Inhibition of fatty acid synthase by ginkgolic acids from the leaves of Ginkgo biloba and their cytotoxic activity

Abstract

Fatty acid synthase (FAS) has been proposed to be a new drug target for the development of anticancer agents because of the significant difference in expression of FAS between normal and tumour cells. Since a n-hexane-soluble extract from Ginkgo biloba was demonstrated to inhibit FAS activity in our preliminary test, we isolated active compounds from the n-hexane-soluble extract and evaluated their cytotoxic activity in human cancer cells. Three ginkgolic acids 1–3 isolated from the n-hexane-soluble extract inhibited the enzyme with IC₅₀ values 17.1, 9.2 and 10.5 µM, respectively, and they showed cytotoxic activity against MCF-7 (human breast adenocarcinoma), A549 (human lung adenocarcinoma) and HL-60 (human leukaemia) cells. Our findings suggest that alkylphenol derivatives might be a new type of FAS inhibitor with cytotoxic activity.

Keywords::

• Ginkgo biloba
• ginkgolic acids
• fatty acid synthase
• cytotoxic activity

Introduction

Ginkgo biloba L. is an ancient gymnosperm species and is extensively distributed around the world¹. Leaves of the tree contain biologically active ingredients, including flavonoid glycosides, terpene trilactones, alkylphenols and catechins^2,3. An extract from the leaves of G. biloba has been widely used for treating cerebral insufficiency, alleviating memory loss and progression of neurodegenerative disorders such as Alzheimer’s disease^1,4. Unfortunately, alkylphenols such as ginkgolic acids in the nonpolar extract of the leaves have been reported to exert undesirable properties including mutagenicity, toxicity^5,6 and allergy⁷. Therefore, the content of ginkgolic acids and related alkylphenols in EGb 761 (standardized leaves extract of G. biloba) has been restricted to a maximal concentration of 5 ppm for clinical application². Since a hexane-soluble extract prepared from the leaves of G. biloba showed in vitro fatty acid synthase (FAS) inhibitory activity in our preliminary experiment, we hypothesized that cytotoxic properties of ginkgolic acids^8,9 might be associated with the inhibitory activity of FAS¹⁰.

De novo synthesis of fatty acids in the cytosol of animal cells is performed by the 250–270-kD multifunctional and homodimeric FAS. FAS consists of two identical multifunctional polypeptides, in which three catalytic domains in the N-terminal section (β-ketoacyl synthase, malonyl/acetyl transferase and dehydrase) are separated from four C-terminal domains (enoyl reductase, β-ketoacyl reductase, acyl carrier protein and thioesterase)¹¹. FAS expression in normal cells and tissues is remarkably down-regulated but it is significantly up-regulated or highly activated in invasive and metastatic lesions of many different human epithelial cancer cells¹². This difference in expression of FAS between normal and tumour cells provides an ideal approach to cancer therapy. Indeed, previous studies with FAS inhibitors such as cerulenin or C75 selectively triggered apoptosis of cancer cells and inhibited tumour growth have verified that FAS activity is necessary for cancer cell survival^12–14.

In our recent study, we found out that a n-hexane-soluble extract of the leaves of G. biloba had FAS inhibitory activity. Three ginkgolic acids, C15:1 (1), C17:2 (2) and C17:1 (3), were isolated from the n-hexane-soluble extract. The isolated compounds 1–3 were evaluated for their FAS inhibitory potency and cytotoxic activities against MCF-7 (human breast adenocarcinoma), A549 (human lung adenocarcinoma) and HL-60 (human leukaemia) cells to determine the cytotoxic action of ginkgolic acids is attributed to their possible inhibition of FAS activity.

Materials and methods

General experimental procedures

TLC was executed on glass plates precoated with silica gel F₂₅₄ (20 × 20 cm, 200 µm, 60 Å, Merck, Darmstadt, Germany). VLC (Vacuum Liquid Chromatography) was performed on Merck silica gel (70–230 mesh). MPLC (Medium Performance Liquid Chromatography) was carried out with Biotage Isolera™ using silica gel SNAP Cartridge KP-Sil (100 g, 53 µm, Biotage, Uppsala, Sweden). HPLC separation was performed using a Gilson system with a UV detector and a Luna C₁₈ column (250 × 21.20 mm, 10 µm). All other chemicals and solvents were of analytical grade. The HPLC consisted of a surveyor system (Thermo Finnigan, San Jose, CA) with LCQ advantage trap mass spectrometer (Thermo Finnigan, San Jose, CA) equipped with an electrospray ionization (ESI) source. The column used was an Atlantis dC18 (2.1 × 50 mm, 3 µm) with a guard cartridge (Security Guard C₁₈, 4 × 3 mm, Phenomenex, Torrance, CA). Nitrogen was used as the sheath gas at 20 L/min with a capillary temperature of 275°C and the spray voltage was set to 4.50 kV. The mass spectrometer was operated in negative ion mode in m/z range of 150–600. Helium was used as the collision gas for the tandem mass spectrometric experiments followed by the isolation of ions over a selected mass window of 1 Da. A gradient HPLC mobile phase of MeOH/H₂O was used from 80:20 to 100:0 for 6 min, then isocratic of 80:20 until 10 min with a flow rate of 6 mL/min. The ¹H (250 MHz) and ¹³C (63 MHz) NMR spectra were recorded on a Bruker 250 MHz (DMX 250) spectrometer equipped with a 5 mm direct detection PFG (Pulsed Field Gradient) probe. All NMR experiments were performed at 294 K, using chloroform-d as the solvent. Chemical shifts were given on the δ scale and referenced by chloroform-d as an internal standard (δ_H = 7.24, δ_C = 77.16). Coupling constants (J) are in Hz. Data processing was carried out with MestReNova v6.0.2 program.

Plant material

The leaves of G. biloba were collected on the campus of Yeungnam University, Gyeongsan, Korea, on October 2009, and identified by Prof. MinKyun Na of the College of Pharmacy, Yeungnam University, Korea. The leaves were dried in a cool place away from direct sunlight for 14 days. A voucher specimen (2009-0026L) was deposited in the herbarium of the College of Pharmacy, Yeungnam University.

Isolation of ginkgolic acid derivatives

The dried leaves of G. biloba were extracted with MeOH at room temperature for 7 days, and the dried MeOH extract was suspended in water and partitioned with n-hexane. A portion of the n-hexane extract (10 g) that showed FAS inhibitory activity was subjected to Si gel VLC (21 × 17.5 cm) and eluted with a stepwise gradient of n-hexane (3 L), CHCl₃/MeOH [100:0, 80:20, 60:40, 40:60, 20:80 (each 2.5 L)], and MeOH (4 L) to give 7 fractions (Fr. 1–7). Of these, Fr. 3 (3.6 g) was suspended in MeOH (1.5 L) and partitioned with n-hexane (1.5 L) three times. The MeOH layer was further divided into five fractions (Fr. 3-M-1 to 3-M-5) using Si gel MPLC (16 × 4 cm, 100 g) with an isocratic solvent of n-hexane/CHCl₃/MeOH (20:200:1). Fr. 3-M-5 (684.7 mg) was applied to C₁₈ HPLC (250 × 21.2 mm, 10 μm) with a gradient of MeOH/H₂O (3 % acetic acid) from 45:55 to 100:0 for 75 min, then maintained at 100:0 until 115 min to afford 11 subfractions (Fr. 3-M-5-1 to 3-M-5-11). Fr. 3-M-5-7 (48.4 mg) was further purified byC₁₈ HPLC (250 × 21.2 mm, 10 μm) with a linear gradient of MeOH/H₂O (95:5 → 100:0, 3% acetic acid) for 60 min to yield 1 (t_R 40 min, 13.6 mg). Compounds 2 (t_R 54 min, 21.4 mg) and 3 (t_R 62 min, 28.7 mg) were obtained from Fr. 3-M-5-11 (123.0 mg) by C₁₈ HPLC (250 × 21.2 mm, 10 μm) with a gradient of MeOH/H₂O (3 % acetic acid) from 80:20 to 100:0 during 80 min.

Compound (1, ginkgolic acid C15:1): ¹H-NMR (250 MHz, CDCl₃) δ 7.34 (1H, t, J = 7.5 Hz), 6.86 (1H, d, J = 7.5 Hz), 6.76 (1H, d, J = 7.5 Hz), 5.32 (2H, m), 2.97 (2H, t-like, J = 7.5 Hz), 2.01 (4H, m), 1.59 (2H, m), 1.32–1.25 (m), 0.89 (3H, t-like, J = 7.0 Hz); ¹³C-NMR (63 MHz, CDCl₃) δ 175.7, 163.6, 147.8, 135.3, 130.1–130.0, 122.8, 115.9, 110.8, 36.6, 32.1, 30.0–29.5, 27.3, 27.1, 22.8, 22.5, 14.2; FAB-MS m/z 347.4 [M + H]⁺, 369.5 [M + Na]⁺.

Compound (2, ginkgolic acid C17:2): ¹H-NMR (250 MHz, CDCl₃) δ 7.34 (1H, t, J = 7.5 Hz), 6.86 (1H, d, J = 7.5 Hz), 6.76 (1H, d, J = 7.5 Hz), 5.35 (4H, m), 2.97 (2H, t-like, J = 7.0 Hz), 2.02 (4H, m), 1.60 (2H, m), 1.32–1.27 (m), 0.89 (3H, t-like, J = 7.0 Hz); ¹³C-NMR (63 MHz, CDCl₃) δ 176.1, 163.6, 147.9, 135.4, 130.0–129.9, 122.9, 115.9, 110.9, 36.6, 32.1, 30.0–29.5, 27.3, 27.1, 22.5, 14.2; ESI-MS m/z 373.2 [M + H]⁺.

Compound (3, ginkgolic acid C17:1): ¹H-NMR (250 MHz, CDCl₃) δ 7.34 (1H, t, J = 7.5 Hz), 6.85 (1H, d, J = 7.5 Hz), 6.76 (1H, d, J = 7.5 Hz), 5.32 (2H, m), 2.97 (2H, t-like, J = 7.5 Hz), 2.01 (4H, m), 1.59 (2H, br s), 1.32–1.27 (m), 0.89 (3H, t-like, J = 7.0 Hz); ¹³C-NMR (63 MHz, CDCl₃) δ 175.9, 163.5, 147.8, 135.3, 130.1–130.0, 122.8, 115.9, 111.0, 36.6, 32.1, 30.0–29.5, 27.3, 27.1, 22.8, 22.5, 14.2; ESI-MS m/z 373.5 [M–H]⁻.

FAS inhibitory activity

FAS was purified from chicken liver by stepwise ammonium sulfate precipitations, gel filtration and anion exchange chromatography as described previously¹⁵. The FAS was 95% pure as estimated from SDS/PAGE with Coomassie blue staining. FAS activity was measured by analysis of the incorporation of [³H] acetyl CoA into palmitate^15,16. To each microtube (final volume: 100 µL), FAS was added (20–30 µg protein) in a buffer containing 100 mM potassium phosphate (pH 7.0), 2.5 mM dithiothreitol and 2.0 mM EDTA with or without test compounds. The mixtures were preincubated at 37°C for 60 min and the reaction was started by the addition of a substrate mixture containing 0.25 mM NADPH, 0.4 nmol of malonyl CoA and 0.02 µCi of [³H] acetyl CoA. Following incubation at 37°C for 10 min, the reaction was terminated with 60% HClO₄. Fatty acids were extracted with 1 mL of hexane and incorporation of radioactivity into the fatty acids was assessed by scintillation counting¹⁶.

Cytotoxicity assays

MCF-7 cancer cells were maintained in a complete tissue culture medium consisting of RPMI 1640 (Sigma, Poole, UK) supplemented with 5% foetal calf serum (Sigma), 100 U/mL penicillin/100 µg/mL streptomycin (Sigma), 2 mM l-glutamine (Sigma), and 5 × 10⁻⁵ M 2-mercaptoethanol (2-ME, Sigma), while HL-60 and A549 cancer cells were maintained in RPMI 1640 that included l-glutamine (JBI) with 10% fetal bovine serum (JBI) and 2% penicillin–streptomycin (GIBCO). Cells were cultured at 37°C in a 5% CO₂ incubator. Cytotoxicity was measured using a modified MTT assay. Viable cells were seeded in growth medium (180 µL) into 96-well microtiter plates (1 × 10⁴ cells per well) and incubated at 37°C in a 5% CO₂incubator. Test samples were dissolved in DMSO and adjusted to final sample concentrations ranging from 5.0 to 100 µM by diluting with growth medium. The final DMSO concentration was adjusted to <0.1%. After standing for 2 h, 20 µL of the test sample was added to each well. The same volume of DMSO was added to the control wells. After 48 h, the test sample was added, 20 µL of MTT was also added to each well (final concentration, 5 µg/mL). After 2 h, the plate was centrifuged for 5 min at 1500 rpm, the medium was removed and the resulting formazan crystals were dissolved in 150 µL of DMSO. The optical density was measured with a microplate reader (Molecular Devices, Sunnylvale, CA). The IC₅₀ value was defined as the concentration of sample that reduced the absorbance by 50% relative to the vehicle-treated control.

Results and discussion

The FAS inhibitory compounds isolated from G. biloba were obtained as semisolid substances and identified as ginkgolic acids C15:1 (1), C17:2 (2) and C17:1 (3) by means of LC-MS and NMR data analyses compared with those in the literatures (Figure 1)^2,17. Compound 1 was identified as two isomers that consist of Δ8 and Δ10 Z double bond with the ratio of 1:2². According to the previous report by van Beek and Wintermans², ginkgolic acid C17:1 (3) could be a mixture of three isomers with the double bond at a different position occurring in a ratio of 2:1:12 of which the main isomer has a double bond at Δ12. All the isolates (1 − 3) were assayed for their in vitro FAS inhibitory activity and all of them dose-dependently inhibited enzyme activity with IC₅₀ values ranging from 9.2 to 17.1 µM (Table 1). Interestingly, compounds 1–3 exerted more potent FAS inhibitory activity than the positive controls, cerulenin (IC₅₀ = 13.0 µM) and luteolin (IC₅₀ = 62.5 µM).Compound 2, which has two double bonds at Δ9 and Δ12, showed the highest inhibitory activity against FAS. Several human cancer celllines (MCF-7, A549 and HL-60) were used to verify if these FAS inhibitory compounds could actually exert cytotoxic activity in several cell types (Table 2). For the MCF-7 cancer cell line, all tested metabolites demonstrated more powerful cytotoxic activity than cyclophosphamide monohydrate. HL-60 was the most susceptible strain in the employed cancer cell lines. Compound 2, which was proven to be the most powerful FAS inhibitor among the tested compounds, showed the strongest cytotoxic activity on HL-60 (IC₅₀4.0 µM), which was consistent with previous studies that the number of double bonds in alkyl chain of alkylphenols was proportional to their cytotoxic potency⁹. A previous study⁸ also verified that 6-[8(Z),11(Z)-pentadecadienyl] salicylic acid demonstrated more potent cytotoxic activity on BT-20 breast and HeLa epithelioid cervix carcinoma cells than 6-[8(Z)-pentadecenyl] salicylic acid and 6-pentadecylsalicylic acid, which supported the direct proportional relationship between the number of double bonds on ginkgolic acids and their cytotoxic potency. Interestingly, antibacterial activity of anacardic acids, which share the structural similarity with the tested compounds, has been proven to be proportional to the number of olefinic bonds on the alkyl side chain as well¹⁸.

Table 1.  Inhibitory activity of ginkgolic acids 1–3 isolated from G. biloba against FAS.

Compound	FAS inhibitory activity
	IC50 (µM)*
Ginkgolic acid C15:1 (1)	17.1 ± 1.5
Ginkgolic acid C17:2 (2)	9.2 ± 1.0
Ginkgolic acid C17:1 (3)	10.5 ± 1.3
Luteolin^†	62.5 ± 4.0
Cerulenin^†	13.0 ± 0.9

*IC₅₀values were determined by regression analyses and expressed as means of three replicates ± standard deviation.

^†Positive control.

FAS, fatty acid synthase.

Table 2.  Cytotoxic activity of ginkgolic acids 1–3 isolated from G. biloba (IC50 µM)*.

Compound	MCF-7	A549	HL-60
Ginkgolic acid C15:1 (1)	146.3	66.8	5.0
Ginkgolic acid C17:2 (2)	108.2	56.4	4.0
Ginkgolic acid C17:1 (3)	93.2	50.5	4.2
Adriamycin^†	3.5	1.8	0.8
Cyclophosphamide monohydrate^†	3792.5	NT	NT

*IC₅₀values were determined by regression analyses and expressed as means of three replicates.

^†Positive controls.

NT, not tested.

Figure 1.  Structures of ginkgolic acids 1–3 isolated from the leaves of G. biloba.

Conclusion

We identified that three ginkgolic acids C15:1 (1), C17:2 (2), and C17:1 (3) isolated from the n-hexane-soluble fraction of G. biloba had in vitro FAS inhibitory activity and cytotoxicity against human cancer cells. The FAS inhibitory activities of ginkgolic acids 1–3 were similar to or more potent than the positive controls, luteolin and cerulenin, suggesting that ginkgolic acids and related alkylphenols could be a new type of FAS inhibitor. In addition, ginkgolic acids 1–3 exerted micromolar range cytotoxic activities toward the employed human cancer cell lines (HL-60, A549 and MCF-7). Therefore, it is possible that the cytotoxic activity of ginkgolic acids might be attributed to their FAS inhibitory effect. Ginkgolic acids and related alkylphenols have been regarded as hazardous compounds with suspected allergenic, cytotoxic, immunotoxic and carcinogenic properties⁶. Although the cytotoxic potential of ginkgolic acid derivatives has been reported^6,19, the mechanism of action has not been clearly characterized. As demonstrated in this study, ginkgolic acids C15:1 (1), C17:2 (2), and C17:1 (3) were capable of inhibiting FAS activity, which might provide a new insight into the mechanism underlying the anticancer potential of ginkgolic acids and related alkylphenols.

Declaration of interest

This research was supported by Basic Science Research Program through the National Research Foundation of Korea funded by the Ministry of Education, Science and Technology (2011-0024389).