Botanical Aspects

Artemisia is a genus belonging to the family Asteraceae and has a very wide global distribution (Abad et al. 2012). According to Plants of the World (2024), there are currently 498 recognized Artemisia spp. with Asia being by far the most diverse with 174 species found in Russia, 150 species found in China, and about 50 species documented in Japan (Taleghani et al. 2020). Several species are also found in South America and Europe (Abad et al. 2012). In Africa, A. afra is widespread in the southern and eastern parts and can be found throughout Kenya, Tanzania, Uganda, and as far north as Ethiopia. It is very widely distributed in South Africa, Namibia, and Zimbabwe (Liu et al. 2009; Kane et al. 2019). Species within this genus are morphologically and anatomically diverse and may be perennial, biennial, or even annual herbs, subshrubs, or shrubs. One commonality is that they are usually aromatic due to the production of volatile oils (Taleghani et al. 2020).

Artemisia afra is an erect, perennial woody multi-stemmed shrub and under ideal conditions can grow up to 2 m tall (Fig. 1A and C). The leaves are small oval shaped in appearance and emit a pleasing sweet-smelling aroma when crushed, as opposed to its rather unpleasant and bitter taste when a tea infusion prepared from the leaves is consumed. In late summer until early winter, it produces small yellow flowers and produces tiny fruits of roughly 1 mm in length with seeds an order of magnitude smaller (Fig. 1B) (Du Toit and Van der Kooy 2019). Artemisia afra is known by many vernacular names across Southern Africa, including African wormwood, wild wormwood (English), wilde als, wilde-alsies, wildealsum, bitterals (Afrikaans), umhlonyane, mhlonyane (Xhosa, Zulu) lengana, and zengana (Sotho, Tswana) (Watt and Breyer-Brandwijk 1962; Viljoen et al. 2021).

Fig. 1

Artemisia afra Jacq. ex Willd., Asteraceae. A Leaves and B inflorescences. C A. afra growing next to a gravel road in the wild (photos: South African National Biodiversity Institute, South Africa)

Ethnopharmacology

An exceptionally valuable resource regarding traditional uses of Southern and Eastern African medicinal and poisonous plants was published by Watt and Breyer-Brandwijk (1962), and due to A. afra’s historical popularity, it is also extensively covered by these authors. According to Watt and Breyer-Brandwijk (1962), the most common way of consuming A. afra is by preparing an infusion or decoction, and occasionally adding sugar when the herb is used for chest problems; this may also be done to improve the bitter taste.

The list of traditional uses in Southern Africa is extensive and includes the treatment of coughs and colds, chills, dyspepsia, loss of appetite, stomachache and other gastric derangements, colic, croup, whooping cough, gout and as purgative, fevers, and blood-poisoning. The preparation is also used topically for the treatment of hemorrhoids and measles and is furthermore gargled to treat boils in the mouth. The vapor produced from boiling the leaves is inhaled to treat respiratory problems and genitalia are steamed after childbirth or for menstrual chills. Fresh leaves are inserted into the nose or hollow teeth to relieve colds, headaches, and toothaches. A dressing of the leaves is used to treat neuralgia, mumps, and infantile colic with the leaves usually moistened with brandy to treat colic or when taken by mouth. The plant is used as a treatment for fever including fever caused by malaria. The leaves are moreover smoked to treat throat irritations, and it has also been used as a vermifuge. It is furthermore used as an enema to treat constipation.

In Botswana, the leaves are used for colds and gastrointestinal problems, and it serves as a tea substitute. In Zimbabwe, it is used as a treatment of diabetes with the Manyika of Mozambique and Zimbabwe using it as a blood tonic and to treat pimples. In Kenya, Tanzania, and Uganda, the plant is mainly used as a cough remedy, emetic, anthelmintic, and febrifuge and for the treatment of colds, fever, and constipation.

Traditional Combination Treatments

Artemisia afra is extensively used in combination with Osmitopsis asteriscoides (L.) Less., Asteraceae, for the treatment of influenza and in combination with Leonotis ocymifolia var. schinzii (Gürke) Iwarsson (syn. L. microphylla Skan.), Lamiaceae, for digestive problems especially when accompanied by fever. A decoction of Allium sativum L., Alliaceae, and A. afra is consumed to treat fever with the treatment producing sweating. A decoction of Agrimonia eupatoria L., Rosaceae, and A. afra is used as a remedy for colds and in combination with the leaves of Zanthoxylum capense Harv. (syn. Fagara capensis Thunb.), Rutaceae, for the treatment of colds and fevers and it is stated that the latter combination treatment was very popular during the influenza epidemic of 1918. It is also used in combination with tobacco as a snuff and in combination with Lippia javanica (Burm.f.) Spreng., Verbenaceae, for the treatment of fevers, influenza, and measles (Watt and Breyer-Brandwijk 1962).

Chemical AnalysisVolatiles

In previous reviews, all volatile and non-volatile compounds that have been identified in A. afra until 2019 have been tabulated (Liu et al. 2009; Du Toit and Van der Kooy 2019). Recent studies include a solvent-free microwave-assisted extraction of volatile oils from A. afra leaves, in which 86 compounds were identified using gas chromatography-mass spectroscopy (GC–MS) and NIST 11 MS library matching (Falowo et al. 2019). Additionally, three A. afra specimens were investigated by Motshudi et al. (2021), and their chemical profiles were compared. They identified another 280 compounds using advanced GC–MS and NIST library matching. Letseka et al. (2022) used headspace GC analysis and identified 11 compounds in A. afra leave material. Selected major volatile compounds identified since 2019 can be found in Table S1. Of note is that none of the authors reported the presence of artemisinin (1), the active component of the sweet wormwood plant, A. annua, an herb used in Chinese traditional medicine in the treatment of malaria.

Non-volatiles

The phytochemical profiles of A. afra collected in five different countries were compared by Kane et al. (2019). No compounds were identified but the total alkaloids, tannins, saponins, terpenoids, and glycosides were reported. Recently, Nortjie et al. (2024) conducted a similar experiment and reported the presence/absence of phenols, flavonoids, quinones, tannins, saponins, terpenoids, and steroids, all of which were present in a methanolic extract of A. afra. Four phenolics, scopoline, rutin, scopoletin, and 3,5-di-O-caffeoylquinic acid, were identified and a further three phenolics, namely acacetin, chrysoeriol, and 4-caffeoylquinic acid, were tentatively identified by Sotenjwa et al. (2020). Using reference standards and high-pressure liquid chromatography-mass spectroscopy (LC–MS), 13 reasonably common phenolics were identified in A. afra tea infusions (Van Loggenberg et al. 2022; Taljaard et al. 2022; Vogel et al. 2023; Olivier et al. 2023; Stevens et al. 2024). Another eight compounds were identified by Molokoane et al. (2023) using standard isolation procedures and spectroscopic techniques. All newly identified compounds can be found in Table S1. Here again, none of the authors listed in Table S1 mentioned the presence of the well-known bioactive molecule artemisinin.

Pharmacological AspectsToxicity

Mekonen et al. (2020) tested the acute and sub-acute toxicity of aqueous extracts prepared from A. afra leaves on the brain, heart, and suprarenal glands of Swiss albino mice. For the acute study, a single dose ranging from 200 to 5000 mg/kg was administered whilst daily doses of 600–1800 mg/kg or 1800 mg/kg were administered for the sub-acute study. The LD50 was reported to be > 5000 mg/kg indicating that the plant can be considered relatively non-toxic. In the sub-acute study, no signs of toxicity were observed in all treatment groups, and upon microscopic examination of the brain, heart, and suprarenal glands, no signs of cellular injury were observed. Motshudi et al. (2021) tested the chloroform, butanol, and water extracts of three A. afra samples for cytotoxic effects against Vero monkey kidney cells. They reported relatively low cytotoxicity of 414.56 ± 4.81 µg/ml with the remaining samples all giving values of > 500 µg/ml. Amoussa et al. (2023) reported no statistical difference in the body weight of rats who received 300 mg/kg of an aqueous extract of A. afra as compared to the control group. Moreover, no difference in hematological, biochemical, and pathological parameters and pathological changes in tissues and organs of rats were observed in comparison with the control group indicating that at the administered dose little to no toxicity was observed.

General Biological Activity

The antioxidant activity of volatile oils extracted from A. afra was reported by Falowo et al. (2019). A 10% volatile oil solution of A. afra provided a radical scavenging activity of 61.08% which was comparable to the positive control rutin. Reddy et al. (2023) tested the antioxidant and anti-hypertensive properties of methanol extracts of A. afra as compared to other plant species. In the 2,2-diphenyl-1-picrylhydrazyl (DPPH) assay, A. afra showed reasonable activity as compared to the positive control, ascorbic acid, at 250 μg/ml. In the nitrous oxide assay, A. afra showed similar activity to the positive control ascorbic acid, again at a relatively high concentration of 250 μg/ml. Stevens et al. (2024) tested the alpha-glucosidase activity of compounds identified in A. afra infusions and reported that two major compounds, 1,5-dicaffeoylquinic acid (1,5-DCQA) and 3,5-DCQA, exhibited inhibition of 88.96% ± 0.02 and 84.69% ± 0.03, respectively, as compared to the positive control acarbose (47.54% ± 0.07). Interestingly, the chemically similar isomers 4,5-DCQA and 3,4-DCQA only showed inhibition of 25.55% ± 0.12 and 43.96% ± 0.58.

In an in vivo study, the anticonvulsant activity of an aqueous ethanolic extract of A. afra was reported by Kediso et al. (2021). They reported a delayed onset and duration of pentylenetetrazol-induced spasms at doses of 250, 500, and 1000 mg/kg in rats. The convulsion onset time was found to be 92.8 s for the negative control as compared to 1001.1 s for diazepam (positive control). Artemisia afra doses of 250, 500, and 1000 mg/kg reduced the onset time to 504.8, 551.8, and 808.3 s, respectively. The duration of convulsions was also significantly reduced from 34.8 s (negative control) to 4.5 s (positive control) and 17.0, 13.0, and 7.9 s with increasing doses of A. afra. Weathers et al. (2024) tested infusions prepared from A. annua and A. afra against human dermal fibroblasts (CRL-2097) and determined the fibroblast viability after 1 and 4 days. They reported that the Artemisia teas reduced the viability by up to 80% compared to solvent controls. They also reported that A. annua had greater antifibrotic activity than pure artemisinin but in turn was less active than A. afra.

Antimicrobial Activity

The antimycobacterial activity of dichloromethane extracts of A. afra and A. annua was tested against Mycobacterium tuberculosis (mc26230 and Erdman) as well as M. smegmatis (mc2155) and M. abscessus (ATCC 19977) (Martini et al. 2020). Dichloromethane extracts of leaves were prepared, and the amount of artemisinin was quantified in each extract. The A. afra material used in this study obtained from Liege, Belgium, is reported to contain detectable levels of artemisinin (1) which was quantified as 0.0077% as compared to A. annua which contained 0.82%. Pure artemisinin was also tested in this study and gave a MIC of 75 μg/ml. The activity of A. annua and A. afra is provided in a somewhat complicated way by stating that the MIC value for both was the extract of 4.81 mg of dried leaf material in 1 ml of medium. Based on the artemisinin content of these extracts, the MIC was further expressed as 39 and < 0.37 μg/ml artemisinin respectively, which implies that artemisinin does not seem to play a pivotal role in the observed activity. The positive control rifampicin and ethambutol provided MIC values of 0.075 and 0.5 μg/ml, respectively. They furthermore tested extracts against M. abscessus and found that only A. annua displayed bacteriostatic activity. Kiani et al. (2023) conducted a similar study and tested DCM extracts of A. annua and A. afra against M. tuberculosis (mc26230). They reported MIC values of 1.9–4.5 mg/ml and 1.3–10 mg/ml (DW/ml) for A. annua and A. afra extracts, respectively. No positive controls were included in this study.

The antimicrobial activity of leaf extracts of A. afra sourced from three different vendors was extracted using water, butanol, and chloroform and tested for activity against Staphylococcus aureus (ATCC 29213), Enterococcus faecalis (ATCC 29212), Escherichia coli (ATCC 25922), and Pseudomonas aeruginosa (ATCC27853) (Motshudi et al. 2021). All samples gave MIC values between 4.03 and 9.88 μg/ml as compared to the positive control, gentamicin, which gave an MIC of 0.01 μg/ml. Methanol, ethanol, and hexane extracts of A. afra were also tested by Haile and Jiru (2022) against multi-antibiotic-resistant E. coli, S. aureus, P. aeruginosa, and Streptococcus pneumoniae, affording quite high MIC values of 6.25–12.5 mg/ml. No positive control was included in this study. Molokoane et al. (2023) tested eight isolated compounds from A. afra against E. coli, E. faecalis, P. aeruginosa, S. aureus, S. Typhimurium, and C. albicans and reported weak activity except for sitosterol-3-O-β-D-glucopyranoside which afforded an IC50 value of 31.25 μg/ml against E. faecalis, compared to the positive control, gentamicin, which afforded an IC50 of 3.9 μg/ml.

An ethnobotanical survey was conducted in Benin by Amoussa et al. (2023) wherein plants used for the treatment of symptoms related to salmonellosis were included. Twenty-four plant species, including A. afra, were collected and extracted with ethanol and tested against Salmonella enterica subsp. enterica serovar Typhi (ATCC 19430), Salmonella enterica subsp. enterica serovar Typhimurium (ATCC 14028), Salmonella enterica subsp. enterica serovar Enteritidis (ATCC 13076), and three isolates of S. Typhimurium (P70, L22, and R309). The most active plant species identified during the in vitro tests were also tested for biofilm inhibition against the clinical isolates of S. Typhimurium (P70, L22, and R309). The best candidates were tested in infected rats with aqueous extracts of the selected plant species. A. afra showed MIC values of > 0.312 μg/ml as compared to A. annua (250 μg/ml) and was therefore further tested for biofilm inhibition. At a concentration of 0.156 mg/ml, A. afra afforded biofilm inhibition greater than 50% against clinical isolates P70 and R309. In vivo rat experiments also revealed that A. afra reduced the numbers of viable S. Typhimurium in feces at 200 mg/kg/bw but not at a lower dose of 100 mg/kg/bw as compared to ciprofloxacin (positive control) which showed complete elimination at 8 mg/kg/bw.

Antiparasitic Activity

The chemical difference of five A. afra samples collected in different countries was compared by Kane et al. (2019). Four solvent extracts (water, ethanol, dichloromethane, and hexane) of each specimen were also tested for antimalarial activity. They found that the hexane extract of A. afra collected in Burundi was the most active and reported IC50 values of 0.71 μg/ml for W2 and IC50 3.18 μg/ml for D6 Plasmodium falciparum strains, respectively, as compared to the positive control chloroquine that afforded IC50 values of 74.58 ± 12.1 ng/ml for W2, and 16.95 ± 3.14 ng/ml for D6. It is important to note that some samples of the hexane extracts of the remaining A. afra samples provided an almost 40-fold weaker activity highlighting the large chemical differences and subsequent activity based on collection location.

A first comparative study of the in vitro antiplasmodial activity of A. afra and A. annua against the asexual erythrocytic stages of P. falciparum and the pre-erythrocytic stages of various Plasmodium species was conducted by Ashraf et al. (2021). Leaves and twigs (2.5 g) were boiled in 25 ml water for 5 min and stock solutions of 100 g/l infusion were used for bioactivity testing. The results were presented as gram plant material/litre and the inhibition was expressed in this unit. Infusions of both plants after exposure of 72 h to P. falciparum rings inhibited their growth in a dose-dependent manner, with A. annua active at slightly lower doses. Exposure of hepatocytes of P. berghei sporozoites led to a significant dose-dependent reduction in the size and number of hepatic schizonts, with A. afra being effective at a slightly lower concentration with 10 g/l A. annua and 4 g/l A. afra completely clearing the parasites from cultures. The same concentrations also cleared all P. falciparum from the cultures with the authors concluding that the infusions quickly and directly affect the parasite’s replication.

Gruessner and Weathers (2021) tested infusions of A. afra and A. annua against the ring stage of P. falciparum. The activity was also expressed as ug plant material per litre or normalized for the artemisinin content in the samples. IC50 values of 401 μg/l and 177 mg/ml (DW/l) for a dichloromethane extract of A. annua and A. afra were given, respectively. Normalized to artemisinin content, both extracts exhibited similar IC50’s, and hence, the authors concluded that the observed activity was strongly linked to artemisinin (1) with other components playing a small role. This study was followed by Snider and Weathers (2021) who assessed artemisinin, methylene blue, A. annua, and A. afra tea infusions (5 g DW/l) against P. falciparum NF54. The parasites were induced to form gametocytes using N-acetylglucosamine and cultures during asexual, early-stage, and late-stage gametocytogenesis were tested. They found that both plant species inhibited gametocytemia and that the observed activity was again strongly linked to artemisinin content. Taljaard et al. (2022) tested A. afra tea infusions and organic extracts of the tea infusions against Schistosoma and found that a hexane and dichloromethane extract prepared from a tea infusion gave IC50 values of 1.8 and 1.7 μg/ml as compared to the positive control praziquantel (1.5 μg/ml).

Cytotoxic Activity

Van Loggenberg et al. (2022) tested tea infusions prepared from A. afra collected at four different sites against a panel of lung cancer cell lines (H69V; H69AR; A549). The most active infusion presented an IC50 value of 6 μg/ml and a selectivity index against Vero cells of 10. However, the positive control, paclitaxel, afforded an IC50 value of 0.0031 μg/ml and an SI of 3.1 against Vero cells. Vogel et al. (2023) tested A. afra extracts, fractions, and pure compounds against MCF-7, BT-20, and MDA-MB-231 breast cancer cell lines. A dichloromethane and hexane fraction prepared from the tea infusion provided IC50 values of 4.5 µg/ml and 4.7 μg/ml, respectively. None of the major compounds in the tea infusion provided this level of activity, and therefore, it was concluded that a minor unknown compound is responsible for the observed activity. The positive control paclitaxel gave IC50 values ranging between 0.002 and 0.005 μg/ml against the different cell lines.

Antiviral Activity and COVID-19

Based on A. afra’s historical popularity as a flu remedy, it is quite surprising how few studies before the COVID-19 pandemic were conducted. Since Lubbe et al. (2012) reported on the antiviral activity of A. afra, which they found to be comparable to A. annua against HIV, very few studies on A. afra’s general antiviral activity were conducted. Osunsanmi et al. (2022) evaluated the neuraminidase inhibition activity of A. afra against three influenza strains: A/PR/8/34 (H1N1), A/Sydney/5/97 (H3N2), and B/Jiangsu//10/2003 at 400 μg/ml. A methanol extract afforded the best inhibition of 52.04 ± 2.08% with a calculated IC50 of 76.11 ± 3.84 μg/ml against the B/Jiangsu//10/2003 strain vs. 0.064 ng/ml for the positive control zanamivir.

Nie et al. (2021) tested various Artemisia spp. against feline coronavirus (ATCC VR-989, WSU 79–1683) which included ten A. annua samples and four A. afra samples collected from different countries. One sample each of A. tridentata and A. absinthium, one ethanolic extract of A. annua, and the (in)famous Covid-Organics from Madagascar were also tested. The plant samples were extracted at 90 °C in water with A. afra from Benin affording the best activity at 330 μg/ml. Based on the original screening, six of the most active samples were also tested against Sars-Cov-2 (BavPat 1 isolate) including four A. annua samples, one A. afra sample, and an ethanolic extract of A. annua as well as pure artemisinin. The A. afra sample afforded the highest activity of the water extracts with 0.09 mg/ml ± 0.03 and a selectivity index of 26.22. However, the ethanolic extract of A. annua was found to be by far the most active at 0.0004 mg/ml ± 0.0001 (0.4 µg/ml). As no positive control existed at the time, the pure compound artemisinin was included which gave an EC50 of 4.23 mg/ml which equates to roughly 10,000-fold less activity than the ethanolic extract of A. annua. The Covid-Organics sample did not demonstrate good activity but the EC50 value was expressed in a rather unintelligible “7.73% to raw drink.” No other literature regarding the antiviral activity of A. afra could be found.

The chemical analyses that have been conducted on A. afra since 2019 reported many new volatile compounds. However, because A. afra is generally prepared by boiling the plant material in water for varying amounts of time, it will likely lead to a (complete) loss of the volatile oils and other volatile components. Another potential pitfall is that MS library matching does have its limitations, and without confirming the identity of the reported volatiles with reference standards, the identity remains rather tentative. Coupled with the biological tests that were conducted, as reported in this overview, which usually evaluate different liquid extracts and subsequent non-volatile extracts/fractions may further limit the usefulness of this data.

Several non-volatiles, mainly relatively well-known phenolic compounds, were also reported. Phenolics, in general, suffer from low bioavailability when orally administered and this again will limit the usefulness of these reports (Grgić et al. 2020). It is however important to note that the presence of artemisinin was not reported in any study, except in studies that made use of A. afra plant material obtained from Liege, Belgium, and Senegal (Kane et al. 2019; Martini et al. 2020; Kiani et al. 2023). When the same laboratory is used to prepare both A. annua and A. afra samples, the possibility exists that cross-contamination may have occurred. This has indeed occurred in our laboratory and great care had to be taken to prevent cross-contamination of our samples. However, Kiani et al. (2023) explain the presence of artemisinin in A. afra as being caused by possible cross-hybridization with A. annua.

The general bioactivity reports that were published since 2019 revealed some interesting results. The anticonvulsant activity of an aqueous ethanolic extract of A. afra tested in rats compared quite well with the positive control (Kediso et al. 2021). The study conducted by Taljaard et al. (2022) also reported IC50 values of 1.8 and 1.7 μg/ml as compared to the positive control praziquantel at 1.5 μg/ml. The study of Kane et al. (2019) afforded IC50 values of 0.71 and 3.18 μg/ml for W2 and D6 Plasmodium falciparum strains, respectively, which compared favorably with the positive control chloroquine. Stevens et al. (2024) also reported good comparative results. The main reasons why positive controls are such an important aspect are primarily twofold. Firstly, in vitro and in vivo bioassays making use of living cells or organisms can be quite troublesome and can easily lead to severe variability in response; therefore, the inclusion of a positive control is an absolute necessity to demonstrate that the test system functions as intended (a negative control, e.g., solvent or carrier only, can also fulfil this requirement). Secondly, a positive control serves as an important comparator to be able to relate apples with apples, e.g., is the IC50 for the positive control similar to before and to published literature? And how does the test sample compare to the positive control? Four studies were highlighted above and, based on the results as compared to their positive controls, warrant further investigation.

Unfortunately, many studies provided results that are many orders of magnitude less active than the positive controls (e.g., Motshudi et al. 2021; Loggenberg et al. 2022; Osunsanmi et al.