Hirsutane-Type Sesquiterpenes with Inhibitory Activity of Microglial

May 17, 2017 - Graduate Institute of Medical Science and Department of Pharmacology, College of Medicine, Taipei Medical University, Taipei 11031, Tai...
0 downloads 0 Views 1MB Size
Download PDF
Article pubs.acs.org/jnp

Hirsutane-Type Sesquiterpenes with Inhibitory Activity of Microglial Nitric Oxide Production from the Red Alga-Derived Fungus Chondrostereum sp. NTOU4196 George Hsiao,† Wei-Chiung Chi,‡,§ Ka-Lai Pang,§ Jih-Jung Chen,⊥ Yueh-Hsiung Kuo,∥,¶ Yu-Kai Wang,□ Hyo-Jung Cha,§ Shen-Chieh Chou,# and Tzong-Huei Lee*,□ †

Graduate Institute of Medical Science and Department of Pharmacology, College of Medicine, Taipei Medical University, Taipei 11031, Taiwan ‡ Department of Food Science, National Quemoy University, Kinmen 89250, Taiwan § Institute of Marine Biology, National Taiwan Ocean University, Keelung 20224, Taiwan ⊥ School of Pharmaceutical Sciences, National Yang-Ming University, Taipei 11221, Taiwan ∥ Department of Chinese Pharmaceutical Sciences and Chinese Medicine Resources and #School of Pharmacy, College of Pharmacy, China Medical University, Taichung 40447, Taiwan ¶ Department of Biotechnology, Asia University, Taichung 41354, Taiwan □ Institute of Fisheries Science, National Taiwan University, Taipei 10617, Taiwan S Supporting Information *

ABSTRACT: The marine red alga Pterocladiella capillacea is an economic alga for the food industry in Taiwan, and its associated highly diversified fungi have not been investigated meticulously thus far. The EtOAc extract of the fermented broth of Chondrostereum sp. NTOU4196, a fungal strain isolated from P. capillacea, was found to exhibit significant nitric oxide (NO) production inhibitory activity in lipopolysaccharide-activated murine RAW 264.7 cells at a concentration of 100 μg/mL in the preliminary screening. Therefore, separation of the active principles from the fermented broths was performed, and that has led to the isolation of eight new 5,5,5-tricyclic hirsutane-type sesquiterpenes, namely, chondroterpenes A−H (1−8), together with seven known analogues. They were identified by analyses of spectroscopic data and comparison with literature values. Among the new isolates, chondroterpene A (1) exhibited more significant NO production inhibitory activity in murine BV-2 microglial cells, and of all the isolated compounds, hirsutanol A (9) exerted limited cytotoxic effects and the most potent inhibitory activity on NO production.

P

commonly found through all seasons, and its abundant polysaccharides have long been used as gelling agents in a variety of food dishes and have also been used widely as fermentation media for microbial cultures, clarifying agents for fermenting wine, and starch substitutes.7 In an attempt to search for anti-inflammatory compounds from marine-derived fungi, we have isolated a number of fungal strains from this alga, and a cell-based nitric oxide production inhibitory platform was adopted for biological screening of their extracts. As a result, the extract of Chondrostereum sp. NTOU4196 with significant bioactivity was selected for chemical investigation, and that has led to the isolation of eight new hirsutane-type sesquiterpenes (1−8) along with seven known analogues. This report herein describes the isolation and structure elucidation of the new compounds together with their nitric oxide production inhibitory activities.

lant-associated microorganisms including endophytes and epiphytes establish a mutually beneficial relationship with their hosts and live inside the plants or just on the surface of the plants, respectively.1 The presence of fungi inside plant structures has been found since the end of the 19th century, and so far, microorganisms have been isolated from various organs of different plant species.2 It is conceivable that microorganisms can play crucial roles in the metabolic processes of the plant hosts, and vice versa. On the basis of the coevolution hypothesis, it was proposed that the associated microorganisms might assist the hosts in chemical defense by producing bioactive secondary metabolites.3 Besides, various potent plant-derived clinically used anticancer drugs are suspected to be the biosynthetic products of the endophytic microorganisms.4−6 Thus, plant-associated microorganisms could be one of the best options for the investigation of bioactive secondary metabolites. Pterocladiella capillacea (Gelidiaceae) distributed in the north and northeast intertidal zone of Taiwan is an economic red alga © 2017 American Chemical Society and American Society of Pharmacognosy

Received: March 7, 2017 Published: May 17, 2017 1615

DOI: 10.1021/acs.jnatprod.7b00196 J. Nat. Prod. 2017, 80, 1615−1622

Journal of Natural Products

Article

at δC 35.5, 58.1, 82.7, 87.1, and 91.9 together with one set of double-bond resonances at δC 108.7 and 161.4 (Table 1). Analysis of the 1H NMR spectrum of 1 showed distinctive signals characteristic of an exomethylene at δH 4.98 (1H, br d, J = 2.4 Hz) and 5.21 (1H, d, J = 2.4 Hz), a carbinoyl proton at δH 4.74 (1H, tdd, J = 2.4, 7.0, 11.4 Hz), and three methyls at δH 1.04, 1.08, and 1.18 attached to quaternary carbons, in addition to complex signals appearing at δH 1.56−2.13, which corresponded to four methylene functionalities (Table 1). Considering the molecular formula of 1, C15H24O4, its doublebond equivalents were four including a double bond. Thus, 1 would have a tricyclic structural feature with three methyls and one terminal double bond, indicating a skeleton analogous to that of hirsutanol A (9). An ABX coupling system at δH 4.74 (H-4) and 1.69, 2.13 (each 1 H, H2-5) in the COSY spectrum together with conspicuous cross-peaks of δH 1.04 (H3-12)/δC 58.1 (C-2), 91.9 (C-1), and 161.4 (C-3), δH 4.98, 5.21(H2-13)/ δC 58.1 (C-2), 71.5 (C-4), and 161.4 (C-3), δH 1.69, 2.13 (H25)/δC 58.1 (C-2), 71.5 (C-4), and 82.7 (C-6), δH 2.07, 2.10 (H2-7)/49.7 (C-5), 54.2 (C-9), 58.1 (C-2), 82.7 (C-6), 87.1 (C-8), and 91.9 (C-1), δH 1.67, 1.85 (H2-9)/δC 35.5 (C-10), 49.3 (C-11), 56.2 (C-7), 87.1 (C-8), and 91.9 (C-1), δH 1.18 (H3-14)/δC 33.3 (C-15), 35.5 (C-10), 49.3 (C-11), and 54.2 (C-9), and δH 1.08 (H3-15)/δC 34.0 (C-14), 35.5 (C-10), 49.3 (C-11), and 54.2 (C-9) in the HMBC spectrum (Figure 1) corroborated that 1 was a linearly fused tricyclopentane hirsutane-type sesquiterpene with four hydroxy groups borne at C-1, -4, -6, and -8 and a terminal double bond assigned at Δ3(13). Subsequently, the 1H NMR spectrum of 1 was measured alternatively in DMSO-d6 to see the deuterium-exchangeable hydroxy proton chemical shifts and coupling patterns (Table S1), which facilitated the determination of configurations of the carbinol stereocenters by a ROESY spectrum. In the ROESY spectrum (in DMSO-d6) of 1 (Figure 1), cross-peaks of δH 4.71 (OH-4)/δH 1.57 (Hα-5), δH 1.90 (Hβ-5)/δH 4.49 (OH-6), δH 4.49 (OH-6)/δH 0.89 (H3-12), δH 0.89 (H3-12)/δH 2.89 (OH1), δH 2.89 (OH-1)/δH 1.56 (Hβ-9), δH 1.56 (Hβ-9)/δH 4.31

■

RESULTS AND DISCUSSION The EtOAc extracts of the fermented broth of Chondrostereum sp. NTOU4196 were concentrated to give a brown residue, which was then separated through Diaion HP-20 column chromatography and repeated HPLC to yield 15 compounds, including eight previously unreported compounds (1−8) and seven known analogues. Compound 1, obtained as a colorless oil, was deduced to have a molecular formula of C15H24O4 by 13C NMR spectroscopy and HRESIMS. The IR spectrum indicated the presence of a hydroxy (3389 cm−1) and an olefinic functionality (1657 cm−1). The 13C NMR data coupled with the HSQC spectrum displayed three methyl carbons at δC 19.4, 33.3, and 34.0, four methylene carbons at δC 49.3, 49.7, 54.2, and 56.2, one methine carbon at δC 71.5, and five nonprotonated carbons

Table 1. 1H and 13C NMR Spectroscopic Data for Compounds 1−4 1 no. 1 2 3 4 5

a

δCa 91.9, 58.1, 161.4, 71.5, 49.7,

C C C CH CH2

6 7

82.7, C 56.2, CH2

8 9

87.1, C 54.2, CH2

10 11

35.5, C 49.3, CH2

12 13

19.4, CH3 108.7, CH2

14 15

34.0, CH3 33.3, CH3

2 δHb

4.74 tdd (2.4, 7.0, 11.4) 1.69 dd (11.4, 11.4) 2.13 dd (7.0, 11.4) 2.07 d (15.6) 2.10 d (15.6) 1.67 d (13.6) 1.85 d (13.6) 1.56 2.00 1.04 4.98 5.21 1.18 1.08

d (14.2) d (14.2) s t (2.4) d (2.4) s s

3

δCa 187.7, 58.3, 151.4, 204.8, 49.3,

C C C C CH2

92.3, C 77.6, CH

δHb

2.42 d (18.9) 2.97 d (18.9)

19.3, CH3 120.0, CH2 25.3, CH3 25.6, CH3

187.8, 59.6, 53.2, 76.2, 50.1,

C C CH CH CH2

90.3, C 42.0, CH2

4.73 m

143.6, C 211.0, C 50.6, C 40.2, CH2

4

δCa

δHb

1.76 3.35 1.79 2.27

dd (7.0, 10.1) m dd (10.4, 12.5) dd (6.4, 12.5)

2.45 m 2.49 m

142.3, C 211.8, C

2.39 2.50 1.38 5.54 6.11 1.15 1.11

dd (1.9, 18.9) dd (3.0, 18.9) s s s s s

50.0, C 43.5, CH2 18.9, CH3 14.0, CH3

2.45 2.47 1.17 1.08

m m s d (7.0)

25.4, CH3 25.8, CH3

1.09 s 1.13 s

δCa 154.1, 52.6, 51.8, 75.1, 47.8,

δHb

C C CH CH CH2

1.61 3.23 1.67 2.24

dd (6.7, 10.7) dt (5.9, 10.7) dd (10.7, 12.0) dd (5.9, 12.0)

94.9, C 126.9, CH

5.36 dd (1.8, 1.8)

157.9, C 77.2, C

4.30 d (1.8)

54.8, C 132.9, CH

5.12 d (1.8)

21.6, CH3 13.3, CH3

1.09 s 1.03 d (6.7)

28.0, CH3 23.6, CH3

1.10 s 1.02 s

Measured in methanol-d4 (125 MHz). bMeasured in methanol-d4 (500 MHz). 1616

DOI: 10.1021/acs.jnatprod.7b00196 J. Nat. Prod. 2017, 80, 1615−1622

Journal of Natural Products

Article

Figure 1. COSY, selected HMBC, and ROESY correlations of compounds 1, 3, and 6.

Figure 2. Selected HMBC and ROESY (NOESY) correlations of compounds 2, 5, and 8.

the presence of a hydroxy group (3421 cm−1), a conjugated carbonyl moiety (1692 cm−1), and an olefinic functionality (1635 cm−1). The 13C NMR spectrum accompanied by a phase-sensitive HSQC spectrum revealed that there were three methyls, three methylenes, one methane, and eight nonprotonated carbons including two distinctive ketone carbonyls at δC 204.8 (C-4) and 211.0 (C-9) (Table 1). In the 1H NMR spectrum of 2, two mutually coupled nonequivalent methylene signals at δH 2.42, 2.97 (each 1 H, H2-5) and 2.39, 2.50 (each 1 H, H2-11) and one set of an exomethylene group at δH 5.54,

(OH-8), and δH 1.56 (Hβ-9)/δH 1.18 (H3-14) established the relative configuration of OH-1, -4, -6, and -8 to be β-, α-, β-, and β-oriented, respectively. The methyl group at C-12 of all the hirsutane-type sesquiterpenes found naturally is always βoriented, as indicated in the literature.8 Due to a biogenetic relationship of 1 and the known analogues isolated from the same fungal strain, the absolute configuration of 1 is suggested to be as shown. The molecular formula for 2, C15H18O4, was determined by 13 C NMR and HRESIMS data. The IR spectrum of 2 indicated 1617

DOI: 10.1021/acs.jnatprod.7b00196 J. Nat. Prod. 2017, 80, 1615−1622

Journal of Natural Products

Article

Table 2. 1H and 13C NMR Spectroscopic Data for Compounds 5−8 5

6 δHb

no.

δCa

1 2 3 4 5

58.7, CH 55.5, C 54.7, CH 222.0, C 51.2, CH2

6 7

86.3, C 51.4, CH2

8 9

88.0, C 58.1, CH2

10 11

43.5, C 44.0, CH2

12 13

16.7, CH3 9.4, CH3

1.52 1.53 1.01 0.94

14 15 16

30.9, CH3 29.1, CH3

1.15 s 1.06 s

a

2.03 m 2.14 dq (1.7, 6.9) 2.33 dd (1.7, 19.5) 3.12 d (19.5) 1.99 d (13.5) 2.08 d (13.5) 1.49 d (13.4) 1.78 d (13.4) m m s d (6.9)

δCa 43.3, 51.5, 52.7, 77.1, 42.9,

CH C CH CH CH2

90.9, C 86.0, CH 46.8, CH 47.1, CH2 43.0, C 44.6, CH2

7 δHb

2.29 dt (7.8, 11.7) 1.66 3.67 1.36 2.63

dd (6.8, 8.0) ddd (7.4, 8.0, 9.0) dd (7.4, 14.9) dd (9.0, 14.9)

3.61 d (9.4) 2.03 m 1.24 dd (8.2, 12.6) 1.73 ddd (1.8, 8.2, 12.6)

18.9, CH3 11.3, CH3

1.25 1.36 0.89 0.94

29.9, CH3 28.1, CH3

1.07 s 0.93 s

8

δCa 52.4, 55.8, 51.3, 78.3, 43.3,

CH C CH CH CH2

2.97 m 1.70 3.88 1.45 2.83

dd (7.0, 7.7) ddd (5.0, 7.7, 8.7) dd (5.0, 15.8) dd (8.7, 15.8)

δCa

4.22 d (1.6)

91.9, C 49.9, CH2

148.2, C 135.7, CH

5.41 dd (1.6, 3.0)

90.3, C 57.1, CH2

14.7, CH3 13.3, CH3

1.48 1.57 0.77 0.96

dd (8.8, 7.7) dd (6.2, 7.7) s d (7.0)

29.8, CH3 27.8, CH3

1.10 s 1.07 s

δHb

62.8, CH 57.0, C 157.4, C 206.0, C 46.6, CH2

95.5, C 76.5, CH

50.6, C 43.8, CH2

dd (7.8, 12.2) dd (11.7, 12.2) s d (6.8)

δHb

2.67 dd (8.6, 11.4)

2.52 d (17.8) 2.70 d (17.8) 1.55 dd (1.8, 14.6) 2.47 d (14.6) 1.68 d (1.8) 1.69 d (1.8)

40.8, C 45.3, CH2 19.2, CH3 118.1, CH2 30.6, CH3 28.2, CH3 52.3, CH3

1.64 1.73 1.07 5.35 6.01 1.05 1.14 3.23

ddd (1.8, 8.6, 12.3) dd (11.4, 12.3) s s s s s s

Measured in methanol-d4 (125 MHz). bMeasured in methanol-d4 (500 MHz).

protons at δH 5.36 (dd, J = 1.8, 1.8 Hz, H-7) and 5.12 (d, J = 1.8 Hz, H-11) in 4 (Table 1), revealing two double bonds located at Δ7 and Δ1(11). The assignments were also reflected in the differences between 13C NMR data of 3 and 4. The two aliphatic carbons C-7 and -11 along with the carbonyl carbon C-9 in 3 were replaced by two olefinic carbons at δC 126.9 and 132.9 and one carbinol carbon at δC 77.2 in 4 (Table 1), respectively. After the COSY and HMBC spectra of 4 were fully assigned, the gross structure of 4 was thus determined and was further supported by a sodium adduct [M + Na]+ at m/z 273.1450 (calcd for C15H22O3Na, 273.1467) in the HRESIMS spectrum, a UV λmax at 246 nm indicating a conjugated doublebond moiety, and no carbonyl absorption observed in the IR spectrum. The OH-4, -6, and -9 of 4 were deduced to be all βoriented based on the cross-peaks of H3-12/OH-6, OH-6/H-7, H-7/OH-9, and OH-9/H3-14 in the ROESY spectrum. The IR spectrum of 5 suggested the presence of a keto group as judged from an adsorption band at 1726 cm−1. The 1H and 13 C NMR data of 5 coincided well with those of 1 except that H2-13, OH-4, and OH-1 in 1 were replaced by a methyl [δH 0.94 (d, J = 6.9 Hz, H3-13); δC 9.4 (C-13)], a ketone carbonyl [δC 222.0 (C-4)], and a proton [δH 2.03 (m, H-1); δC 58.7 (C1)] in 5 (Table 2), respectively. The gross structure of 5 was further confirmed by an HMBC spectrum, in which key signals of H3-12/C-1, -2, -3, and -6, H3-13/C-2, -3, and -4, H2-5/C-2, -4, and -6, H2-7/C-2, -5, -6, and -8, and H2-9/C-1, -8, -11, and -14 were observed (Figure 2). The cross-peaks of H3-12/OH-6, OH-6/Hβ-7 (δH 1.84), Hα-7 (δH 1.94)/OH-8, OH-8/H-1, H1/H3-13, and OH-8/H3-15 in the ROESY spectrum (in DMSO-d6) (Figure 2) established OH-6, OH-8, and H3-13 to be β-, α-, and α-oriented. The structure of compound 5 was determined to have a cis−trans−cis-tricyclopentane ring system. Compound 6, afforded as an amorphous, white powder, was assigned a molecular formula of C15H26O3 by 13C NMR and HRESIMS data. Its IR spectrum indicated the presence of a hydroxy group (3300 cm−1). The 1H NMR data of 6 were in

6.11 (each 1 H, H2-13) were observed in addition to three methyls at δH 1.11 (H3-15), 1.15 (H3-14), and 1.38 (H3-12) as well as one carbinol proton at δH 4.73 (H-7) (Table 1). Key correlations of H2-13/C-2, -3, and -4, H2-5/C-2, -3, -4, and -6, H-7/C-1, -6, and -8, H2-11/C-1, -8, -9, and -10, H3-14/C-9, -10, -11, and -15, and H3-15/C-9, -10, -11, and -14 assigned in the HMBC of 2 (Figure 2) unambiguously established the gross structure to be a hirsutane-type sesquiterpene with an exomethylene at Δ3(13), two conjugated ketone carbonyls at C-4 and -9, two hydroxy groups attached at C-6 and -7, and a double bond at Δ1(8). Conspicuous cross-peaks of H3-12/OH6, OH-6/H-7, H-7/H3-12, H3-12/Hβ-11 [δH 2.29 (dd, J = 1.6, 18.8 Hz)], and Hβ-11/H3-14 in the NOESY spectrum (in DMSO-d6) confirmed that OH-6, OH-7, and H3-14 were β-, α-, and β-oriented (Figure 2), respectively. Compound 3 was assigned a molecular formula of C15H22O3 by 13C NMR and HRESIMS data. The IR spectrum of 3 indicated the presence of a hydroxy group (3373 cm−1), a conjugated carbonyl moiety [1681 cm−1 and UV λmax 243 nm (log ε 3.9)], and an olefinic functionality (1627 cm−1). The 1H and 13C NMR data of 3 were almost compatible with those of 2 except that a carbonyl at C-4 in 2 was replaced by a hydroxy (δH 3.35; δC 76.2) in 3, the exomethylene at Δ3(13) in 2 was replaced by a methyl (δH 1.08; δC 14.0) in 3, and the carbinol at C-7 in 2 was substituted by a methylene [δH 2.45 and 2.49 (each 1 H); δC 42.0] in 3 (Table 1). The differences were further confirmed by its COSY and HMBC assignments (Figure 1). The configurational analysis of the ROESY spectrum (in DMSO-d6) of 3 revealed key signals of H3-12/ OH-6, OH-6/Hβ-5 [δH 1.65 (dd, J = 10.2, 12.2 Hz)], Hβ-5/ OH-4, and Hα-4/H3-13, indicating the H3-13, OH-4, and OH-6 to be α-, β-, and β-oriented (Figure 1). The 1H NMR data of 4 were consistent with those of 3 except for one additional carbinol proton appearing at δH 4.30 (d, J = 1.8 Hz, H-9) and the two methylenes at C-7 and -11 in 3 were substituted by two long-range mutually coupled olefinic 1618

DOI: 10.1021/acs.jnatprod.7b00196 J. Nat. Prod. 2017, 80, 1615−1622

Journal of Natural Products

Article

accordance with those of 5 except that a methylene at δH 1.99 and 2.08 (each 1 H, d, J = 13.5 Hz, H2-7) in 5 was substituted by a carbinol proton at δH 3.61 (1 H, d, J = 9.4 Hz, H-7) in 6, and two additional methines were observed at δH 3.67 (ddd, J = 7.4, 8.0, 9.0 Hz, H-4) and 2.03 (m, H-8) in 6 (Table 2). These changes were also reflected in the 13C NMR spectrum (Table 2). These assignments were evidenced from COSY cross-peaks of H-4/H2-5, H-7/H-8, H-8/H2-9, and H2-11/H-1 and key HMBC correlations of H3-12/C-1, -2, -3, and -6, H3-13/C-2, -3, and -4, H-5/C-6 and -7, H-7/C-6 and −-9, and H-8/C-1 and C-11 (Figure 1). The orientations of H-1, OH-4, OH-6, OH-7, and H-8 in 6 were established as shown by the key cross-peaks of H-3/OH-4, H-3/H3-12, H3-12/OH-6, OH-6/H-7, H3-13/H1, H-1/H-8, and H-1/H3-15 in the ROESY spectrum (in DMSO-d6) (Figure 1). The HRESIMS spectrum of 7 showed a prominent sodium adduct ion [M + Na]+ at m/z 275.1615 (calcd for C15H24O3Na, 275.1623), 2 Da less than that of 6. Comparison of 13C NMR data of 7 with those of 6 revealed that the one quaternary C-8 at δC 46.8 and the secondary C-9 at δC 47.1 in 6 were absent and were substituted by two olefinic carbons at δC 148.2 (C-8) and 135.7 (C-9) in 7 (Table 2) attributable to an olefinic functionality. One additional olefinic proton signal at δH 5.41 (dd, J = 1.6, 3.0 Hz, H-9) in the 1H NMR spectrum of 7 (Table 2) was located between C-8 and C-9 by key cross-peaks of H9/C-1, -7, -8, -10, -11, and -14 in the HMBC spectrum. After further interpretation of all of the two-dimensional NMR data, including COSY, HSQC, HMBC, and ROESY, the structure of 7 was thus determined. Compound 8 was obtained as a colorless oil, and its IR absorptions at 3412, 1722, and 1643 cm−1 indicated the presence of a hydroxy, a carbonyl, and a double bond, respectively. The UV λmax at 235 nm (log ε 4.4) suggested the presence of a conjugated olefinic functionality. When comparing 1H and 13C NMR data of 8 with those of 1 (Table 2), major differences observed included replacement of the OH-1 in 1 by a hydrogen at δH 2.67 (dd, J = 8.6, 11.4 Hz, H-1)/δC 62.8 (C-1) in 8, a ketone carbonyl at δC 206.0 (C-4) in 8 in place of the OH-4 in 1, and an additional O-methyl group at δH 3.23 (s, H3-16)/δC 52.3 positioned at C-6 in 8, as corroborated by key HMBC correlations of H-1/C-3, -6, -8, -9, and -11, H3-12/C-6, H2-13/C-4, and H3-16/C-6 (Figure 2). The relative configurations of β-oriented H-1 and OH-8 and αoriented H3-16 were deduced by key cross-peaks of H3-12/Hβ7 (δH 1.53), Hβ-7/OH-8, Hα-7 (δH 2.26)/H3-16, OH-8/H-1, OH-8/H3-14, and H-1/H3-12 in the ROESY spectrum (in DMSO-d6) (Figure 2). Hirsutanols A (9) and C and ent-gloeosteretriol, three hirsutane-type sesquiterpenes, were originally isolated from an unidentified marine sponge-derived fungus, and of these, hirsutanol A and ent-gloeosteretriol were found to exhibit antimicrobial activity against Bacillus subtilis previously.9 Hirsutanol E, with a cis−trans−cis-tricyclopentane ring system, was isolated for the first time from a soft coral-derived Chondrostereum sp. fungus.10 The spectroscopic data for chondrosterins A (10) and B (11) were compatible with published data, and chondrosterin A showed significant cytotoxicity against the cancer cell lines A549, CNE2, and LoVo.8 The structure of arthrosporone was determined by spectroscopic data when compared to those reported.11 Dysregulation of neuroinflammation plays an important role in the progression of neurodegenerative diseases.12 The microglial overproduction of nitric oxide (NO) is a pathological

condition in neuroinflammatory processes.13 The in vitro antineuroinflammatory activities of new compounds 1−8 and the known compounds were assessed by microglial NO production. As listed in Table 3, compounds 1, 2, 8, hirsutanol Table 3. Inhibitory Effects of Compounds 1−8, Hirsutanol A (9), Hirsutanol C, Hirsutanol E, Chondrosterin A (10), Chondrosterin B (11), Arthrosporone, and entGloeosteretriol in LPS-Induced NO Production in Microglial BV-2 Cells compound (20 μM)

NO (μM)a

resting vehicle (+LPS)b 1 2 3 4 5 6 7 8 hirsutanol A (9) hirsutanol C hirsutanol E chondrosterin A (10) chondrosterin B (11) arthrosporone ent-gloeosteretriol

5±2 15 ± 1c 8 ± 2e 10 ± 2d 11.6 ± 3 11 ± 1 12 ± 4 12 ± 2 14 ± 3 4 ± 1f 8 ± 2f 12 ± 2 13 ± 2 4 ± 1f 4 ± 1f 11 ± 3 13 ± 1

Values represent mean ± SD based on three individual experiments. The activation of BV-2 cells treated with LPS (150 ng/mL). cp < 0.001. dp < 0.05. ep < 0.01. fp < 0.001.

a b

A (9), and chondrosterins A (10) and B (11) exhibited significant inhibitory effect on lipopolysaccharide (LPS)induced NO production in BV-2 microglial cells at a concentration of 20 μM. According to the cytotoxic activity by MTT assay (Table 4), the new compound 8 and chondrosterins A (10) and B (11) exhibited potent cytotoxic effects on BV-2 cells at a concentration of 20 μM. The new compound 1 and hirsutanol A (9) (5−20 μM) had lesser Table 4. Effects of Compounds 1−8, Hirsutanol A (9), Hirsutanol C, Chondrosterin A (10), and Chondrosterin B (11) on Microglial Viabilitya viability (%) compound 1 2 3 4 5 6 7 8 hirsutanol A (9) hirsutanol C chondrosterin A (10) chondrosterin B (11)

2 μM 93 ± 95 ± ntb nt nt nt nt 47 ± 94 ± nt ntb nt

3 5

6d 1

5 μM 82 95 nt nt nt nt nt 30 87 nt nt nt

± 5d ±4

± 5d ± 3c

10 μM 73 89 nt nt nt nt nt 11 77 nt 22 23

± 4d ±9

± 2d ± 8d ± 3d ± 7d

20 μM 67 69 97 93 83 83 87 11 39 71 12 11

± ± ± ± ± ± ± ± ± ± ± ±

3d 5d 8 9 13 9 12 2d 5d 12c 1d 1d

Values are mean ± SD of three independent experiments. bnt: not tested. cp < 0.01. dp < 0.001. a

1619

DOI: 10.1021/acs.jnatprod.7b00196 J. Nat. Prod. 2017, 80, 1615−1622

Journal of Natural Products

Article

effects on cell viability. At the concentration of 20 μM, compounds 2−7 and hirsutanol C had little effect on viabilities. The results were consistent with the findings of Asai et al.;14 only hirsutanol A (9), with an α-exomethylene carbonyl moiety, was more potent than hirsutanol C on inducing cellular toxicity. It was reported that cysteine-targeting inhibitors possessing an α-exomethylene carbonyl group had anticellular activities.15 At a concentration of 20 μM, compound 8 and chondrosterin A (10) reduced cellular viability more than hirsutanol A (9). It was proposed that compound 8 and chondrosterin A (10), with OH-8 and an α-exomethylene carbonyl group, could enhance cell death. Interestingly, the similar structures of hirsutanol C and chondrosterin B (11) had a carbonyl group at C-4 but no conjugation with exomethylene. It revealed that only chondrosterin B (11)’s conjugated dicarbonyl (C-4−C-9) caused the markedly lower viability. As shown in Figure 3, new compound 1 significantly abrogated NO production in a concentration-dependent manner (2−20 μM) in microglial BV-2 cells. Compound 1 (20 μM) provided 47 ± 5% inhibition against NO production. The known compound hirsutanol A (9) also markedly inhibited NO production in a concentration-dependent manner. At a concentration of 5 μM, hirsutanol A (9) exerted 53 ± 6% inhibition of NO production with a cellular viability of 87 ± 3% (n = 3). Additionally, the IC50 value of curcumin, the control compound for inhibition of NO production, was determined to be 6.00 ± 0.29 μM (n = 3). The results clearly showed that compound 8, chondrosterin A (10), and chondrosterin B (11) (20 μM) all decreased the NO levels to the resting condition. At the same concentration, these compounds were also found to cause the lowest cellular viability. These results implied that cell death predominantly caused decreasing NO production. However, the decreased NO levels by compound 1 and hirsutanol A (9) (5 μM) was dependent on inhibition of NO production rather than cell death.

■

EXPERIMENTAL SECTION

General Experimental Procedures. Optical rotations were measured on a JASCO P-2000 polarimeter. UV spectra were measured on a Thermo Helios α spectrometer. IR spectra were recorded on a JASCO FT/IR 4100 spectrometer. 1H and 13C NMR were acquired on a Bruker DRX-500 SB and an AVIII-800 spectrometer. Low- and highresolution mass spectra were obtained using a VG Platform electrospray ESIMS and a high-definition mass spectrometry system with an ESI interface and a TOF analyzer (Waters Corp.), respectively. Diaion HP-20 (Mitsubishi Chemical, Tokyo, Japan) was used for open-column chromatography. A refractive index detector (Bischoff) was used for HPLC purifications. Fermentation of Chondrostereum sp. NTOU4196. Chondrostereum sp. (strain no. NTOU4196 from Keelung, Taiwan) was isolated from a sterilized seawater-washed thallus of the edible marine red alga Pterocladiella capillacea collected from the north coast of Taiwan and was identified by sequencing of the internal transcribed spacer regions of the rDNA (ITS). A BLAST search of the sequence (GenBank accession no. KY681773) led to the best match as Chondrostereum purpureum (sequence coverage 100%, sequence identity 84%). In a phylogenetic analysis of the ITS using the closely related sequences in the BLAST search and other phylogenetically related taxa obtained from published studies, NTOU4196 is phylogenetically related to C. purpureum and Gloeostereum incarnatum, but it is not monophyletic to both (results not shown). Therefore, NTOU4196 is designated as a Chondrostereum sp. The fungal strain was deposited at the Institute of Marine Biology, National Taiwan Ocean University, Keelung, Taiwan. The mycelium of Chondrostereum sp. NTOU4196 was inoculated into 5 L serum bottles, each containing PDY medium (peptone 6 g; dextrose 30 g; yeast 3 g) (Becton, Dickinson and Company, Sparks,

Figure 3. Effects of compound 1 (A) and hirsutanol A (B) on NO production in LPS-activated BV-2 cells. Data are shown as mean ± SD (n = 3). ###p < 0.001, compared with the resting group; *p < 0.05, **p < 0.01, ***p < 0.001, compared with the LPS- and vehicle-treated group. MD, USA) and 3 L of deionized water containing 20% natural seawater acquired from the northern coast of Taiwan. The fermentation was conducted with aeration at 25 °C for 20 days. Extraction and Isolation. The filtered fermented broth (15.0 L) of Chondrostereum sp. NTOU4196 was partitioned three times with 20 L of recycled EtOAc, then concentrated under vacuum to dryness (12.5 g). Subsequently, the extract was precoated onto 20.0 g of Diaion HP-20 gel, then applied onto a Diaion HP-20 column (2.8 cm i.d. × 25 cm) eluting with mixtures of H2O−MeOH in a stepwise gradient mode with a flow rate of 2.0 mL/min. Each fraction (1 L) was collected and evaporated to dryness. Fraction II, which eluted with 10% MeOH, was rechromatographed on a semipreparative reversedphase column (Phenomenex Luna 5 μm PFP, 10 × 250 mm) with MeCN−H2O (7:13, v/v) as eluent, 2 mL/min, to afford 1 (3.4 mg, tR = 13.06 min). Fraction IV, which eluted with 30% MeOH, was further 1620

DOI: 10.1021/acs.jnatprod.7b00196 J. Nat. Prod. 2017, 80, 1615−1622

Journal of Natural Products

Article

as previously described with some modification.18 The culture supernatants were reacted with reconstituted cofactor solution and reconstituted nitrate reductase solution for 1 h at room temperature in the dark according to the instructions of the nitrate/nitrite colorimetric assay kit (Cayman). Absorptions were measured at 550 nm using a microplate reader (MRX). Nitrite concentrations were calculated from the standard solutions of sodium nitrite. Curcumin was used as a positive control. Statistical Analysis. Data analyses were performed using the software SigmaStat 3.5 (SYSTAT Software). Data were expressed as the mean ± SD. The statistical analysis was performed using a one-way ANOVA and assessed using the Student−Newman−Keuls test. A pvalue of