/56.8). The added CH 3 O-17 was located at the C-7 position based on the NOE correlation between H-17 and H-8. One hydroxyl group was identified at the C-6 position based on the chemical shift of C-6 (δ c 134.7) as well as the HMBC correlation between H-8 and C-6.

Mar. Drugs 2016, 14, 215 4 of 11
H-17 and H-8. One hydroxyl group was identified at the C-6 position based on the chemical shift of C-6 (δc 134.7) as well as the HMBC correlation between H-8 and C-6.  The relative stereochemistry of 2 was established by its NOESY. The NOE correlation between H-2 and H3-16, similar to those of 1, indicated the relative stereochemistry of 2 as shown in Figure 3.
Compounds 2 and 1 have identical chiral spheres, just opposite in the signs of their specific rotation data; their ECD spectra were symmetric ( Figure 5). The ECD spectrum of 2 showed negative Cotton effect at 318 (Δε −0.48) nm as well as positive one at 291 (Δε +0.53) nm. Meanwhile, the ECD spectrum of 1 displayed opposite Cotton effects at the same wavelengths. For the above reasons, the absolute configuration of 2 was suggested as 2R, 13S.  H-17 and H-8. One hydroxyl group was identified at the C-6 position based on the chemical shift of C-6 (δc 134.7) as well as the HMBC correlation between H-8 and C-6.  The relative stereochemistry of 2 was established by its NOESY. The NOE correlation between H-2 and H3-16, similar to those of 1, indicated the relative stereochemistry of 2 as shown in Figure 3.
Compounds 2 and 1 have identical chiral spheres, just opposite in the signs of their specific rotation data; their ECD spectra were symmetric ( Figure 5). The ECD spectrum of 2 showed negative Cotton effect at 318 (Δε −0.48) nm as well as positive one at 291 (Δε +0.53) nm. Meanwhile, the ECD spectrum of 1 displayed opposite Cotton effects at the same wavelengths. For the above reasons, the absolute configuration of 2 was suggested as 2R, 13S.  The relative stereochemistry of 2 was established by its NOESY. The NOE correlation between H-2 and H 3 -16, similar to those of 1, indicated the relative stereochemistry of 2 as shown in Figure 3.
Compounds 2 and 1 have identical chiral spheres, just opposite in the signs of their specific rotation data; their ECD spectra were symmetric ( Figure 5). The ECD spectrum of 2 showed negative Cotton effect at 318 (∆ε −0.48) nm as well as positive one at 291 (∆ε +0.53) nm. Meanwhile, the ECD spectrum of 1 displayed opposite Cotton effects at the same wavelengths. For the above reasons, the absolute configuration of 2 was suggested as 2R, 13S. H-17 and H-8. One hydroxyl group was identified at the C-6 position based on the chemical shift of C-6 (δc 134.7) as well as the HMBC correlation between H-8 and C-6.  The relative stereochemistry of 2 was established by its NOESY. The NOE correlation between H-2 and H3-16, similar to those of 1, indicated the relative stereochemistry of 2 as shown in Figure 3.
Compounds 2 and 1 have identical chiral spheres, just opposite in the signs of their specific rotation data; their ECD spectra were symmetric ( Figure 5). The ECD spectrum of 2 showed negative Cotton effect at 318 (Δε −0.48) nm as well as positive one at 291 (Δε +0.53) nm. Meanwhile, the ECD spectrum of 1 displayed opposite Cotton effects at the same wavelengths. For the above reasons, the absolute configuration of 2 was suggested as 2R, 13S.  Phomopsichin C (3, Figure 1) was obtained as a white solid and had a molecular formula of C 16 H 16 O 7 based on HRESIMS data (observed m/z 319.08200 M − , calculated 319.08233). The 1 H-NMR, 13 C-NMR, 1 H-1 H COSY, and HMBC spectra of 3 were very similar to those of compound 2 (Table 1, Figure 2), except for the changes of CH-13 signals (δ H/C 5.57/94.5) in 2 to CH 2 -13 signals (δ H/C 4.82 d; 4.48 d/62.5) in 3. These results suggested that compound 3 is lacking a methoxy group at the C-13 position. The absolute configuration of compound 3 was determined as 2R by the result that the experimental ECD and calculated ECD spectrum for 2R isomer matched exactly ( Figure 6). Phomopsichin C (3, Figure 1) was obtained as a white solid and had a molecular formula of C16 H16O7 based on HRESIMS data (observed m/z 319.08200 M − , calculated 319.08233). The 1 H-NMR, 13 C-NMR, 1 H-1 H COSY, and HMBC spectra of 3 were very similar to those of compound 2 (Table 1, Figure  2), except for the changes of CH-13 signals (δH/C 5.57/94.5) in 2 to CH2-13 signals (δH/C 4.82 d; 4.48 d/62.5) in 3. These results suggested that compound 3 is lacking a methoxy group at the C-13 position. The absolute configuration of compound 3 was determined as 2R by the result that the experimental ECD and calculated ECD spectrum for 2R isomer matched exactly ( Figure 6). Phomopsichin D (4, Figure 1) had a molecular formula of C15H16O7 based on HRESIMS data (observed m/z 307.08194 M − , calculated 307.08233), requiring eight degrees of unsaturation. The 1 H-NMR, 13 C-NMR, 1 H-1 H COSY, and HMBC spectra of 4 were very similar to those of 1 (Table 1 and Figure 2), except for the change of CH-13 signals (δH/C 5.40/95.2) in 1 to CH2OH-13 signals (δH/C 4.55/55.0) and the absence of a methoxy group signal in 4. A dicyclic 4H-chromen-4-one segment of 4 was decided based on its eight degrees of unsaturation, which was supported by the absence of HMBC correlation between H-13 and C-2. The hydroxymethyl group was located at C-12 based on the HMBC correlation between H-13 and C-11. A 2-hydroxypropyl group was located at C-4 based on the HMBC correlations between H-1 and C-4, as well as, between H-3 and C-12.
The absolute configuration of 4 was confirmed as 2S based on the result that the experimental data and calculated ECD spectrum for the 2S isomer matched exactly (Figure 7).  Phomopsichin D (4, Figure 1) had a molecular formula of C 15 H 16 O 7 based on HRESIMS data (observed m/z 307.08194 M − , calculated 307.08233), requiring eight degrees of unsaturation. The 1 H-NMR, 13 C-NMR, 1 H-1 H COSY, and HMBC spectra of 4 were very similar to those of 1 (Table 1 and Figure 2), except for the change of CH-13 signals (δ H/C 5.40/95.2) in 1 to CH 2 OH-13 signals (δ H/C 4.55/55.0) and the absence of a methoxy group signal in 4. A dicyclic 4H-chromen-4-one segment of 4 was decided based on its eight degrees of unsaturation, which was supported by the absence of HMBC correlation between H-13 and C-2. The hydroxymethyl group was located at C-12 based on the HMBC correlation between H-13 and C-11. A 2-hydroxypropyl group was located at C-4 based on the HMBC correlations between H-1 and C-4, as well as, between H-3 and C-12.
The absolute configuration of 4 was confirmed as 2S based on the result that the experimental data and calculated ECD spectrum for the 2S isomer matched exactly (Figure 7).

Mar. Drugs 2016, 14, 215 5 of 11
Phomopsichin C (3, Figure 1) was obtained as a white solid and had a molecular formula of C16 H16O7 based on HRESIMS data (observed m/z 319.08200 M − , calculated 319.08233). The 1 H-NMR, 13 C-NMR, 1 H-1 H COSY, and HMBC spectra of 3 were very similar to those of compound 2 (Table 1, Figure  2), except for the changes of CH-13 signals (δH/C 5.57/94.5) in 2 to CH2-13 signals (δH/C 4.82 d; 4.48 d/62.5) in 3. These results suggested that compound 3 is lacking a methoxy group at the C-13 position. The absolute configuration of compound 3 was determined as 2R by the result that the experimental ECD and calculated ECD spectrum for 2R isomer matched exactly ( Figure 6). Phomopsichin D (4, Figure 1) had a molecular formula of C15H16O7 based on HRESIMS data (observed m/z 307.08194 M − , calculated 307.08233), requiring eight degrees of unsaturation. The 1 H-NMR, 13 C-NMR, 1 H-1 H COSY, and HMBC spectra of 4 were very similar to those of 1 (Table 1 and Figure 2), except for the change of CH-13 signals (δH/C 5.40/95.2) in 1 to CH2OH-13 signals (δH/C 4.55/55.0) and the absence of a methoxy group signal in 4. A dicyclic 4H-chromen-4-one segment of 4 was decided based on its eight degrees of unsaturation, which was supported by the absence of HMBC correlation between H-13 and C-2. The hydroxymethyl group was located at C-12 based on the HMBC correlation between H-13 and C-11. A 2-hydroxypropyl group was located at C-4 based on the HMBC correlations between H-1 and C-4, as well as, between H-3 and C-12.
The absolute configuration of 4 was confirmed as 2S based on the result that the experimental data and calculated ECD spectrum for the 2S isomer matched exactly (Figure 7).  Compound 5 was identified as phomoxanthone A (5, Figure 1) by comparison of its spectral data with that of the literature [16,17]; both compound 5 and phomoxanthone A had the same NMR, MS, ECD data ( Figure S27, in Supplementary Materials) and specific rotation data.

Biological Evaluation
The various bioactivities of compounds 1-5 were evaluated in vitro. The five compounds displayed low inhibitory activities on acetylcholinesterase (AchE) as well as α-glucosidase, weak radical scavenging effects on DPPH as well as OH, and low antimicrobial activity against 13 pathogenic bacteria strains (Tables S8 and S9, in Supplementary Materials). Compounds 1-4 showed no cytotoxic activity against MDA-MB-435 breast cancer cells. It was reported that phomoxanthone A (5) has strong pro-apoptotic activity and immunostimulatory activity [17], so we did not consider its cytotoxicity assays in the study.
Optical rotation measurements were carried out using a Bellingham-Stanley 37-440 polarimeter (Bellingham Stanley Ltd., Kent, UK). UV spectra were determined using a UV-240 spectrophotometer (Shimadzu, Tokyo, Japan). ECD spectra were measured using a Chirascan Circular Dichroism Spectrometer (Applied Photophysics, London, UK). IR spectra were measured on a TENSOR37 spectrometer (Bruker Optics, Ettlingen, Germany). The 1 H-NMR and 13 C-NMR data were acquired using a Bruker Avance 400 spectrometer at 400 MHz for 1 H nuclei and 100 MHz for 13 C nuclei (Bruker Biospin, Rheinstetten, Germany). Tetramethylsilane (TMS) was used as an internal standard, and the chemical shifts (δ) were expressed in ppm. The HRESIMS were obtained using a LTQ-Orbitrap LC-MS (Thermo Fisher, Frankfurt, Germany). Single-crystal data were carried out on an Agilent Technologies Gemini A Ultra system (Agilent Tech, Santa Clara, CA, USA). HPLC was performed using a 515 pump with a UV 2487 detector (Waters, Milford, CT, USA) and an Ultimate XB-C-18 column (250 mm × 10 mm, 5 µm; Welch, Maryland, USA). Normal pressure preparative column chromatography was carried out on RP-18 gel (25-40 µm, Daiso Inc., Osaka, Japan), silica gel (200-400 mesh, Qingdao Marine Chemical Inc., Qingdao, China), or a Sephadex-LH-20 (GE Healthcare, Stockholm, Sweden) for reverse and direct phase elution modes, respectively. The thin-layer chromatography was performed over F 254 glass plates (Qingdao Marine Chemical Inc.) and analyzed under UV light (254 and 366 nm).

Fungal Material
Endophytic fungus Phomopsis sp. 33# was isolated with PDA medium from the bark of the mangrove plant Rhizophora stylosa, collected in the intertidal region of Zhanjiang, in Guangdong Province, China, and identified according to its morphological characteristics and internal transcribed spacer (ITS) region [18]. A voucher specimen is deposited in our laboratory at −20 • C.

Fermentation, Extraction, and Isolation
Small agar slices bearing mycelia were placed in 1000 mL Erlenmeyer flasks containing rice medium (composed of 60 g rice, 80 mL distilled water, and 0.24 g sea salt) and incubated for 30 days at 28 • C. In total, 140 flasks of culture were obtained. Cultures were extracted with EtOAc. In total, Mar. Drugs 2016, 14, 215 7 of 11 250 g crude extract was obtained by evaporation of EtOAc. The crude extract was suspended in H 2 O (3 L) and partitioned with n-hexane (5 L × 2) and EtOAc (5 L × 2) to give n-hexane (90 g) and EtOAc (110 g) extracts, respectively.
The EtOAc extract was subjected to a silica gel column, eluted with a n-hexane-EtOAc gradient (from 100:0 to 0:100) to obtain six fractions (Fractions 1-6). Fraction 2 (15 g) was extracted with 300 mL of chloroform to give dark yellow liquid phases and solid. The chloroform-soluble fraction was evaporated to dryness and washed with methanol (50 mL × 6) to give compound 5 (a light yellow solid, 5.

Computational Analyses
All of the theoretical methods and the basis set used for optimization and spectrum calculation were recommended in previous studies [19,20]. All of the theoretical calculations, including geometry optimization, frequency analysis, and ECD spectrum prediction, were carried out with the density functional theory (DFT) and time-dependent density functional theory (TDDFT) methods in the Gaussian 09 software package (Gaussian Inc., Wallingford, CT, USA) [21]. The geometry optimizations were performed at the B3LYP/6-31+G (d) level in the gas phase. Based on the final optimized structure, the ECD spectra were calculated at the PBE1PBE-SCRF/6-311++g (d, p) level using the Polarized Continuum Model (PCM) with methanol as a solvent. The theoretical predicted ECD spectra were fitted in the SpecDis 1.6 software package (University of Würzburg, Würzburg, Germany) [22].

X-ray Crystallographic Analysis of Compound 1
Single crystals of compound 1 were obtained from CH 3 OH-EtOAc. A suitable crystal was selected and all crystallographic data were collected at 150 K with Cu/Kα radiation (λ = 1.54178 Å). Using Olex2 (OlexSys Ltd., Durham University, Durham, UK), the structure was solved with the SIR2004 structure solution program using direct methods and refined with the XL refinement package using least squares minimization [23][24][25].

AchE Inhibitory Assay
The inhibitory activities against AchE of compounds 1-5 were investigated in vitro using the modified Ellman method [26]. The substrates were S-acetylthiocholine iodide and 5,5 -dithio-bis-(2-nitrobenzoic acid). Huperzine A was used as a positive control.

OH-Radical-Scavenging Assay
Radical scavenging effect on OH of compounds 1-5 were carried out according to previously reported methods [29,30]. The indicator used was 1,10-phenanthroline-Fe 2+ ; vitamin C was used as a positive control.

α-Glucosidase Inhibitory Assay
The inhibitory activities against α-glucosidase of compounds 1-5 were investigated in vitro using the modified method described by Moradi-Afrapoli et al. [31]; p-nitrophenyl-α-D-glucopyranoside was used as the substrates, and trans-resveratrol was used as a positive control.

Antibacterial Experiment
The antibacterial activity of compounds 1-5 were investigated in vitro using the modified 96 well microtiter-based method described by Pierce et al. [32].

Conclusions
Mangrove endophytic fungi from the South China Sea provide rich fungal diversity, and are promising sources of structurally-unprecedented bioactive natural products [33][34][35][36][37]. Five chromone derivatives were isolated from the mangrove endophytic fungus Phomopsis sp. 33#, four of them are new compounds (1-4). Compounds 1-5 showed weak inhibitory activity of AchE as well as α-glucosidase, radical scavenging effects on DPPH as well as OH, and low antimicrobial activity. The compounds (1)(2)(3)(4) showed no cytotoxic activity against MDA-MB-435 breast cancer cells. Their other bioactivities are worthy of further study, considering their unique tricyclic molecular structures, in which a dihydropyran ring is fused with the chromone ring.