Despite being a highly virulent and drug-resistant pathogen, P. aeruginosa has somehow, not received much attention from SA ethnobotanists as evidenced by the scantly available information in the accessed literature (Figure 1). In over two decades, only 31 ethnobotanical studies relating to the pathogen were published, averaging a meagre 1.6 publications per annum. Interestingly, no such publications were made within the first 3 years (2018–2020) after the pathogen was classified as a critical research priority by the World Health Organization (Figure 1) (WHO 2017). It was also quite interesting to note that none of the studies conducted were dedicated solely to P. aeruginosa. Apparently, little efforts were made to screen SA medicinal plants against antibiotic-resistant strains of the pathogen. Of the 31 reports published, only a study by Soyingbe et al. (2013) included the drug-resistant strain of the bacterium as a test organism.

A total of 152 plant species belonging to 78 families were screened against the bacterium (Tables 1, 2, 3, 4). The most represented families were Fabaceae (16.7%), Euphorbiaceae (12.8%) and Anacardiaceae (9%). The most investigated species were Cussonia spicata, Ricinus communis, Sclerocarya birrea and Zizipus mucronata. It was encouraging to note that almost half (45%) of all plant species evaluated demonstrated noteworthy antibacterial activities (minimum inhibitory concentration [MIC] < 1 mg/mL) against the pathogen. Barringtonia racemosa, Croton megalobotrys, Erythrina caffra, Leucosidea sericea, Maesa lanceolata, Morella serrata and Trichilia emetica yielded potent antibacterial activities (MIC range: 0.02–0.09 mg/mL, Table 1) and as such, warrants further investigations. Using the disc diffusion assay, Mongalo, Opoku and Zobolo (2012) also demonstrated that the acetone root extracts of Waltheria indica possess significant growth inhibitory properties against the pathogen. As the aforementioned plants belong to eight different families (Euphorbiaceae, Fabaceae, Lecythidaceae, Maesaceae, Meliaceae, Myricaceae, Rosaceae and Sterculiaceae), the principal bactericidal compounds in them are likely both structurally and functionally diverse. This presents encouraging prospects of finding an assortment of clinically relevant anti-pseudomonas compounds within them.

In addition to crude plant extracts, some researchers investigated the antibacterial properties of semi-purified extracts or isolated phyto compounds. Magama et al. (2003), for instance, screened semi-purified leaf extracts of Euclea crispa against the bacterium and reported weak-moderate antibacterial activities. However, the compounds isolated from the leaves exhibited poor antibacterial activities against P. aeruginosa in a separate study by Pretorius, Magama and Zietsman (2003).

Whilst some researchers screened medicinal plants from a variety of families, others focused on specific medicinal plant species including Antidesma madagascariense, Erythrina caffra, E. crispa, Lavandula angustifolia, Morella serrata, S. schinus, Tulbaghia violacea and many others (Gundidza et al. 2008; Pretorius et al. 2003; Seebaluck-Sandoram et al. 2017; Soyingbe et al. 2013). In some studies, only medicinal plants used by people within a specific geographical location were evaluated. The study areas covered were predominantly in the Eastern Cape, Limpopo, KwaZulu-Natal and Mpumalanga (Masika & Afolayan 2002; Masoko 2013; Mongalo et al. 2012; Oyedeji, Afolayan & Eloff 2005; Selowa et al. 2010).

In a study by Samie et al. (2005) on 14 medicinal plants used by the Venda people of Limpopo to manage a wide range of infectious diseases, only S. cordatum and A. johnsonii demonstrated noteworthy antibacterial activities against P. aeruginosa (MIC = 0.35 and 0.62 mg/mL, respectively). Kelmanson, Jäger and Van Staden (2000) examined the antibacterial activities of 14 Zulu medicinal plant species and reported that Crinumviridis, Dioscorea dregeana and Vernonia colorata possess moderate growth inhibitory activities against the bacterium (MIC range: 0.36–0.63 mg/mL). These two were, to the best of our knowledge, the only reports in which the medicinal plants used by a given ethnic group in SA were screened against the pathogen. This could probably stem from a limited number of relevant ethnobotanical surveys available. Generally, the indigenous knowledge gathered through such surveys informs scientists as to which plants to screen for antibacterial or any other pharmacological properties.

South Africa has a vast array of aromatic plant species some of which are part of the materia medica utilised by local traditional healers. Asteraceae (2300 species), Lamiaceae (235 species) and Rutaceae (290) are classic examples of prominent aromatic families found in the country (Lawrence 2006). Owing to the importance of aromatherapy worldwide, an increasing number of local aromatic plants are being screened for a variety of therapeutic properties.

As shown in Table 2, only a few studies (7) focused on screening essential oils against P. aeruginosa. The investigated oils were extracted from different parts of Annona segegalensis, Callistemon citrinus, Callistemom viminalis, L. angustifolia, Leonotis leonurus, L.ocymifolia, Lippia javanica, Myrothamnus flabellifolius, Schinus terebinthifolius and Tulbaghia violacea. These plants belonged to seven families namely, Alliaceae, Anacardiaceae, Annonaceae, Lamiaceae, Myrothamnaceae, Myrtaceae and Verbenaceae. However, none of the evaluated oils demonstrated promising antibacterial activities except those from the leaves of L. angustifolia, Leonotis ocymifolia (MIC: 0.3 mg/mL) and the aerial parts of M. flabellifolius (Table 2).

Essential oils are generally volatile and lipophilic making them much more difficult to assess than crude plant extracts (Van Vuuren 2008). This could have possibly resulted in the former being less favoured by researchers than the latter (Tables 1 and 2). It is also important to note that some hydro-distilled essential oils are generally less efficacious than ‘full-bodied’ crude plant extracts primarily because some bioactive compounds in the purified oils work synergistically with non-volatile phyto-compounds to elicit therapeutic effects. Isolating the oils might therefore disrupt key synergistic interactions. This could possibly explain why most of the essential oils evaluated were not effective against P. aeruginosa (Table 2).

As already alluded to, the empirical basis of herbal medicine lies in the existence of bioactive phyto compounds. The multi-step process of plant-based drug development often starts with the accurate identification of potential sources of therapeutic phyto-compounds. As such, phytochemical analysis has become an important part of ethnopharmacology. In addition to antibacterial screening, some of the plant extracts screened against P. aeruginosa were subjected to either qualitative or quantitative phytochemical analysis.

Using bio-guarded isolation techniques, Mabona et al. (2013) managed to isolate and identify a compound known as plumbagin (Figure 2), from the leaves of Aristea ecklonii (Asteraceae). Interestingly, the compound displayed potent bactericidal effects against P. aeruginosa (8 µg/ mL). To the best of our knowledge, this was the only successful attempt at isolating potent anti-P. aeruginosa compounds from South African medicinal plants documented over the past 20 years. It is, however, worth noting that plumbagin was previously isolated from other medicinal plants such as Plumbago scandens and Plumbago zeylanica (Jeyachandran et al. 2009; Paiva et al. 2003).

Based on gas chromatography-mass spectroscopy (GC-MS) data analysis, Gundidza et al. (2008) attributed the antibacterial properties displayed by S. terebinthifolius essential oil against P. aeruginosa (Table 2) to a wide range of bioactive compounds including camphene, m-cymene, 1-β-pinene, α-pinene and γ-terpinene. Camphene, m-cymene, α-pinene and β-pinene were also putatively detected (GC-MS) in the essential oil extracted from Myrothamnus flabellifolius by Viljoen et al. (2002), which, however, demonstrated poor antibacterial properties when assessed using the time-kill assay (Table 2). Other putative antibacterial compounds found in the oils were pinocarvone, myrtenol and trans-pinocarveol. Gas chromatography-mass spectroscopy was also used to determine the phytochemical profiles of antibacterial essential oil extracted from L. leonurus and L. ocymifolia (Oyedeji et al. 2005). Although a comprehensive phytochemical analysis of the oils was conducted, no mention of possible antibacterial compounds was made in the report.

Using a thin-layer chromatography-based bioautography, some researchers identified medicinal plants with promising anti-P. aeruginosa compounds (Table 3). Interestingly, some of the medicinal plants evaluated contained more than one active compound against the pathogen, particularly Ricinus communis, Senna italica and Sclerocarya birrea (5, 4 and 3 active inhibition bands, respectively, Table 3). It is worth noting that even though some of the bands significantly inhibited the growth of the bacterium (Table 3), the principal antibacterial compounds were not isolated and identified. Some preliminary phytochemical studies tested the presence of alkaloids, anthraquinones, flavonoids, phenolics, glycosides, saponins and tannins (Chikoto et al. 2005; Mabona et al. 2013; Shikanga, Combrinck & Regnier 2010). These compounds are known to have a wide range of pharmacological properties and as such could have contributed to the observed antibacterial activities.

One effective way to combat drug-resistant pathogens is by using combination therapies. As such, modern antibacterial therapies often include combinations of antibiotics with other antibiotics, plant extracts or different chemical entities. Combination therapies widen the antibacterial spectrum, improve the efficacy of clinically infective drugs and generally delay the development of antibiotic resistance (Chukwujekwu & Van Staden 2016; Tripodi et al. 2007; Zhao et al. 2002). Systematic evaluation of combination therapies used in folk medicine as well as those involving plant extracts and antibiotics could lead to the discovery of new therapeutic compounds.

It was encouraging to note that some South African medicinal plant extracts interacted synergistically with conventional antibiotics against P. aeruginosa (Table 4). Nearly all investigations were conducted using the checkerboard bioassay (Rand et al. 1993) in which the efficacy of each pairwise combination was determined using the fractional inhibitory concentration index (FICI). The FICI was obtained by using the formulae, FICI = FIC_X + FIC_Y, where FIC_X was the MIC of antibacterial agent X when used in combination with antibacterial agent Y, divided by the MIC of antibacterial agent X when used alone. The results were interpreted, thus FICI ≤ 0.5 (synergy), 0.5 < FICI ≤ 1.0 (additive) and 1.0 < FICI ≤ 4.0 (no interaction) and FICI > 4.0 (antagonism) according to Van Vuuren and Viljoen (2011).

In a study by Hübsch et al. (2014), P. aeruginosa was subjected to combinations of L. javanica essential oils with each of the antibiotics ciprofloxacin, gentamicin and tetracycline. Only two of the combinations (L. javanica + gentamicin and L. javanica + tetracycline) yielded noteworthy antibacterial synergism (FICI range: 0.32–0.5), whilst the combination of ciprofloxacin and L. javanica resulted in additive effects (FICI = 0.56, Table 4). In a separate study by De Rapper et al. (2016), the combination of L. angustifolia essential oils and chloramphenicol also resulted in notable antibacterial synergism (FICI = 0.29). Further investigations by the same authors using isobolograms reaffirmed the results. Seebaluck-Sandoram et al. (2017) observed a 53-fold decrease in the MIC value of ciprofloxacin when used in combination with essential oils extracted from the leaves of Antidesma madagascariense against the pathogen. Combinations of the same oils with chloramphenicol or streptomycin also yielded outstanding synergistic effects (FICI range: 0.08–0.19) against the bacterium.

Overall, these findings seem to suggest that some chemical entities in the evaluated extracts inhibited efflux pumps or beta-lactamase enzymes in P. aeruginosa ultimately leading to the intracellular accumulation of antibiotics to lethal levels. This possibly explains the significant reduction in the MICs of antibiotics when used in combination with plant extracts, especially in the pairwise combination of A. madagascariense essential oil with either ciprofloxacin or streptomycin (FICI: 0.08–0.11). Another possibility is that the combined bactericidal effect of the antibiotics and some phytocompounds made the pathogen more susceptible to death, especially if the two antibacterial agents used had different target sites.

As advised by the WHO, the search for new and effective anti-P. aeruginosa agents must be an utmost research priority, especially in countries endowed with plant species richness like SA (WHO 2017). Local researchers could widen the search for such agents by screening plants (medicinal or non-medicinal) closely related to those who previously exhibited noteworthy activities against the pathogen (Table 1, 2, 3). Such taxonomic-based antibacterial screening should perhaps mainly target plants within the same genus. Future antibacterial screening should also include a wide range of drug-resistant mutants of the pathogen. Currently, the efficacy of South African medicinal plants against antibiotic-resistant pathogens is largely unknown (Van Vuuren 2008). The principal antibacterial compounds in medicinal plants that demonstrated potent antibacterial activities (MIC: 0.02–0.09 mg/mL, Table 1) should be isolated and unequivocally identified. Additionally, as the chemical profiles of antibacterial essential oils were putatively determined by GC-MS, consideration should also be given to identifying the principal antibacterial compounds in the oils using spectroscopic and X-ray data analysis techniques.

Furthermore, precedence should not only be given to the search for bactericidal phyto-compounds, but concerted efforts should also be made to discover effective indigenous plant-based resistance modifying compounds. The use of drug-resistance modifying agents (RMAs) in combination therapies could potentially improve the efficacy and hence allow the possible reintroduction of some clinically ineffective antibiotics. From a financial point of view, this approach seems more appealing than developing novel therapeutic agents which customarily have to undergo extensive efficacy and safety evaluations before approval. In a bid to systematically screen SA medicinal plants for RMAs, synergistic evaluations between conventional antibiotics and plant extracts that possess noteworthy antibacterial activities (Tables 1, 2, 3, 4) should be evaluated against various drug-resistant mutants of P. aeruginosa. It should, however, be noted that the checkerboard assay is not always reliable and as such all synergistic interactions detected by the method should be confirmed by other more sensitive techniques such as the time-kill bioassay.