Artemisia afra : A potential flagship for African medicinal plants ?

The genus Artemisia consists of about 500 species, occurring throughout the world. Some very important drug leads have been discovered from this genus, notably artemisinin, the well known anti-malarial drug isolated from the Chinese herb Artemisia annua. The genus is also known for its aromatic nature and hence research has been focussed on the chemical compositions of the volatile secondary metabolites obtained from various Artemisia species. In the southern African region, A. afra is one of the most popular and commonly used herbal medicines. It is used to treat various ailments ranging from coughs and colds to malaria and diabetes. Although it is one of the most popular local herbal medicines, only limited scientific research, mainly focussing on the volatile secondary metabolites content, has been conducted on this species. The aim of this review was therefore to collect all available scientific literature published on A. afra and combine it into this paper. In this review, a general overview will be given on the morphology, taxonomy and geographical distribution of A. afra. The major focus will however be on the secondary metabolites, mainly the volatile secondary metabolites, which have been identified from this species. In addition all of the reported biological activities of the extracts derived from this species have been included as well as the literature on the pharmacology and toxicology. We aim at bringing together most of the available scientific research conducted on this species, which is currently scattered across various publications, into this review paper. © 2008 SAAB. Published by Elsevier B.V. All rights reserved.


Introduction
Traditional Chinese Medicine (TCM) is a well-known and well-studied research field. Many academic and pharmaceutical institutions around the World fund research into TCM with the aim of finding new pharmaceutical entities, establishing quality control protocols and to prove the efficacy of these traditional medicines. Compared to TCM, the traditional African medicine research is lagging behind. One of the main reasons for this is that the available information on medicinal plants is not always well documented. Furthermore, Africa encompasses a huge area with diverse cultures and languages, which makes sharing of information difficult. The lack of well-funded research institutions or pharmaceutical companies on the continent is also adding to the problem. African researchers should therefore aim at establishing their own "Traditional African Medicine" (TAM), which on the long run should bring to Africa the same economical and social benefits that TCM has brought to China. In order to achieve this goal, it would be extremely beneficial to the cause if one can identify a flagship species which will epitomize the end goal of establishing a TAM. For this reason, we have decided to review most of the scientific literature available on the endemic Artemisia afra Jacq. (Asteraceae) species to establish if it might have the same medicinal potential as the Chinese A. annua species.
Artemisia afra is one of the most popular and commonly used herbal medicines in southern Africa. However, only limited research has been conducted on this species. According to a recent publication by Van Wyk (2008), only 42 scientific publications and 2 patents related to A. afra are available of which 27 publications and 1 patent were published between 2001 to 2008. This indicates that interest in A. afra has increased significantly during recent years, prompting us to gather all available research on A. afra and combine it into this review paper. We could not find any previous reviews written on this species and therefore we thought it important to take stock of what has been done in the past, in order to provide a solid foundation and direction to any future research that may be conducted on this species.

Morphology
Artemisia afra is a perennial woody shrub, which grows up to 2 m tall with a leafy, hairy and ridged stem (Van Wyk et al., 1997). Its leaves are of soft texture, dark green on the adaxial surface and a lighter green on the abaxial surface, reaching a length of 8 cm and a width of 4 cm. Artemisia afra blossoms from January to June, producing yellow, butter-coloured flowers with abundant bracts (Hilliard, 1977). The plant has an easily identifiable aromatic odour and smells pungent and sweet after bruising. The fruit is approximately 1 mm in length, covered with a silvery-white coating and the shape is slightly 3-angled and curved. In winter, the branches rapidly regenerate from the base directly after dying back (Hilliard, 1977;Van Wyk et al., 1997).

Distribution
Artemisia afra occurs in the mountain regions of Kenya, Tanzania, Uganda and as far north as Ethiopia. It is also widely distributed in southern Africa, such as South Africa, Namibia, and Zimbabwe. In South Africa, it grows in the northern provinces of Gauteng and Limpopo along the eastern parts of South Africa, including Swaziland and Lesotho, to the Western Cape in the south. It is also abundantly found in KwaZulu-Natal province from the coast to the Drakensberg (Van Wyk et al., 1997;Hilliard, 1977) at altitudes of 20 to 2440 m on damp slopes, along streamsides and forest margins.

Ethnopharmacology and utilization
The many different common or local names describing this plant may be ascribed to the widespread use by different ethnic groups (Watt and Breyer-Brandwijk, 1932). In the Xhosa language it is known as "Umhlonyane", the Zulu language "Mhlonyane", in Sotho "Lanyana", Tswana "Lengana", English "African wormwood" and in Afrikaans "Wilde als". It is usually employed for treating a variety of ailments such as coughs, colds, headaches, chills, dyspepsia, loss of appetite, gastric derangements, colic, croup, whooping-cough, gout, asthma, malaria, diabetes, bladder and kidney disorders, influenza, convulsions, fever, heart inflammation, rheumatism and is also used as a purgative (Watt and Breyer-Brandwijk, 1932;Thring and Weitz, 2006). These uses indicate that A. afra possesses antiviral, antibacterial and anti-inflammatory activities.
Many different preparations of this plant are employed to treat the various symptoms and ailments. A syrup prepared from A. afra is used for bronchial troubles, while infusions or decoctions can be applied as a lotion to bathe hemorrhoids and for earache. An infusion of leaves or roots of this species is also used for the treatment of diabetes in the Eastern Cape Province of South Africa (Erasto et al., 2005;Mahop and Mayet, 2007). Respiratory infections are treated through inhaling the vapour from boiling leaves and this vapour is also used to treat menstrual chill. Fresh tips are inserted into the nose for colds and headaches and into hollow teeth to treat toothache. The poultice of the leaves can be applied to relieve neuralgia, to treat the swellings in mumps and is placed on the abdomen to treat infantile colic. It is also helpful to reduce colic by administering a tincture made of the leaves wetted by brandy (Watt and Breyer-Brandwijk, 1932).
Artemisia afra is also used as an infusion. Usually, a quarter cup of fresh leaves is added to a cup of boiling water and the infusion is allowed to draw for 10 min. The mixture is then strained and the resulting infusion is sweetened with honey. This preparation is taken orally for relief of most of the abovementioned ailments (Roberts, 1990).
The use of other medicinal plants or substances in combination with A. afra is also documented in African ethnopharmacology. In South Africa, preparations of A. afra are often made in combination with brandy, sugar, ginger, thyme, rosemary, mint, chamomile, Osmitopsis asteriscoides or Eucalyptus globulus. A combination of A. afra and E. globulus is employed to treat influenza, and the infusion of the leaves and stems of Lippia asperifolia and A. afra is used as a formula for fevers, influenza, measles, and as a prophylactic against lung inflammations (Watt and Breyer-Brandwijk, 1932). Decoctions of the leaves of A. afra and Agrimonia bracteata are used for colds in southern Africa while decoctions of Tetradenia riparia and umhlonyane (Xhosa name for A. afra) with salt are used to treat coughs in the Transkei region of the Eastern Cape Province of South Africa (Hutchings et al., 1996).

Chemistry
A major problem for using this plant by traditional means is claimed to be variations in the phytochemical composition (Dube, 2006). The volatile components are easily lost during the production of traditional preparations as was shown by Asfaw et al. (2005). Eight sesquiterpenes were lost during hydrodistillation, which aimed at simulating the traditional preparation method. Most work focussed on the volatile secondary metabolites obtained from A. afra, while only relatively few publications reported on other types of secondary metabolites identified in this plant. In the following sub-sections, secondary metabolites identified from A. afra will be discussed with their corresponding references.
For the identification of the major components in the oil, the only reported analytical equipment used was gas chromatography coupled with either a flame ionization detector or mass spectroscopy detector (Asfaw et al., 2005). Until now, 131 compounds have been identified from the oil of A. afra. Most of these compounds can be classified as monoterpenes and sesquiterpenes. Several other components have also been found during various other studies and are listed in Table 1 with the corresponding references.

Non-volatile secondary metabolites
Early in the 1920s, Goodson (1922) found that A. afra contained a wax-ester (probably ceryl cerotate), triacontane, scopoletin and quebrachitol. During the following decades, a few other publications reported on new secondary metabolites identified from this species. Table 2 lists the non-volatile secondary metabolites identified in A. afra and the corresponding references.

Chemical variation in A. afra
The volatile secondary metabolite content in A. afra vary frequently and is mainly due to the following factors: (a) Geographical variation: The main components of the volatile secondary metabolites identified in A. afra fluctuate enormously in plants collected from different geographical regions. It is reported that artemisyl acetate (24.4-32.1% of the total oil) was found to be the major constituent in Ethiopian oil (Worku and Rubiolo, 1996) while 1,8-cineole (67.4%) was found to be the major constituent in Kenyan oil (Mwangi et al., 1995). In Zimbabwean oil, αand β-thujone (52%) was the major constituent (Graven et al., 1992) while only α-thujone (54.2%) was found to be the major constituent of South African oil (Libbey and Sturtz, 1989). Chagonda et al. (1999) found that the major constituents of the volatile secondary metabolites from A. afra grown in Zimbabwe, showed a large variation between cultivated and wild populations.  Table 2 Non-volatile secondary metabolites identified from Artemisia afra. Dube (2006) 1999; Burits et al., 2001;Viljoen et al., 2006;Vagionas et al., 2007). Analysis of oils obtained from different plant parts showed that the chemical components varied depending on the plant part used e.g. Goodson (1922) found camphor, a wax ester, triacontane, scopoletin and quebrachitol in the flowering tops of A. afra. Bohlmann and Zdero (1972) revealed that the roots contained isomeric coumarins and five acetylenes, while the aerial parts contained thujone and umbelliferone-derivatives, but no acetylenes. The same variation occurred in the volatile secondary metabolite composition between the leaves of A. afra (Mwangi et al., 1995) and the whole plant (Worku and Rubiolo, 1996). The results showed that the oil obtained from the leaves consisted mainly of 1,8cineole (67.4%), while yomogi alcohol (21.6-26.8%) and artemisyl acetate (24.4-32.1%) predominated in the oil extracted from the whole plant. (c) Drying methods: Asekun et al. (2007) reported that the yields and compositions of oil from A. afra varied according to the drying method used. Yields obtained were 0.18%, 0.88%, 1.54% and 1.88% for fresh, ovendried, air-dried and sun-dried material respectively. They also found that artemisia ketone was present in the oil extracted from the fresh material but was absent in the oil extracted from the air-dried and sun-dried material. (d) Variation within the natural population: Viljoen et al. (2006) collected 16 individual A. afra samples from four natural populations, Setibeng (Lesotho), Giant's Castle (KwaZulu-Natal), Qwa-qwa (Free State) and Klipriversberg (Gauteng) and found qualitative and quantitative differences within and between populations.

Biological effect of A. afra extracts
Considering the many traditional uses, the plant has been studied for various biological activities. These activities will be discussed in the following paragraphs with their corresponding references.

Biological activities
A. afra is rich in terpenes and is therefore likely to have valuable biological activities. From the available literature the results of conducted tests indicated that this plant has a broad spectrum of inhibitory activity against the organisms listed in Table 3. Most of the activities were expressed in zone of inhibition, while relatively few studies determined the minimum   Poswal and Witbooi (1997) found that, of the five plant species evaluated, the oils of A. afra and Lavandula angustifolia were the most effective against Pseudomonas syringae pv. Syringae, while the A. afra extract was least active against P. solanacearum. Different strains of the same species also showed different susceptibilities to A. afra extracts. P. solanacearum Ps46 proved more susceptible than P. solanacearum Ps62 to the plant oil. The activity of A. afra oil was determined against different strains of yeasts. Eventually, Hanseniaspora vinae and Saccharomyces cerevisiae were the most susceptible to the oil, while Zygosaccharomyces bailii and Dekkera bruxellensis were proved to be the least sensitive (Mangena and Muyima, 1999). Rabe and Van Staden (1997) tested the antimicrobial activities of the extracts of A. afra against Staphylococcus aureus, S. epidermis and Bacillus subtilis, the results showed that the lowest MIC was achieved against S. aureus. They also reported that the majority of the antibacterial activity of A. afra extracts was present in the methanol extracts instead of water extracts. Mangena and Muyima (1999) also found that of 29 species of bacteria and 12 strains of yeast tested, only three species of bacteria (Escherichia coli, P. aeruginosa and P. fluorescens) and one strain of yeast (Torulaspora delbruckii) did not show obvious dose-effect relationship when three different concentrations of A. afra oil were used (undiluted, 1:1 and 1:2 dilutions). Muyima et al. (2002) did a similar experiment and the results showed that in most cases only the highest concentrations of A. afra oil inhibited the tested bacteria, except against P. aeruginosa ATCC 27853 and strains of P. aeruginosa isolated from the environment. Asres et al. (2001) indicated that the crude methanolic extract of A. afra did not inhibit Mycobacterium tuberculosis. Kraft et al. (2003) found that the lipophilic extracts of the aerial parts of A. afra showed the highest activity against the chloroquine-sensitive Plasmodium falciparum strain PoW and against the chloroquine-resistant clone Dd2 of P. falciparum, compared to the lipophilic extracts of Cussonia spicata, Clutia hirsute, Vernonia natalensis, Parinari curatellifolia, Flueggea virosa, Adenia gummifera, Hymenodictycon floribundum and the hydrophilic extracts of all above species including A. afra itself and V. colorata. They also indicated that the antiplasmodial effect of A. afra was probably due to the additive or even synergistic activity of the complex mixture of substances found in the extract. Clarkson et al. (2004) reported that the dichloromethane extract exhibited the strongest antiplasmodial activity against P. falciparum D10 (IC 50 value of 5 µg/ml) compared to the dichloromethane/methanol (1:1), methanol and water extracts (IC 50 value of 7.3, 8.0 and N100.0 µg/ml, respectively). Gathirwa et al. (2007) proved that the methanolic extract of A. afra showed more activity against P. falciparum D6 and P. falciparum W2 (IC 50 values of 9.04 ± 0.54 and 3.98 ± 0.98, respectively), compared to the water extract (IC 50 values of 11.23 ± 1.98 and 4.65 ± 0.64, respectively). In addition, the P. berghei infected mice treated with the methanolic extract had a longer survival time (19.21 ± 3.57 days) than those treated with the water extract (17.85 ± 2.81). Both studies indicated that the antiplasmodial compounds in this plant are more soluble in lipophilic solvents than hydrophilic solvents. Jenett-Siems et al. (2002) indicated that A. afra was not very suitable for antiplasmodial treatment because the activity of the isolated compound (e.g. a sesquiterpene lactone) was lower than the toxicity against ECV-304 cells. However, Kraft et al. (2003) discovered that two types of compounds were responsible for the antiplasmodial effect of this plant, including flavonoids (7-methoxyacacetin, acacetin, genkwanin and apigenin) and sesquiterpene lactones (1-desoxy-1α-peroxy-rupicolin A-8-O-acetate, rupicolin A-8-Oacetate and 1α,4α-dihydroxybishopsolicepolide).  Gathirwa et al., 2007 Results are expressed with standard deviations only if included in the original publication. a Antimicrobial activity of undiluted essential oil of A. afra was evaluated by diameter of zone of inhibition in millimeters (including disc diameter of 6 mm) against selected microorganism; b No activity at highest concentration tested; c Data from Mangena and Muyima, 1999;d Data from Muyima et al., 2002;e Streptomycin resistant strain; f Pseudomonas aeruginosa isolated from environment and identified by API identification system; g Wild type; h Chloroquine sensitive; i The highest activity appeared in methanolic extract; j High activity: IC 50 (10 μg/ml; k High activity: IC 50 ≤ 5 μg/ml; l Chloroquine resistant. The University of Zululand started to combine A. afra with standard forms of treatment for HIV positive patients (Mulholland and Drewes, 2004). Positive results obtained suggest that A. afra exhibits some antiviral activity or immune boosting properties. Mulatu and Mekonnen (2007) tested the traditional folk use of A. afra for stomach pains and intestinal cramps. The results showed that the ethanolic extract of A. afra leaves, significantly reduced spontaneous rhythmic and agonist-induced contractions of isolated mouse duodenum and guinea pig ileum.

Cardiovascular effect
The aqueous A. afra extracts displayed a hypotensive effect in vivo and a concentration-dependent biphasic effect on the heart in vitro. Low concentrations induced an initial cardio stimulation followed by cardio depression, while higher concentrations mainly showed cardio depression. In addition, a compound isolated from A. afra, scopoletin, induced a dose-dependent decrease in inotropic activity and a decrease in heart rate, especially at higher doses (Guantai and Addae-Mensah, 1999).

Antioxidant activity
Artemisia afra has been used to treat various inflammatory diseases such as rheumatism, fever and diabetes (Halliwell and Gutteridge, 1989). It was shown that the anticoccidial potential of this plant is also due to its antioxidant activity (Naidoo et al., 2008). Burits et al. (2001) used a rapid TLC screening method for the evaluation of the antioxidant activity of A. afra. It was claimed that the volatile oils of A. afra could play a role as a nonspecific donor for hydrogen atoms or electrons in the diphenylpicrylhydrazyl assay and it was further shown to be an effective hydroxyl radical scavenging agent in the deoxyribose degradation assay. The oil of A. afra also showed antioxidant potential in the bioassay for non-enzymatic lipid peroxidation in liposomes. According to the results of various activity tests and GC-MS analysis, the sesquiterpene, chamazulene, was suggested to be responsible for the radical scavenging effect of the oil.

Sedative and CNS-acting activities
Artemisia species have been proven to have anti-convulsive and GABA A -benzodiazepine receptor-binding activity (Sayyah et al., 2004;Salah and Jäger, 2005). Stafford et al. (2005) discovered sedative and CNS-acting activities of A. afra according to the GABA A -benzodiazepine receptor-binding assay. The results showed that the ethanolic extract of the plant had a good dose-dependent activity (binding % of 5.0 ± 0.9, 45.3 ± 3.5 and 81.8 ± 4.2 at concentrations of 0.455 mg/ml, 0.046 mg/ml and 0.005 mg/ml respectively). It was also claimed that flavonoids were the active compounds related to the sedative and CNS-acting activities.

Antidepressant activity
According to the monoamine hypothesis, some transmitters, such as serotonin, noradrenalin, dopamine, GABA and L-glutamate, are related to depression (Rang et al., 1999). Therefore, several antidepressants available on the market today are based on selective inhibition of serotonin reuptake. Nielsen et al. (2004) reported that the ethanolic extract of A. afra showed a low affinity to the serotonin transporter protein (Less than 50% [ 3 H] citalopram binding at the highest concentration of 5 mg/ml). The affinity of this extract exhibited a concentration-dependent relationship, which indicated that A. afra possesses a certain potential to treat depression. -Siems et al. (2002) found that a sesquiterpene lactone isolated from A. afra was responsible for the observed cytotoxicity. The crude ethanolic extract of A. afra also showed cytotoxicity in Vero cells with an IC 50 value of 113.0 ± 2.05 µg/ml (Mativandlela et al., 2008). Acute and chronic toxicity of this plant was also exhibited through in vivo tests in rodents. In the acute toxicity assay, single intraperitoneal injections (1. 5-5.5 g/kg) and single oral doses (2-24 g/kg) of an aqueous extract of A. afra were administered to observe the toxicity of this plant in rodents. The intraperitoneal route of administration induced a regular dose-dependent increase in the mortality and incidence of general behaviour adverse effects, while the oral route showed a dose-independent result. Meanwhile, the LD 50 of these two routes of administration were found to be 2.45 and 8.96 g/kg, respectively. A chronic toxicity assay was also performed in rodents by administering oral doses of an A. afra extract (0.1 or 1 g/kg/day) for 3 months. The result showed that no significant changes occurred in general behaviour, body weight, haematological and biochemical parameters, organ weights and morphology of organs (Mukinda and Syce, 2007).

Quality control
Artemisia afra has been traditionally used to treat symptoms which can be related to malaria. Although there are only few scientific publications available to support this, a commercial company is currently selling an A. afra product which they claim to be effective in the treatment of malaria. The claim is further supported by the fact that the activity of their product can be ascribed to the well known antiplasmodial compound, artemisinin. Due to these claims a study was performed to establish a quality control protocol and metabolomic differentiation between A. afra and the well known A. annua (which contains artemisinin). This study was conducted in order to establish a rapid metabolomics approach and to carry out quality control on these herbal formulations ( Van der Kooy et al., 2008). The claim that artemisinin (so far only found in A. annua) is responsible for the antiplasmodial effects of A. afra was investigated and disproved. The results indicated that these two plant species (A. afra and A. annua) could easily be identified based on their metabolomic profiles. The separation between the two species was mainly due to the three methyl signals of artemisinin in the NMR spectrum. Additional LC-MS analysis also demonstrated that artemisinin could not be detected in the A. afra samples.

Discussion and conclusions
Artemisia afra is widely distributed in the southern parts of Africa. It is therefore one of the most utilized plants in the African ethnopharmacology. This review aimed at describing the different uses of A. afra, either alone or in combination with different substances or species, in treating various ailments. The main focus of this review was however to bring together all available scientific research that has been conducted on this species. The identification of the volatile secondary metabolites obtained from this species received a lot of attention as is reflected by the number of papers that has been published on this matter.
The differences in the chemical profiles of the volatile constituents of A. afra can be ascribed to geographical variations, different plant parts used, different sample preparation and extraction methods and genetic differences within populations. The obtained variation makes quality control of this species extremely difficult. Although three of the above-mentioned causes for variation can be controlled, the natural variation between individuals cannot be controlled in the wild. Future research should therefore focus on establishing reproducible quality control methods in order to minimize the variation found in the volatile secondary metabolites, but also the non-volatile secondary metabolites, between individuals. In addition, all the secondary metabolites reported in literature are listed in this review in order to be able to conduct future dereplication studies.
Artemisia afra has also been tested against a multitude of microorganisms as can be seen from Table 3. Future research should however first focus on establishing quality control protocols, followed by testing the extracts against the microorganisms which showed the highest sensitivities. Although A. afra has been tested for its antiplasmodial activity (Jenett-Siems et al., 2002;Kraft et al., 2003;Clarkson et al., 2004;Gathirwa et al., 2007), this has not yet been fully demonstrated. It has also been proven that this species does not contain the antiplasmodial component, artemisinin (in detectable levels), which occurs in A. annua (Van der Kooy et al., 2008). Due to the traditional use and commercial claims that A. afra is effective as a prophylactic as well as treating uncomplicated malaria, this aspect should be researched further. Research in this area should focus on determining the antiplasmodial activity of A. afra samples from various regions as well as at different growth stages. If compared to the well known antiplasmodial plant, A. annua, it was shown that the active component, artemisinin, is present at its highest concentration just before flowering. The concentration of this compound can fluctuate between 0.01%-1.00% (Liu et al., 2006). If a similar antiplasmodial compound is present in A. afra, the growth stage at which to produce an extract to perform the bioassays, becomes extremely important.
Based on the data presented in this review it can be concluded that A. afra might become a future flagship species for TAM, if the problems associated with quality control could be solved and more importantly, if we can identify the active component(s), especially the antiplasmodial secondary metabolites, in this species.