South Africa’s Best BARK Medicines Prescribed at the Johannesburg Muthi Markets for Skin, Gut, and Lung Infections: MIC’s and Brine Shrimp Lethality

Indigenous trade of medicinal plants in South Africa is a multi-million-rand industry and is still highly relevant in terms of primary health care. The purpose of this study was to identify today’s most traded medicinal barks, traditionally and contemporaneously used for dermatological, gastrointestinal, and respiratory tract infections; then, to investigate the antimicrobial activity and toxicity of the respective extracts and interpret outcomes in light of pharmacokinetics. Thirty-one popularly traded medicinal barks were purchased from the Faraday and Kwa Mai-Mai markets in Johannesburg, South Africa. Information on the medicinal uses of bark-based medicines in modern commerce was recorded from randomly selected traders. The minimum inhibitory concentration (MIC) method was used for antimicrobial screening, and brine shrimp lethality was used to determine toxicity. New medicinal uses were recorded for 14 bark species. Plants demonstrating some broad-spectrum activities against tested bacteria include Elaeodendron transvaalense, Erythrina lysistemon, Garcinia livingstonei, Pterocelastrus rostratus, Rapanea melanophloeos, Schotia brachypetala, Sclerocarya birrea, and Ziziphus mucronata. The lowest MIC value of 0.004 mg/mL was observed against Staphylococcus epidermidis for a dichloromethane bark extract of E. lysistemon. The tested medicinal barks were shown to be non-toxic against the Artemia nauplii (brine shrimp) bioassay, except for a methanol extract from Trichilia emetica (69.52% mortality). Bacterial inhibition of bark extracts with minimal associated toxicity is consistent with the safety and valuable use of medicinal barks for local muthi market customers. Antimicrobial outcomes against skin and gastrointestinal pathogens are feasible because mere contact-inhibition is required in vivo; however, MIC values against respiratory pathogens require further explaining from a pharmacokinetics or pharmacodynamics perspective, particularly for ingested rather than smoked therapies.


Introduction
Bark is the outer protective layer of woody plants. It consists of all tissues outside the vascular cambium, an actively dividing layer of cells responsible for the production of xylem and phloem tissues [1,2]. Bark contains a high mass of polyphenols with medicinal applications [3,4]. Biological studies of medical bark remain scarce, which has led to it being undervalued as a suitable organ [3]. The use of bark in traditional medicine is also common in other parts of the world outside of Africa. A study by Pagliosa et al. [5] conducted in Brazil reported potent antioxidant activity from the methanol extract of bark from Ilex paraguariensis A. St. Hil. Furthermore, medicinal bark of different plant species from the British Moracin family were reported to contain bioactive benzofuran derivatives [6]. From south and southeast Asian countries, species from the genus Dillenia documentation of the basic characters and character states of bark even though DNA barcoding and metabolomics methods may be convenient and accurate methods to confirm and verify identifications.
Commercial medicinal uses of 31 barks purchased from the Johannesburg Muthi Markets are presented in Table 1. Medicinal uses recorded from the traders were similar to those recorded from the literature. In some cases, new medicinal plant uses were recorded, and these are shown in bold in Table 1. New uses were recorded for the first time for 14 species (Albizia adianthifolia, Bersama lucens, Croton sylvaticus, Cryptocarya latifolia, Dombeya rotundifolia, Elaeodendron transvaalense, Erythrina lysistemon, Harpephyllum caffrum, Prunus africana, Schotia brachypetala, Syzygium cordatum, Trichilia emetica, Vachellia natalitia, and Vachellia robusta). Medicinal barks are packaged in small A4 size plastic bags or wrapped in old newspapers. These medicinal barks are used by customers to prepare infusions, decoctions, powders, or pastes. Decoctions and infusions are often used as emetics ("ukhuphalaza") and enemas ("ukucha-tha") for treatment of internal disorders, chest complaints, and sometimes used as a wash ("ukugeza") for treating skin disorders. Powdered bark material is often licked from the palm of the hand ("khotha") or taken as snuff ("umbemiso") for treating ailments such as headache, food poisoning, and stroke. Finely powdered bark may also be incorporated into complex mixtures for use as "insizi" (powdered mixtures that are rubbed into scarification in the skin to alleviate pain).
The present study showed that plant species are often used for the treatment of more than one ailment. This has also been reported from various surveys across the country [13,14,42,43]. Mixtures of species are sometimes used as a combination, such as the bark of Harpephyllum caffrum, Schotia brachypetala, and Syzygium cordatum, taken as an emetic to treat respiratory complaints.
The respiratory system is the most vulnerable site for bacterial infections. Environmental instabilities and frequent antimicrobial use are reportedly among the major contributors to bacterial accumulation in the respiratory tract [44]. Results from the present study ( Figure 1) reveal that most of the studied medicinal barks are used for the treatment of coughs (23%), chest complaints (20%), tuberculosis (14%), and unspecified respiratory ailments (14%). Most of the South African population, from rural areas in particular, depend on traditional medicinal plants for the treatment of respiratory ailments [45]. McGaw et al. [46] documented ca. 180 medicinal plants used for treatment of tuberculosis and related symptoms such as chronic cough, respiratory complaints, and fever. In South Africa, infections of the respiratory tract associated with high mortality rates include tuberculosis, pneumonia, and influenza [47,48]. While there are multiple species that are successful in traditional healthcare at treating these bacterial infections, it is still necessary to increase our understanding of the primary mechanisms and magnitude of effects [49].
Different ailments of the gastrointestinal tract that are commonly treated with bark are shown in Figure 2. Most of the medicinal barks are used against diarrhea (14%), intestinal worms (12%), dysentery (8%), and food poisoning (8%). These findings are in accordance with those reported by other authors [16,50,51] who found that lay people through ethnobotanical surveys use mostly bark of trees for treating diarrhea. In 2005, there were more than 25,000 deaths caused by diarrheal diseases in South Africa [52]. Medicinal plants used traditionally for the treatment of diarrhea are reported to have antispasmodic properties, delay gastrointestinal transit, suppress gut motility, stimulate water adsorption, and/or reduce electrolyte secretion [53]. However, the main reason for using bark against diarrhea is the presence of tannins, which are known to be non-specific protein poisons [34,54]. According to Hutchings [55], tannins are frequently present in those parts of plants used in the treatment of dysentery and diarrhea, or for respiratory ailments. The tannins form a protective, impermeable layer over mucosa and also prevent the development of bacteria by denaturing the proteins within these unicellular organisms. Moreover, of importance is the use of bark against intestinal worms but in this case, the efficacy probably depends on the chemical compounds with specific activity against parasites. Other authors [13,51] also reported the frequent use of bark against internal parasites.
Gastrointestinal infections are predominantly associated with pathogens transmitted from fecal-contaminated water, food, or environment, as well as poor hygiene associated with poverty and a poor nutritional status [56,57]. South African rural communities, including northern Maputaland (KwaZulu-Natal province) and Bizana (Eastern Cape province), have been reported to still lack access to safe drinking water, good hygiene, and sanitation [50,58]. Therefore, antimicrobial studies of traditional medicinal plants used for the treatment of gastrointestinal tract ailments are still very relevant in South Africa.
Skin ailments and the cosmetic use of bark are shown in Figure 3. Many barks are used to promote wound healing (24%), to remove pimples or acne (9%), and to treat skin rashes (7%). De Wet et al. [38] found that leaves were mostly used for treatment of general skin ailments (31%), closely followed by bark with 28%, and then the roots with 13%.
Grace et al. [13] also found that wound healing is among the most common external applications of bark. The Fabaceae are especially popular for treating wounds [38], as is also shown in the present study.

Antimicrobial Outcomes
A new criterion of interpreting MIC values of medicinal plant extracts stipulates that activity is "noteworthy to moderate" if the MIC values are between 0.1 to 0.6 mg/mL [45]. The antimicrobial activity of methanol and dichloromethane bark extracts tested against pathogens associated with dermatological, gastrointestinal, and respiratory tract infections isshown in Table 2. The MIC values ranged from 0.004 mg/mL to above 8 mg/mL. Considering that MIC values below 1 mg/mL are regarded by some contemporary researchers as qualifying as "active" [59,60], then many of the extracts produced in the current study are active and warrant further investigation. Medicinal barks demonstrating broad-spectrum activities include Elaeodendron transvaalense, Erythrina lysistemon, Garcinia livingstonei, Pterocelastrus rostratus, Rapanea melanophloeos, Schotia brachypetala, Sclerocarya birrea, Warburgia salutaris, Prunus africana, and Ziziphus mucronata. The lowest MIC value of 0.004 mg/mL was noted against S. epidermidis from the dichloromethane extract of E. lysistemon, which is considered acceptable by the strict criteria set forth by another group of authors [61]. Garcinia livingstonei and Kigelia africana exhibited interestingly low MIC values of 0.06 mg/mL (methanol extract) against B. cereus and 0.16 mg/mL for dichloromethane extracts, respectively, which is close to the limits for good activity [59,62]. Bacillus cereus was found to be the most susceptible bacterium screened. The antimicrobial activities of medicinal barks seem to support the traditional and contemporary medicinal uses as reported by traders on the Johannesburg Muthi Markets.

Erythrina lysistemon
Bark extracts demonstrated interesting antimicrobial activity against the bacterial pathogens screened. Bark of E. lysistemon is widely used in traditional medicine for the treatment of skin ailments (Table 1). Staphylococcus epidermidis was the most susceptible bacteria against both extracts, yielding the lowest MIC value of 0.12 mg/mL (methanol) and 0.004 mg/mL (dichloromethane). Bacterial inhibition against S. aureus (MIC of 0.20 mg/mL) demonstrates the efficacy of its traditional use against wounds. Staphylococcus aureus is a bacterial strain associated with wound infection [45,63]. Reports from earlier studies [64] showed that methanol extracts of E. lysistemon bark did not exhibit activity against S. aureus. The ethyl acetate, ethanol, and water extracts of E. lysistemon bark [63] also did not show antimicrobial activity against S. epidermidis, P. aeruginosa, and E. coli on the disc diffusion assay, but activity was only observed against S. aureus. The same authors observed that bark of the Erythrina species is more active in terms of antibacterial activity than the leaves. The discrepancy in the results could be due to the type of solvent extract used and the different antimicrobial techniques employed. However, more recent results by Mabona et al. [63] demonstrated comparable noteworthy activity from the leaf extract against S. aureus with an MIC value of 0.20 mg/mL. In addition, Mukandiwa et al. [65] reported comparable MIC values against S. aureus (0.31 mg/mL) from methanol and dichloromethane extracts of E. lysistemon leaves.
Antimicrobial screening of bark extracts against the newly reported traditional medicinal use for stomach ailments (Table 1) resulted in broad-spectrum activity against gastrointestinal tract pathogens (Table 2). Noteworthy antimicrobial activity with an MIC value of 0.10 mg/mL (dichloromethane extract) and 0.16 mg/mL (methanol extract) against E. faecalis and 0.16 mg/mL for both extracts was noted against S. sonnei (Table 2). A follow-up study of these MIC values irrefutably demonstrated that antimicrobial effects were largely attributed to the presence of prenylated isoflavones and pterocarpans. A structure-activity relationships analysis demonstrated that prenylation on rings A and B, with a hydroxyl group on the same ring as the prenyl group, conferred the most potent antimicrobial outcomes [66].

Garcinia livingstonei
The methanol extract of G. livingstonei demonstrated the best antimicrobial activity against most of the bacteria responsible for gastrointestinal tract infections screened. Garcinia livingstonei is widely used for the treatment of diarrhea [13,50]. Noteworthy activity, with MIC values of 0.25 mg/mL against S. typhimurium and 0.41 mg/mL against E. coli (methanol extracts), was noted. Excellent activity, with the lowest MICs of 0.06 mg/mL against B. cereus and 0.18 mg/mL against E. faecalis, was noted from the methanol extract. A recent study by Van Vuuren and co-workers in 2015 [67] showed similar results from the organic extract of bark (1:1 methanol: dichloromethane). The reported MIC values were 0.12 mg/mL against B. cereus, 0.19 mg/mL against S. typhimurium, 0.34 mg/mL against E. faecalis, and 0.38 mg/mL against Shigella flexneri [67]. Furthermore, isolated chemical compounds from the leaf extract of G. livingstonei demonstrated strong activity, with MICs of 0.006 mg/mL and 0.008 mg/mL against E. faecalis and E. coli, respectively [68,69]. Corresponding biological activity of bark against diarrheal diseases was also reported from Garcinia buchananii bark aqueous extract and its fractions, which reversed high-lactose-diet diarrhea-induced weight loss and reduced bloating as well as fecal fluid content on the tested rats [70].

Schotia brachypetala
Bark and roots of S. brachypetala are widely used in traditional medicine as a remedy for dysentery and diarrhea [13,14,43]. Noteworthy activity was observed against E. coli with the lowest MIC value of 0.33 mg/mL for the dichloromethane extract ( Table 2). The dichloromethane extract from the present study was much more active against E. coli, in contrast to the ethanol bark extract with an MIC value of 3.31 mg/mL from an earlier study [94]. Methanol extract of the bark has also been screened against other Shigella species from other studies and demonstrated interesting activity against Shigella dysenteriae (MIC of 0.15 mg/mL) and against Shigella flexneri (MIC of 0.31 mg/mL) [16]. Organic extracts from the present study demonstrated noteworthy activity, which differed from results obtained by Van Vuuren and co-workers in 2015 [67]. The (methanol:dichloromethane) extracts from their study exhibited poor antimicrobial activity against B. cereus, E. coli, and S. typhimurium with MIC values of 8 mg/mL. In addition, corresponding activity was observed against E. faecalis with a MIC value of 0.63 mg/mL. The major reason for such variation in the MIC values could be attributed to the different choice of solvent extract [95]. Successful extraction of phytochemicals from plant materials may vary significantly depending on the type of solvent extract used [96].

Sclerocarya birrea
The stem bark of S. birrea is widely used in traditional medicine for the treatment of stomach ailments, including diarrhea [14,40,50], gastritis, peptic ulcers, and stomach cancer [97]. Bark extracts demonstrated overall good antimicrobial activity against five bacterial strains screened with MIC values mostly ≤1 mg/mL ( Table 2). Antimicrobial activity with a similar MIC value of 0.83 mg/mL against Gram-negative S. typhimurium and S. sonnei was recorded from the methanol extract. The dichloromethane extract exhibited similar activity with a MIC value of 0.50 mg/mL against B. cereus, E. faecalis, and E. coli. The results from this study differ with those obtained by Moyo et al. in 2011 [98] against E. coli (MIC of 3.12 mg/mL) reported from the dichloromethane extract of the twig bark from S. birrea. These differences in the MIC values may be attributed to different bacterial strains used and also to the age of the plant; thus, 83% of traders from other muthi markets [98] preferred the use of mature bark.

Pterocelastrus rostratus
Bark decoctions are taken as emetics for respiratory ailments and administered orally for a blocked and wheezing chest [8,28]. Good antimicrobial activity was demonstrated from the extracts against the four bacterial strains tested with MIC values ranging from 0.20 mg/mL to 0.83 mg/mL (Table 2). Moderate activity against K. pneumoniae and M. catarrhalis with a MIC value of 0.66 mg/mL was noted from the methanol extract and 0.50 mg/mL against S. aureus from both extracts. Interesting activity with MIC values of 0.25 mg/mL (methanol) and 0.20 mg/mL (dichloromethane) was detected against P. aeruginosa. The broad-spectrum noteworthy antimicrobial activity of P. rostratus is not surprising, as Pterocelastrus bark species have been previously reported to be used for respiratory ailments in traditional medicine [43,46]. To the best of our knowledge, this is the first report on antimicrobial activity of the plant against respiratory tract pathogens.

Rapanea melanophloeos
The bark is popularly used for skin, stomach, and respiratory tract infections (Table 1). Moderate activity against S. aureus with a MIC value of 0.41 mg/mL was recorded from the dichloromethane extract ( Table 2). The lowest MIC value of 0.25 mg/mL against S. epidermidis was also demonstrated from the methanol extract (Table 2). In a study by Madikizela and co-workers [58], the dichloromethane leaf extract of R. melanophloeos exhibited weak activity against S. aureus, with a MIC value of 3.12 mg/mL. The bark of R. melanophloeos is used in many parts of sub-Saharan Africa for the treatment of respiratory tract infections [99]. Moderate activity of bark extracts with MIC values of 0.50 mg/mL was detected against K. pneumoniae, M. catarrhalis, and P. aeruginosa. The MIC value of 0.41 mg/mL was noted from the dichloromethane extract against M. catarrhalis. Acetone and water extracts of the bark were also screened against pathogens coincident with TBrelated symptoms. The acetone extract exhibited a MIC value of 5 mg/mL, and water extracts displayed no activity against Mycobacterium tuberculosis [46,80]. Although the traditional use focused on sore throat and TB, the results presented here convey that R. melanophloeos could possibly be used for bronchial types of chest ailments, particularly if active compounds can be smoked or inhaled.

Warburgia salutaris
The peppery scented bark is powdered and mixed with fats and applied on to the skin to treat infections, acne, and sores. The powdered bark is also smoked, sometimes mixed with the Cannabis sativa leaf, to treat colds and various chest complaints [13,14,39,40,42,43,71]. The dichloromethane extract inhibited S. aureus (0.5 mg/mL), P. aeruginosa (0.25), and M. catarrhalis (0.41 mg/mL), demonstrating non-discriminate inhibition across Gram-types. Further study conveyed that the drimane sesquiterpenes were important in achieving these effects [100].

Pharmacokinetics Perspective
The medicinal applications of bark in the present study generally require ingestion, topical application, or inhalation as smoke or vapor. For topical applications, the choice of extract solvent is relevant to the solubility of active compounds in the source material. Aqueous extracts will dissolve saponins, tannins, flavonoid glycosides, and so forth, which will be present in the methanol extracts that we made in the current study. Alternatively, fat extracts dissolve lipid soluble (lipophilic) ingredients, which will be present in the dichloromethane extract.
Due to topical application, direct contact-inhibition is all that is necessary to inhibit microbes associated with skin infection [101]. This means that transdermal penetration (or absorption) is not a limiting factor in achieving the higher MIC values demonstrated in the current study. There are many exceptions if consideration is given to subcutaneous infections, such as boils or acne, and in such cases lipophilic compounds are more suited to achieving these effects. However, saponins can deeply penetrate pores and the hair follicle bulb because of direct passage that is achieved by clearing lipophilic sebum (oil) due to the soap-like effects.
For gastrointestinal applications, in most cases the bark is boiled in water, and the aqueous solution is ingested. Although this conveys that aqueous solubility is an important precursor to achieve therapeutic effects, during the boiling process lipophilic ingredients are temporarily liquified and are driven into the infusion where they flocculate as the drink cools. The aqueous mixture is consumed with the flocculants, meaning lipophilic ingredients may also be participating or solely responsible for therapeutic outcomes. Nevertheless, as in the first example, contact-inhibition is all that is necessary to achieve bacterial inhibition [101].
For respiratory complaints that involve bronchial pathogens, there are two types of therapy that are enacted, with one involving inhalation of vapors or smoke, and the other involving oral ingestion, with the therapeutic effects expected to occur non-locally (or generally). In this latter case, the MIC values achieved in the current study are not low enough to be of any benefit, particularly because the systemic blood plasma concentration of these extracts (or active compounds) is expected to be in the order of hundreds of times lower than the MIC values in Table 2. However, many of the therapies are smoked or the emitted vapors inhaled, and, in these cases, merely contact inhibition is necessary to inhibit the pathogen. For example, W. salutaris is smoked, so the MIC values sufficiently validate that contact inhibition is the primary mechanism, and the effects are not necessarily enacted indirectly through the immune response [101].

Brine Shrimp Toxicity Assay
The % mortality of methanol and dichloromethane extracts from 31 medicinal barks induced in Artemia nauplii following 24 and 48 h of exposure, respectively, is shown in Table 3. All the tested medicinal barks showed less than 50% mortality except for the methanolic extract of Trichilia emetica, which exhibited 69.52% mortality at 48 h. The negative control demonstrated 0% mortality and, the solvent control demonstrated a maximum mortality rate of 5.34% after 48 h.  Differences in polarities of solvent extracts used for similar plant species will yield a diverse range of quantitative differences in phytochemicals, which may result in a different response to the brine shrimp assay. The methanol and dichloromethane extract of A. adianthifolia, B. lucens, and T. emetica displayed varying toxicity values. Bussmann et al. [102] also observed varying toxicity values between the same plant species due to the use of different solvent extracts. Despite these limitations, the results presented here seem to agree with other studies. Van Vuuren et al. [67], for example, also found that methanol and water extracts from the stem bark of Warburgia salutaris and Syzygium cordatum exhibited less than 50% mortality. Furthermore, the leaf methanol and water extracts of Kigelia africana, Pittosporum viridiflorum, Schotia branchypetala, Sclerocarya birrea, Vachellia karroo, and Ziziphus mucronata also demonstrated low toxicity in the brine shrimp assay, with some of these plants inducing mortality below 20% following 24 and 48 h of exposure [67].
The methanolic bark extract of Trichilia emetica displayed the highest level of toxicity, with 69.52% mortality, while the dichloromethane extract exhibited 19.60% mortality at 48 h (Table 3). These findings are in accordance with a study by Oryema et al. [103], who reported that methanol and water extracts of the roots were more toxic than ether extracts. The first report of toxic properties of T. dregeana (closely related to T. emetica) was observed in 1899, when an African woman died as a result of drinking a decoction of the bark for use as a laxative [104]. The bark of T. emetica is widely used in traditional medicine for treatment of gastrointestinal ailments [14,43]. The dichloromethane extract of the bark exhibited some degree of cytotoxicity in the monkey kidney cell with an ID 50 of 50 µg/mL [105]. Furthermore, the root aqueous extract and ethyl ether fraction did not display toxicity (LC 50 > 1000 µg/mL) in the brine shrimp bioassay [106]. To the best of our knowledge, this is the first report of brine shrimp lethality test results for the bark of T. emetica.
Both extracts from Erythrina lysistemon in this study did not exhibit toxicity against the brine shrimp (Table 3). A study by Prozesky et al. [105] also showed that acetone bark extract from E. lysistemon exhibited very low cytotoxicity when tested against the monkey kidney cell line.
Surprisingly, the bark of Erythrophloeum lasianthum appeared to be non-toxic in the brine shrimp assay. All parts of this plant are popularly known to be highly toxic. The active toxic principles include two diterpenoid alkaloids, cassaine and erythrophleine [28,40,41]. Erythrophloeum lasianthum is an important Zulu traditional medicine used as a snuff for treating headache [14].
The bark from Erythrophloeum lasianthum has been harvested for muthi market trade from as early as 1946 [107] as cited by Williams et al. [19], and it is still actively traded on the Johannesburg Muthi Markets. These findings are in agreement with Lewis [108], who also found that aqueous extract of the bark exhibited very low toxicity values in the brine shrimp assay. Aqueous extract of the pods and seeds, however, were slightly toxic. The mode of administration is a very critical aspect of traditional medicine that should be carefully handled, as there could be a thin line between toxicity and safety, depending on the dosage and method of preparation of the plant material. According to the information given by traders at the Johannesburg Muthi Markets, a pinch of finely powdered bark of E. lasianthum is taken as snuff for headaches but anything above that quantity results in severe bleeding from the nostrils and causes dizziness. The Faraday Muthi Market is located in Johannesburg, Gauteng Province ( Figure 4). The Faraday Muthi Market has in the past decade become the biggest and most popularly known market in the southern African region and perhaps throughout Africa [109]. Furthermore, an estimate of 220 market traders were recorded at Faraday in October 2015, which decreased from over 300 traders recorded in October 2014 [109]. The trading stalls used for storing or packing plant products are usually outdoors, slightly exposed or covered, with very few traders that are found selling within the old and neglected public buildings [33]. There is also an active trade of animal products at the Faraday market such as snake and crocodile skin, and bones of carnivorous animals including lion, leopard, and different kinds of birds. These are usually sold as finely powdered material in very small amounts (teaspoon) and incorporated in plant mixtures to enhance the healing effect [23]. Purchasing of plant products from informal markets is a common day-to-day activity among the southern African population. The customers include local people, healers, and consumers from townships and rural and urban areas around Gauteng and other provinces, as well as neighboring countries such as Swaziland, Lesotho, and Mozambique. Most of these plant products are used to treat a variety of ailments including Alzheimer's disease, sexually transmitted infections, cancer, malaria, inflammatory pain, and cardiovascular dis-orders [23]. Plants used for charm and magic are also popular, purchased most frequently for protection against enemies and to prevent or remove bad luck [13,23].

The Kwa Mai-Mai Muthi Market
The Kwa Mai-Mai Muthi Market is located south of Johannesburg under a bridge on the corner of Anderson and Berea Streets ( Figure 5). The Kwa Mai-Mai Muthi Market is known as one of the oldest markets in Johannesburg, having opened about 50 years ago [23]. It is also known as "Ezinyangeni", which means "the place of traditional healers", as the majority of the traders are either traditional healers or "sangomas". The city of Johannesburg municipality manages the muthi market. Unlike the Faraday Muthi Market, which sells mainly plant and animal products, the Kwa Mai-Mai Muthi Market has a wider diversity of products such as traditional attires, brooms that are often used in wedding ceremonies, and traditional mats (made from plant material). The trading stalls at the Kwa Mai-Mai Muthi Market are indoors and have private consultation rooms where customers can be offered "medical" service ( Figure 5). There are various packaging and distribution methods adapted by traders when selling harvested bark material from the wild. These include raw, dried, solid pieces of material, or partially processed (chopped or pounded) products ( Figure 4). Bark may be ground into fine powder, usually for use as snuffs or for incorporation into complex mixtures for use as insizi (powdered mixtures rubbed into scarification in the skin to alleviate pain). Packaging of the plant material is normally in the form of newspapers, recycled liquor bottles, or plastic bags ( Figure 6). Medicines are commonly prepared with water as an extraction solvent to form a decoction or infusion.

Trader Information and Data Recorded
The selection of traders was random and based on the willingness of an individual to participate, after clearly explaining in their home languages the purpose of conducting the research study. No formal interviews were constructed, as traders were giving out commercial information that a normal customer would require to use the product. For each package of bark sample purchased, the trader was asked the following information: (i) to give a vernacular name or names ("uyalazi igama lesintu lexolo") to aid in the correct identification of the plant, (ii) main uses of the product (bark) ("lisetshenziselwani leli xolo"), (iii) method of preparation ("lisetshenziswa ngaluphi uhlobo"), and (iv) mode of administration ("lisebenza kanjani, uyaphuza, uyaphalaza noma uyachatha"). There were no ethical issues, since only commercial information of consumer and trader was recorded. No bark samples were collected from nature. One sample (Warburgia salutaris) was collected from a cultivated garden tree with permission from the owner (Prof van Wyk). Therefore, no permits were necessary. Plant identification was verified by a botanist, Prof B.-E van Wyk and Dr. E.L. Kotina, a bark anatomist from the University of Johannesburg. Voucher specimens have been deposited at the University of Johannesburg Herbarium. Bark samples often vary considerably, depending on the provenance or the age of the trunk from which it was sampled. This variability may result in errors and misidentifications when bark is described, and it therefore seems very difficult to create a reliable key for the identification of bark samples. In this study, there were cases whereby several bark samples of the same species were purchased from different traders in order to get a more representative sample and to account for natural variation. As a first step towards an identification guide for commercially relevant medicinal barks of southern Africa, the outer and inner bark appearance descriptions were prepared for all 31 barks recorded in this study ( Table 1, Column 1). For accurate color identification, the Munsell Soil Chart was used to record the color codes.

Preparation of Plant Extracts
Dried bark material was cleaned with a brush to remove lichens, moss, and sand or dust particles. Dried bark samples were ground into powder using a high-speed Fritsch Pulverisette grinder (Labotec). Approximately 28 g of each powdered bark material was soaked in 150 mL of either methanol or dichloromethane for 48 h at room temperature. Methanol was chosen as an ideal solvent because it is likely to extract a wider diversity of polar compounds than water. It is often suggested that water extracts should be used when studying traditional medicine, because infusions and especially decoctions are popular traditional dosage forms. However, bark often contains high amounts of triterpenoid glycosides (saponins) and other compounds that may act as emulsifying agents [101]. As a result, traditional decoctions will also contain non-polar compounds. Furthermore, powdered bark is often directly ingested by licking it from the hand (a custom known as "khotha" in Zulu). Dichloromethane was selected to extract non-polar components. Extracts were filtered using a Whatman No. 1 filter paper and left for evaporation in a fumehood. The dried samples were re-dissolved in acetone or dimethyl sulfoxide (DMSO) for antimicrobial and toxicity assays, respectively.

Test Pathogens
Bacterial test organisms were selected based on the conditions that barks are reported to treat, i.e., infections of the skin, stomach, and respiratory tract.

Antimicrobial Assay
A serial micro-dilution assay was used to determine the minimum inhibitory concentration (MIC) values of the bark extracts. The technique involves the use of a 96-well micro-titre plate and tetrazolium salts as an indicator of microbial growth [95]. Using aseptic manipulation, 100 µL of TSB was aliquoted into each well of a sterile 96 well micro-titre plate. A volume of 100 µL of bark extracts at a starting concentration of 32 mg/mL in acetone or DMSO, as well as positive and negative controls, were transferred to the first row of the micro-titre plate. Ciprofloxacin (starting concentration 0.01 mg/mL) was used as a positive control while acetone and 10% DMSO were used as negative controls. Serial dilutions were performed longitudinally by transferring 100 µL of the well content and subsequently diluting the extracts and controls with 50% each time. Crude extracts were initially tested at 8.00 mg/mL and serially diluted two-fold 0.06 mg/mL. A 100 µL of the sub-culture (bacteria) was added to all the wells of each micro-titre plate. The culture was first diluted in broth (0.5 McFarland standard) and diluted 1:100 to give a density of approximately 1 × 10 6 colony forming units/mL (CFUs/mL). Each micro-titre plate was then sealed with a sterile adhesive seal and incubated at 37 • C for 24 h. To confirm the purity of the cultures used, each diluted pathogen-broth mixture was also streaked onto TSA and incubated overnight. After incubation, 40 µL of p-iodonitrotetrazolium (INT: Sigma Aldrich, Johannesburg, South Africa) dissolved in sterile water was added to all micro-titre plate wells and used as a microbial growth indicator. The tests were performed in triplicate, and the MIC values were averaged.

Sample Preparation
Dried powdered stem bark, approximately 1 g, was soaked in 2-3 mL of methanol and dichloromethane. After the solvent extracts had evaporated, the dried solid material was dissolved in 2% DMSO to obtain a concentration of 2 mg/mL. The brine shrimp lethality was tested at this concentration because of the estimated low dermal penetration of active ingredients.

Cytotoxic Brine Shrimp Assay
A total of 31 medicinal barks were subject to toxicity screening against brine shrimps hatched in seawater solution [110]. The brine shrimp eggs were incubated at 25 • C for 24 h. A mass of 16 g sea salt was dissolved in 500 mL distilled water for the preparation of artificial seawater. A sealed conical-shaped plastic container was placed in an inverted position in an empty glass beaker. The dried brine shrimp (Artemia franciscana) eggs (1 g) were then added to the prepared artificial seawater on the inverted plastic bottle. A rotary pump was placed at the bottom of the container to create aeration, in the presence of a constant light source. This was done in order to mimic the natural seawater environment to ensure suitable conditions for the eggs to hatch. This mixture was then poured on the micro-titre well lid and placed upside down with full exposure to the light source to channel or promote accumulation of the nauplii in clusters. This was done to increase the number of brine shrimp for collecting (pipetting). Then, 48-well micro-titre plates were prepared by adding 400 µL of the salt-water solution containing about 40-60 live nauplii to each well along with 400 µL of the bark extract. The same volume of 400 µL was used for the addition of positive control, solvent control, and negative control separately with the brine shrimp eggs. The positive control used was 1.6 mg/mL potassium dichromate, a highly toxic, frequently used, and recommended reference compound for aquatic organisms. The solvent control was DMSO, which was kept at 2%. The number of dead brine shrimp was counted by viewing the plates under a light microscope at 40× magnification following 24 and 48 h of exposure to the test samples at room temperature. After counting at 48 h, a lethal dose of acetic acid (Saarchem; 100% (v/v); 50 µL) was added to each well, and then a final death count was undertaken to calculate the percentage mortality. A mortality percentage of 50% and above was regarded as toxic [102].

Conclusions
The predominant language spoken on the Faraday and Kwa Mai-Mai Muthi Markets in Johannesburg is isiZulu, and it appears that most of the traders have isiZulu as their home language. The dominant medicinal culture on the markets is also Zulu, and very few traditional medicines from other cultures are sold.
Individual species of bark are commonly used for the treatment of more than one ailment. Many of the popularly traded medicinal barks used against skin ailments are explained anecdotally to promote wound healing and achieve cosmetic effects, such as to remove pimples. Common gastrointestinal ailments were diarrhea and intestinal worms. Commonly treated respiratory tract ailments include coughs, tuberculosis, and colds. In general, the studied medicinal barks demonstrated good antimicrobial activity against dermatological, gastrointestinal, and respiratory tract pathogens thereby validating the traditional and commercial uses reported by the Johannesburg Muthi Market traders. Based on the toxicity results from this study, it can be concluded that the moderate activity exhibited by bark extracts is not strongly influenced by the presence or absence of toxic properties of the bark material, since extracts displayed mortality rates of less than 50%, which is regarded as non-toxic. These results support the safety and efficacy of traded medicinal barks from the informal muthi markets of South Africa. The most potent bark species were E. transvaalense, E. lysistemon, Garcinia livingstonei, Pterocelastrus rostratus, Rapanea melanophloeos, S. brachypetala, Sclerocarya birrea, and Ziziphus mucronata. Some of these species have already been examined for phytochemical composition, but the connection between chemistry and antimicrobial activity is not fully elucidated for all the species examined here. Hence, a follow-up bioactivity guide fractionation process may help to identify antimicrobial actives and create leads for the nutricosmetics industry [111].