Journal of Ethnopharmacology 97 (2005) 285291

Phytochemical and pharmacological screening of Sterculiaceae species and isolation of antibacterial compounds
K.A. Reida , A.K. J gera , M.E. Lighta , D.A. Mulhollandb , J. Van Stadena, a
a

Research Centre for Plant Growth and Development, School of Botany and Zoology, University of KwaZulu-Natal Pietermaritzburg, Private Bag X01, Scottsville 3209, South Africa b Natural Products Research Group, School of Pure and Applied Chemistry, University of KwaZulu-Natal Durban, Durban 4041, South Africa Received 22 March 2004; received in revised form 7 October 2004; accepted 12 November 2004 Available online 1 January 2005

Abstract Little previous phytochemical investigation has been conducted on South African Sterculiaceae species used in traditional medicine. In this study, ve species, varying in growth type (small herbs, shrubs and large trees) and traditional usage were investigated. The species screened were Cola greenwayi Brenan, Cola natalensis Oliv., Dombeya burgessiae Gerr. ex Harv., Dombeya cymosa Harv. and Hermannia depressa N.E.Br. Extracts were screened for alkaloids, cardiac glycosides, cyanogenic glycosides, saponins and tannins. The probable presence of bufadienolides in the leaf material of Dombeya burgessiae and Dombeya cymosa was determined. Alkaloids, cyanogenic glycosides and saponins were absent in all the plant material investigated. Tannins were detected in the leaf extract of Cola greenwayi and in the leaves, stems and roots of Hermannia depressa. Extracts were screened for anti-inammatory and antibacterial activity using the cyclooxygenase-1 (COX-1) inhibition assay and the microdilution antibacterial assay. The ethanol and dichloromethane extracts of Cola greenwayi, Dombeya burgessiae and Dombeya cymosa, and the dichloromethane extracts of Hermannia depressa showed the highest levels of COX-1 inhibition. It is possible that the high levels observed may be due to the presence of tannins in some of the extracts. Generally, all the aqueous extracts exhibited low activity. Similarly, no antibacterial activity was observed with the aqueous extracts, although some mild activity was exhibited with some of the ethanol and ethyl acetate extracts. Following the general phytochemical and pharmacological screening, extracts showing antibacterial activity were further puried using bioassay-guided fractionation. Dombeya rotundifolia (Hochst.) Planch., which was screened in a previous study, was also included in the isolation of active compounds. A bioautographic assay, using Staphylococcus aureus, was used to detect the presence of the antibacterial compounds. These were isolated and identied as fatty acids. 2004 Elsevier Ireland Ltd. All rights reserved.
Keywords: Sterculiaceae; Cola sp.; Dombeya sp.; Hermannia sp.; Antibacterial; Anti-inammatory; Fatty acids

1. Introduction In developing countries, bacterial infections are widespread, especially in informal settlements, due to poor sanitation and unhygienic conditions. Furthermore, diseases such as AIDS, malaria and tuberculosis, result in high mortality rates (Rasoanairo and Ratsimamanga-Urverg, 1993). Diarrhoea is a prominent clinical feature of childhood malnu

Corresponding author. Tel.: +27 33 260 5130; fax: +27 33 260 5897. E-mail address: vanstadenj@ukzn.ac.za (J.V. Staden).

trition and is largely due to gastrointestinal infections and infestations by bacteria (Gracey, 1985; Sleigh and Timbury, 1998). In many of these developing countries, and in South Africa, traditional medicine is widely used to treat many of these common ailments. The treatments, in most cases, are administered by traditional healers and generally consist of crude plant material and extracts. This study was conducted as part of an effort to validate the use of traditional medicine in South Africa, and investigate plants that are used locally, particularly for antibacterial activity. Five Sterculiaceae species, Cola greenwayi Brenan,

0378-8741/$ see front matter 2004 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.jep.2004.11.010

286

K.A. Reid et al. / Journal of Ethnopharmacology 97 (2005) 285291

Cola natalensis Oliv., Dombeya burgessiae Gerr. ex Harv., Dombeya cymosa Harv. and Hermannia depressa N.E.Br. were selected for screening in phytochemical tests and for biological activity. Extracts were tested for alkaloids, cardiac glycosides, cyanogenic glycosides, saponins and tannins. The extracts were also tested for anti-inammatory activity in the cyclooxygenase-1 (COX-1) inhibition assay, and for antibacterial activity using a microdilution assay. Following the initial screening, active antibacterial components were isolated and identied from the species that showed antibacterial activity. Cola greenwayi (hairy cola or coshwood) is a small understorey tree that grows to a height of 57 m. It is recognizable by the furry, dark brown swelling just below the leaf blade (Pooley, 1993) and occurs in evergreen and sand forests (Coates Palgrave, 1977). Although Cola greenwayi is used in traditional medicine, its use has not been specied (Hutchings et al., 1996; Pooley, 1993). A related species, Cola natalensis (coshwood), is found in the coastal forests of KwaZulu-Natal. It is a medium sized tree of 710 m tall, recognizable by a crooked or bent main stem (Pooley, 1993). No mention has been made of the use of Cola natalensis in traditional medicine. Other related species, however, are used in traditional medicine in other parts of Africa. Cola acuminata (Pal.) Schott & Endl., from the mountains of Angola (Morton, 1992) and Cola nitida (Beauv.) Schott & Endl., native to the Ivory Coast, but now occurring throughout the west coast of Africa (Morton, 1992) are used in traditional medicine (Watt and Breyer-Brandwijk, 1962). The fruit of Cola acuminata are used as a stimulant and topical analgesic (Iwu, 1993). No alkaloids were found in Cola acuminata and aqueous extracts gave a negative antibiotic result (Watt and Breyer-Brandwijk, 1962). Cola nitida stems are used to treat arthritis and rheumatism (Ebana et al., 1991) and the fruit and leaves are used as stimulants and in healing rituals (Iwu, 1993). Cola nitida yielded alkaloids, cardiac glycosides, polyphenols and reducing compounds (Ebana et al., 1991). Cola acuminata and Cola nitida both have kola nuts that contain purines, caffeine, theobromine (Maillard et al., 1986), koltin and kolatein, catechols, ()epicatechol, kolanin and a red anthocyanin pigment (Hutchings et al., 1996). Dombeya burgessiae (pink dombeya) is a shrub that grows to a height of 24 m. It inhabits forest margins or woodlands in KwaZulu-Natal, Swaziland, Mpumalanga and the Northern Province, extending to east Africa (Von Breitenbach, 1965). The leaf decoction is drunk as an antimalarial, crushed leaves are applied over the affected area against leprosy and roots are used for stomach complaints (Chhabra et al., 1993). Dombeya cymosa (Natal dombeya) is a small slender tree, growing up to 8 m tall (Coates Palgrave, 1977). It inhabits coastal shrub, riparian vegetation and wooded kloofs from the eastern Cape through KwaZulu-Natal, Swaziland, Mpumalanga to southern Mozambique (Von Breitenbach, 1965). The roots are used in traditional medicine, but no particular use has been specied (Hutchings et al., 1996). Dombeya cymosa and Dombeya burgessiae are reported

to contain no alkaloids (Raffauf, 1996), and the leaves of Dombeya burgessiae are known to contain polyuronoids and steroids (Chhabra et al., 1984). Hermannia depressa (creeping red hermannia) is a prostrate herb. It is sparsely hairy and slightly glandular, occurring in grassland and marshes from the Cape through to Zimbabwe. It is used by the Zulu people as an emetic, and the leaf sap in water is used to treat stomach-ache, having a purgative and diaphoretic effect. Decoctions of the plant are taken to soothe coughs. Hermannia depressa is also used in mixtures with other plants in the treatment against diarrhoea (Watt and Breyer-Brandwijk, 1962; Hutchings et al., 1996; Pooley, 1998). No other studies relating to the chemical composition of this species have previously been reported. For the antibacterial compound isolation, another species Dombeya rotundifolia (Hochst.) Planch. was included. Phytochemical and pharmacological properties of this plant have previously been investigated (Reid et al., 2001). It is commonly used in traditional medicine to treat intestinal ulcers, stomach complaints (Coates Palgrave et al., 1985), haemorrhoids and diarrhoea (Watt and Breyer-Brandwijk, 1962; Van Wyk et al., 1997).

2. Materials and methods 2.1. Plant material All plant material was collected in the KwaZulu-Natal Province, South Africa. Voucher specimens were deposited in the University of KwaZulu-Natal Herbarium Pietermaritzburg. Cola greenwayi (Reid 6 NU), Dombeya burgessiae (Reid 4 NU) and Dombeya cymosa (Reid 3 NU) were obtained from the National Botanical Gardens in Pietermaritzburg. Cola natalensis (Reid 5 NU) was collected from the coastal Hawaan Forest in Umhlanga, Dombeya rotundifolia (Reid 1 NU) from Umgeni Valley Nature Reserve in Howick, and Hermannia depressa (Reid 7 NU) from the Agricultural campus of the University of KwaZulu-Natal, Pietermaritzburg. Plant material was dried at 50 C for 72 h, ground and stored in airtight containers at room temperature in the dark. 2.2. Extract preparation Dried, ground material (1 g) was extracted with water, ethanol (EtOH), dichloromethane (CH2 Cl2 ) or ethyl acetate (EtOAc) (10 ml) in an ultrasonic bath for 30 min. The extracts were ltered and air-dried overnight. For isolation of the antibacterial compounds, twig material from Cola greenwayi (176 g) and leaf material from Dombeya burgessiae (80 g), Dombeya rotundifolia (250 g) and Hermannia depressa (116 g) were used. Plant material was extracted with ethanol (for Hermannia depressa) or ethyl acetate (for Cola greenwayi, Dombeya burgessiae and Dombeya rotundifolia) using Soxhlet apparatus.

K.A. Reid et al. / Journal of Ethnopharmacology 97 (2005) 285291

287

2.3. Phytochemical screening Partitioning of the extracts, for alkaloid analysis, was performed according to Brimer et al. (1989). Dragendorfs reagent and Mayers reagent were used to indicate the presence of alkaloids by the formation of a precipitate (Wagner et al., 1984). Extracts were screened for cardiac glycosides by testing for 2-deoxy-sugars using the KellerKilliani test (Evans, 1989; J ger and Van Staden, 1995). The a presence of unsaturated lactone rings were tested for by spraying identical TLC plates with Keddes reagent (for cardenolide detection), chloramine Ttrichloroacetic acid reagent and antimony(III) chloride reagent (for bufadienolide detection) (Wagner et al., 1984; Evans, 1989; J ger and Van Staden, 1995). Extracts were a screened for cyanogenic glycosides using the Grignard test (J ger et al., 1995). Amygdalin (0.1 mg) was used as a stana dard. Extracts were tested for saponins using the haemolysis test (Fong et al., 1974) with a Columbia Blood Agar Base (Oxoid CM 331) (Reid et al., 2001). Saponaria ofcinalis extract was used as a positive control. The procedure for testing for tannins was performed using the Gelatinsalt Block test (Duncan et al., 1999). 2.4. Cyclooxygenase-1 inhibition assay The inhibition of prostaglandin synthesis by the plant extracts was determined using the cyclooxygenase-1 inhibition assay, according to the method of J ger et al. (1996). a Percentage inhibition of the test solutions was obtained by analysing the amount of radioactivity present in the test solutions relative to the radioactivity in a solvent blank (McGaw et al., 1997). The extracts were tested at a concentration of 0.5 g/ l in the nal assay volume and indomethacin (5 M) was used as a positive control. 2.5. Microdilution antibacterial assay The microdilution antibacterial assay was used to determine the minimum inhibitory concentration (MIC) values of the extracts against four bacterial strains (Eloff, 1998). The Gram-positive bacteria tested were Bacillus subtilis (ATCC 6051) and Staphylococcus aureus (ATCC 12600), and the Gram-negative bacteria were Escherichia coli (ATCC 11775) and Klebsiella pneumoniae (ATCC 13883). Extracts were tested at a concentration range from 12.5 to 0.098 mg/ml in the assay. The antibiotic neomycin was used as a positive control. 2.6. Bioautography on TLC plates Bioautographic evaluation of chromatography fractions, at various steps in the isolation procedure, was conducted using Staphylococcus aureus (ATCC 12600) as a test organism. A saturated culture was prepared and centrifuged at 3000 g for 10 min. The supernatant medium was dis-

carded and the pellets were redissolved in 10 ml fresh MuellerHinton broth. This culture was sprayed onto a developed TLC plate and incubated overnight at 37 C in 100% relative humidity. After incubation, the plate was sprayed with a 2 mg/ml solution of iodonitrotetrazolium chloride (INT) and incubated for a further 6 h. Inhibition of bacterial growth was indicated by white spots against a pink background on the TLC plate (Begue and Kline, 1972). 2.7. Chromatographic purication The crude extracts, for each species, were initially fractionated by vacuum liquid chromatography (VLC). The VLC was performed over silica gel 60 (0.0400.063 mm, Merck) and eluted with a hexane:ethyl acetate gradient of increasing polarity (10:00:10). Fractions were taken to dryness under vacuum and evaluated (0.5 mg of each fraction) using TLC and bioautography (see Section 2.6). The TLC plates (Merck Silica gel 60 F254 , 0.25 mm) were developed in hexane:ethyl acetate (1:1). Anisaldehyde spray reagent was used to determine the position of any additional compounds not visible with UV light (254 or 366 nm). Fractions that showed spots of antibacterial activity were further puried using preparative TLC (PTLC). For the PTLC, aliquots of the fractions were strip-loaded on TLC plates (Merck Silica gel 60 F254 , 0.25 mm) and developed in hexane:ethyl acetate (1:1). Bioautography of a small portion of the TLC plate was used to conrm the position of compounds showing antibacterial activity. Active fractions from Hermannia depressa were puried using HPLC. A Varian 5000 Liquid Chromatograph with a C18 column (Phenomenex, 260 mm 10 mm, 5 m) was used with a Spectrasystem UV3000 HR detector (detection at 210, 305 and 400 nm). The mobile phase, running at 2.5 ml/min, consisted of a gradient elution of 5080% methanol for 15 min followed by 80100% methanol for 25 min. 2.8. Spectroscopy The isolated compounds were identied using NMR spectroscopy and GCMS analysis. NMR spectra were recorded in CDCl3 on a 500 MHz Varian INOVA spectrometer. GCMS analysis was performed using an Agilent Model 6890 gas chromatograph coupled to an Agilent Model 5973 mass spectrometer. A J&W-5MS column, 30 m 250 m diameter and 0.30 m lm thickness, using a temperature program starting at 50 C for 2 min and ramping at 20 C/min to 300 C after which the temperature was held at 300 C for 20 min, was used. Compounds were identied using the NIST Spectral Search Program (Vers 1.6, build 06/24/98).

288

K.A. Reid et al. / Journal of Ethnopharmacology 97 (2005) 285291

3. Results and discussion 3.1. Phytochemical screening No alkaloids, cyanogenic glycosides or saponins were detected in any of the extracts from the species tested. In screening for cardiac glycosides, yellow bands were observed in leaf extracts from Dombeya burgessiae and Dombeya cymosa after spraying with antimony(III) chloride reagent. These results indicate the possible presence of bufadienolides. Cardenolides were not detected, although they are known to occur in many Sterculiaceae species (Trease and Evans, 1983). Cola greenwayi, Cola natalensis and Hermannia depressa extracts showed no indication of the presence of cardiac glycosides. In small doses, cardiac glycosides are used medicinally for controlling congestive heart failure (Harborne and Baxter, 1993). Generally, concentrations of cardiac glycosides in plants are very low, lower than 1% (Bruneton, 1995). Thus, it may be possible to use plant material without any fatality. It is important, however, that the concentration of the cardiac glycoside is known, for proper treatment. Tannins were detected in the leaf extracts of Cola greenwayi, and in the leaf, stem and root extracts of Hermannia depressa. From the amount of precipitate formed and degree of colour change, it was deduced that the leaf extract of Hermannia depressa yielded the lowest concentration of tannins and the root the highest. 3.2. COX-1 inhibitory activity The results for the COX-1 assay are shown in Fig. 1. Generally, the aqueous extracts showed no noteworthy activity, with the exception of the leaf extract of Cola greenwayi and the Hermannia depressa extracts. However, these results could be possible false-positive due to the presence of tannins in these extracts (see Section 3.1). The extracts from Cola greenwayi exhibited good inhibition against COX-1 for both the ethanol extracts of the leaves (75%) and twigs (78%),

as did the dichloromethane extract of the twigs (78%). The extracts of Cola natalensis showed low levels of inhibition against COX-1. For the Dombeya burgessiae extracts, high activity was obtained in both the ethanolic (76% and 89%) and dichloromethane (81% and 87%) extracts from the leaf and twig material, respectively. Dombeya cymosa extracts also exhibited high activity in the ethanol extracts of the leaf (78%) and twig (84%) material. For Hermannia depressa, the highest levels of COX-1 inhibition were obtained with the dichloromethane extracts of the stem (78%) and root (81%) material. 3.3. Antibacterial activity For all the species tested, the aqueous extracts exhibited no antibacterial activity against the test bacteria. The ethyl acetate extracts of Cola greenwayi leaf material showed moderate activity against Klebsiella pneumoniae (MIC 0.78 mg/ml) and Staphylococcus aureus (MIC 0.39 mg/ml) (Table 1). For Cola natalensis, none of the extracts exhibited any activity against the bacteria tested. Similarly, the water and ethanol extracts of Dombeya burgessiae leaf material showed no activity, and very low levels of activity were obtained with the ethyl acetate extract (MIC 12.5 mg/ml). The ethanol extracts of Dombeya cymosa leaf material gave MIC values of 1.56 mg/ml against Bacillus subtilis and Escherichia coli. Greater activity was observed against Staphylococcus aureus and Klebsiella pneumoniae with MIC values of 0.195 and 0.78 mg/ml, respectively. In comparison, extracts of the twig material showed low levels of activity. It is interesting to note that although some antibacterial activity was observed with the ethanolic leaf extracts of Dombeya cymosa, and not the root extracts, it is generally the roots that are used in traditional medicine (Hutchings et al., 1996). For Hermannia depressa, the ethanol and ethyl acetate extracts of the leaves, stems and roots exhibited moderate antibacterial activity, with MIC values ranging from 0.195 to 3.125 mg/ml. The ethanol extract of the

Fig. 1. COX-1 inhibition by water, ethanol and dichloromethane extracts of various plant parts from the Sterculiaceae species tested. Extracts were tested at a nal concentration of 0.5 mg/ml in the nal assay volume. Values represent the mean of double determinations. The indomethacin standard (5 M) gave 68% COX-1 inhibition.

K.A. Reid et al. / Journal of Ethnopharmacology 97 (2005) 285291

289

Table 1 The minimal inhibitory concentrations (mg/ml) of plant extracts from Sterculiaceae species against test bacteria. Water extracts gave no activity and are not included in the table Species Plant part Extract Bacteriuma B.s. Cola greenwayi Leaf Twig Dombeya burgessiae Dombeya cymosa Leaf Leaf Twig Hermannia depressa Leaf Stem Root Neomycin standard EtOH EtOAc EtOH EtOAc EtOH EtOAc EtOH EtOAc EtOH EtOAc EtOH EtOAc EtOH EtOAc EtOH EtOAc 12.5 6.25 12.5 6.25 12.5 1.56 12.5 12.5 1.56 1.56 0.39 0.78 0.195 0.78 1.56 102 S.a. 12.5 0.39 12.5 12.5 12.5 0.195 12.5 12.5 12.5 3.125 0.78 0.78 3.125 3.125 3.125 6.25 102 E.c. 12.5 6.25 12.5 6.25 12.5 1.56 12.5 12.5 3.125 0.78 3.125 NT 1.56 3.125 1.56 102 K.p. 12.5 0.78 12.5 12.5 12.5 0.78 12.5 6.25 3.125 0.78 3.125 NT 3.125 0.78 3.1 102

NT: not tested. a Abbreviations: B.s. Bacillus subtilis; S.a. Staphylococcus aureus; E.c. Escherichia coli; K.p. Klebsiella pneumoniae.

Hermannia depressa root material gave the highest activity with a MIC value of 0.195 mg/ml against Bacillus subtilis. 3.4. Isolation and identication of antibacterial compounds A summary of the compounds showing antibacterial activity is given in Table 2. Most of the active fractions contained a mixture of fatty acids and related compounds. McGaw et al. (2002) has recently reviewed the antibacterial effects of fatty acids from plants. Palmitic (hexadecanoic) acid (CH3 (CH2 )14 COOH) is one of the most common saturated fatty acids in leaf lipids and also occurs in some seed oils (Gurr and James, 1971; Harborne and Baxter, 1993). It has previously been reported to be the major antibacterial com-

pound in a mixture of fatty acids from Diplotaxis harra and Erucaria microcarpa (Hashem and Saleh, 1999), the fruit of Kigelia africana (Grace et al., 2002) and Pentanisia prunelloides (Yff et al., 2002). Lauric (dodecanoic) acid (CH3 (CH2 )10 COOH) and myristic (tetradecanoic) acid (CH3 (CH2 )12 COOH) are saturated fatty acids which are widely distributed in plants and are major components of some seed fats (Gurr and James, 1971). Stearic (octadecanoic) acid (CH3 (CH2 )16 COOH), another saturated fatty acid, is less prominent in leaf lipids, but is found in seed fats and vegetable oils and fats (Gurr and James, 1971; Harborne and Baxter, 1993). Eicosanes are derived from arachidonic acid, a 20-carbon polyunsaturated fatty acid, which is found in the phospholipid membrane.

Table 2 Summary of the isolation and identication of compounds from Sterculiaceae species showing antibacterial activity Species Cola greenwayi Twig EtOAc Chromatographic techniques VLC PTLC Compounds identied Eicosane Myristic acid Palmitic acid Stearyl alcohol Palmitic acid Lauric acid Myristic acid Palmitic acid Stearic acid Lauric acid Myristic acid Palmitic acid Stearyl alcohol

Dombeya burgessiae Leaf EtOAc Dombeya rotundifolia Leaf EtOAc

VLC PTLC VLC PTLC

Hermannia depressa Leaf EtOH

VLC PTLC HPLC

290

K.A. Reid et al. / Journal of Ethnopharmacology 97 (2005) 285291 Evans, C.W., 1989. Trease and Evans Pharmacognosy, 13th ed. Bailli` re e Tindall, London. Fong, H.H.S., Tin-Wa, M., Farnsworth, N.R., 1974. Practical Manual for Phytochemical Screening. College of Pharmacy, University of Illinois. Grace, O.M., Light, M.E., Lindsey, K.L., Mulholland, D.A., Van Staden, J., J ger, A.K., 2002. Antibacterial activity and isolation of aca tive compounds from fruit of the traditional African medicinal tree Kigelia africana. South African Journal of Botany 68, 220 222. Gracey, M., 1985. Diarrhoeal Disease and Malnutrition: A Clinical Update. Churchill Livingstone, U K. Gurr, M.I., James, A.T., 1971. Lipid Biochemistry: An Introduction, second ed. Chapman and Hall, London, pp. 1884. Harborne, J.B., Baxter, H., 1993. Phytochemical Dictionary: A Handbook of Bioactive Compounds from Plants. Taylor and Francis, London. Hashem, F.A., Saleh, M.M., 1999. Antimicrobial components of some Cruciferae plants (Diplotaxis harra Forsk and Erucaria microcarpa Boiss). Phytotherapy Research 13, 329332. Hutchings, A., Scott, A.H., Lewis, G., Cunningham, A.B., 1996. Zulu Medicinal Plants: An Inventory. University of Natal Press, Pietermaritzburg. Iwu, M.M., 1993. Handbook of African Medicinal Plants. CRC Press, Boca Raton, ISBN 0-8493-4266-X. J ger, A.K., Van Staden, J., 1995. Screening for cardiac glycosides in a Schizobasis intricata. South African Journal of Botany 61, 101 103. J ger, A.K., McAlister, B.G., Van Staden, J., 1995. Cyanogenic glycosides a in leaves and callus cultures of Schlechterina mitosemmatoides. South African Journal of Botany 61, 274275. J ger, A.K., Hutchings, A., Van Staden, J., 1996. Screening of Zulu a medicinal plants for prostaglandin-synthesis inhibitors. Journal of Ethnopharmacology 52, 95100. Maillard, C., Ollivier, B., Balansard, G., De Meo, M., 1986. Dosage de caf ine, de th obromine, de cat chine et d picat chine par chroe e e e e ` matographic liquide a haute performance dans un extrait de graines fraches stabilis es de Cola. Annales Pharmaceutiques Francaises 44, e 495500. McGaw, L.J., J ger, A.K., Van Staden, J., 1997. Prostaglandin synthesis a inhibitory activity in Zulu, Xhosa and Sotho medicinal plants. Phytotherapy Research 11, 113117. McGaw, L.J., J ger, A.K., Van Staden, J., 2002. Antibacterial effects of a fatty acids and related compounds from plants. South African Journal of Botany 68, 417423. Morton, J.F., 1992. Widespread tannin intake via stimulants and masticatories, especially guaran , kola nut, betel vine, and accessories. a In: Hemingway, R.W., Laks, P.E. (Eds.), Plant Polyphenols: Synthesis, Properties, Signicance. Plenum Press, New York, pp. 739 765. Pooley, E., 1993. The Complete Field Guide to Trees of Natal, Zululand and Transkei. Natal Flora Publication Trust, Durban. Pooley, E., 1998. A Field Guide to Wild Flowers: KwaZuluNatal and Eastern region. Natal Flora Publication Trust, Durban. Raffauf, R.F., 1996. Plant Alkaloids: A Guide to their Discovery and Distribution. Food Products Press, New York. Rasoanairo, P., Ratsimamanga-Urverg, S., 1993. Biological evolution of plants with reference to Malagasy ora. In: Monograph for the IFS NAPRECA Workshop on Bioassays, Antananarivo, Madagascar. Reid, K.A., J ger, A.K., Van Staden, J., 2001. Pharmacology and phytoa chemical properties of Dombeya rotundifolia. South African Journal of Botany 67, 349353. Sleigh, D.J., Timbury, M.C., 1998. Notes on Medicinal Bacteriology, fth ed. Churchill Livingstone, New York. Trease, G.E., Evans, W.C., 1983. Pharmacognosy, 12th ed. Bailli` re Tine dall, London, p. 508.

4. Conclusions The plants investigated in this study showed some inhibitory activity against COX-1 and antibacterial activity. It is interesting to note that Cola natalensis, which has no recorded medicinal usage, showed little inhibition against COX-1 and no observable antibacterial activity. Highest levels of COX-1 inhibition were observed with the ethanol and dichloromethane extracts of Dombeya burgessiae. The treatment of wounds and stomach-ache may be attributable to these properties. The Hermannia depressa extracts exhibited the highest overall antibacterial activity, which may play a role in combating coughs, diarrhoea and stomach-ache caused by bacterial infections. With the ever-increasing problem of HIV and AIDS, it is important that inexpensive treatments against opportunistic infections be developed. Plants that possess antibacterial activity can potentially be used to treat ailments brought about by bacterial infections, and alleviate many of the associated symptoms.

Acknowledgements The National Research Foundation and the University of KwaZulu-Natal Research Fund are gratefully acknowledged for nancial assistance. Brian Tarr and Janet Gates (National Botanical Gardens, Pietermaritzburg) are thanked for the provision of plant material.

References
Begue, W.J., Kline, R.M., 1972. The use of tetrazolium salts in bioautographic procedures. Journal of Chromatography 64, 182184. Brimer, L., Lorentzen, B., Wagner Smitt, V., 1989. velser I Farmakognosi K-25/9. Royal Danish School of Pharmacy, Copenhagen. Bruneton, J., 1995. Pharmacognosy, Phytochemistry, Medicinal Plants. Intercept Ltd., Andover. Chhabra, S.C., Uiso, F.C., Mshiu, E.N., 1984. Phytochemical screening of Tanzanian medicinal plants. Journal of Ethnopharmacology 11, 157179. Chhabra, S.C., Mahunnah, R.L.A., Mshiu, E.N., 1993. Plants used in traditional medicine in Eastern Tanzania. VI. Angiosperms (Sapotaceae to Zingiberaceae). Journal of Ethnopharmacology 39, 83103. Coates Palgrave, K., 1977. Trees of Southern Africa. Struik Publishers, Cape Town. Coates Palgrave, K., Coates Palgrave, P., Coates Palgrave, M., 1985. Everyones Guide to Trees of South Africa. National Book Printers, Cape Town. Duncan, A.C., J ger, A.K., Van Staden, J., 1999. Screening of Zulu medica inal plants for angiotensin converting enzyme (ACE) inhibitors. Journal of Ethnopharmacology 68, 6370. Ebana, R.U.B., Madunagu, B.E., Ekpe, E.D., Otung, I.N., 1991. Microbiological exploitation of cardiac glycosides and alkaloids from Garcinia kola, Borreria ocymoides, Cola nitida and Citrus aurantifolia. Journal of Applied Bacteriology 71, 398401. Eloff, J.N., 1998. A sensitive and quick microplate method to determine the minimal inhibitory concentration of plant extracts for bacteria. Planta Medica 64, 711713.

K.A. Reid et al. / Journal of Ethnopharmacology 97 (2005) 285291 Van Wyk, B., Van Oudtshoorn, B., Gericke, N., 1997. Medicinal Plants of South Africa. Briza Publications, Pretoria. Von Breitenbach, F, 1965. The Indigenous Trees of Southern Africa, IV (Descriptive Text, Part 3). The Government Printer, Pretoria, pp. 725737. Wagner, H., Bladt, S., Zgainski, E.M., 1984. Plant Drug Analysis. Springer-Verlag, Berlin.

291

Watt, J.M., Breyer-Brandwijk, M.G., 1962. The Medicinal and Poisonous Plants of Southern and Eastern Africa. Livingstone, Edinburgh. Yff, B.T.S., Lindsey, K.L., Taylor, M.B., Erasmus, D.G., J ger, A.K., a 2002. The pharmacological screening of Pentanisia prunelloides and the isolation of the antibacterial compound palmitic acid. Journal of Ethnopharmacology 79, 101107.