Aromatase Inhibitory Activity of Geranylated Coumarins, Mammeasins C and D, Isolated from the Flowers of Mammea Siamensis

880 Chem. Pharm. Bull. 64, 880–885 (2016) Vol. 64, No. 7 Special Collection of Papers

This article is dedicated to Professor Satoshi Ōmura in celebration of his 2015 Nobel Prize.

Regular Article

Aromatase Inhibitory Activity of Geranylated Coumarins, Mammeasins C and D, Isolated from the Flowers of Mammea siamensis

Kiyofumi Ninomiya,a Kanae Shibatani,a Mayumi Sueyoshi,a Saowanee Chaipech,a,b Yutana Pongpiriyadacha,c Takao Hayakawa,a Osamu Muraoka,a and Toshio Morikawa*,a a Pharmaceutical Research and Technology Institute, Kindai University; 3–4–1 Kowakae, Higashi-osaka, Osaka 577–8502, Japan: b Faculty of Agro-Industry, Rajamangala University of Technology Srivijaya; Thungsong, Nakhonsithammarat 80110, Thailand: and c Faculty of Science and Technology, Rajamangala University of Technology Srivijaya; Thungsong, Nakhonsithammarat 80110, Thailand. Received March 2, 2016; accepted March 17, 2016

A methanol extract of the ﬂowers of Mammea siamensis (Calophyllaceae) was found to inhibit enzymatic

activity against aromatase (IC50 16.5 µg/mL). From the extract, two new geranylated coumarins, mammeasins C (1) and D (2), were isolated together with seven coumarins: 8-hydroxy-5-methyl-7-(3,7-dimethyl- octa-2,6-dienyl)-9-(2-methyl-1-oxobutyl)-4,5-dihydropyrano[4,3,2-de]chromen-2-one (9), 8-hydroxy-5-methyl-7-(3,7-dimethyl-octa-2,6-dienyl)-9-(3-methyl-1-oxobutyl)-4,5-dihydropyrano[4,3,2-de]chromen-2-one (10), mammeas A/AA (14), A/AB (15), A/AA cyclo D (18), E/BA (23), and E/BC cyclo D (25). The structures of 1 and 2 were elucidated on the basis of spectroscopic evidence. Among the isolates including 17 previously

reported coumarins, 1 (IC50 2.7 µM), 2 (3.6 µM), and mammea B/AB cyclo D (21, 3.1 µM) showed relatively strong inhibitory activities comparable to the activity of the synthetic nonsteroidal aromatase inhibitor aminoglutethimide (2.0 µM). Key words Mammea siamensis; mammeasin; aromatase inhibitor; geranylated coumarin; Calophyllaceae

Mammea siamensis (MIQ.) T. ANDERS. is a species of flower- estrogens and estrogen receptors are widely studied molecular ing plant in the Calophyllaceae family and is widely distrib- targets.28–30) The presence of high concentrations of estrogen uted in Thailand, Laos, Cambodia, Vietnam, and Myanmar. in breast tissue increases the risk of developing breast cancer The flowers of this plant have been used for preparing a heart and the ability of immature breast tissue cells to strongly bind tonic in Thai traditional medicine (“Sarapi” in Thai).1–10) to carcinogens, decreasing their DNA repair capacity.31,32) Several coumarins,1–7) xanthones,8,9) triterpenoids,10) and Aromatase, a CYP19 enzyme, is the rate-limiting enzyme steroids10) have been isolated from the flowers,1,2,6,7,10) seeds,3,9) in the conversion of testosterone and androstenedione to the twigs,4,8) and bark5) of this plant. In the course of our char- estrogens, estrone and estradiol.26–30,32–34) It is involved in the acterization studies on bioactive constituents in Thai natural final step of the estrogen biosynthetic pathway and its selec- medicine,1,11–25) we reported that the methanol extract of the tive inhibition will not affect the production of other steroids flowers of M. siamensis and its coumarin constituents showed in the pathway.32,35–37) The source of estrogen production in inhibitory effects on nitric oxide production in lipopolysac- breast cancer tissues is intratumoral aromatase, and thus, inhi- charide-activated RAW264.7 cells.1) Further studies revealed bition of aromatase may inhibit the growth stimulation effect that the methanol extract inhibited enzymatic activity against of estrogens in breast cancer tissues. Therefore, aromatase is aromatase. Separation of the active constituents in the extract considered a useful therapeutic target in the treatment and allowed us to isolate two new geranylated coumarins, mam- prevention of estrogen-dependent breast cancer.32) measins C (1) and D (2). This paper describes the isolation The dried flowers of M. siamensis (collected from Na- and structure elucidation of these new coumarins (1, 2) and khonsithammarat Province, Thailand) were extracted with the inhibitory effects of the coumarin constituents (1–26) on methanol under reflux (25.66% from the dried flowers). The 26) aromatase. methanol extract was partitioned into an EtOAc–H2O (1 : 1, v/v) mixture to furnish an EtOAc-soluble fraction (6.84%) Results and Discussion and an aqueous phase. The aqueous phase was subjected to Effects of the Methanol Extract from the Flowers of M. Diaion HP-20 column chromatography (H2O→MeOH) to give siamensis against Human Recombinant Aromatase Breast H2O- and MeOH-eluted fractions (13.50, 4.22%, respectively), cancer is one of the most common reasons for mortality in as described previously.1) As shown in Table 1, the methanol 26,27) women. Estrogens and estrogen receptors are well known extract had an inhibitory effect on aromatase (IC50=16.5 µg/ to play an important role in the development and progres- mL). A bioassay-guided fractionation revealed that the EtOAc- sion of hormone-dependent breast cancer; for this reason, soluble and MeOH-eluted fractions also showed aromatase in-

Table 1. Inhibitory Effects of the Methanol Extract from the Flowers of M. siamensis and Its Fractions against Human Recombinant Aromatase

Inhibition (%) IC50 (µg/mL) 3 µg/mL 10 µg/mL 30 µg/mL 100 µg/mL

MeOH Extract 13.2±3.2 41.8±1.6** 80.0±3.1** 97.5±0.6** 16.5 EtOAc-Soluble fraction 46.3±4.4** 89.0±1.2** 100.1±0.6** 99.9±0.5** 2.9 MeOH-Eluted fraction 2.0±3.9 70.7±1.2** 93.0±1.0** 96.9±0.9** 8.5

H2O-Eluted fraction −4.7±4.8 1.2±2.2 −1.7±2.1 −0.1±4.0 >100 Each value represents the mean±S.E.M. (N=3). Signiﬁcantly different from the control, ** p<0.01.

Fig. 1. Coumarin Constituents (1–26) from Flowers of M. siamensis

hibitory activities (IC50=2.9, 8.5 µg/mL, respectively), whereas dienyl)-9-(2-methyl-1-oxobutyl)-4,5-dihydropyrano[4,3,2- 4) the H2O-eluted fraction showed no noticeable activity. de]chromen-2-one (9, 0.0015%), 8-hydroxy-5-methyl-7- Isolation of Coumarin Constituents from the Methanol (3,7-dimethyl-octa-2,6-dienyl)-9-(3-methyl-1-oxobutyl)-4,5- Extract In our previous report we described the isola- dihydropyrano[4,3,2-de]chromen-2-one 4) (10, 0.0012%), mam- tion of 17 coumarins: mammeasins A (3, 0.0293%) and meas A/AA38) (14, 0.0494%), A/AB38) (15, 0.0048%), A/AA B (4, 0.0115%), surangins B (5, 0.0271%), C (6, 0.0571%), cyclo D38) (18, 0.0035%), E/BA39) (23, 0.0045%), and E/BC and D (7, 0.0632%), kayeassamins A (8, 0.0578%), E (11, cyclo D2) (25, 0.0058%), using normal-phase silica gel and 0.0113%), F (12, 0.0390%), and G (13, 0.0171%), mam- reversed-phase octadecylsilane (ODS) column chromatogra- measins A/AC (16, 0.1056%), A/AD (17, 0.0022%), A/AB phy, and ﬁnally, HPLC (Fig. 1). cyclo D (19, 0.0097%), A/AC cyclo D (20, 0.0109%), B/AB Structures of Mammeasins C (1) and D (2) Mammea - cyclo D (21, 0.0016%), B/AC cyclo D (22, 0.0062%), and E/ sin C (1) was obtained as pale yellow oil. Its IR spectrum BB (24, 0.0194%), and deacetylmammea E/BC cyclo D (26, showed absorption bands at 1748 and 1634 cm−1 assignable to 0.0073%), β-amyrin (0.0072%), and benzoic acid (0.0043%).1) an α,β-unsaturated γ-lactone moiety and a chelated aryl keto In the present study we additionally isolated two new gera- group.1,2,4) The UV spectrum exhibited absorption maxima nylated coumarins, mammeasins C (1, 0.0008%) and D (2, at 221, 292, and 328 nm, similar to those of 5,7-dioxygenated 0.0047%), from the active EtOAc-soluble fraction together with coumarins.1,2,4) The electron ionization (EI)-MS spectrum of 1 seven coumarins: 8-hydroxy-5-methyl-7-(3,7-dimethyl-octa-2,6- showed a molecular ion peak at m/z 424 (M+), and the molecu- 882 Chem. Pharm. Bull. Vol. 64, No. 7 (2016) lar formula was determined as C26H32O5 by high-resolution of 1 were quite similar to those of 9 and 10, except for the sig- (HR)-EI-MS measurement. The 1H- and 13C-NMR spectra of 1 nals due to the 1-oxo-alkyl moiety.4) The 1H–1H COSY experi-

(Tables 2, 3, CDCl3) were assigned with the aid of distortion- ment on 1 indicated the presence of partial structures, as in- less enhancement by polarization transfer (DEPT), 1H–1H cor- dicated by the bold lines in Fig. 2. In the HMBC experiment, relation spectroscopy (COSY), 1H-detected heteronuclear mul- long-range correlations were observed between the following tiple quantum coherence (HMQC), and heteronuclear multiple proton and carbon pairs: 3-H and 2,4,6b-C; 4-H and 3,6b-C; bond connectivity (HMBC) experiments (Fig. 2). The spectra 1″-H2 and 6a,7,8-C; 2″-H and 7,3″,5″-C; 4″-H2 and 3″-C; 5″-H3 showed signals assignable to three secondary and three vinyl and 2″–4″-C; 7″-H2 and 9″,10″-C; 9″-H3 and 7″,8″,10″-C; 10″-H3 methyls [δ 1.26, 1.27 (3H each, both d, J=6.6 Hz, 3‴-H3, and 7″–9″-C; 2‴-H and 1‴-C. On the basis of comprehensive 4‴-H3), 1.54 (3H, d, J=6.2 Hz, 1′-H3), 1.57 (3H, s, 10″-H3), 1.63 two dimensional (2D)-NMR experiments, we assigned the (3H, d, J=0.7 Hz, 9″-H3), 1.78 (3H, s, 5″-H3)]; four methylenes structure of 1 as 8-hydroxy-5-methyl-7-(3,7-dimethyl-octa-2,6- {δ 1.96 (2H, m, 4″-H2), 2.05 (2H, m, 6″-H2), [2.78 (1H, ddd, dienyl)-9-(2-methyl-1-oxopropyl)-4,5-dihydropyrano[4,3,2-de]- J=1.4, 11.0, 17.2 Hz), 2.91 (1H, dd, J=2.6, 17.2 Hz), 4-H2], chromen-2-one. 3.35 (2H, d, J=7.2 Hz, 1″-H2)}; two methines [δ 4.01 (1H, qq, Mammeasin D (2) was also isolated as pale yellow oil. Its J=6.6, 6.6 Hz, 2‴-H), 4.36 (1H, m, 5-H)]; and two oleﬁnic pro- molecular formula, C26H32O5, was found to be the same as tons [δ 5.06 (1H, qt, J=0.7, 6.9 Hz, 7″-H), 5.20 (1H, br t, J=ca. that of 1 by HR-EI-MS measurement. The 1H- and 13C-NMR 1 13 7 Hz, 2″-H)]. The H- and C-NMR spectroscopic properties spectroscopic properties (Tables 2, 3, CDCl3) of 2 were simi-

1 Table 2. H-NMR (500 MHz, CDCl3) Data for Mammeasins C (1) and D (2)

1 2 Position δH (J Hz) δH (J Hz) 3 5.94 (1H, br s) 5.93 (1H, br s) 4 2.78 (1H, ddd, 1.4, 11.0, 17.2) 2.78 (1H, ddd, 1.5, 10.9, 16.9) 2.91 (1H, dd, 2.6, 17.2) 2.90 (1H, dd, 2.6, 16.9) 5 4.36 (1H, m) 4.36 (1H, m) 1′ 1.54 (3H, d, 6.2) 1.54 (3H, d, 6.3) 1″ 3.35 (2H, d, 7.2) 3.34 (2H, d, 6.9) 2″ 5.20 (1H, br t, ca. 7) 5.19 (2H, br t, ca. 7) 4″ 1.96 (2H, m) 1.96 (2H, m) 5″ 1.78 (3H, s) 1.78 (3H, s) 6″ 2.05 (2H, m) 2.05 (2H, m) 7″ 5.06 (1H, qt, 0.7, 6.9) 5.05 (1H, qt, 0.9, 6.9) 9″ 1.63 (3H, d, 0.7) 1.63 (3H, d, 0.9) 10″ 1.57 (3H, s) 1.57 (3H, s) 2‴ 4.01 (1H, qq, 6.6, 6.6) 3.26 (2H, br t, ca. 7) 3‴ 1.26 (3H, d, 6.6)a) 1.79 (2H, m) 4‴ 1.27 (3H, d, 6.6)a) 1.05 (3H, t, 7.5) 8-OH 14.55 (1H, s) 14.51 (1H, s) a) Assignments may be interchangeable within the same column.

13 Table 3. C-NMR (125 MHz, CDCl3) Data for Mammeasins C (1) and D (2)

1 2 1 2 Position Position δC δC δC δC 2 159.8 159.7 1″ 21.4 21.3 3 105.7 105.8 2″ 121.2 121.2 3a 149.2 149.1 3″ 135.8 135.8 4 35.1 35.0 4″ 39.8 39.8 5 72.6 72.7 5″ 16.1 16.1 6a 156.6 156.6 6″ 26.7 26.7 6b 99.6 99.6 7″ 124.3 124.3 7 113.1 113.1 8″ 131.3 131.1 8 167.6 167.2 9″ 25.6 25.7 9 103.2 104.0 10″ 17.7 17.7 9a 154.3 154.6 1‴ 210.3 205.8 1′ 20.7 20.7 2‴ 40.1 46.3 3‴ 19.15a) 18.0 4‴ 19.21a) 13.8 a) Assignments may be interchangeable within the same column. Vol. 64, No. 7 (2016) Chem. Pharm. Bull. 883

Fig. 2. 1H–1H COSY and HMBC Correlations of 1 and 2

Table 4. Inhibitory Effects of Constituents from the Flowers of M. siamensis and Related Compounds against Human Recombinant Aromatase

IC50 (µM) IC50 (µM) Mammeasin C (1) 2.7 Mammea A/AA cyclo D (18) 7.2 Mammeasin D (2) 3.6 Mammea A/AB cyclo D (19) 24.1 Mammeasin A (3) 8.7 Mammea A/AC cyclo D (20) 35.0 Mammeasin B (4) 4.1 Mammea B/AB cyclo D (21) 3.1 Surangin B (5) 9.8 Mammea B/AC cyclo D (22) 24.6 Surangin C (6) 8.8 Mammea E/BA (23) 16.6 Surangin D (7) 18.1 Mammea E/BB (24) 18.6 Kayeassamin A (8) 10.0 Mammea E/BC cyclo D (25) 11.5 9 7.5 Deacetylmammea E/BC cyclo D (26) 16.6 10 8.8 β-Amyrin >100 (24.4)a) Kayeassamin E (11) 14.9 Benzoic acid >100 (−3.4)a) Kayeassamin F (12) 19.7 Kayeassamin G (13) 27.8 Umbelliferone >100 (2.6)a) Mammea A/AA (14) 6.9 Scopoletin >100 (−15.9)a) Mammea A/AB (15) 8.6 4-Hydroxycoumarin >100 (10.9)a) Mammea A/AC (16) 13.7 Mammea A/AD (17) 11.3 Aminoglutethimide 2.0

Each value represents the mean±S.E.M. (N=3). a) Values in parentheses present inhibition % at 100 µM. lar to those of 1, except for the signals due to an 1-oxobutyl of the coumarin skeleton is essential for the inhibitory activ- moiety in the 9-position [δ 1.05 (3H, t, J=7.5 Hz, 4‴-H3), 1.79 ity. In particular, mammeasins C (1, 2.7 µM) and D (2, 3.6 µM), (2H, m, 3‴-H2), 3.26 (2H, br t, J=ca. 7 Hz, 2‴-H2)] instead of and mammea B/AB cyclo D (21, 3.1 µM) show relatively potent 40,41) the 2-methyl-1-oxopropyl moiety of 1. As shown in Fig. 2, activity, comparable to that of aminoglutethimide (2.0 µM). the connectivities of the quaternary carbons in 2 were eluci- Therefore, these coumarins may be useful in the treatment of dated on the basis of 1H–1H COSY and HMBC experiments. hormone-dependent breast cancer. Detailed structural require- Thus, the structure of 2 was elucidated to be 8-hydroxy-5- ments of coumarins for aromatase inhibitory activity and the methyl-7-(3,7-dimethyl-octa-2,6-dienyl)-9-(1-oxobutyl)-4,5- mechanism of action, however, need further exploration. dihydropyrano[4,3,2-de] chromen-2-one. Possible biogenetic pathway for the formation of the pyran ring in 9 and 10, Experimental having the same moiety with those of 1 and 2, from 6 and 7 The following instruments were used to obtain physical have been reported previously.4) These new compounds (1, 2) data: UV spectra, UV-1600 spectrometer (Shimadzu Co., might be derived through the same pathway. Further studies, Kyoto, Japan); IR spectra, FTIR-8100 spectrometer (Shimadzu e.g. total syntheses of 1 and 2, would be needed to elucidate Co.); EI-MS and HR-EI-MS, JMS-GCMATE mass spec- the absolute stereochemistry and to verify whether these were trometer (JEOL Ltd., Tokyo, Japan); 1H-NMR spectra, JNM- the artifacts. ECA500 (500 MHz), and JNM-ECS400 (400 MHz) spectrom- Effects of Coumarin Constituents of the Flowers of M. eters (JEOL); 13C-NMR spectra, JNM-ECA500 (125 MHz), siamensis and Related Compounds on Human Recombi- and JNM-ECS400 (100 MHz) spectrometers (JEOL Ltd.) with nant Aromatase To characterize the active constituents tetramethylsilane as an internal standard; HPLC detector, of this plant material, the inhibitory effects of 28 isolates SPD-10Avp UV-VIS detector (Shimadzu Co.); HPLC column, including 26 coumarins (1–26) against aromatase were exam- Cosmosil 5C18-MS-II (Nacalai Tesque, Inc., Kyoto, Japan), ined. As shown in Table 4, the coumarin constituents (1–26, 4.6×250 mm i.d. and 20×250 mm i.d. for analytical and pre-

IC50=2.7–35.0 µM) possess inhibitory activity, whereas com- parative studies, respectively. mercially available coumarins such as umbelliferone, scopole- The following experimental conditions were used for chro- tin, and 4-hydroxycoumarin do not. These results suggest that matography (CC): ordinary-phase silica gel column chro- the presence of the side chain at the 4-, 6-, and/or 8-positions matography, silica gel 60N (Kanto Chemical Co., Tokyo, 884 Chem. Pharm. Bull. Vol. 64, No. 7 (2016)

Japan; 63–210 mesh, spherical, neutral); reverse-phase silica by HPLC [Cosmosil 5C18-MS-II, MeOH–1% aqueous AcOH gel CC, Diaion HP-20 (Nippon Rensui, Tokyo, Japan) and (90 : 10, v/v)] to give mammeas A/AA (14, 17.0 mg, 0.0076%) Chromatorex ODS DM1020T (Fuji Silysia Chemical, Aichi, and A/AB (15, 10.7 mg, 0.0048%) together with mammeasins Japan; 100–200 mesh); normal-phase TLC, pre-coated TLC A (3, 65.8 mg, 0.0293%) and B (4, 21.6 mg, 0.0096%), suran- plates with silica gel 60F254 (Merck, Darmstadt, Germany; gin B (5, 58.2 mg, 0.0259%), and 16 (112.6 mg, 0.0501%) as 0.25 mm); reversed-phase TLC, pre-coated TLC plates with reported previously.1) Fraction 6-5 (128.8 mg) was puriﬁed silica gel RP-18 F254S (Merck, 0.25 mm); reversed-phase by HPLC [Cosmosil 5C18-MS-II, MeOH–1% aqueous AcOH HPTLC, pre-coated TLC plates with silica gel RP-18 WF254S (90 : 10, v/v)] to give 4 (24.1 mg, 0.0019%) and 5 (15.1 mg, (Merck, 0.25 mm), detection was carried out by spraying 1% 0.0012%) as reported previously.1) Fraction 6-6 (487.1 mg) was

Ce(SO4)2–10% aqueous H2SO4 on the plates, followed by heat- purified by HPLC [Cosmosil 5C18-MS-II, MeOH–1% aque- ing. ous AcOH (90 : 10, v/v)] to give mammeasins C (1, 10.4 mg, Plant Material The flowers of Mammea siamensis were 0.0008%) and D (2, 60.9 mg, 0.0047%), 8-hydroxy-5-methyl-7- collected from Nakhonsithammarat Province, Thailand, in (3,7-dimethyl-octa-2,6-dienyl)-9-(2-methyl-1-oxobutyl)-4,5- September 2006, as described previously.1) The plant material dihydropyrano[4,3,2-de]chromen-2-one (9, 19.8 mg, 0.0015%), was identified by one of the authors (Y.P.). A voucher speci- and 8-hydroxy-5-methyl-7-(3,7-dimethyl-octa-2,6-dienyl)-9-(3- men (2006.09. Raj-04) for this plant has been deposited in our methyl-1-oxobutyl)-4,5-dihydropyrano[4,3,2-de]chromen-2-one laboratory. (10, 16.3 mg, 0.0012%) together with 16 (6.0 mg, 0.00050%) as Extraction and Isolation Dried flowers of M. siamensis reported previously.1) (1.8 kg) were extracted three times with MeOH under reflux Mammeasin C (1) + for 3 h. Evaporation of the combined extracts under reduced Pale yellow oil. HR-EI-MS: Calcd for C26H32O5 (M ): pressure afforded the MeOH extract (463.7 g, 25.66%). An 424.2250. Found: 424.2243. UV [MeOH, nm (log ε)]: 221 aliquot (413.7 g) of the extract was partitioned into an EtOAc– (4.32), 292 (4.22), 328 (4.05). IR (film): 1748, 1717, 1634, 1601, −1 1 H2O (1 : 1, v/v) mixture to furnish an EtOAc-soluble fraction 1456, 1404, 1385, 1237, 1194, 1171, 1132, 1109 cm . H-NMR 13 (110.34 g, 6.84%) and an aqueous phase. The aqueous phase (500 MHz, CDCl3) δ: see Table 2. C-NMR data (125 MHz, + was subjected to Diaion HP-20 CC (2.4 kg, H2O→MeOH, CDCl3) δC: see Table 3. EI-MS m/z: 424 (M , 37), 301 (100). twice) to give H2O-eluted (217.70 g, 13.50%) and MeOH-eluted Mammeasin D (2) + (68.10 g, 4.22%) fractions, respectively. An aliquot (89.45 g) Pale yellow oil. HR-EI-MS: Calcd for C26H32O5 (M ): of the EtOAc-soluble fraction was subjected to normal-phase 424.2250. Found: 424.2243. UV [MeOH, nm (log ε)]: 221 silica gel CC [3.0 kg, n-hexane–EtOAc (10 : 1→7 : 1→5 : 1, (4.39), 292 (4.29), 324 (4.09). IR (film): 1734, 1717, 1636, 1617, v/v)→EtOAc→MeOH] to give 11 fractions [Fr. 1 (3.05 g), Fr. 2 1456, 1387, 1235, 1192, 1115, 1049 cm−1. 1H-NMR (500 MHz, 13 (2.86 g), Fr. 3 (11.71 g), Fr. 4 (1.62 g), Fr. 5 (4.15 g), Fr. 6 (6.29 g), CDCl3) δ: see Table 3. C-NMR data (125 MHz, CDCl3) δC: Fr. 7 (2.21 g), Fr. 8 (2.94 g), Fr. 9 (10.23 g), Fr. 10 (11.17 g), and see Table 3. EI-MS m/z: 424 (M+, 30), 301 (100). Fr. 11 (21.35 g)] as reported previously.1) Fraction 5 (4.15 g) was Bioassay subjected to reversed-phase silica gel CC [120 g, MeOH–H2O Reagents (80 : 20→85 : 15, v/v)→MeOH→acetone] to afford six frac- Dibenzylfluorescein (DBF) and Human CYP19 + P450 Re- tions [fr. 5-1 (115.7 mg), fr. 5-2 (2789.8 mg), fr. 5-3 (515.4 mg), ductase SUPERSOMES (human recombinant aromatase) were fr. 5-4 (430.0 mg), fr. 5-5 (119.2 mg), and fr. 5-6 (110.0 mg)]. purchased from BD Biosciences (Heidelberg, Germany). The Fraction 5-2 (517.0 mg) was purified by HPLC [Cosmosil other chemicals used in this study were purchased from Wako

5C18-MS-II, MeOH–1% aqueous AcOH (85 : 15, v/v)] to give Pure Chemical Industries, Ltd. (Osaka, Japan). mammeas A/AA (14, 101.2 mg, 0.0418%), A/AC (16, 112.9 mg, Inhibitory Effects against Human Recombinant Aro- 0.0466%), A/AA cyclo D (18, 2.7 mg, 0.0011%), and E/BC matase The experiments were performed according to the cyclo D (25, 14.0 mg, 0.0058%). Fraction 5-3 (515.4 mg) was method described previously but with a slight modiﬁcation.42) puriﬁed by HPLC [Cosmosil 5C18-MS-II, MeOH–1% aque- A test sample was dissolved in dimethyl sulfoxide (DMSO) ous AcOH (85 : 15, v/v)] to give 16 (45.6 mg, 0.0035%), 18 and the solution was diluted with potassium phosphate buffer

(14.9 mg, 0.0011%), mammeas A/AB cyclo D (19, 46.4 mg, (50 m M, pH 7.4) containing MgCl2 (0.5 mM) to afford the test 0.0035%), and A/AC cyclo D (20, 30.1 mg, 0.0023%). Fraction sample solution (concentration of DMSO: 2%). An enzyme/

5-4 (430.0 mg) was puriﬁed by HPLC [Cosmosil 5C18-MS-II, substrate solution in the buffer (20 µL, 1.6 µM DBF, 8 nM MeOH–1% aqueous AcOH (90 : 10, v/v)] to give 18 (17.4 mg, human recombinant aromatase) and the test sample solution 0.0013%), 19 (19.1 mg, 0.0015%), 20 (11.5 mg, 0.0009%), and (20 µL) were mixed into a 96-well half area black microplate mammea B/AC cyclo D (22, 9.5 mg, 0.0007%). Fraction 6 (Greiner Bio-One, Frickenhausen, Germany) at 37°C for (6.29 g) was subjected to reversed-phase silica gel CC [200 g, 10 min. The enzymatic reaction was initiated by the addi-

MeOH–H2O (80 : 20→90 : 10→95 : 5, v/v)→MeOH→acetone] tion of reduced nicotinamide adenine dinucleotide phosphate to afford 10 fractions [fr. 6-1 (44.7 mg), fr. 6-2 (157.2 mg), (NADPH) solution (40 µL, 500 µM) at 37°C for 30 min. After fr. 6-3 (928.8 mg), fr. 6-4 (3117.0 mg), fr. 6-5 (128.8 mg), fr. 30 min incubation, NaOH (30 µL, 2 mM) was added, and the 6-6 (487.1 mg), fr. 6-7 (230.8 mg), fr. 6-8 (280.5 mg), fr. 6-9 reaction mixture was incubated at 37°C for 2 h to induce the (102.9 mg), and Fr. 6-10 (96.5 mg)]. Fraction 6-3 (514.6 mg) was fluorescent signals. Fluorescence was measured using a fluo- purified by HPLC [Cosmosil 5C18-MS-II, MeOH–1% aque- rescence microplate reader (SH-9000, CORONA) at excitation ous AcOH (80 : 20, v/v)] to give mammea E/BA (23, 32.7 mg, wavelength of 485 nm and emission wavelength of 535 nm.

0.0045%) together with 16 (35.6 mg, 0.0049%), mammeas A/ Experiments were performed in triplicate, and IC50 values AD (17, 15.8 mg, 0.0022%) and E/BB (24, 140.1 mg, 0.0194%) were determined graphically. The aromatase inhibitor amino- as reported previously.1) Fraction 6-4 (536.2 mg) was puriﬁed glutethimide was used as a reference compound. Vol. 64, No. 7 (2016) Chem. Pharm. Bull. 885

Statistics Values are expressed as the mean±standard yadacha Y., Hayakawa T., Muraoka O., Bioorg. Med. Chem., 20, error of the mean (S.E.M.). One-way ANOVA, followed by 832–840 (2012). Dunnett’s test, was used for statistical analysis. Probability (p) 18) Chaipech S., Morikawa T., Ninomiya K., Yoshikawa M., Pongpiri- yadacha Y., Hayakawa T., Muraoka O., J. Nat. Med., 66, 486–492 values less than 0.05 were considered significant. (2012). 19) Morikawa T., Chaipech S., Matsuda H., Hamao M., Umeda Y., Sato Acknowledgments This study was supported in part by H., Tamura H., Ninomiya K., Yoshikawa M., Pongpiriyadacha Y., the MEXT-Supported Program for the Strategic Research Hayakawa T., Muraoka O., J. Nat. Med., 66, 516–524 (2012). Foundation at Private Universities, 2014–2018 (T.M.), and 20) Tanabe G., Nakamura S., Tsutsui N., Balakishan G., Xie W., Tsuchi- a Grant-in-Aid for Scientific Research [KAKENHI, Grant ya S., Akaki J., Morikawa T., Ninomiya K., Nakanishi I., Yoshi- Numbers 15K08008 (T.M.), 15K08009 (K.N.), and 16K0813 kawa M., Muraoka O., Chem. Commun., 48, 8646–8648 (2012). (O.M.)]. The authors also acknowledge the financial support 21) Xie W., Tanabe G., Xu J., Wu X., Morikawa T., Yoshikawa M., extended by Kobayashi International Scholarship Foundation Muraoka O., Mini Rev. Org. Chem., 10, 141–159 (2013). (T.M.). 22) Akaki J., Morikawa T., Miyake S., Ninomiya K., Okada M., Tanabe G., Pongpiriyadacha Y., Yoshikawa M., Muraoka O., Phytochem. Anal., 25, 544–550 (2014). Conflict of Interest The authors declare no conflict of 23) Morikawa T., Akaki J., Ninomiya K., Kinouchi E., Tanabe G., interest. Pongpiriyadacha Y., Yoshikawa M., Muraoka O., Nutrients, 7, 1480–1493 (2015). References 24) Nakamura K., Akaki J., Ishibushi F., Tani K., Morikawa T., 1) Morikawa T., Sueyoshi M., Chaipech S., Matsuda H., Nomura Y., Pongpiriyadacha Y., Muraoka O., Hayakawa T., Kakutani K., Yabe M., Matsumoto T., Ninomiya K., Yoshikawa M., Pongpiri- Shoyakugaku Zasshi, 69, 53–58 (2015). yadacha Y., Hayakawa T., Muraoka O., Bioorg. Med. Chem., 20, 25) Ninomiya K., Matsumoto T., Chaipech S., Miyake S., Katsuyama 4968–4977 (2012). Y., Tsuboyama A., Pongpiriyadacha Y., Hayakawa T., Muraoka O., 2) Mahidol C., Kaweetripob W., Prawat H., Ruchirawat S., J. Nat. Morikawa T., J. Nat. Med., 70, 179–189 (2016). Prod., 65, 757–760 (2002). 26) Benson J. R., Ravisekar O., Curr. Cancer Ther. Rev., 3, 67–79 3) Laphookhieo S., Maneerat W., Kiattansakul R., Can. J. Chem., 84, (2007). 1546–1549 (2006). 27) Monteiro R., Faria A., Azevedo I., Calhau C., J. Steroid Biochem. 4) Prachyawarakorn V., Mahidol C., Ruchirawat S., Phytochemistry, Mol. Biol., 105, 124–130 (2007). 67, 924–928 (2006). 28) Osborne C. K., Schiff R., Fuqua S. A., Shou J., Clin. Cancer Res., 7 5) Ngo N. T. N., Nguyen V. T., Vo H. V., Vang O., Duus F., Ho T.-D. (Suppl.), 4338s–4342s, discussion, 4411s–4412s (2001). H., Pham H. D., Nguyen L.-H. D., Chem. Pharm. Bull., 58, 1487– 29) Balunas M. J., Su B., Brueggemeier R. W., Kinghorn A. D., J. Nat. 1491 (2010). Prod., 71, 1161–1166 (2008). 6) Kaweetripob W., Mahidol C., Prawat H., Ruchirawat S., Pharm. 30) Balunas M. J., Kinghorn A. D., Planta Med., 76, 1087–1093 (2010). Biol., 38 (Suppl. 1), 55–57 (2000). 31) Gray J., Evans N., Taylor B., Rizzo J., Walker M., Int. J. Occup. 7) Mahidol C., Prawat H., Kaweetripob W., Ruchirawat S., Nat. Prod. Environ. Health, 15, 43–78 (2009). Commun., 2, 557–564 (2007). 32) Ahmad I., Shagufta, Eur. J. Med. Chem., 102, 375–386 (2015). 8) Poobrasert O., Constant H. L., Beecher C. W. W., Farnsworth N. 33) Kuller L. H., Matthews K. A., Meilahn E. N., J. Steroid Biochem. R., Kinghorn A. D., Pezzuto J. M., Cordell G. A., Santisuk T., Reu- Mol. Biol., 74, 297–309 (2000). trakul V., Phytochemistry, 47, 1661–1663 (1998). 34) Simpson E. R., Clyne C., Rubin G., Boon W. C., Robertson K., Britt 9) Laphookhieo S., Promnart P., Syers J. K., Kanjana-Opas A., Pongli- K., Speed C., Jones M., Annu. Rev. Physiol., 64, 93–127 (2002). manont C., Karalai C., J. Braz. Chem. Soc., 18, 1077–1080 (2007). 35) Muftuoglu Y., Mustata G., Bioorg. Med. Chem. Lett., 20, 3050– 10) Subhadhirasakul S., Pechpongs P. Songklanakarin J., Sci. Tech., 27 3064 (2010). (Suppl. 2), 555–561 (2005). 36) Mokbel K., Int. J. Clin. Oncol., 7, 279–283 (2002). 11) Muraoka O., Morikawa T., Miyake S., Akaki J., Ninomiya K., 37) Maiti A., Cuendet M., Croy V. L., Endringer D. C., Pezzuto J. M., Yoshikawa M., J. Pharm. Biomed. Anal., 52, 770–773 (2010). Cushman M., J. Med. Chem., 50, 2799–2806 (2007). 12) Morikawa T., Yakugaku Zasshi, 130, 785–791 (2010). 38) Verotta L., Lovaglio E., Vidari G., Finzi P. V., Neri M. G., Rai- 13) Morikawa T., Xie Y., Ninomiya K., Okamoto M., Muraoka O., mondi A., Parapini S., Taramelli D., Riva A., Bombardelli E., Phy- Yuan D., Yoshikawa M., Hayakawa T., Chem. Pharm. Bull., 58, tochemistry, 65, 2867–2879 (2004). 1276–1280 (2010). 39) Yang H., Protiva P., Gil R. R., Jiang B., Baggett S., Basile M. J., 14) Muraoka O., Morikawa T., Miyake S., Akaki J., Ninomiya K., Pong- Reynertson K. A., Weinstein I. B., Kennelly E. J., Planta Med., 71, piriyadacha Y., Yoshikawa M., J. Nat. Med., 65, 142–148 (2011). 852–860 (2005). 15) Xie W., Tanabe G., Akaki J., Morikawa T., Ninomiya K., Mine- 40) Yue W., Brodie A. M., J. Steroid Biochem. Mol. Biol., 63, 317–328 matsu T., Yoshikawa M., Wu X., Muraoka O., Bioorg. Med. Chem., (1997). 19, 2015–2022 (2011). 41) Martínez-Campa C., González A., Mediavilla M. D., Alonso- 16) Chaipech S., Morikawa T., Ninomiya K., Yoshikawa M., Pongpiri- González C., Sánchez-Barceló E. J., Cos S., Breast Cancer Res. yadacha Y., Hayakawa T., Muraoka O., Chem. Pharm. Bull., 60, Treat., 94, 249–254 (2005). 62–69 (2012). 42) Jacobsen N. W., Halling-Sørensen B., Birkved F. K., Toxicol. In 17) Morikawa T., Chaipech S., Matsuda H., Hamao M., Umeda Y., Sato Vitro, 22, 146–153 (2008). H., Tamura H., Kon'i H., Ninomiya K., Yoshikawa M., Pongpiri-