Biflavonoids with Cytotoxic and Antibacterial Activity from Ochna macrocalyx

• Abstract
• Introduction
• Material and Methods
• General
• Plant material
• Extraction and isolation of bark constituents
• Isolates
• MTT reduction assays
• NF-κB assays
• Antibacterial broth dilution minimum inhibitory
  concentration assay
• Results
• Isolated compounds
• NF-κB activity
• Cytotoxic activity
• Anti-bacterial activity
• Discussion
• Acknowledgements
• References

Abstract

Six biflavonoid and related compounds were isolated from the ethanolic extract of Ochna macrocalyx bark. One is a new compound, the isoflavanone dimer dehydroxyhexaspermone C (1). Previously isolated compounds obtained from the bark are biisoflavonoid hexaspermone C (2), tetrahydrofuran derivative ochnone (3), furobenzopyran derivative cordigol (4), and biflavonoids calodenin B (5) and dihydrocalodenin B (6). Although 3 has already been isolated, its spectral data are presented here for the first time. Isolated compounds were tested for cytotoxic activity on MCF-7 breast cancer cells using the MTT reduction assay method. Compound 5 showed cytotoxic activity (7 ± 0.5 μM) and 6 showed moderate cytotoxicity (35 ± 7 μM). In antibacterial assays performed using three strains of multi-drug resistant (mdr) Staphylococcus aureus (RN4220, XU212 and SA-1199-B) compounds 5 and in particular 6 showed strong antibacterial activity (MICs 5 : 64, 8, 16 μg/mL 6 : 8, 8, 8 μg/mL, respectively). The ethanolic extract of the bark also showed NF-κB inhibitory activity.

Introduction

Ochna macrocalyx Oliv. (Ochnaceae) is a tree with a characteristic yellow bark, used medicinally by the Washambaa of the Western Usambara Mountains of Tanzania [1]. Its bark is used for gynaecological (dysmennorrheoa and fertility problems) and gastrointestinal (diarrhoea, haemorrhoids, stomach pains) disorders.

The Ochnaceae is a small family of trees and shrubs, spread over warm parts of the world. Several members are known for their toxic and pharmacological properties and previous phytochemical studies have shown the family to be rich in biflavonoids and related chalcones. In our continued search for bioactive natural products with anti-inflammatory and antibacterial activity, Ochna macrocalyx was selected for investigation, as the crude bark extract was found to have NF-κB inhibitory activity in an electrophoretic mobility shift assay (EMSA). In addition, to our knowledge no phytochemical nor biological work has been performed on this species. In the NF-κB assays the bark extracts were observed to have cytotoxic effects on HeLa cells, which led to cytotoxicity investigations being undertaken. Isolated compounds were included in antibacterial assays.

Here we report on the isolation and structure elucidation of one new compound and five previously known polyphenolic compounds, as well as the antibacterial, cytotoxic and NF-κB inhibitory activities of the bark extract and isolated compounds.

Material and Methods

General

Analytical TLC plates: Merck silica gel 60 F₂₅₄ on aluminium sheets. Preparative TLC plates: Merck silica gel 60 F_{254 + 366} 2 mm. Spots were visualised by spraying with vanillin 4 % in H₂SO₄ and heating. HPLC: C-18 analytical and semi-preparative column (Waters Radially compressed model 25 mm) using a Waters 600 controller with 996 photo-diode array detector (detection 8 = 210 - 394 nm). FABMS: positive ion mode, matrix MNOBA + Na, EIMS: direct inlet with 70 eV ionisation, FAB and EIMS obtained using a ZAB-SE instrument from VG-Analytical. ¹H- and ¹³C-NMR experiments: Bruker DRX-400. Melting point: Gallenkamp apparatus. Optical rotation: Bellingham and Stanley ADP 220 polarimeter.

Plant material

Ochna macrocalyx was collected from the foot of the Western Usambara mountains in 1997 - 1998 (research permit no. 97 - 165-NA-97-35) and identified by Christina Schlage. A voucher specimen (collection no. CS 42,2) is held in the Centre for Pharmacognosy and Phytotherapy at the School of Pharmacy, London.

Extraction and isolation of bark constituents

Fifty grams of the powdered bark were refluxed with 500 mL of 96 % ethanol (30 min) and twice with 70 % ethanol, yielding a residue of 18 g. This was partitioned with water, hexane and ethyl acetate. The ethyl acetate extract (15 g) was fractionated using Sephadex LH-20 (300 g) in MeOH, producing 9 fractions. Compound 1 (30 mg) and hexaspermone C (2; 20 mg) ([α]_D ^22.2: -25.9, c 0.06, MeOH) were isolated together from fraction 5 (820 mL) by TLC (CHCl_3, R_f 1 : 0.06, 2 : 0.08, both red when sprayed), and then separated by semi-preparative HPLC (isocratic: 55 % acetonitrile in H₂O, 5 mL/min, t_R 1 : 65 min, 2 : 86 min). Compounds 3 (180 mg) and 4 (trace amounts) were isolated together from fraction 5 using Biotage^TM flash chromatography and then separated on analytical TLC plates (CHCl₃/butan-2-ol, 8 : 2, R_f 2 : 0.5 orange when sprayed, R_f 4 : 0.4 pink when sprayed). Calodenin B (5) (135 mg), an orange compound, was purified from fractions 7 (640 mL) and 8 (400 mL) using preparative TLC (PTLC). Dihydrocalodenin B (6) (320 mg) ([α]_D ^24.9: 123.5, c 0.03, MeOH), a yellow compound, was isolated from fractions 6 (200 mL) and 7 using PTLC and purified using semi-preparative HPLC (isocratic, 55 % MeOH in H₂O, 5 mL/min, t_R 6 : 55 min). TLC solvent system used for 5 and 6 was DCM/MeOH/H₂O, 80 : 20 : 2, R_f 5 : 0.46, 6 : 0.44.

Isolates

Rel-4′,7-dimethoxy-4-oxo-2,3-trans-isoflavanyl-(2→ 2″)-4″,5″-dihydroxy-7″-methoxy-2″,3″-trans-isoflavanone (1): White/colourless solid; m. p. 133 - 135 °C; [α]_D ^23.7: -117.0 (c 0.05, MeOH); ¹H- and ¹³C-NMR: see Table [1]. EIMS: m/z = 285, 284, 283, 252, 167, 151, 121; FABMS: m/z = 569 [M + H]⁺.

Rel-4α-(2,4-dihydroxybenzoyl)-3β-(4-hydroxybenzoyl)-2α-(2,4-dihydroxyphenyl)-5α-(4-hydroxyphenyl)tetrahydrofuran (3): White solid; m. p. 164 - 166 °C; [α]_D ^23.1: -96.7 (c 0.21, MeOH); ¹H- and ¹³C-NMR: see Table [2]; EIMS: m/z = 169, 149, 132; FABMS: m/z = 551 [M + Na]⁺.

MTT reduction assays

MCF-7 breast cancer cells were seeded (1000 cells/well) in 96 well plates (media: MEMα with Glutamax-1 supplemented with 5 % FCS - both from GibCo) and allowed to attach for 48 hours (at 5 % CO₂, 37 °C) before treatment. Test compound solution (300 μg/mL) was prepared by dissolving in media with 0.5 % v/v DMSO and allowed to solubilise overnight under refrigeration before filter sterilisation. The cells were treated in replicates of 6 with the test compound at a series of different concentrations ranging from 1 - 100 μg/mL and incubated for 72 hours (final volume of 150 μL per well). Blanks (wells containing no cells) and controls (untreated cells) were run in parallel. Twenty μL of MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] solution (5 mg/mL in PBS) were added to each well and incubated for 3 hours; 100 μL of MTT solubilisation solution (500 mL H₂O, 200 g SDS, 500 mL dimethylformamide, 20 mL glacial acetic acid, 10 mL 2 M HCl) were added to each well and incubated overnight. Optical density of each well was measured at 570 nm with a plate reader (Labsystems Multiskan Multisoft) [2]. The mean optical density obtained for the blanks was subtracted from the readings obtained, and relative cell viability in each well expressed as a percentage of the controls. Cell viability was plotted against log concentration, and the IC₅₀ obtained graphically. Results are shown as the mean and standard deviations obtained from assays repeated three times. Control experiments were performed using doxorubicin as a reference compound.

NF-κB assays

a) Electrophoretic mobility shift assay (EMSA): The method used is described in [3], with the following additions: Preparation of test plant extracts: 5 mg of plant extract were mixed with 25 mL of media and 125 μL of DMSO. HeLa cells grown overnight on 10 cm dishes were treated with the extract by removal of the media which was replaced with 5 mL of the prepared plant extract (1 hour incubation time). The plant extract media was warmed and centrifuged prior to being added to the cells. Cells were stimulated using phorbol 12-myristate-13-acetate (PMA) for 15 minutes.

b) IL-6 Luciferase reporter gene method: The method used is described in [3] with the following modifications: Transfected HeLa cells were grown overnight in 12 well plates. Preparation of plant extract: Extracts were dissolved in DMSO (10 mg/mL). Incubation of cells with plant extract (1 hour): 10 μL of plant extract in DMSO were added to the media in the well (1 mL of media per well) to obtain a final concentration of 100 μg/mL. Each test was performed in replicates of three. Cells were stimulated with PMA for 7 hours.

For both assays controls using stimulated and unstimulated cells were run in parallel. Parthenolide was used as a positive control [3].

Antibacterial broth dilution minimum inhibitory
concentration assay

Staphylococcus aureus (SA) strain XU212 [Tet(K) efflux system, tetracycline resistant] was cultured from clinical isolates from the Adnan hospital (Kuwait). SA strain RN 4220 [Msr(A) efflux system, macrolide resistant] was provided by Dr. Jon Cove. SA strain 1199-B [Nor(A) efflux system, norfloxacin resistant] was provided by Dr. Glenn W. Kaatz. Bacteria were subcultured on nutrient agar (Oxoid) 24 hours prior to assay.

The method used for the minimum inhibitory concentration assay is described in [4], with the following additions: Preparation of test compound solution: Test and antibiotic compounds were dissolved in DMSO and diluted out in Mueller-Hinton broth (Oxoid) (adjusted to contain 20 mg/mL and 10 mg/mL of Ca²⁺ and Mg²⁺ , respectively) to give a final concentration of DMSO of 3.125 %. Each test was performed in duplicate. Assays using standard antibiotics (tetracycline, erythromycin and norfloxacin), DMSO, growth and sterile controls were run in parallel.

Results

Isolated compounds

Compounds 1 and 2 are classical biisoflavonoids, the ¹H-NMR spectra of both giving characteristic signals as seen in [5]. Both are composed of two isoflavanone subunits joined together through a 2→2″ link. The FABMS of 1 gave a molecular ion at 568. Accurate mass calculations gave a molecular formula of C₃₃H₂₈O₉ . 1D ¹H- and ¹³C-NMR spectra, 2D ¹H-¹H COSY and ¹H-¹³C HMQC spectra in acetone-d ₆ were obtained. The compound slowly decomposed in acetone-d ₆, therefore an ¹H-¹³C HMBC spectra could not be obtained. The ¹³C- and ¹H-NMR spectra showed the presence of three methoxy groups (δ_C = 55.5, 56.2, 56.3, δ_H = 3.77, 3.81, and 3.83), two carbonyl groups (δ_C = 194.6 and 200.1), four aliphatic protons (δ_H = 4.58, 4.67, 5.62 and 5.68), one chelated hydroxy group (visible with CDCl₃) at δ_H = 11.91 and four ring systems, two AA′BB′ (δ_H = 6.82, 7.39 and 6.89, 7.42), one ABX (δ_H = 6.62, 6.71 and 7.58) and one meta coupled system (δ_H = 6.12 and 6.17). Compound 1 was isolated together with 2 (C₃₃H₂₈O₁₀). Compound 2 was identified as a substance previously isolated from Ouratea hexasperma (Ochnaceae) [5] by comparison of spectral data; 2 is an almost symmetrical compound, its two isoflavanone subunits only differing at positions 4″′ and 4′. The ¹H-NMR spectrum in CDCl₃ of 2 was similar to that of 1, but simpler with many of the signals superimposed due to its high degree of symmetry; 2 also has two AA"BB" ring systems (δ_H = 6.80, 7.33 and 6.87, 7.38), but differs from 1 with two meta coupled systems [δ_H = 6.13 (2H) and 6.15 (2H)] and no ABX system. Signals from two chelated hydroxy groups came at δ_H = 11.93, and two of the methoxyl groups (positions 7 and 7¿) at δ_H = 3.76 with the third methoxy at δ_H = 3.80. Four aliphatic protons showed signals at δ_H = 4.51 (2H), 5.68 and 5.69. The molecular formula of 2 differs from 1 by an additional oxygen, and due to the similarity of their ¹H- and ¹³C-NMR spectra, it was concluded that the two differ only by the absence of one OH group in 1. In addition only one chelated hydroxy group could be seen in the ¹H-NMR spectrum for 1. The position of this OH group was established using EIMS, where the compounds fragment into their isoflavanone subunits. The EIMS for 1 showed two major signals at m/z = 283 (1b) and 285, for 2 the signals occurred at 285 and 299 (2b). The signal at m/z = 285, consistent with C₁₆H₁₃O₅, corresponds to the isoflavanone fragment 1a/2a, which remains the same for both compounds. This led to the deduction that the OH group is absent from position 5 in 1. Structures drawn show the isoflavanone subunits in their reduced forms, which correspond to the m/z of the signals seen in the EIMS. Assignments of the carbons and protons for 1 (Table [1]) were based on HMQC data and by similarity with 2.

The J-values between H3-H2, H2-H2″, H2″-H3″ were 10.8, 8.0, and 10.8 Hz respectively, consistent with three trans linkages, making the relative stereochemistry of the two isoflavonone subunits of both compounds the same.

In contrast to the assignments in [5], ¹H signals at δ_H = 4.51 - 4.67 were assigned to positions 2 and 2″ because of the multiplicity of the signals (Table [1]). ¹H signals at δ_H = 5.62 - 5.69 as doublets were thus assigned to positions 3 and 3″, although HMQC showed them to be attached to carbons at δ_C = 84.5 - 86.4, which would seem further downfield than expected.

Ochnone (3) (C₃₀H₂₄O₉) has been previously isolated from Brackenridgea zanguebarica [6], although its spectral data were not reported. Here we present for the first time the full NMR data obtained for the compound (Table [2]). The FABMS of compound 3 gave a molecular ion at 528. The ¹³C-NMR data showed thirty carbons. Most of the carbons were in the aromatic region, 6 were oxygen-bearing quaternary carbons and two were carbonyls. The ¹H-NMR showed the presence of two AA′BB′ (δ_H = 6.61, 7.22 and 6.76, 7.69) and two ABX (δ_H = 6.10, 6.32, 7.75 and 6.32, 6.38, and 7.35) ring systems, four aliphatic protons (δ_H = 4.93, 4.99, 5.37, 5.56) and two deshielded hydroxy groups (δ_H = 8.26 and 12.36) indicative of H-bonding to a carbonyl group. ¹H-¹H COSY, ¹H-¹³C HMQC and ¹H-¹³C HMBC enabled the assignment of the carbons and protons. The relative stereochemistries of the protons at positions 2 - 5 were deduced from ¹H-¹H NOESY data (Table [3]).

Spectral data for compounds 4 - 6 were compared to previously published data. Compound 4, calodenin B (5) and 2,3-dihydrocalodenin B (6) have been isolated from Brackenridgea zanguebarica (Ochnaceae) [7], [8], [9]. Compound 5 has also been isolated from Ochna calodendron [10], while 4 was first reported in [7] as a compound isolated from Brackenridgea zanguebarica which was inadvertently ascribed to Cordia goetzei (Ioset and Hostettmann, personal comment March 2002).

Compound 4 was isolated as a minor contaminant with 3, and biological testing of 4 was not possible due to insufficient quantities obtained.[] [*]

NF-κB activity

The crude ethanolic bark extract and subfractions (obtained by Sephadex separation) which later yielded compounds 1 - 6 were tested and found to have inhibitory activity at 200 μg/mL using EMSA. The crude and ethyl acetate extract were tested using the IL-6 luciferase reporter gene assay method at 100 μg/mL, and reduced NF-κB activity to 42 % ± 8 % and 30 % ± 9 % (mean and S.D. of 9 replicates obtained from three experiments), respectively. Some cytotoxicity in HeLa cells was observed in most of the extracts, which was more evident in 5 and 6. Both were tested at reduced concentrations where cytotoxicity was removed, but along with the other isolated compounds showed no significant NF-κB inhibitory activity.

Cytotoxic activity

Compound 5 showed cytotoxic activity against MCF-7 breast cancer cells, and 6 showed moderate toxicity. Compound 3 exhibited mild cytotoxicity, and also produced a particularly steep dose-response curve, reducing cell viability from 100 % at 42 ± 7 μM to 10 % at 78 ± 3 μM (Fig. [1]). Compounds 1 and 2 did not show toxicity below 100 μM. IC₅₀ values obtained were 3 : 56 ± 7 μM, 5 : 7 ± 0.5 μM, and 6 : 35 ± 7 μM. The crude extract gave an IC₅₀ of 52 ± 10 μg/mL. In control experiments doxorubicin gave an IC₅₀ of 0.1 ± 0.01 μM.

Anti-bacterial activity

Anti-bacterial activity against the three strains of multi-drug resistant Staphylococcus aureus was found for compounds 5 and 6, the minimum inhibitory concentrations (MICs) of which are presented in Table [4], alongside those for control antibiotics. Substances 1, 2 and 3 showed no activity at 64 μg/mL; 6 showed good activity, with an MIC of 8 μg/mL (15 μM) for the three strains, and 5 showed good activity for one strain. The crude bark extract was also tested against Escherichia coli, but showed no activity at 512 μg/mL.

Discussion

While biflavonoids and related compounds are well known from the genus and the family, this report adds to our knowledge about their biological and pharmacological effects.

Calodenin B (5) was found to have strong cytotoxic effects against MCF-7 breast cancer cells. It is interesting to see that although 5 and 6 are similar with the exception of a double bond in 5, 5 is significantly more toxic. The cytotoxicity dose-response curve for 3 which shows it reducing cell viability from 100 % to 0 % within a narrow range of concentrations is interesting, although the cause is unknown.

Some isoflavonoids and lignans are referred to as phytoestrogens, with known oestrogenic effects. Therefore the presence of compounds like calodenin B, ochnone, and derivatives with a chalcone structure in Ochna macrocalyx may account for its use in gynaecological problems by the Washambaa. Phytoestrogens are also thought to have anti-oestrogenic properties [11], which may be a mode of action for 5 and 6 in the cytotoxicity assay.

For the anti-bacterial assays, compound 6 showed good activity against the three strains of mdr S. aureus, and 5 showed good activity against one strain. Again it is interesting to see that although these two compounds are structurally similar, there is improved activity in 6 in two of the mdr S. aureus strains (RN 4220 and 1199-B). The strong antibacterial activity of 6 against mdr S. aureus may make it a good candidate for further investigation, and though moderate cytotoxic activity was seen in the MCF-7 breast cancer cells, it was not cytotoxic at the minimum inhibitory concentration (15 μM). The crude bark extract was also tested against E. coli, which is known to cause gastrointestinal problems, especially diarrhoea, one of the traditional uses of the bark, but was found to have no activity against the organism.

The inhibitory effect of the crude extract in the NF-κB assay led us to investigate this botanical drug [12], however none of the pure compounds isolated were shown to have any inhibitory activity. The reason for this remains speculative, it may be that the observed NF-κB inhibition was related to the cytotoxic effects of the extract, or that synergistic or additive effects are of importance [13]. In the latter case further phytochemical work may lead to active minor compounds not detected in this study.

In conclusion, our results provide data which support some of the traditional uses of Ochna macrocalyx, and indicate that the compounds isolated have some interesting biochemical effects.

Acknowledgements

We would like to thank the Washambaa who provided information on the traditional use of as well as the bark of the Ochna. Research in our laboratory greatly benefits from funding by the European Union (Framework 5: The Cell Factory). S. Tang would also like to thank the School of Pharmacy for providing the funding for her research.

References

• 1 Schlage C, Mabula C, Mahunnah R LA, Heinrich M. Medicinal plants of the Washambaa (Tanzania): Documentation and ethnopharmacological evaluation.  Plant Biology. 2000;  2 83-92
  Thieme ConnectPubMedSearch in Google Scholar
• 2 Hansen M B, Nielsen S E, Berg K. Re-examination and further development of a precise and rapid dye method for measuring cell growth/cell kill.  J Immunol Methods. 1989;  119 203-10
  CrossrefPubMedSearch in Google Scholar
• 3 Bork P M, Bacher S, Schmitz M L, Kaspers U, Heinrich M. Hypericin as a non-antioxidant inhibitor of NF-κB.  Planta Medica. 1999;  65 297-300
  Thieme ConnectPubMedSearch in Google Scholar
• 4 Gibbons S, Udo E E. The effect of reserpine, a modulator of multidrug efflux pumps, on the in vitro activity of tetracycline against clinical isolates of methicillin resistant Staphylococcus aureus (MRSA) possessing the tet (K) determinant.  Phytother Res. 2000;  14 139-40
  CrossrefPubMedSearch in Google Scholar
• 5 Moreira I C, Sobrinho D C, De Carvalho M G, Braz-Filho R. Isoflavonone dimers hexaspermone A, B and C from Ouratea hexasperma .  Phytochemistry. 1994;  35 1567-72
  CrossrefPubMedSearch in Google Scholar
• 6 Hostettmann K, Marston A. Countercurrent chromatography in the preparative separation of plant derived natural products.  J Liq Chromatogr RT. 2001;  24 1711-21
  PubMedSearch in Google Scholar
• 7 Marston A, Zagorski M G, Hostettmann K. Antifungal polyphenols from Cordia goetzei .  Helv Chim Acta. 1988;  71 1210-9
  CrossrefPubMedSearch in Google Scholar
• 8 Drewes S E, Hudson N A, Bates R B, Linz G S. Novel dimeric chalcone based pigments from Brackenridgea zanguebarica .  Tetrahedron Lett. 1984;  25 105-8
  CrossrefPubMedSearch in Google Scholar
• 9 Drewes S E, Hudson N A, Bates R B, Linz G S. Medicinal plants of Southern Africa. Part 1. Dimeric chalcone-based pigments from Brackenridgea zanguebarica .  J Chem Soc Perkin Trans. 1987;  I 2809-13
  CrossrefPubMedSearch in Google Scholar
• 10 Messanga B, Tih R G, Sondengam B L, Martin M T. et al . Biflavonoids from Ochna calodendron .  Phytochemistry. 1994;  35 791-4
  CrossrefPubMedSearch in Google Scholar
• 11 Ingram D, Sanders K, Kolybaba M, Lopez D. Case-control study of phytoestrogens and breast cancer.  Lancet. 1997;  350 990-4
  CrossrefPubMedSearch in Google Scholar
• 12 Bremner P, Heinrich M. Natural products as modulators of the NF-κB-pathway.  Journal of Pharmacy and Pharmacology. 2002;  54 453-72
  CrossrefPubMedSearch in Google Scholar
• 13 Williamson E M. Synergy and other interactions in phytomedicines.  Phytomedicine. 2001;  8 401-9
  CrossrefPubMedSearch in Google Scholar

Prof. Dr. M. Heinrich

Centre for Pharmacognosy and Phytotherapy

The School of Pharmacy

29-39 Brunswick Square

London, WC1N 1AX

United Kingdom

Fax: +44 0207 753 5909

Email: phyto@ulsop.ac.uk

References

• 1 Schlage C, Mabula C, Mahunnah R LA, Heinrich M. Medicinal plants of the Washambaa (Tanzania): Documentation and ethnopharmacological evaluation.  Plant Biology. 2000;  2 83-92
  Thieme ConnectPubMedSearch in Google Scholar
• 2 Hansen M B, Nielsen S E, Berg K. Re-examination and further development of a precise and rapid dye method for measuring cell growth/cell kill.  J Immunol Methods. 1989;  119 203-10
  CrossrefPubMedSearch in Google Scholar
• 3 Bork P M, Bacher S, Schmitz M L, Kaspers U, Heinrich M. Hypericin as a non-antioxidant inhibitor of NF-κB.  Planta Medica. 1999;  65 297-300
  Thieme ConnectPubMedSearch in Google Scholar
• 4 Gibbons S, Udo E E. The effect of reserpine, a modulator of multidrug efflux pumps, on the in vitro activity of tetracycline against clinical isolates of methicillin resistant Staphylococcus aureus (MRSA) possessing the tet (K) determinant.  Phytother Res. 2000;  14 139-40
  CrossrefPubMedSearch in Google Scholar
• 5 Moreira I C, Sobrinho D C, De Carvalho M G, Braz-Filho R. Isoflavonone dimers hexaspermone A, B and C from Ouratea hexasperma .  Phytochemistry. 1994;  35 1567-72
  CrossrefPubMedSearch in Google Scholar
• 6 Hostettmann K, Marston A. Countercurrent chromatography in the preparative separation of plant derived natural products.  J Liq Chromatogr RT. 2001;  24 1711-21
  PubMedSearch in Google Scholar
• 7 Marston A, Zagorski M G, Hostettmann K. Antifungal polyphenols from Cordia goetzei .  Helv Chim Acta. 1988;  71 1210-9
  CrossrefPubMedSearch in Google Scholar
• 8 Drewes S E, Hudson N A, Bates R B, Linz G S. Novel dimeric chalcone based pigments from Brackenridgea zanguebarica .  Tetrahedron Lett. 1984;  25 105-8
  CrossrefPubMedSearch in Google Scholar
• 9 Drewes S E, Hudson N A, Bates R B, Linz G S. Medicinal plants of Southern Africa. Part 1. Dimeric chalcone-based pigments from Brackenridgea zanguebarica .  J Chem Soc Perkin Trans. 1987;  I 2809-13
  CrossrefPubMedSearch in Google Scholar
• 10 Messanga B, Tih R G, Sondengam B L, Martin M T. et al . Biflavonoids from Ochna calodendron .  Phytochemistry. 1994;  35 791-4
  CrossrefPubMedSearch in Google Scholar
• 11 Ingram D, Sanders K, Kolybaba M, Lopez D. Case-control study of phytoestrogens and breast cancer.  Lancet. 1997;  350 990-4
  CrossrefPubMedSearch in Google Scholar
• 12 Bremner P, Heinrich M. Natural products as modulators of the NF-κB-pathway.  Journal of Pharmacy and Pharmacology. 2002;  54 453-72
  CrossrefPubMedSearch in Google Scholar
• 13 Williamson E M. Synergy and other interactions in phytomedicines.  Phytomedicine. 2001;  8 401-9
  CrossrefPubMedSearch in Google Scholar

Prof. Dr. M. Heinrich

Centre for Pharmacognosy and Phytotherapy

The School of Pharmacy

29-39 Brunswick Square

London, WC1N 1AX

United Kingdom

Fax: +44 0207 753 5909

Email: phyto@ulsop.ac.uk