Abstract
As part of our continuing investigation of polymethoxyflavone (PMF) derivatives as potential anticancer substances, a series of PMF derivatives was synthesized. The synthesized compounds were evaluated for cytotoxicity against the promyelocytic leukemic HL60 cell line, and structure–activity relationship correlations were investigated along with previously isolated PMFs from the peel of king orange (Citrus nobilis). 7,3’-Dimethoxyflavone demonstrated the most potent activity among the synthetic PMFs. Consideration of correlation between the methoxylation pattern and antiproliferative activity revealed the importance of the 3’-methoxyl group and the higher degree of methoxylation on the A-ring moiety of PMFs.
Flavonoids are diverse plant-derived chemicals that are produced by various higher plants (1), which can therefore be found in numerous food sources such as fruits, vegetables, legumes, and whole grains (2). Citrus plants are rich sources of flavonoid and they contain a wide range of flavonoid constituents. Polymethoxyflavones (PMFs) are found almost exclusively in the Citrus genus, particularly in the peel of king orange (Citrus nobilis), sweet orange (Citrus sinensis), and mandarin orange (Citrus reticulata). They are of particular interest due to their broad spectrum of biological activity, including anti-inflammatory, anticarcinogenic, antiatherogenic, neurotropic and memory-enhancing properties (3-10). PMFs have been shown to inhibit the growth of the human leukemia HL60 cell line in vitro and to suppress proliferation, while promoting apoptosis (11).
Their biological properties have led not only to horticultural breeding attempts to produce new citrus cultivars containing high amounts of PMFs (12), but also to development of a rapid, large-scale method for the preparation of PMF-rich fraction from peel of Citrus sunki (13).
In previous studies, we found potent differentiation-inducing activity towards HL-60 leukemia cells from the juice of C. nobilis. The active principles were isolated and identified as four PMFs, namely tangeretin (8), nobiletin (9), heptamethoxyflavone (10) and natsudaidain (11) (Figure 2) and these PMFs inhibited the growth of several cancer cell lines (14). Among these PMFs, 11 demonstrated the most potent activity (15, 16).
In the course of our investigation on the structure–activity relationship of flavonoids, the correlation between the number/position of methoxyl groups and the antiproliferative activity of PMFs drew our attention. In order to study the effect of the methoxylation pattern of PMFs on their bioactivity, we synthesized 15 PMFs with variations of methoxylation of the A-ring moiety, since most natural PMFs have a highly methoxylated A-ring moiety.
Materials and Methods
General procedures. Chemicals and solvents from commercial sources were used without further purification unless specified. Reactions were carried out under argon and monitored by thin-layer chromatography on silica gel (mesh size 60, F254) with visualization under UV light. Standard and flash column chromatography procedures were not optimized. Nuclear magnetic resonance (NMR) spectra were recorded on a 400-MHz JEOL ECP-400 spectrometer (JOEL, Tokyo, Japan), and chemical shifts values are expressed in ppm (δ) relative to the residual 1H signal of the solvents. Unless otherwise specified, compounds were dissolved in 2HCCl3. Electrospray ionization (ESI) and atmospheric pressure chemical ionization (APCI) mass spectrometry were performed on Thermo Exactive (Thermo Fisher Scientific K.K, Yokohama, Japan) and Hitachi M8000 instruments (Hitachi, Tokyo, Japan), respectively.
Synthesis of 7,3’-dimethoxyflavone (Figure 1). General procedure for synthesis of PMFs (Figure 2, 6b-6o). To a suspension of dicyclohexylcarbodiimide (1.00 g, 4.85 mmol) and N,N-dimethyl-4-aminopyridine (107 mg, 0.876 mmol) in dry dichloromethane (12 ml), 2’-hydroxy-4’-methoxyacetophenone (718 mg, 4.32 mmol) and then 3-methoxybenzoic acid (977 mg, 6.42 mmol) were added. The reaction mixture was stirred at room temperature for 20 hours. The mixture was filtered to remove dicyclohexylurea as white precipitate. The solvent of the filtrate was removed under reduced pressure, and the residue was chromatographed over silica gel [hexane/ dichloromethane (DCM); 2:8] to afford 4’-methoxy-2’-(3-methoxybenzoyloxy)acetophenone (4g) as a white solid. To a suspension of KOH (212 mg, 3.79 mmol) in dry pyridine (3 ml) 4g (573 mg, 2.12 mmol) was added and the mixture was stirred at 100°C for 10 min. After being cooled to room temperature, the reaction mixture was neutralized with acetic acid (approximately 3.0 ml). The mixture was added to ethanol (3.0 ml) and deionized water (3.0 ml). The resulting precipitate was filtered and washed with cold ethanol to give 1-(3-methoxyphenyl)-3-(2’-hydroxy-4’-methoxyphenyl) propan-1,3-dione (5g, 350 mg, yield from 4g: 61%). To a solution of 5g (431 mg, 1.30 mmol) dissolved in acetic acid (6 ml) kept at 100°C, 20% H2SO4/acetic acid (1 ml) was added. The mixture was stirred at 100°C for 10 min. After cooling to room temperature, deionized water was added to the mixture. The resulting precipitate was filtered and washed with water, and was then chromatographed over silica gel (hexane/DCM; 2:8) to afford 7,3’-dimethoxyflavone (6g, 379 mg, yield from 5g: 93%).
For 4g: 1H-NMR (CDCl3) δ 2.49 (s, 3H), 3.87 (m, 6H), 6.71 (d, 1H, J=2.2 Hz), 6.87 (dd, 1H, J=8.8, 2.2 Hz), 7.19 (dt, 1H, J=7.7, 1.5 Hz), 7.42 (t, 1H, J=7.7 Hz), 7.71 (t, 1H, J=1.5 Hz), 7.81 (dt, 1H, J=7.7, 1.5 Hz), 7.90 (d, 1H, J=8.8 Hz). HRMS (M + H)+: calcd for C17H16O5, 300.0994; found 300.0997.
For 5g: 1H-NMR (CDCl3) δ 3.86 (m, 8H), 6.47 (m, 2H), 6.69 (s, 1H), 7.07 (dt, 1H), 7.38 (t, 1H, J=8.8 Hz), 7.44 (st, 1H, J=2.2 Hz), 7.48 (dt, 1H), 7.68 (d, 1H, J=8.8 Hz). HRMS (M + H)+: calcd for C17H16O5, 300.0994; found 300.0994.
By combination of 3 2’-hydroxyacetophenone derivatives, namely 2’-hydroxy- (1a), 2’-hydroxy-4’-methoxy- (1b), and 2’-hydroxy-4’,6’-acetophenone (1c), and benzoic acid (2a) and its derivatives, namely 3-methoxy- (2b), 4-methoxy- (2c), 3,4-dimethoxy- (2d), and 3,4,5-tirmethoxybenzoic acids (2e), 15 PMFs were synthesized. Spectral data for 6g are shown in Table I along with data for other synthetic PMFs.
Flavone (6a) and sinensetin (7) were purchased from Tokyo Chemical Industry Co. (Tokyo, Japan) and Funakoshi (Tokyo, Japan). Tangeretin (8), nobiletin (9), heptamethoxyflavone (10) and natsudaidain (11) were isolated from king orange juice. Fruits were juiced by hand, and 200 ml of juice was absorbed on 250 g of polystyrene resin (Diaion HP-20; Mitsubishi Chemical, Tokyo, Japan) and eluted with ethanol. The combined eluents were concentrated and subjected to silica-gel column chromatography (Wako Gel C-200; Wako Pure Chemicals, Tokyo, Japan) eluted with 10% ethanol in hexane, chloroform, and then 50% chloroform in methanol. The chloroform eluates were further purified by a reversed-phase high performance liquid chromatography (HPLC), giving 8 (yield; 11.8 mg), 9 (yield; 12.7 mg), 10 (yield; 30.4 mg), 11 and (yield; 12.0 mg).
Cell proliferation assay. HL60 cells were maintained in RPMI-1640 medium supplemented with 10% fetal bovine serum. The level of cellular proliferation for HL60 cells grown in a 96-well microplate was measured by using alamar blue (Life Technologies Ltd., Tokyo, Japan). To each well, 100 μl of HL60 cell suspension (1.0×104 cells/100 μl) was inoculated and 100 μl of medium containing serial dilutions of the samples to be assayed was aded. After three days of incubation, 20 μl of alamar blue was aseptically added to each well, and cells were further incubated for approximately 20 hours. Cellular proliferation (as a percentage that of the untreated control) was calculated with the following equation: Proliferation (%)= where A570 and A595 are the absorbance at 570 nm and 595 nm, respectively.
Results and Discussion
Chemistry. In this study, 15 PMFs were synthesized. The synthetic route to 7,3’-dimethoxyflavone (6g) is illustrated in Figure 1. The starting 2’-hydroxyacetophenone derivatives were first treated with methyl iodide and K2CO3 in dry acetone to afford their corresponding methoxylated derivatives, namely 2’-hydroxy- (2a), 2’-hydroxy-4’-methoxy-(2b), 2’-hydroxy-4’,6’-dimethoxyactophenones (2c). Methylation selectively occurred at the 4’-position of 2’,4’-dihydroxyacetophenone (1b) and at the 4’- and 6’-positions of 2’,4’,6’-trihydroxyacetophenone (1c). Fortunately, 2’-hydroxyl groups did not react with the methylation reagent, presumably because of the weakened nucleophilicity of the 2’-hydroxyl oxygen atom by its hydrogen bonding to the oxygen atom of the acetyl carbonyl group.
2’-Hydroxyacetophenone derivatives (2a-2c) were then reacted with a series of methoxylated derivatives of benzoic acids to give the corresponding esters. The intermediate esters were subjected to the Baker-Venkataraman rearrangement in dry pyridine in the presence of KOH, to afford the corresponding 1,3-diarylpropan-1,3-diones. Acid-catalyzed cyclization was carried out in acetic acid in the presence of H2SO4, which yielded the target products in yields of 25-35%.
Antiproliferative activity. The antiproliferative activity of synthetic 15 PMFs along with several PMFs isolated from the peel of king orange was determined using promyelocytic leukemic HL60 cells based on the alamar blue assay. The results are summarized in Figure 2.
Among the compounds tested, 3-hydroxy-5,6,7,8,3’,4’-hexamethoxyflavone (11, trival name: natsudaidain) demonstrated the most potent activity (IC50=5.0 μM), followed by 7,3’-dimethoxyflavone (6g, IC50=8.0 μM). Furthermore, the 5,7-dimethoxylated derivatives, namely 5,7,4’-trimethoxyflavone (6m, IC50=23 μM), 5,7,3’-trimethoxyflavone (6l, IC50=24 μM) and 5,7-dimethoxyflavone (6k, IC50=31 μM) also demonstrated significant activity.
Structure and activity relationship. Since these results showed that there was not a simple correlation between the number of methoxyl groups on PMFs and their antiproliferative activity, the pattern of methoxylation, i.e. the number and the position of methoxylation, shoult be considered. Thus 20 PMFs tested in this study were classified based on their methoxylation pattern and were subjected to the consideration of the structure–activity relationship.
Increase of the number of methoxyl groups on the B-ring moiety tended to lower the activity among the PMFs with the same A-ring methoxylation pattern. When the number of A-ring methoxyl groups was 0-3, attachment of more than two methoxyl groups on the B-ring moiety completely diminished the activity (IC50 values of compounds 6d, 6e, 6i, 6j, 6n, 6o, and 7 were higher than 400 μM). Although each of the PMFs with four methoxyl groups on their A-ring moiety demonstrated significant activity, the tendency of activity reduction by the increase of B-ring methoxyl groups was maintained.
On the contrary, an increase in the number of A-ring methoxyl groups tended to enhance the activity. Typically, only compound 9, which had the tetramethoxylated A-ring structure, demonstrated significant activity among a series of PMFs with 3’,4’-dimethoxylated B-ring structure, namely compounds 6d, 6i, 6n, 7, and 9. Similarly, compounds 6m and 8 demonstrated potent activity among a series of PMFs with the 4’-methoxylated B-ring structure.
The position of methoxyl group on B-ring moiety had a significant effect on the activity. The activity was reduced when a methoxyl group was attached to the 4’-position; most of the 4’-methoxylated PMFs tested had no activity. Similar activity-weakening phenomena by 4’-substitution on the B-ring were reported for 1-azaflavanone (18). In contrast, many 3’-methoxylated PMFs demonstrated significant activity. For example, compounds 6b and 6g exhibited more potent activity than their corresponding 4’-methoxylated derivatives, namely compounds 6c and 6h, respectively. The importance of the 3’-methoxyl group was reduced when two methoxyl groups were attached to the A-ring moiety.
The free rotation of the C2-C1’ single bond causes two typical conformers, namely conformers A and B as indicated in Figure 3. In order to determine which conformer contributes to the activity, the nuclear Overhauser effect (NOE) spectroscopy was measured. The observed NOE correlation between H3 and H2’ indicated the predominance of conformer A. While these results suggested the involvement of conformer A for activity, further studies are necessary.
In summary the data presented here demonstrated that an increase in the number of methoxyl groups on the A-ring enhanced the activity of PMFs, whereas the increase of B-ring methoxyl groups reduced the activity.
- Received September 25, 2012.
- Revision received October 26, 2012.
- Accepted October 29, 2012.
- Copyright© 2012 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved