Phytochemicals and Estrogen-Receptor Agonists from the Aerial Parts of Liriope platyphylla

One new benzofuran, (2R)-(2',4'-dihydroxybenzyl)-6,7-methylenedioxy-2,3-dihydrobenzofuran (1), one new phenylisocoumarin, 3-(2'-hydroxyphenyl)-6,8-dihydroxy-7-methoxy-isocoumarin (2), and one new benzofuroisocoumarin, platyphyllarin C (3), were isolated from the ethanolic extract of Liriope platyphylla aerial parts, along with seventeen known compounds. The structures of the isolates were established by spectroscopic analysis and comparison with the literature data. The results indicated that structures 1–3 are uncommon in Nature. Benzofuroisocoumarin 4, flavonoids 9, 10, and 13–15, and homoisoflavonoids 19 and 20 exhibited significant binding activity to estrogen-receptor α and/or β as demonstrated by the SEAP reporter assay system in an MCF-7 cell-line.

Recently, we reported that certain components isolated from L. platyphylla roots showed significant anti-platelet and estrogenic effects [12]. In our continuing investigation to search for novel phytoestrogenic resources, different plants used in Asian folk medicines were assayed for their estrogenic activity. Among the examined plants, the ethanolic extract of L. platyphylla aerial parts was found to be significantly active in the transgenic Arabidopsis pER8:GUS assay [12] (Arabidopsis pER8:GUS possessing a human estrogen ERα receptor activating assay system) at 12.5 μg/mL. Therefore, the extract was selected for further phytochemical and pharmacological investigation. In the current investigation, the isolation and structure elucidation of secondary metabolites isolated from L. platyphylla aerial parts were discussed. The binding activity of the isolates to the estrogen-receptors (ER) α and/or β in a SEAP reporter assay system using the MCF-7 cell-line was also investigated.

Structural Elucidation of the Isolated Compounds
The ethanolic extract of L. platyphylla aerial parts was selected for further investigation and subsequently partitioned with n-hexane, ethyl acetate (EtOAc), n-BuOH, and H2O to yield four fractions.
Compound 3 was isolated as a white amorphous powder. The molecular formula was calculated as C17H12O6 from the analysis of its HRESIMS (m/z 335.05254 [M + Na] + , calcd. for 335.05261), indicating 12 degrees of unsaturation. The IR spectrum indicated the presence of hydroxy (3350 cm −1 ), carbonyl (1708 cm −1 ), and aromatic (1594 cm −1 ) functionalities. From the 13 C-NMR and DEPT spectra, two methoxy groups, five methines, six oxygenated aromatic carbons, one ester carbon, and three aromatic/olefinic quaternary carbons were observed. The NMR data of 3 suggested a double bond, two aromatic rings, and an ester functionality in the skeleton, similar to the functional groups in 2.  indicated the presence of a 1,2-disubstituted aromatic ring. Moreover, the one degree of unsaturation more than 2 and the HMBC correlations between H-9/C-3 and C-12a and H-12/C-9a (Figure 2), suggested the presence of a benzofuran moiety. This proposal was further confirmed by comparing the NMR data of 3 with the reported literature [27]. Through connecting the two moieties, the scaffold of 3 was established as a benzofuroisocoumarin [28]. The position of the methoxy groups (δH 3.94 and 3.97) were confirmed by the HMBC correlations to C-7 and C-8, respectively, as well as by the absence of any NOESY cross peak to H-5. Based on the aforementioned discussion, compound 3 was established as 6-hydroxy-7,8-dimethoxy-benzofuroisocoumarin, and named platyphyllarin C. 2-Benzylbenzofuran [5], phenylisocoumarin [29], and benzofuroisocoumarin [12] structures are uncommon in Nature. 2-Benzyl-2,3-dihydrobenzofuran structures such as compound 1, have never been reported from natural sources. Compound 2 is only the second example of a natural 3-phenyl-isocoumarin and 3 is the third example of a natural benzofuroisocoumarin.

The Binding Activity of the Isolates to ERα and/or β
Due to the limitated sample availability of some of the isolates, compounds 3, 4, 9, 10, 13-15, 19, and 20 were selected to test their binding potential to the estrogen-receptors α and β at several concentrations (1 nM-100 μM). The basic medium without the addition of the tested compounds (basal) and 17β-estradiol (E2, 100 nM) were used as the blank and positive control, respectively. Cell viability was determined by MTT assay, and the estrogenic activity was evaluated using the SEAP reporter assay system. Both assays were expressed as a percentage based on the basal control, which SEAP% and MTT% of blank were defined as 100% (Figures 3 and 4). For the SEAP% of ERα, compounds 4, 9, 10, 13, and 15 displayed significant binding activity. After excluding the false results from the cell survival ratio, 3-O-methylquercetin (10) and 6-C-methylquercetin-3-methyl ether (13) exhibited the most potent effect in a dose-dependent manner.

Plant Material
The plant material was collected in July 2012 from Taichung County, Taiwan, and was identified by a specialist in herbal medicine, Dr. Ming-Hong Yen. A voucher specimen (LP011) was deposited at the Graduate Institute of Natural Products, Kaohsiung Medical University, Kaohsiung, Taiwan. The fresh aerial parts were crushed using a mill (Rong Tsong, Taichung City, Taiwan) before extraction.

Bioassay Procedure-Estrogenic Activity
The estrogenic activity was determined through adapting the method described by Lai et al. (2011) [30]. The assay was done using SEAP (secreted alkaline phosphatase) reporter gene assay system. Human breast adenocarcinoma cells MCF-7 obtained from Bioresource Collection and Research Center were cultured in phenol-red free minimum essential medium Eagle (MEM) supplemented with dextran-charcoal treated serum, 2 mM L-glutamine, 10% foetal bovine serum (GIBCO BRL, Gaithersburg, MD, USA), penicillin, and streptomycin. The cells were incubated in an atmosphere of 5% CO2 at 37 °C; 0.2 µg of pERE-TA-SEAP plasmid (Clontech, Palo Alto, CA, USA) were transfected into 2 × 104 cells in 100 µL of growth medium per well and incubated for 6 h. Cells were treated with samples of interest in growth medium for 48 h. Aliquots of culture media were analyzed for secreted alkaline phosphatase activity using the Phospha-Light reporter chemiluminescence assay kit (Applied Biosystems, Foster City, CA, USA). The MTT colorimetric assay was performed on the cells for assessing their corresponding cytotoxicity. The estrogenic data were determined by the formula, final SEAP/MTT % = SEAP/cell viability × 100%, to avoid false results from cytotoxicity. Tests were done in triplicate. For the raw data please see supplementary data. The purity of tested compounds was more than 95%.

Conclusions
In the current investigation, the estrogenic activity of different fractions obtained from the ethanolic extract of L. platyphylla areal parts was evaluated. Bioassay-guided isolation of the most active fraction (EtOAc) led to the purification of one new 2-benzyl-2,3-dihydrobenzofuran 1, one new phenylisocoumarin 2, and one new benzofuroisocoumarin 3, along with seventeen known compounds. The spectroscopic analyses revealed that 1-3 structures possess unique skeletons which were rarely isolated from nature. 3-O-methylquercetin (10) and 6-C-methylquercetin-3-methyl ether (13) exhibited the most potent estrogenic effect in a dose-dependent manner as demonstrated by the SEAP reporter assay system. Moreover, flavonoids 9, 10 and, 13, and homoisoflavones 19 and 20 expressed high binding ability in the ERβ binding assay. Selective binding to ERβ receptor is an interesting phenomenon from the therapeutic point of view because such binding mimics the beneficial effect of estrogen without the side effects associated with unspecific receptor targeting.

Supplementary Materials
1D and 2D NMR spectra, HR-ESIMS of the new isolates and the raw data of the estrogenic activity can be found in the Supplementary materials at: http://www.mdpi.com/1420-3049/20/04/6844/s1.