A New 9,10-Dihydrophenanthrene and Cell Proliferative 3,4-δ-Dehydrotocopherols from Stemona tuberosa

A new compound, 9,10-dihydro-5-methoxy-8-methyl-2,7-phenanthrenediol (1), was isolated from the roots of Stemona tuberosa Lour. (Stemonaceae) together with two new optically active compounds, (2S,4'R,8'R)-3,4-δ-dehydrotocopherol (2) and (2R,4'R,8'R)-3,4-δ-dehydrotocopherol (3). The structures of compounds 1–3 were determined on the basis of spectroscopic data analysis. Compounds 2 and 3 were each purified from a stereoisomeric mixture of 2 and 3 by preparative HPLC using a chiral column for the first time. The absolute configurations at C-2 of 2 and 3 were determined by Circular Dichroism (CD) experiments. As a part of the research to find natural wound healing agents, all isolates and the mixture of 2 and 3 were evaluated for their cell proliferative effects using a mouse fibroblast NIH3T3 and a HeLa human cervical cancer cell line. As a result, 1, 2, 3, or the mixture of 2 and 3 showed 41.6%, 78.4%, 118.6%, 38.2% increases of cell proliferation in the mouse fibroblast NIH3T3 respectively, compared to 28.4% increase of δ-tocopherol. Moreover, none of them induced cancer cell proliferation. Therefore, 3,4-δ-dehydrotocopherols, especially pure isomers 2 and 3 can be suggested as potential wound healing agents.

In the general wound healing processes, fibroblasts play important roles in the replacement of damaged cells by synthesizing collagen proteins [9], so natural compounds found to induce cell proliferation in fibroblasts can be suggested as potential wound healing agent candidates.

Chiral Separation and Structure Determination of Compounds 2 and 3
A subfraction was obtained by repeated column chromatography from the EtOAc extract of S. tuberosa. This subfraction appeared as a single peak in normal phase HPLC as described in the Experimental Section. However, the corresponding 13 C-NMR spectrum showed overlapped peaks, which suggested that the subfraction was a mixture of isomers. This mixture was applied to a chiral column (ChiralPak IA, Daicel, Osaka, Japan, 5 μm, 250 × 10 mm), and effectively separated by preparative HPLC to afford two pure stereoisomers, 2 (tR 18.1 min, 15 mg) and 3 (tR 21.1 min, 15 mg) as shown in Figure 3.  The planar structures of 2 and 3 were confirmed to correspond to 3,4-δ-dehydrotocopherol by the detailed analysis of their HMBC and NOESY data ( Figure 2). However, the 1 H-, 13 C-NMR data of 3 were slightly different from those of 2. The 1 H-NMR spectrum of 3 was comparable with that of 2, except for the aliphatic protons of C-2' appearing at δH 1.41 (2H, m) in 3 instead of δH 1.46 and 1.36 (each 1H, m) in 2. Two methyl groups, CH3-4' and CH3-8' resonated at δH 0.834 (6H, d, J = 6.4 Hz) in 3 instead of δH 0.841 (3H, d, J = 6.4 Hz) and 0.830 (3H, d, J = 6.6 Hz) in 2. Moreover, the carbon signals of C-1' and CH3-4' in 3 appeared at δC 40.97 and 19.6 respectively, whereas they were observed at δC 41.01 and 19.7 in 2.
To solve the absolute configurations of 2 and 3, Circular dichroism (CD) experiments were performed. The Cotton effect of the 260-270 nm transition has been used to determine the absolute configurations of chiral centers near a styrene chromophore, and subgroups on the chromophore do not affect the sign of the Cotton effect [11]. The CD curve of 2 displayed a positive Cotton effect at 278 nm (as shown in Figure 4), thus the absolute configuration of C-2 was determined to be S [12]. Conversely, a negative Cotton effect at 275 nm was shown in the CD data of 3 and the absolute configuration of C-2 was confirmed as R. As a result, the structures were assigned as (2S,4′R,8′R)-3,4δ-dehydrotocopherol (2) and (2R,4′R,8′R)-3,4-δ-dehydrotocopherol (3), respectively. This is the first report on the separation of 2 and 3 as optically active compounds, together with determination of the absolute configurations for both isomers.

Cell Proliferative Effects
The cell proliferation effects of all isolates and the mixture of 2 and 3 were evaluated on mouse fibroblasts NIH3T3 and human cervical cancer cells HeLa ( Figure 5). In terms of chemical relevance with the tested compounds, δ-tocopherol was used for comparison in the present biological study. Cells were treated with 10 µM of all isolates, the mixture of 2 and 3, δ-tocopherol, or DMSO alone and cell counts were performed every 24 h. Daily microscopic examination of the cells showed no differences in cell attachment or morphology between compound-treated and control cells during the incubation times. As a result, all isolates or the mixture of 2 and 3 increased the number of NIH3T3 cells (41.6%, 78.4%, 118.6%, or 38.2% respectively) compared to the DMSO-treated control cells after 4 days incubation with compounds. Incubation with δ-tocopherol also led to a significant but smaller increase of NIH3T3 cells (28.4%). Interestingly, none of these compounds had a proliferative effect on HeLa cells, indicating their ideal wound healing possibility. The compounds promoted fibroblast growth dose-dependently. NIH3T3 cells were treated with the indicated concentrations of δ-tocopherol, all isolates, and the mixture of 2 and 3. After 4 days of incubation, cell numbers were counted. Growth promotion rates in NIH3T3 cells by 0.5, 1, 5, and 10 µM concentrations of compound 3 were 1.0%, 50.6%, 89.5%, and 119.6%, respectively (* p < 0.01 vs. control) ( Figure 6). Similar dose-dependent growth promotion by compound 2 was also observed. Growth promotion rates in NIH3T3 cells by 0.5, 1, 5, and 10 µM concentrations of compound 2 were 0.5%, 35.7%, 49.6%, and 78.7%, respectively (* p < 0.01 vs. control) Incubation with compounds 1, the mixture of 2 and 3, and δ-tocopherol also led to a significant but smaller dose-dependent growth promotion effect.

Plant Material
The dried roots of S. tuberosa were purchased from the Insan Oriental Herbal Market in Seoul, Korea. The sample was identified by Prof. Je-Hyun Lee (College of Oriental Medicine, Dongguk University, Gyeongju, Korea). A voucher specimen (No. EA322) has been deposited at the Natural Product Chemistry Laboratory, College of Pharmacy, Ewha Womans University (Seoul, Korea).

Cell Culture
The mouse fibroblasts NIH3T3 and HeLa human cervical cancer cells were obtained from the American Type Culture Collection (Manassas, VA, USA). The NIH3T3 cells displayed fibroblastic morphology and HeLa cells were polygonal. Both cells were tightly adherent to the flask, and were highly proliferative. NIH3T3 cells were cultured in Dulbecco's Modified Eagle's Medium (DMEM, Welgene, Gyeongsan, Korea) supplemented with 10% bovine calf serum (HyClone, Logan, UT, USA) and 1% penicillin-streptomycin (HyClone), HeLa cells were maintained in DMEM supplemented with 10% fetal bovine serum (HyClone) and 1% penicillin-streptomycin. Cells were maintained at 37 °C with 5% CO2 in a humidified atmosphere. The cells were subcultured when 80% confluence was reached according to the manufacturer's recommendation. The morphology of the cells was examined under a microscope.

Cell Proliferation Assay
For proliferation studies, cell number count was conducted in monolayer culture in 12-well culture plates using a hemocytometer. NIH3T3 or HeLa cells were seeded at an initial density of 1 × 10 5 cells per mL medium containing DMSO only, 10 μM of all isolates, the mixure of 2 and 3, or δ-tocopherol. Cells were trypsinized and then viable cell numbers were counted in triplicates for each group every 24 h. The data was recorded as the average of three independent experiments. For the dose-dependency of cell proliferative effects of compounds, NIH3T3 cells were seeded at an initial density of 1 × 10 5 cells per mL medium containing all isolates, the mixture of 2 and 3, or δ-tocopherol (0.5 μM-10 μM; control DMSO only) and the effect on cell growth was determined after 4 days by counting cell numbers. Cell numbers were counted in triplicate for each group, and three independent experiments were performed.

Statistical Analysis
All results were expressed as mean ± s.d. Student's t-test was used for comparisons involving two groups. p < 0.05 was considered to indicate a statistically significant difference.

Conclusions
In the present phytochemical study, a new compound, 9,10-dihydro-5-methoxy-8-methyl-2,7phenanthrenediol (1) was isolated from the roots of S. tuberosa Lour. The structure of 1 was elucidated by the interpretation of its spectroscopic data. In addition, (2S,4'R,8'R)-3,4-δ-dehydrotocopherol (2) and (2R,4'R,8'R)-3,4-δ-dehydrotocopherol (3) were separated for the first time from the stereoisomeric mixtures of 2 and 3 by preparative HPLC using a chiral column. CD experiments were utilized to determine the absolute configurations at C-2 of 2 and 3. When all isolates and the mixture of 2 and 3 were evaluated for their cell proliferative effects, the pure isomers 2 and 3 exhibited remarkable effects, suggesting the two compounds as powerful wound healing agent candidates.