Available online at www.sciencedirect.com PHYTOCHEMISTRY Phytochemistry 69 (2008) 982–987 www.elsevier.com/locate/phytochem Constituents of the stem bark of Discopodium penninervium and their LTB4 and COX-1 and -2 inhibitory activities Abraham Abebe Wube a, Eva-Maria Wenzig a, Simon Gibbons b, Kaleab Asres c, Rudolf Bauer a, Franz Bucar a,* a b Institute of Pharmaceutical Sciences, Department of Pharmacognosy, Karl-Franzens University Graz, Universitaetsplatz 4/1, A-8010 Graz, Austria The School of Pharmacy, Centre for Pharmacognosy and Phytotherapy, University of London, 29-39 Brunswick Square, London WC1N 1AX, UK c The School of Pharmacy, Department of Pharmaccognosy, Addis Ababa University, P.O. Box 1176, Addis Ababa, Ethiopia Received 21 September 2007; received in revised form 5 November 2007 Available online 20 December 2007 Abstract The stem bark of Discopodium penninervium afforded a withanolide, 6a,7a-epoxy-1-oxo-5a,12a,17a-trihydroxywitha-2,24-dienolide (1) and a coloratane sesquiterpene, 7a,11a-dihydroxy-4(13),8-coloratadien-12,11-olide (4) along with five known compounds, withanone (2), 5a,17b-dihydroxy-6a,7a-epoxy-1-oxowitha-2,24-dienolide (3), 7a,11a-dihydroxy-8-drimen-12,11-olide (5), withasomnine (6), and (E,Z)-9-hydroxyoctadeca-10,12-dienoic acid (7). The identity of the compounds was established on the basis of spectroscopic data analysis. All compounds were assessed for inhibition of leukotriene metabolism in an in vitro bioassay using activated human neutrophile granulocytes, and for in vitro cycloxygenase-1 and -2 inhibition from sheep cotyledons and seminal vesicles, respectively. In the leukotriene biosynthesis assay all compounds tested at a concentration of 50 lM exhibited activity with percentage inhibitions ranging from 11.5 to 36.6. The withanolide, 1, displayed a 46.4% inhibition of COX-2 and a 22.9% inhibition of LTB4 formation at 50 lM concentration. Compounds 4 and 6 inhibited LTB4 biosynthesis but showed minor inhibition of COX-1 and COX-2. The remaining compounds, on the other hand, were found to be inactive on COX enzymes. Ó 2007 Elsevier Ltd. All rights reserved. Keywords: Discopodium penninervium; Solanaceae; LTB4; Cyclooxygenase; Withanolide; Coloratane sesquiterpene; 6a,7a-Epoxy-1-oxo-5a,12a,17a-trihydroxywitha-2,24-dienolide; 7a,11a-Dihydroxy-4(13),8-coloratadien-12,11-olide 1. Introduction Discopodium penninervium Hochst (Solanaceae), a shrub or small tree, is the only species in the genus and endemic to Ethiopia. The leaves and barks are used for the treatment of schistosomiasis, leprosy and stomach-ache in Ethiopian folk medicine (Geyid et al., 2005). Previous phytochemical investigation of D. penninervium revealed the presence of 5a,17b-dihydroxy-6a,7a-epoxy-1oxowitha-2,24-dienolide, withanone and withanolide A in the roots (Habtemariam and Gray, 1998), and 5,6-epoxy* Corresponding author. Tel.: +43 316 3805531; fax: +43 316 3809860. E-mail address: franz.bucar@uni-graz.at (F. Bucar). 0031-9422/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.phytochem.2007.11.001 16-oxygenated withanolides, jaborosalactone-L, and 17epiacnistin-A in the leaves (Habtemariam et al., 1993, 2000). The latter Discopodium withanolides have been shown to exhibit cytotoxic activities to both human and murine cancer cell lines and immunosuppressive activity in rat spleen cells in vitro (Habtemariam et al., 2000). In the continuation of a search for biologically active constituents from Ethiopian medicinal plants, we report herein the isolation and structural elucidation of a new withanolide, 1, and a new coloratane sesquiterpene, 4, together with five known compounds from the stem bark of D. penninervium (see Fig. 1). The anti-inflammatory activity of the compounds was assessed in vitro using leukotriene formation, and cycloxygenase-1 and -2 inhibitory assays. 983 A.A. Wube et al. / Phytochemistry 69 (2008) 982–987 28 21 H R1 O 2 1 3 4 18 12 19 10 5 11 9 H 20 R2 17 16 15 13 14 24 23 22 O 25 O 11 12 O H 7 2 3 1 10 9 5 4 H O 8 7 6 14 H OH H HO 15 8 6 27 26 OH H O H Ha 13 4 1: R1 = OH; R 2 = α-OH 2: R1 = H; R2 = α-OH 3: R1 = H; R 2 = β-OH HO Hb N N H O O OH H H 5 O 6 agreement with a previous report by Habtemariam and Gray (1998).The presence of an epoxide between C-6 and C-7 was inferred from the low field resonances (56.9 and 57.3 ppm) observed for these two oxygenated carbons in the 13C NMR spectrum. A broad triplet at d 4.36 was assigned to a methine proton attached to C-12, as it showed a correlation with this carbon at d 76.3 in the HMQC spectrum, although a methine carbon resonance was absent in the 13C NMR spectrum. This assignment was further substantiated by HMBC cross peaks observed between C-12 and the methyl protons at position 18 and the 1H–1H COSY correlation between H-11 and H-12. The remaining methine proton resonances were placed by analysis of HMQC, HMBC and COSY spectra. The upfield carbon resonance at d 28.3 was assigned to C-9 on the basis of HMBC cross peaks observed for H-12/C-9 and H-19/C-9. All the resonances for methylene protons were assigned unambiguously by HMQC, HMBC and COSY NMR experiments. The relative configuration of the hydroxyl group at position 12 was established on the basis of the small coupling constant of H12 and by a NOESY NMR experiment (Fig. 2), showing OH OH 7 H OH H Fig. 1. The structures of compounds 1–7. OH O 2. Results and discussion H Compound 1 was obtained as amorphous powder and its molecular formula was deduced as C28H38O7 by the LCESI-MS data at m/z 487 [M+H]+. The HRMS showed an ion [MH] peak at m/z 485.2553 corresponding to the molecular formula, C28H38O7. An absorption maximum at 223 nm in the UV spectrum was indicative of an a,bunsaturated d-lactone structural feature, which was further confirmed by absorption band at 1694 cm1 observed in the IR spectrum. The broad absorption band at 3493 cm1 in the IR spectrum was attributed to hydroxyl stretching. The 1H NMR spectrum of 1 showed resonances for five methyl groups, with four tertiary methyls at d 0.90, 1.18, 1.79, 1.91 and one secondary methyl at d 1.14 (d, J = 6.7 Hz). The first two methyls resonances were assigned to CH3-18 and CH3-19, respectively, whereas the latter two downfield resonances were assigned in turn to CH3-27 and CH3-28, and were attached to olefinic carbons. The methyl signal resonated as a doublet was assigned to CH3-21. Among the ten methine protons that appeared in the 1H NMR spectrum, the deshielded doublet at d 5.67 and the double of doublets at d 6.62, which were coupled to one another, were assigned to H-2 and H-3 respectively. This further confirmed the presence of a,b-unsaturated ketone moiety. A 1H doublet at d 3.07 and a multiplet at d 3.33 were assigned to H-6 and H-7, respectively, and the small coupling (3.8 Hz) observed for H-6/H-7 indicated that these hydrogens are axial and equatorial, respectively. This is in H O O H H H Heq OH Hax O H 1 HO Heq H O Hax Heq O Hax H OH H Heq H Hax Hb Ha 4 Fig. 2. Major NOE correlations in compounds 1 and 4. 984 A.A. Wube et al. / Phytochemistry 69 (2008) 982–987 NOE enhancement between H-12 and CH3-18. The relative stereochemistry of C-17, C-20 and C-22, on the other hand, are the same as in compound 2. Comparison of the above data with those in the literature indicated that the structure of 1 is very closely related to that of withanone (2) (Habtemariam, 1997), except for the presence of an additional hydroxyl group at position 12 in 1. Thus, the structure of 1 was established as 6a,7a-epoxy-1-oxo-5a,12a,17a-trihydroxywitha-2,24-dienolide. Compound 4 was obtained as white needles in n-hexane/ EtOAC. Its molecular formula C15H20O4 was determined by HRMS (m/z; measured 263.1291 [MH]; calc. 263.1289). The proton and 13C NMR data were suggestive of a coloratane skeleton for 4 (Wube et al., 2005). The IR spectrum showed a strong absorption band at 1762 and 1740 cm1, indicating the presence of an a,b-unsaturated lactone (Sakio et al., 2001). This structural feature was further supported by an absorption maximum at 210 nm in the UV spectrum and double bond carbon resonances at d 129.5 (C-8) and 170.0 (C-9), and a signal for a lactone carbonyl carbon at d 170.7 (C-12) in the 13C NMR spectrum. In addition a hydroxyl stretch band at 3436 cm1 was observed in the IR spectrum. The 13C and DEPT NMR spectra also revealed the presence of two methyls, four methylenes, four methines, and five quaternary carbons. Two 3H signals, a singlet and a doublet, were observed in the 1H NMR spectrum of 4. The singlet at d 0.98, which showed HMBC cross peaks with C-1 at d 35.1 and C-9 at d 170.0, was assigned to CH3-15. The methyl doublet gave HMBC correlations with C-2 at d 32.6, C-3 at d 39.1 and C-4 at d 153.5 was assigned to CH3-14. The signals for the H-13a and H-13b protons were clearly assigned based on NOE correlations observed between CH3-14 and H13a and between H-6a and H-13b in the NOESY spectrum. Four 1H signals at d 2.13, 2.53, 4.48, and 6.22 were assigned to H-3, H-5, H-7, and H-11, respectively. The chemical shifts for the remaining methylene protons were assigned by a detailed analysis of HMQC and HMBC spectra. Exocyclic double bond carbon resonances at d 153.5 (C-4) and 105.1 (C-13), observed in the 13C NMR spectrum, were characteristic of coloratane sesquiterpenes. The carbon resonances at d 97.8 and 59.7 were assigned to a hemiketal C-11 and the hydroxyl bearing C-7, respectively. The relative configuration of the five asymmetric centres in the coloratane skeleton was determined by analysis of the coupling constants and NOESY experiments. The NOESY spectrum of 4 showed correlations between the protons of H-3 and H-1a as well as the H-3 and H-5. These correlations require that H-3 and H-5 are cis to each other, which would enable the C-3 and C-5 protons to have a-orientation. Similarly, H-11 exhibited a NOE correlation (Fig. 2) with CH3-15, and this implied that H-11 and CH315 are on the same relative axial face as each other. This would imply that hydroxyl group at C-11 be equatorial and alpha. A small coupling constant value of H-7 (d 4.48, d, J = 3 Hz) suggested an a-orientation for the hydro- xyl group. On the basis of these observations, compound 4 was established structurally as 7a,11a-dihydroxy-4(13),8coloratadien-12,11-olide. The known compounds were identified by analysis of their physical and spectral data and by comparison with published values as withanone (2) (Habtemariam and Gray, 1998), 5a,17b-dihydroxy-6a,7a-epoxy-1oxowitha-2, 24-dienolide (3) (Nittala and Lavie, 1981), 7a,11a-dihydroxy-8-drimen-12,11-olide (5) (Sakio et al., 2001), withasomnine (6) (Houghton et al., 1994), and (E,Z)-9hydroxyoctadeca-10,12-dienoic acid (7) (Murakami et al., 1992). The drimane sesquiterpene, 7a,11a-dihydroxy-8-drimen-12,11-olide (5), was previously isolated from the animal species, Dendrodoris carbunculosa (Sakio et al., 2001) and this is the first report of its occurrence in the plant kingdom. Pyrazole alkaloids appear to be rare in plants of the Solanaceae, although they have been found to occur, for example in Newbouldia laevis (Bignoniaceae) (Aladesanmi et al., 1998), Elytraria acaulis (Acanthaceae) (Ravikanth et al., 2001) and Withania somnifera (Solanaceae) (Schroeter et al., 1966). The occurrence of withasomnine (6) in D. penninervium is interesting from a chemotaxonomic point of view because it was also isolated from the Indian medicinal plant W. somnifera. Moreover, this is the first report of the occurrence of drimane and coloratane sesquiterpenes in the family Solanaceae. All compounds were evaluated for inhibition of leukotriene metabolism as well as COX-1 and -2 enzymes in vitro. The results presented in Table 3 show that all compounds tested in the leukotriene biosynthesis assay showed some inhibitory effects on LTB4 formation. The sesquiterpene 4 displayed the highest activity against LTB4. On the other hand, compound 5, the biogenetic precursor of compound 4 was found to be three times less active than compound 4. Thus, the presence of an exocyclic methylene group at position 4 might enhance the LTB4 inhibitory activity of compound 4. This is the first report on LTB4 inhibitory activity of withanolides. Recently, we have reported the leukotriene biosynthesis inhibitory effect of drimane and coloratane sesquiterpenes obtained from the stem bark of Warburgia ugandensis (Wube et al., 2006) and results of our study revealed that compounds having a dialdehyde structural feature displayed superior inhibitory effects on the stable metabolite of 5-LOX, LTB4. The in vitro COX-1 and -2 test results revealed that at a concentration of 50 lM, compounds 4 and 6 showed dual COX-1 and -2 inhibition, whereas compounds 2, 3, 5, and 7 produced no inhibitory effects. The new withanolide, compound 1, displayed a selective inhibition of the COX-2 enzyme with an IC50 value close to 50 lM. Unfortunately, due to lack of compound higher amounts could not be tested. This is the second report on the cycloxygenase inhibitory activity of withanolides and the selective COX2 inhibitory potency of compound 1 is much higher than the withanolide obtained from W. somnifera (Jayaprakasam and Nair, 2003). As compared to compounds 2 and A.A. Wube et al. / Phytochemistry 69 (2008) 982–987 3, the COX-2 inhibitory potency of compound 1 might be enhanced by the hydroxyl group at position 12. Concerning anti-inflammatory drugs, selective COX-2 inhibitors are considered to have advantages over nonselective NSAIDs by a lower risk of gastrointestinal side effects (Smith et al., 2000). Furthermore, it has been evident that COX-2 and 5-LOX pathways are both involved in cell proliferation and angioneogenesis. Therefore the withanolide 1 represents interesting features of a dual inhibitor of COX-2 and leukotriene formation and may serve for the development of anti-inflammatory and cancer chemo preventive agents. However, further structural variations should be evaluated in order to improve the inhibitory potency of 1 in the COX-2 and LTB4 assays. 3. Experimental 3.1. General experimental procedures Melting points were determined with Kofler microscope and are uncorrected. Optical rotations were measured on a Perkin–Elmer 241 MC polarimeter. UV spectra were recorded on Shimadzu UV-160A spectrophotometer. Perkin–Elmer 881 infrared spectrophotometer was used in recording the IR spectra. NMR spectra were recorded at 500 MHz for 1H and 125 MHz for 13C on a Bruker AVANCE 500 spectrometer. All spectra were measured in CDCl3, except for compounds 1 and 4, which were dissolved in (CD3)2CO. HRMS were determined with Micromass QTOF Ultima using the internal standard TCA, which had [MH] = m/z 514.2839. Mass spectra were also obtained by LC-ESI-MS analysis on a Thermo-Finnigan LCQ Deca XP Plus mass spectrometer connected to a Surveyor LC-system (Thermo-Finnigan). The absorbance for LTB4 quantification was conducted using a Tecan RAIN BOW photometric ELISA plate reader. Chromatographic separation were performed by analytical TLC on Si gel 60 F254 (0.2 mm thick), column chromatography on Si gel 60 (70–240 mesh), size exclusion chromatography on Sephadex LH-20, solid phase separation on Isolute C18 (10 g) columns, and semipreparative HPLC with LiChrospherÒ RP-18 (10 lm, 250 times 10 mm i.d.) column. Türks solution, Na2EDTA, CaCl2  2H2O p.a., anhydrous D-glucose, MgCl2  6H2O  KCl, Tris p.a., formic acid and ethanol p.a. were purchased from Merck. Trypan blue solution and eicosatetraenoic acid were obtained from Sigma Chemicals. Ca ionophor A 23187 and epinephrinehydrogentartrate were bought from Fluka. A LTB4 EIA kit, purified PGHS-1 and -2, indomethacin, NS-398 and arachidonic acid were obtained from Cayman Chemicals, Ann Arbor, MI, USA. A PGE2-EIA kit was obtained from R&D systems, Minneapolis, MN, USA. TRIS/HCl buffer (pH 8.0) was bought from Roth. Hematin was obtained from porcine, ICN, Aurora, OH, USA. Zileuton was purchased from Sequoia Research Products Ltd., Oxford, UK. 985 3.2. Plant material The stem bark of D. penninervium was collected in April 2001 from Dinsho towards Addis Ababa and Dodola in Bale zone, Ethiopia and identified by Mr. Melaku Wondafrash, the National Herbarium, Department of Biology, Addis Ababa University. A voucher specimen (collection number 1487) has been deposited in the National Herbarium, Department of Biology, Addis Ababa University, Addis Ababa, Ethiopia. 3.3. Extraction and isolation The air-dried powdered stem barks of D. penninervium (600 g) were extracted successively with petroleum ether and dichloromethane in a Soxhlet apparatus. The dichloromethane extract was concentrated at reduced pressure to give 5 g of a greenish residue. After removal of chlorophyll with Sephadex LH-20 eluting with CH2Cl2–MeOH (1:1), the residue (1.2 g) was subjected to solid-phase separation using a H2O/MeOH (100:0 ? 0:100) gradient elution and ten fractions of 100 ml each were collected. Fraction 4, eluted with H2O-MeOH (7:3), was purified by semi-prep. HPLC using MeCN/H2O (35:65 ? 55:45) gradient elution for 25 min to afford 5 (2 mg). Fr. 8, eluted with H2O/MeOH (3:7), was subjected to semi-prep. HPLC using MeCN/H2O (4:6 ? 1:1) gradient elution for 40 min to yield 1 (3 mg), 3 (6.5 mg) and 2 (9 mg) at 24, 32 and 34.7 min, respectively. Fr. 9 and 10 were combined and applied to a Sephadex-20 column using CH2Cl2 as eluent to give 56 subfractions of 20 ml each. Subfr. 19–27 gave 4 (11 mg) after purification by semi-prep. HPLC using a MeCN/H2O (35:65 ? 65:35) gradient system as eluents. Subfr. 31–46 were further chromatographed on semi-prep. HPLC using a MeCN/H2O (45:55 ? 8:2) gradient elution for 45 min to afford 6 (2 mg) and 7 (14 mg) at 9 and 37 min, respectively. 3.4. In vitro leukotriene metabolism, cycloxygenase-1, and -2 inhibitory assays The leukotriene metabolism inhibitory assay was conducted as described previously (Adams et al., 2004) using human neutrophile granulocytes with some modification. After isolation of human neutrophile granulocytes from human blood, cell vitality test, cell concentration determination, and incubation with test sample solution the samples were diluted 40-fold and the free LTB4 concentration was measured using a competitive LTB4 EIA kit. LTB4 biosynthesis inhibition was quantified by measuring the absorption at 405 nm after addition of 50 ll aqueous Na3PO4 as stop solution. Inhibition was expressed in percent in relation to a control using abs. EtOH. Inhibition values are means of three experiments and each sample was tested in duplicate. Zileuton was used as a positive control. The cycloxygenase-1 and -2 assays were done with purified PGHS-1 from ram seminal vesicle for COX-1 and purified PGHS-2 from sheep placental cotyledons for COX-2 986 A.A. Wube et al. / Phytochemistry 69 (2008) 982–987 as reported previously (Fiebich et al., 2005; Rollinger et al., 2005). The percent inhibitions at 50 lM concentration of test compounds were determined for both enzymes. Inhibition values were means of three experiments and each sample was tested in duplicate. Indometacin and NS-398 were used as positive control for COX-1 and -2, respectively. 3.5. 6a,7a-Epoxy-1-oxo-5a,12a,17a-trihydroxywitha-2,24dienolide (1) Colourless optically inactive amorphous powder, UV kMeOH nm (log e) 223 (3.86), IR mMeOH cm1: 3493, 1694, max max 1 1385, 1130, 1094, H NMR (500 MHz (CD3)2CO) and 13 C NMR (125.8 MHz (CD3)2CO) data, see Table 1, ESIMS (70 eV) m/z (rel. Int.): 487 [M+H]+ (34), 469 [MOH]+ (100), 452 (62), 171 (14); HRMS found 485.2553; C28H38O7, calc. 485.2545. 3.6. 7a,11a-Dihydroxy-4(13),8-coloratadien-12,11-olide (4) White needles from n-hexane-EtOAC mixture, m.p. 28 143–145 °C, ½aD +90.0 (MeOH; c0.90), UV kMeOH nm max Table 1 1 H, 13C NMR and HMBC correlations data for compound 1 in acetoned6a Position 1 2 3 4a 4b 5 6 7 8 9 10 11a 11b 12 13 14 15a 15b 16a 16b 17 18 19 20 21 22 23a 23b 24 25 26 27 28 dH (J in Hz) 5.67 6.62 2.56 2.77 (dd) (10.1, 2.2) (ddd) (10.1, 5.2, 2.2) (dd) (14.5, 3.5) (m) 3.07 3.33 1.98 2.01 (d) (3.8) (m) (m) (m) 2.73 (m) 1.60 (m) 4.36 (t) (2.2) 1.84 2.07 1.36 2.37 1.45 (m) (m) (m) (m) (m) 0.90 1.18 2.13 1.14 4.55 2.21 2.45 (s) (s) (m) (d) (6.7) (dt) (12.5, 3.5) (m) (dd) (18.5, 9.5) 1.79 (s) 1.91 (s) Table 2 1 H, 13C NMR and HMBC correlations data for compound 4 in acetoned6a Position dH (J in Hz) 1a 1b 2a 2b 3 4 5 6a 6b 7 8 9 10 11 12 13a 13b 14 15 1.78 1.99 1.35 1.80 2.13 2.53 1.85 1.89 4.48 6.22 4.86 4.67 1.09 0.98 dC, multiplicity HMBC (dt) (12.6, 4.0) 35.1, CH2 (m) (m) 32.6, CH2 (m) (m) 39.1, CH 153.5, C (d) (12.0) 44.5, CH (m) 32.1, CH2 (m) (d) (3.0) 59.7, CH 129.5, C 170.0, C 39.4, C (s) 97.8, CH 170.7, C (bs) 105.1, CH2 (bs) (d) (6.5) 18.5, CH3 (s) 17.2, CH3 H-15 H-14 H-13a, H-14 H-3, H-6a, H-14 H-6b, H-7, H-13b, H-15 H-6a H-6 a, H-7, H-11 H-15 H-6a, H-15 H-11 H-3 a Chemical shifts are in ppm relative to TMS; 500 MHz for 1H and 125 MHz for 13C. dC, multiplicity HMBC Table 3 Leukotriene metabolism, COX-1 and -2 inhibitory activities of compounds isolated from D. penninervium at 50 lM test concentration and IC50 values of the positive controls zileuton, indomethacin and NS-398 203.6, C 129.3, CH 141.3, CH 37.8, CH2 H-19 Compound 74.1, 56.9, 57.3, 37.2, 28.3, 51.6, 31.4, C CH CH CH CH C CH2 H-19 76.3, 50.2, 40.8, 23.1, CH C CH CH2 H-2 H-12, H-14 H-19 H-18 H-18 H-18 37.7, CH2 87.1, C 16.4, CH3 15.1, CH3 43.7, CH 9.2, CH3 79.7, CH 33.4, CH2 151.4, C 121.7, C 166.9, C 12.6, CH3 20.5, CH3 H-18, H-21 H-21 % Inhibition ± SD at 50 lM test concentration LTB4 COX-1 COX-2 1 2 3 4 5 6 7 22.9 ± 1.98 28.7 ± 1.54 18.9 ± 6.63 36.6 ± 1.38 11.5 ± 3.98 25.6 ± 3.24 22.5 ± 6.92 NA NA NA 2.1 ± 9.15 NA 6.8 ± 4.59 NA 46.4 ± 2.23 NA NA 11.83 ± 6.86 NA 6.4 ± 2.31 NA Zileuton Indomethacin NS-398 10.0 lM (IC50) ND ND ND 1.25 lM (IC50) ND ND ND 5.0 lM (IC50) NA, not active and ND, not determined. (log e): 210 (3.96), IR mMeOH cm1: 3436, 1762, 1740, 1177, max 1 1036, 940, 904, H NMR (500 MHz (CD3)2CO) and 13C NMR (125.8 MHz (CD3)2CO) data, see Table 2, HRMS found 263.1291; C15H20O4, calc. 263.1289. Acknowledgements H-21 H-27, H-28 H-27, H-28 H-27 a Chemical shifts are in ppm relative to TMS; 500 MHz for 1H and 125 MHz for 13C. We thank Mrs. Elke Prettner, Department of Pharmaceutical Chemistry, Institute of Pharmaceutical Sciences, University of Graz for recording the IR and UV spectroscopic and optical rotation measurements and Mr. M. Wondafrash, the National Herbarium, Department of Biology, Addis Ababa University for plant identification. Financial support by the Austrian Exchange Service OEAD is gratefully acknowledged. A.A. Wube et al. / Phytochemistry 69 (2008) 982–987 References Adams, M., Kunert, O., Haslinger, E., Bauer, R., 2004. Inhibition of leukotriene biosynthesis by quinolone alkaloids from the fruit of Evodia rutaecarpa. Planta Med. 70, 904–908. Aladesanmi, A.J., Nia, R., Nahrstedt, A., 1998. New pyrazole alkaloids from the root bark of Newbouldia laevis. Planta Med. 64, 90–91. Fiebich, B.L., Grozdeva, M., Hess, S., Hüll, M., Danesch, U., Bodensieck, A., Bauer, R., 2005. Petasites hybridus extracts in vitro inhibit COX-2 and PGE2 released by direct interaction with enzyme and by preventing p42/44 MAP kinase activation in rat primary microglial cells. Planta Med. 71, 12–19. Geyid, A., Abebe, D., Debella, A., Mekonnen, Z., Aberra, F., Teka, F., Kebede, T., Urga, K., Yersaw, K., Biza, T., Hailemariam, B., Guta, M., 2005. Screening of some medicinal plants of Ethiopia for their anti-microbial properties and chemical profiles. J. Ethnopharmacol. 97, 421–427. Habtemariam, S., 1997. Cytotoxicity and immunosuppressive activity of withanolides from Discopodium penninervium. Planta Med. 63, 15–17. Habtemariam, S., Gray, A.I., 1998. Withanolides from the roots of Discopodium penninervium. Planta Med. 64, 275–276. Habtemariam, S., Gray, A.I., Waterman, P.G., 1993. 16-Oxygenated withanolides from the leaves of Discopodium penninervium. Phytochemistry 34, 807–811. Habtemariam, S., Skelton, B.W., Waterman, P.G., White, A.H., 2000. 17Epiacnistin-A, a further withanolide from the leaves of Discopodium penninervium. J. Nat. Prod. 63, 512–513. Houghton, P., Pandey, R., Hawkes, J.E., 1994. Naphthaquinones and an alkaloid from roots of Newbouldia laevis. Phytochemistry 35, 1602– 1603. 987 Jayaprakasam, B., Nair, M.G., 2003. Cyclooxygenase-2 enzyme inhibitory withanolides from Withania somnifera leaves. Tetrahedron 59, 841– 849. Murakami, N., Shirahashi, H., Nagatsu, A., Sakakibara, J., 1992. Two unsaturated 9R-hydroxy fatty acids from the cyanobacterium Anabaena flos-aquae f. flos-aquae. Lipids 27, 776–778. Nittala, S.S., Lavie, D., 1981. Chemistry and genetics of withanolides in Withania somnifera hybrids. Phytochemistry 20, 2741–2748. Ravikanth, V., Ramesh, P., Diwan, P.V., Venkateswarlu, Y., 2001. Pyrazole alkaloids from Elytraria acaulis. Biochem. Syst. Ecol. 29, 753–754. Rollinger, J.M., Bodensieck, A., Seger, C., Ellmerer, E.P., Bauer, R., Langer, T., Stuppner, H., 2005. Discovering COX-2 inhibiting constituents of Morus root bark: activity guided versus computer aided methods. Planta Med. 71, 399–405. Sakio, Y., Hirano, Y.J., Hayashi, M., Komiyama, K., Ishibashi, M., 2001. Dendocarbins A–N, new drimane sesquiterpenes from the nudibranch Dendrodoris carbunculosa. J. Nat. Prod. 64, 726–731. Schroeter, H.B., Neumann, D., Katritzky, A.R., Swinboume, F.J., 1966. Withasomnine. Tetrahedron 22, 2895–2897. Smith, W.L., DeWitt, D.L., Garavito, R.M., 2000. Cyclooxygenase: structural, cellular and molecular biology. Anu. Rev. Biochem. 69, 145–182. Wube, A.A., Bucar, F., Gibbons, S., Asres, K., 2005. Sesquiterpenes from Warburgia ugandensis and their antimycobacterial activity. Phytochemistry 66, 2309–2315. Wube, A.A., Gibbons, S., Asres, K., Streit, B., Adams, M., Bauer, R., Bucar, F., 2006. In vitro 12(S)-HETE and leukotriene metabolism inhibitory activity of sesquiterpenes from Warburgia ugandensis. Planta Med. 72, 754–756.