Antiangiogenic Tocotrienol Derivatives from Garcinia amplexicaulis

Nov 18, 2013 - Fax: 33 241 226 634. Abstract. Abstract Image. Phytochemical investigation of a dichloromethane extract from Garcinia amplexicaulis ste...
0 downloads 4 Views 1MB Size
Article pubs.acs.org/jnp

Antiangiogenic Tocotrienol Derivatives from Garcinia amplexicaulis Alexis Lavaud,†,‡ Pascal Richomme,† Marc Litaudon,§ Ramaroson Andriantsitohaina,‡ and David Guilet*,† †

Laboratoire SONAS, Université d’Angers, IFR Quasav, 49100 Angers, France INSERM UMR U694, IBS-IRIS, Université d’Angers, 49100 Angers, France § Institut de Chimie des Substances Naturelles (ICSN), CNRS, Labex LERMIT, 91198 Gif sur Yvette Cedex, France ‡

S Supporting Information *

ABSTRACT: Phytochemical investigation of a dichloromethane extract from Garcinia amplexicaulis stem bark led to the isolation of four new tocotrienols (1−4); two known tocotrienols, two triterpenes, and a xanthone were also isolated. Their structures were mainly established using NMR and MS methods. The main compounds isolated, δamplexichromanol (1) and γ-amplexichromanol (2), were evaluated on VEGF-induced angiogenesis using a Matrigel assay. Compounds 1 and 2 inhibited in vitro angiogenesis of VEGF-induced human primary endothelial cells in the low nanomolar range. Their capacity to inhibit VEGF-induced proliferation of endothelial cells partially explained this activity, although δ-amplexichromanol (1) also prevented adhesion and migration processes.

associated with the formation of macroscopic tumors.14 Angiogenesis blockade is therefore a key approach for cancer treatment and prevention. As part of our ongoing search for novel antiangiogenic compounds from Clusiaceae,15 a biological evaluation of major tocotrienol derivatives from a CH2Cl2 extract of G. amplexicaulis, δ-amplexichromanol (1) and γ-amplexichromanol (2), was undertaken.

Garcinia is the largest genus of the Clusiaceae family, with about 200 species widely distributed in tropical Asia, Africa, and Polynesia, and consists of 180 species.1 Most Garcinia species are known for their brownish-yellow gum resin (due to the presence of xanthones), which is used as a purgative. Many xanthones, coumarins, benzophenones, and biflavonoids featuring a range of biological activitiesantiviral, cytotoxic, anti-inflammatory, antioxidant, etc.have been identified in plants of the genus Garcinia.2,3 In our ongoing phytochemical investigation of Garcinia species,4,5 we examined a dichloromethane extract from stem bark of Garcinia amplexicaulis Vieill. ex Pierre (Clusiaceae), an endemic shrub from New Caledonia. The present study led to the isolation of four new tocotrienol derivatives (1−4) and five known compounds (5−9, see Supporting Information for structures). Studies on the biological activities of tocotrienols (from vitamin E) revealed their ability to suppress proliferation in a variety of tumor cells.6−8 This antiproliferative activity is partially mediated through modulation of growth factors such as VEGF and so their capacity to inhibit angiogenesis.9,10 Angiogenesis is the growth and remodeling process of new blood vessels from an existing vascular network.11 The induction of angiogenesis is characterized by an imbalance between pro-angiogenic factors, such as vascular endothelial growth factor (VEGF), and antiangiogenic factor production, such as thrombospondin-1 (TSP-1). Among angiogenesis-stimulating molecules, VEGF appears to have a central role in the angiogenic process. Pathological angiogenesis is involved in the pathogenesis of many diseases including cancer, atherosclerosis, rheumatoid arthritis, and diabetic retinopathy.12,13 Most interestingly, angiogenesis is critical for tumor development, and neovascularization leads to the rapid spread of tumor cells © 2013 American Chemical Society and American Society of Pharmacognosy

Received: July 24, 2013 Published: November 18, 2013 2246

dx.doi.org/10.1021/np400598y | J. Nat. Prod. 2013, 76, 2246−2252

Journal of Natural Products



Article

RESULTS AND DISCUSSION The stem bark from G. amplexicaulis was collected in southern New Caledonia in the “Forêt Cachée” area. Dried and powdered stem bark samples were extracted with dichloromethane. The crude extract was then fractionated using centrifugal partition chromatography under optimized conditions. Further fractionations were performed by silica gel normal-phase vacuum flash chromatography and C18 reversephase preparative HPLC to afford compounds 1−9. δ-Amplexichromanol (1) was isolated as an amorphous, optically active compound, [α]23D −14.7 (MeOH, c 0.14), which analyzed for C27H40O4 by HREIMS. The 13C NMR spectrum displayed 12 carbon signals in the aromatic/olefinic region (δC 150−110). This, coupled with two distinct proton signals at δH 6.47 (1H, d, J = 3.0 Hz) and 6.37 (1H, d, J = 3.0 Hz) in the 1H NMR spectrum, revealed the presence of a hydroquinone-type moiety and three double bonds in the molecule. Moreover, in conjunction with the eight degrees of unsaturation inherent to the molecular formula, δ-amplexichromanol (1) was found to possess an additional ring. Also detected were signals of a primary alcohol at [δH 4.20 (2H, s); δH 4.30 (2H, s)] and [δC 67.6 (CH2); δC 60.0 (CH2)] in the 1H and 13C NMR data, respectively. The structure of compound 1 was then determined by detailed interpretation of 2D NMR data. Long-range correlations of the aromatic protons at δH 6.47 and 6.37 and upfield protons at δH 2.67 (2H, t, J = 6.7 Hz) and 2.12 (3H, s) with aromatic carbons in the HMBC data revealed a 3-alkyl-5-methylhydroquinone moiety (Figure 1 and

compound were highly compatible with those obtained for 1 (Table 1), but the usual singlet for H-7 around δH 6.47 had disappeared. Instead of this aromatic proton, a methyl group (δH 2.13; δC 11.9) appeared, which was the most noticeable difference. Long-range correlations of the methyl proton signals at δ 2.13 (H-27) with the aromatic carbons at δC 146.3 (C-6), 121.7 (C-7), and 125.7 (C-8) indicated a γ-chromanol moiety. γ-Amplexichromanol (2) was thus identified as (2R)-2,7,8trimethyl-2-[(3E,7E)-4,8-dimethyl-13-hydroxy-12-hydroxymethyldeca-3,7,11-trienyl]chroman-6-ol. γ-(Z)-Deoxyamplexichromanol (3) analyzed for C28H42O3 by HREIMS and 13C NMR spectrometry. The chromanol nature of this compound was evident from the characteristic proton and carbon signals in the NMR spectra. The spectra were very similar to those of 2. However, only one singlet signal (at δH 4.10) was present in the 1H NMR spectrum, while a methyl signal appeared (δH 1.79; δC 21.2). Upfield shifts in the vinyl methyl carbons at δC 21.2 (C-21) and 16.0 (C-23), coupled with the NOESY cross-peaks H-19/H-21, assigned the Z geometry for this double bond. Therefore, 3 was identified as (2R)-2,8-dimethyl-2-[(3E,7E,11Z)-4,8-dimethyl-12-hydroxymethyldeca-3,7,11-trienyl]chroman-6-ol, or γ-(Z)-deoxyamplexichromanol. (γ,δ)-Bi-O-amplexichromanol (4) is a dimer of 1 and 2, and HREIMS analysis and 13C NMR spectrometry gave the molecular formula C55H80O8. Moreover, considering the integration of aromatic proton signals in the 1H NMR spectrum, four oxymethylene signals were present. The aromatic region of the 1H NMR spectrum of 4 showed singlets at δH 6.36 and 6.53, thus indicating the presence of a δamplexichromanol monomer. The connection of the two units thus involved the oxygen of the OH group of 1 and an aromatic carbon of the other monomer (2). Unsubstituted aromatic carbons of monomers could function as bridgeheads in 4. NOESY cross-peaks H-4/H-7′ confirmed the linkage of both monomers. Dimers of δ-tocotrienol were previously isolated from natural sources such as Iryanthera grandis fruits, and the 13 C NMR data of 4 were in compliance with those reported for dimeric structures.17 Especially, the shielding values of C-4a (δC 112.2), C-6 (δC 140.0), C-8 (δC 121.9), and C-4 (δC 17.5) measured in dimer 4 compared with the monomeric unit 2 (δC 118.2, 146.3, 125.7, and 22.3 respectively) were also reported in 13 C NMR spectra for known dimeric tocotrienols.17,18 The structure of (γ,δ)-bi-O-amplexichromanol (4) was assigned as (2R)-2,7,8-trimethyl-2-[(3E,7E)-4,8-dimethyl-13-hydroxy-12hydroxymethyldeca-3,7,11-trienyl]-5-[[(2R)-2,8-dimethyl-2[(3E,7E)-4,8-dimethyl-13-hydroxy-12-hydroxymethyldeca3,7,11-trienyl]-6-chromanyl]oxy]chroman-6-ol. The known tocotrienol derivative δ-tocotrienilic alcohol (5)19 could be named δ-(Z)-deoxyamplexichromanol. Garcinoic acid (6)20 was also isolated from the stem bark of G. amplexicaulis, along with 6-deoxyisojacareubin (7)21 and the triterpenes euphol (8)22 and euphan-8-ene-3β,24,25-triol (9).23 The structural determination of these compounds was based on an analysis of their 1D and 2D NMR and MS data and comparison with the literature data. It should be noted that this is the first report of natural tocotrienols with two primary alcohol functions located at the terminal part of the prenyl chain. Natural tocotrienols sensu lato constitute a homogeneous group with around 40 different structures associating the chroman-6-ol skeleton and at least two prenyl units. The main part of the structural diversity originates from brown seaweed, Sargassum species, with around

Figure 1. Key 2D NMR correlations of compound 1.

Table 1). The 1H COSY data showed direct spin couplings between benzylic methylene protons at δH 2.67 and methylene protons at δH 1.83 and 1.74. HMBC correlations of the carbon bearing these protons at δC 31.4 and those at δC 75.3 and 39.0 with the methyl proton at δH 1.27 allowed the construction of a chromanol moiety. Similarly, long-range correlations of the vinyl methyl protons at δH 1.55 and 1.57 with neighboring carbons, combined with 1 H COSY correlations of the olefinic protons with upfield protons, defined the structure of the linear prenyl portion. Both alcohol functions were located at the terminal isopropyl part of the prenyl chain on the basis of long-range correlations between the oxymethylene protons and neighboring carbons. Compound 1 possessed double bonds at C-11, C-15, and C-19. Upfield shifts in the vinyl alcohol carbons at δC 67.6 (C-21) and 60.0 (C-22), coupled with the NOESY cross-peaks H-19/H-21, assigned the positions of these oxymethylene carbons. In conclusion, the asymmetric C-2 configuration was defined as R according to the literature since naturally occurring tocotrienols as a rule exclusively possess the 2R-configuration.16 Therefore, 1 was identified as (2R)-2,8-dimethyl-2-[(3E,7E)-4,8-dimethyl13-hydroxy-12-hydroxymethyldeca-3,7,11-trienyl]chroman-6ol, or δ-amplexichromanol. The molecular formula of γ-amplexichromanol (2) was deduced as C28H42O4 by HREIMS. The NMR spectra of this 2247

dx.doi.org/10.1021/np400598y | J. Nat. Prod. 2013, 76, 2246−2252

Journal of Natural Products

Article

Table 1. NMR Spectroscopic Data (500 MHz, CDCl3) for Compounds 1−3 1

a

no.

δC, type

2 3 4 4a 5 6 7 8 8a 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

75.3, C 31.4, CH2 22.4, CH2 121.2, C 112.6, CH 147.9, C 115.7, CH 127.2, C 145.7, C 39.0, CH2 22.2, CH2 124.5, CH 134.1, C 39.4,a CH2 26.0, CH2 124.9, CH 134.7, C 39.3,a CH2 26.3, CH2 131.2, CH 136.6, C 67.6, CH2 60.0, CH2 15.8, CH3 15.9, CH3 24.3, CH3 16.1, CH3

2 δH (J Hz) 1.74−1.83, m 2.67, t (6.7) 6.37, d (3.0) 6.47, d (3.0)

1.50−1.62, m 2.09, m 5.10, t (6.5) 1.97, m 2.06, m 5.05, t (6.5) 1.97, m 1.96−2.13, m 5.51, t (6.5) 4.20, 4.30, 1.57, 1.55, 1.27, 2.12,

s s s s s s

δC

3 δH (J Hz)

75.2, C 31.4, CH2 22.3, CH2 118.2, C 112.1, CH 146.3, C 121.7, C 125.7, C 145.5, C 39.3, CH2 22.2, CH2 124.5, CH 134.1, C 39.4, CH2 26.3, CH2 124.9, CH 134.7, C 39.3, CH2 26.0, CH2 131.1, CH 136.8, C 67.7, CH2 60.1, CH2 15.9,a CH3 15.8,a CH3 24.2, CH3 11.9, CH3 11.9, CH3

1.72−1.78, m 2.67, t (6.7) 6.37, s

1.52−1.63, m 2.10, m 5.11, t (7.0) 1.98, m 1.96−2.06, m 5.08, t (7.0) 1.98−2.07, m 2.15, m 5.52, t (7.0) 4.20, 4.30, 1.56, 1.58, 1.27, 2.11, 2.13,

s s s s s s s

δC 75.2, C 31.2, CH2 22.2, CH2 118.2, C 111.9, CH 146.2, C 121.6, C 125.8, C 145.6, C 39.5, CH2 22.2, CH2 124.2, CH 134.9, C 39.8,a CH2 26.5, CH2 124.4, CH 134.4, C 39.5,a CH2 26.2, CH2 128.3, CH 134.2, C 21.2, CH3 61.6, CH2 16.0, CH3 15.8, CH3 24.1, CH3 11.9, CH3 11.9, CH3

δH (J Hz) 1.73−1.78, m 2.67, t (6.7) 6.36, s

1.58−1.62, m 2.11, m 5.10, t (7.0) 1.98, m 2.05, m 5.08, t (7.0) 1.98−2.07, m 2.11, m 5.28, t (7.0) 1.79, 4.10, 1.57, 1.59, 1.26, 2.11, 2.13,

s s s s s s s

Signals could be interchanged.

2.5 μM (Figure 2A). Regardless of the concentrations used, capillary network formation was strongly inhibited, indicating a powerful effect of both compounds. These results suggest that tocotrienol derivatives 1 and 2 possess in vitro antiangiogenic effects in the nanomolar range. To ensure the absence of cytotoxicity of both compounds 1 and 2, viability measurements were thus performed by MTT assay. After 24 and 48 h stimulation, no cytotoxic effect was observed in HUVEC cells treated with the two compounds at any concentration from 25 nM to 25 μM (data not shown). The ability of endothelial cells to form capillary tubes is a specialized function of this cell type resulting from a finely tuned balance between cell migration, proliferation, and adhesion.28 Compounds 1 and 2 were thus assessed on VEGF-induced adhesion, migration, and proliferation of HUVEC cells. After 24 h treatment, only δ-amplexichromanol (1) decreased the adhesion of VEGF-induced HUVEC cells at 25 nM and 2.5 μM, whereas γ-amplexichromanol (2) had no significant effect, as determined with the adhesion assay using crystal violet (Figure 3A). To form new vessels, endothelial cells need to be disseminated via migration, which contributes to angiogenesis. We studied the effects of both compounds on endothelial cell migration using a Transwell in vitro migration assay. Here again, only δ-amplexichromanol (1) decreased endothelial cell migration at 2.5 μM, while compound 2 did not affect this cellular process at 25 nM and 2.5 μM (Figure 3B). Endothelial cell proliferation is one of the critical steps in angiogenesis. Compounds 1 and 2 significantly reduced VEGFinduced HUVEC proliferation (Figure 3C) at 25 nM and 2.5

20 analogues described.24 This study confirms that in the angiosperm group the Clusiaceae family is an important contributor in terms of structural diversity of tocotrienol derivatives. Angiogenesis is a very complex process, and endothelial cell tube formation is one of the key steps.25 The different isoforms of tocotrienols from vitamin E were assessed on the basis of their antiangiogenic effects on VEGF-induced HUVECs, i.e., primary endothelial cells from human umbilical veins. The better δ- and γ-isoforms showed significant activity at concentrations up to 5 μM.10,26 However, no tocotrienol derivatives from natural sources had been evaluated on angiogenesis. To determine the potential antiangiogenic effects of tocotrienol derivatives from G. amplexicaulis on capillary-like structure formation by endothelial cells, HUVECs were treated with a combination of compound 1 or 2 (25 nM and 2.5 μM) and then VEGF (20 ng/mL). The role of VEGF post-treatment is to promote the formation of a capillary-like structure on endothelial cells and angiogenesis. Under basal conditions, HUVECs were able to organize and form capillary-like structures on ECM gel. The ability of endothelial cells to form tubular structures was assessed by calculating the length of tubes with an inverted photomicroscope (Figure 2B). This protocol is currently used to test the pro- or the antiangiogenic property of any given compound in our laboratory.27 In the presence of VEGF, an increase in the length of tubelike structures was observed as expected. After 24 h treatment, the ability of VEGF to increase capillary formation was completely prevented by tocotrienols 1 and 2 at 25 nM and 2248

dx.doi.org/10.1021/np400598y | J. Nat. Prod. 2013, 76, 2246−2252

Journal of Natural Products

Article

Figure 2. Antiangiogenic properties of δ-amplexichromanol (1) and γ-amplexichromanol (2) on in vitro VEGF-induced HUVEC tube formation. (A) HUVEC cells were cultured in medium supplemented with 10% FBS and treated with 1 or 2 (25 nM, 2.5 μM) + VEGF (20 ng/mL) for 24 h. VEGF (20 ng/mL) was used to promote angiogenesis. (B) Capillary length of in vitro HUVEC experiments was used to quantify angiogenesis. The results are means ± SEM from four independent experiments. **p < 0.01 versus the control group, $$p < 0.01 versus the VEGF group.

μM. As expected, VEGF treatment induced an increase in endothelial cell adhesion, migration, and proliferation. Taken together, these data suggest that tocotrienol derivatives 1 and 2 act on VEGF-induced angiogenesis, targeting especially tubule formation and endothelial cell proliferation. Importantly, δamplexichromanol (1) prevented the ability of VEGF to activate adhesion, migration, and proliferation processes. The two compounds were able to inhibit tube formation to the same extent, but they mediated this effect through different mechanisms. Compared with tocotrienols from vitamin E, the presence of OH groups at the terminal prenyl chain enhanced in vitro antiangiogenic activity in the low nanomolar range. However, further investigations on antiangiogenic mechanisms using Western blot analysis are now required to explain this increased activity. These results suggest that chemical modification of the terminal prenyl chain of δ-amplexichromanol (1) and γ-amplexichromanol (2) could be developed to increase their therapeutic potential against diseases involving angiogenesis.



equipped with a pneumatically assisted electrospray ionization (ESI) source. Chromatographic separations such as centrifugal partition chromatography with a 1L rotor (Armen Instrument, Vannes, France) equipped with a dual HPLC pump, degasser, and UV−visible detector, flash chromatography with an IntelliFlash 310 (Analogix, Burlington, USA) using a prepacked C18 (Interchim, Montluçon, France) or silica gel Chromabond flash RS column (Macherey-Nagel, Dü ren, Germany), and also preparative chromatography with a Varian ProStar 210 and a PrepStar 218 solvent delivery module (Agilent, Santa Clara, CA, USA) with a C18 Varian column (5 μm; 250 × 21.4 mm) were used to purify the compounds. HPLC Apparatus and Chromatography Conditions. The extract and fractions were passed through 0.20 μm filters before chromatographic separation using a Waters Alliance HPLC system (Milford, CT, USA) equipped with a quaternary HPLC pump, degasser, autosampler, and PDA diode array detector (Milford, CT, USA). The HPLC mobile phase consisted of water + 0.1% acetic acid (solvent A) and methanol (solvent B). The solvent gradient was as follows (starting with 100% solvent A): 0 min, 0% B; 30 min, 100% B; 40 min, 100% B. The flow rate was 1 mL/min, the injection volume was 20 μL, and the eluent was detected at 290 nm. All HPLC analyses were performed at 30 °C on a Lichrospher C18 column (5 μm; 150 × 4.6 mm). All purifications using preparative HPLC were previously optimized on a Varian C18 column using a Waters HPLC system. Plant Material. Stem bark from Garcinia amplexicaulis was collected in July 1998 in the “Forêt Cachée” area of southern New Caledonia and identified by one of the authors (M.L.). A specimen (LIT-0554) was deposited at the Laboratoire des Plantes Médicinales (CNRS), Noumea, New Caledonia. Extraction and Isolation. Dried G. amplexicaulis stem bark (270 g) was successively extracted, using a Soxhlet apparatus, for 24 h with 3 L of CH2Cl2 and then with 3 L of MeOH. The solvents were removed

EXPERIMENTAL SECTION

General Experimental Procedures. Optical rotations were recorded on a P-2000 digital polarimeter (Jasco, Great Dunmow, UK). UV spectra were recorded on a Varian Cary 50 Bio spectrophotometer (Varian France, Les Ulis, France). IR spectra were recorded on a Bruker FT IR Vector 22 using liquid films. 1H and 13 C NMR along with 2D NMR data were obtained on a Bruker Avance DRX 500 MHz (500 and 125 MHz, respectively) spectrometer in methanol-d3 with TMS as internal standard. Mass spectrometry analyses were performed on a JMS-700 (JEOL LTD, Akishima, Tokyo, Japan) double focusing mass spectrometer with reversed geometry, 2249

dx.doi.org/10.1021/np400598y | J. Nat. Prod. 2013, 76, 2246−2252

Journal of Natural Products

Article

Figure 3. Properties of δ-amplexichromanol (1) and γ-amplexichromanol (2) in endothelial cellular processes involved in angiogenesis: endothelial cell adhesion (A), migration (B), and proliferation (C). (A) Effect of 1 and 2 on VEGF-induced adhesion of HUVECs. Only compound 1 decreased it at 25 nM and 2.5 μM. The results are means ± SEM of four independent triplicate experiments. (B) Effect of 1 and 2 on VEGF-induced migration of HUVECs. The results are means ± SEM of four independent experiments. (C) Effect of 1 and 2 on VEGF-induced proliferation of HUVECs. The results are means ± SEM of four independent triplicate experiments. VEGF (20 ng/mL) was used as positive control for all experiments. *p < 0.05 versus the control group; $p < 0.05 versus the VEGF group. and 13C NMR (see Table 1); HREIMS m/z 465.2976 [M + Na]+ (calcd 465.2979 for C28H42O4Na). γ-(Z)-Deoxyamplexichromanol (3): pale yellow oil, [α]23D −10.0 (MeOH, c 0.03); UV (MeOH) λmax nm (log ε) 296.9 (3.55), 261.0 (2.75), 206.9 (4.57); 1H and 13C NMR (see Table 1); HREIMS m/z 449.3023 [M + Na]+ (calcd 449.3021 for C28H42O3Na). (γ,δ)-Bi-O-amplexichromanol (4): pale yellow oil, [α]23D +18.4 (MeOH, c 0.05); UV (MeOH) λmax nm (log ε) 293.0 (3.73), 260.0 (3.35), 203.0 (4.80) nm; 1H and 13C NMR (see Table 2); HREIMS m/z 891.5724 [M + Na]+ (calcd 891.5745 for C55H80O8Na). Cell Material and Cell Culture. Trypsin EDTA and culture media were obtained from Lonza (Basel, Switzerland). VEGF was obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Freshly delivered umbilical cords were obtained from a nearby hospital. HUVECs were obtained as previously described29 and grown on plastic flasks in MCDB 131 medium (Invitrogen) containing 1% Lglutamine, 1% streptomycin/penicillin, 500 ng/L epidermal growth factor, and 2 μg/L basic fibroblast growth factor, supplemented with 10% heat-inactivated fetal bovine serum (FBS) (Invitrogen, Cergy Pontoise, France). HUVECs were used at the second to fourth passage. Cells were grown for 24 h in the absence or presence of 25 nM or 2.5 μM tested compound or VEGF (20 ng/mL). Cell Viability Assay. HUVECs were seeded at 104 cells/well on 96-well plates. Cells were treated with 25 nM to 50 μM of the two compounds for 24 to 48 h. Viability was assessed by colorimetric analysis of MTT (Sigma-Aldrich). Absorbance values were obtained at 570 nm wavelength on a microplate reader (Synergy HT, Biotek). In Vitro Capillary Network Formation on ECM Gel. After 24 h incubation with biflavonoids with VEGF (20 ng/mL) or VEGF alone, HUVECs were detached with trypsin EDTA. Cells were seeded with a density of 15 × 104 cells per well precoated with ECM gel (Sigma-

under reduced pressure to yield 29.7 and 31.0 g of extracts, respectively. The CH2Cl2 extract (6 g) was subjected to centrifugal partition chromatography (1200 rpm; flow rate = 15 mL/min; P = 90 bar) with an ascendant elution using a quaternary mixture of heptane/ ethyl acetate/MeOH/water (2:1:2:1), to yield 20 fractions (F1−20). After this first fractionation step, fractions F18 and F16 respectively yielded compounds 1 (250 mg) and 2 (53 mg), with a relative purity of 80%. Each compound was then purified through a 15 g silica gel column using flash chromatography with a mixture of cyclohexane/ EtOAc (15:85 to 1:1) and was obtained with a relative purity of at least 97%. Fraction F1 (1.45 g) was loaded onto silica gel and eluted with a mixture of cyclohexane/EtOAc (1:0 to 1:1) through a 40 g silica gel column using flash chromatography to yield compound 8 (860 mg). Fractions F4 (45 mg) and F6 (25 mg) were then purified using preparative HPLC with an isocratic mode [MeOH/H2O (9:1)] to yield compound 3 (8.5 mg). Fraction F10 (72 mg) was purified by preparative HPLC using a mixture of MeOH/H2O (88% MeOH) to afford compounds 5 (28 mg) and 7 (4 mg). Fractions F11 (70 mg) and F12 (34 mg) were respectively purified by preparative HPLC using MeOH/H2O (88:12) to yield compounds 6 (10 mg) and 9 (2.5 mg). Fraction F14 (50 mg) was also purified by preparative HPLC using MeOH/H2O (93:7) to obtain compound 4 (3 mg). δ-Amplexichromanol (1): pale yellow oil, [α]23D −14.7 (MeOH, c 0.14); UV (MeOH) λmax nm (log ε) 296.9 (3.60), 258.0 (2.40), 206.9 (4.64); IR (film) νmax (cm−1) 3333, 2970, 2927, 2853, 1666, and 1610; 1 H and 13C NMR (see Table 1); HREIMS m/z 451.2813 [M + Na]+ (calcd 451.2813 for C28H42O4Na). γ-Amplexichromanol (2): pale yellow oil, [α]23D −30.5 (MeOH, c 0.07); UV (MeOH) λmax nm (log ε) 296.9 (3.58), 259.1 (2.63), 206.0 (4.64); IR (film) νmax (cm−1) 3350, 2970, 2920, 2853, 1666, 1620; 1H 2250

dx.doi.org/10.1021/np400598y | J. Nat. Prod. 2013, 76, 2246−2252

Journal of Natural Products

Article

Table 2. NMR Spectroscopic Data (500 MHz, CDCl3) for Compound 4 no. 2 3 4 4a 5 6 7 8 8a 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

δC 75.0, 30.5, 17.5, 112.2, 136.6, 140.0, 122.0, 121.9, 144.8, 39.3, 22.1, 124.3, 135.0, 39.5, 26.4, 124.9, 134.1, 39.2, 25.9, 130.9, 137.0, 67.7, 60.1, 16.0, 15.8, 24.2, 11.6, 12.0,

C CH2 CH2 C C C C C C CH2 CH2 CH C CH2 CH2 CH C CH2 CH2 CH C CH2 CH2 CH3 CH3 CH3 CH3 CH3

δH (J Hz)

no. 2′ 3′ 4′ 4a′ 5′ 6′ 7′ 8′ 8a′ 9′ 10′ 11′ 12′ 13′ 14′ 15′ 16′ 17′ 18′ 19′ 20′ 21′ 22′ 23′ 24′ 25′ 26′ 27′

1.62−1.72, m 2.44, m

1.51−1.64, m 2.10, m 5.12, m 2.03, m 1.95−2.04, m 5.09, m 1.95, m 2.16, m 5.52, m 4.19, 4.29, 1.58, 1.58, 1.24, 2.13, 2.20,

s s s s s s s

Aldrich). Briefly, 150 μL of ECM gel substrate diluted with FBS-free medium (1:1 dilution) was added into a four-well plate and allowed to solidify for 1 h at 37 °C. Then cells were incubated with medium containing 10% FBS and allowed to adhere for 1 h, after which the different stimuli were added. Tube formation was examined by phasecontrast microscopy (400×; MOTIC AE21) after 4 and 24 h and was quantified using ImageJ software. The capillary length was counted in three randomly selected microscopic fields for each experiment. To determine the pro- or antiangiogenic properties of the tested compounds, HUVECs were pretreated with the tested compounds or VEGF + tested compounds, as described above, and culture media were removed. Then HUVECs were treated with the conditioned medium for 24 h, and tube formation was determined as described above. Adhesion Assay on HUVECs. Evaluation of adherent cells was performed using crystal violet staining. For adhesion experiments, 5 × 103 cells per well were seeded into 96-well plates for 24 h before addition of the test compounds. After 24 h incubation, the plate was shaken for 15 s. The supernatant with nonadherent cells was removed by three washes with washing buffer (0.1% BSA in medium without serum). Attached cells were fixed with 4% paraformaldehyde for 15 min at room temperature. Cells were rinsed twice with washing buffer, stained with crystal violet (Sigma-Aldrich) (1 mg/mL in 2% of ethanol) for 10 min at room temperature, protected from light, and extensively washed with distilled water. Sodium dodecyl sulfate (2%) was then added, and the mixture was incubated for 30 min at room temperature. Absorbance was then evaluated using a microplate reader at 550 nm (Sinergy HT Biotek, Winooski, VT, USA). Proliferation Assay on HUVECS. Effects of compounds on proliferation on HUVECs were analyzed using a CyQUANT cell proliferation assay kit (Molecular Probes, Eugene, OR, USA). Briefly, 5 × 103 cells per well were seeded into 96-well plates and allowed to attach overnight, and then cells were treated with a mixture of the compounds added 30 min before VEGF (20 ng/mL) for 24 h. After growth medium removal, dye-binding solution was added to each

δC 75.5, 31.2, 22.5, 121.3, 111.9, 149.7, 115.1, 127.6, 147.0, 39.3, 22.1, 124.3, 135.0, 39.5, 26.4, 124.9, 134.1, 39.2, 25.9, 130.9, 137.0, 67.6, 60.0, 16.0, 15.8, 23.8, 16.2,

C CH2 CH2 C CH C CH C C CH2 CH2 CH C CH2 CH2 CH C CH2 CH2 CH C CH2 CH2 CH3 CH3 CH3 CH3

δH (J Hz) 1.72−1.80, m 2.67, m 6.36, d (2.6) 6.53, d (2.6)

1.51−1.64, m 2.10, m 5.12, m 2.03, m 1.95−2.04, m 5.09, m 1.95, m 2.16, m 5.52, m 4.19, 4.29, 1.58, 1.58, 1.26, 2.11,

s s s s s s

microplate well, and cells were incubated at 37 °C for 30 min. The fluorescence levels were read on a fluorescent microplate reader (Synergy HT, Biotek) with filters for 485 nm excitation and 530 nm emission. Migration Assay on HUVECs. Transwell cell culture chambers (Corning Costar 3422, Corning, Cambridge, MA, USA) were used for the cell migration assay. Enriched medium with 20% FBS was injected into the lower chamber, and 7.5 × 104 HUVEC cells were added to the upper compartment (8 μm pore size) in 250 μL of starvation medium containing 0.5% FBS with a mixture of tocotrienol derivatives (25 nM, 2.5 μM) added 30 min before VEGF (20 ng/mL), vehicle, or VEGF (20 ng/mL). After 24 h incubation at 37 °C, nonmigrated cells were removed from the upper surface of the membrane by wiping with a cotton swab. The membrane was then fixed with 4% paraformaldehyde for 15 min, stained with crystal violet solution for 10 min at room temperature, protected from light, and extensively washed with distilled water. Then sodium dodecyl sulfate 2% was added, and the mixture was incubated for 30 min at room temperature. Absorbance was subsequently evaluated using a microplate reader at 550 nm (Sinergy HT Biotek, Winooski, VT, USA). Data Analysis. For the cellular assay, data were represented as mean ± SEM, with n representing the number of experiments repeated at least in triplicate. Statistical analyses were performed by Mann− Whitney U-tests (nonparametric). All tests were two-tailed, and p < 0.05 was considered to be statistically significant.



ASSOCIATED CONTENT

S Supporting Information *

Structures of known compounds (5−9), NMR spectra of compounds 1−4. This material is available free of charge via the Internet at http://pubs.acs.org. 2251

dx.doi.org/10.1021/np400598y | J. Nat. Prod. 2013, 76, 2246−2252

Journal of Natural Products



Article

(28) Soeda, S.; Kozako, T.; Iwata, K.; Shimeno, H. Biochim. Biophys. Acta, Mol. Cell Res. 2000, 1497, 127−134. (29) Favot, L.; Martin, S.; Keravis, T.; Andriantsitohaina, R.; Lugnier, C. Cardiovasc. Res. 2003, 59, 479−487.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Tel: 33 241 226 676. Fax: 33 241 226 634. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We thank Angers Loire Métropole for granting a Ph.D. scholarship to A.L. We thank Dr. I. Freuze and B. Siegler from Plateforme d’Imagerie et d’Analyses Moléculaires (PIAM), Université d’Angers, for their assistance in HREIMS and NMR analysis.



REFERENCES

(1) Bennett, G. J.; Lee, H.-H. Phytochemistry 1989, 28, 967−998. (2) Castardo, J. C.; Prudente, A. S.; Ferreira, J.; Guimaraes, C. L.; Delle Monache, F.; Cechinel, V.; Otuki, M. F.; Cabrini, D. A. J. Ethnopharmacol. 2008, 118, 405−411. (3) Liu, Z. B.; Antalek, M.; Nguyen, L.; Li, X. S.; Tian, X. J.; Le, A.; Zi, X. L.S. Nutr. Cancer 2013, 65, 68−77. (4) Merza, J.; Aumond, M.-C.; Rondeau, D.; Dumontet, V.; Le Ray, A.-M.; Seraphin, D.; Richomme, P. Phytochemistry 2004, 65, 2915− 2920. (5) Hay, A. E.; Aumond, M.-C.; Mallet, S.; Dumontet, V.; Litaudon, M.; Rondeau, D.; Richomme, P. J. Nat. Prod. 2004, 67, 707−709. (6) Srivastava, J. K.; Gupta, S. Biochem. Biophys. Res. Commun. 2006, 346, 447−453. (7) Xu, W. L.; Liu, J. R.; Liu, H. K.; Qi, G. Y.; Sun, X. R.; Sun, W. G.; Chen, B. Q. Nutrition 2009, 25, 555−566. (8) Yu, W.; Simmons-Menchaca, M.; Gapor, A.; Sanders, B. G.; Kline, K. Nutr. Cancer 1999, 33, 26−32. (9) Miyazawa, T.; Inokuchi, H.; Hirokane, H.; Tsuzuki, T.; Nakagawa, K.; Igarashi, M. Biochemistry 2004, 69, 67−69. (10) Shibata, A.; Nakagawa, K.; Sookwong, P.; Tsuduki, T.; Oikawa, S.; Miyazawa, T. J. Agric. Food Chem. 2009, 57, 8696−8704. (11) Folkman, J.; Klagsbrun, M. Science 1987, 235, 442−447. (12) Folkman, J. Nature 1995, 1, 27−31. (13) Virmani, R.; Kolodgie, F. D.; Burke, A. P.; Finn, A. V.; Gold, H. K.; Tulenko, T. N.; Wrenn, S. P.; Narula, J. Arterioscl. Throm. Vas. 2005, 25, 2054−2061. (14) Hanahan, D.; Weinberg, R. A. Cell 2000, 100, 57−70. (15) Lavaud, A.; Soleti, R.; Hay, A. E.; Richomme, P.; Guilet, D.; Andriantsitohaina, R. Biochem. Pharmacol. 2012, 83, 514−523. (16) Drotleff, A. M.; Ternes, W. J. Chromatogr. A 2001, 909, 215− 223. (17) Viera, P. C.; Gottlieb, O. R.; Gottlieb, H. E. Phytochemistry 1983, 22, 2281−2286. (18) Goh, S. H.; Hew, N. F.; Ong, A .S. H.; Choo, Y. M.; Brumby, S. J. Am. Oil Chem. Soc. 1990, 67, 250−254. (19) Teixeira, J. S. R.; Moreira, L. M.; Guedes, M. L. S.; Cruz, F. G. J. Brazil Chem. Soc. 2006, 17, 812−815. (20) Terashima, K.; Shimamura, T.; Tanabayashi, M.; Aqil, M.; Akinniyi, J. A.; Niwa, M. Heterocycles 1997, 45, 1559−1566. (21) Owen, P. J.; Scheinmann, F. J. Chem. Soc., Perkin Trans. 1 1974, 1018−1021. (22) Lin, J. H.; Ku, Y. R.; Lin, Y. T.; Teng, S. F.; Wen, K. C.; Liao, C. H. J. Food Drug Anal. 2000, 8, 278−282. (23) Compagnone, R. S.; Suarez, A. C.; Leitao, S. G.; Monache, F. D. Rev. Bras. Farmacogn. 2008, 18, 6−10. (24) Velazquez, O. C. J. Vasc. Surg. 2007, 45, A39−47. (25) Jang, K. H.; Lee, B. H.; Choi, B. W.; Lee, H. S.; Shin, J. J. Nat. Prod. 2005, 68, 716−723. (26) Nakagawa, K.; Shibata, A.; Yamashita, S.; Tsuzuki, T.; Kariya, J.; Oikawa, S.; Miyazawa, T. J. Nutr. 2007, 137, 1938−1943. (27) Duluc, L.; Jacques, C.; Soleti, R.; Iacobazzi, F.; Simard, G.; Andriantsitohaina, R. Int. J. Biochem. Cell Biol. 2013, 45, 783−791. 2252

dx.doi.org/10.1021/np400598y | J. Nat. Prod. 2013, 76, 2246−2252