Antiplasmodial, antimalarial activities and toxicity of African medicinal plants: a systematic review of literature

Background Malaria still constitutes a major public health menace, especially in tropical and subtropical countries. Close to half a million people mainly children in Africa, die every year from the disease. With the rising resistance to frontline drugs (artemisinin-based combinations), there is a need to accelerate the discovery and development of newer anti-malarial drugs. A systematic review was conducted to identify the African medicinal plants with significant antiplasmodial and/or anti-malarial activity, toxicity, as wells as assessing the variation in their activity between study designs (in vitro and in vivo). Methods Key health-related databases including Google Scholar, PubMed, PubMed Central, and Science Direct were searched for relevant literature on the antiplasmodial and anti-malarial activities of African medicinal plants. Results In total, 200 research articles were identified, a majority of which were studies conducted in Nigeria. The selected research articles constituted 722 independent experiments evaluating 502 plant species. Of the 722 studies, 81.9%, 12.4%, and 5.5% were in vitro, in vivo, and combined in vitro and in vivo, respectively. The most frequently investigated plant species were Azadirachta indica, Zanthoxylum chalybeum, Picrilima nitida, and Nauclea latifolia meanwhile Fabaceae, Euphorbiaceae, Annonaceae, Rubiaceae, Rutaceae, Meliaceae, and Lamiaceae were the most frequently investigated plant families. Overall, 248 (34.3%), 241 (33.4%), and 233 (32.3%) of the studies reported very good, good, and moderate activity, respectively. Alchornea cordifolia, Flueggea virosa, Cryptolepis sanguinolenta, Zanthoxylum chalybeum, and Maytenus senegalensis gave consistently very good activity across the different studies. In all, only 31 (4.3%) of studies involved pure compounds and these had significantly (p = 0.044) higher antiplasmodial activity relative to crude extracts. Out of the 198 plant species tested for toxicity, 52 (26.3%) demonstrated some degree of toxicity, with toxicity most frequently reported with Azadirachta indica and Vernonia amygdalina. These species were equally the most frequently inactive plants reported. The leaves were the most frequently reported toxic part of plants used. Furthermore, toxicity was observed to decrease with increasing antiplasmodial activity. Conclusions Although there are many indigenous plants with considerable antiplasmodial and anti-malarial activity, the progress in the development of new anti-malarial drugs from African medicinal plants is still slothful, with only one clinical trial with Cochlospermum planchonii (Bixaceae) conducted to date. There is, therefore, the need to scale up anti-malarial drug discovery in the African region. Supplementary Information The online version contains supplementary material available at 10.1186/s12936-021-03866-0.


Background
Malaria still constitutes a major public health menace, especially in tropical and subtropical countries. Various species of Plasmodium, transmitted through the bite of an infected female Anopheles mosquito, cause malaria, including Plasmodium falciparum, Plasmodium malariae, Plasmodium ovale, Plasmodium vivax, and Plasmodium knowlesi. Among these species, P. falciparum is the most virulent, responsible for the highest morbidity and mortality. It is also the predominant species in sub-Saharan Africa (SSA), a region with the highest number of malaria cases and deaths in the world. According to the World Health Organization (WHO), there were 228 million cases, and 405,000 malaria attributed deaths in 2018 [1]. In SSA, children and pregnant women are the most at-risk groups [1][2][3].
Malaria can be treated using chemotherapy but there is widespread resistance to many of the drugs. The first case of resistance to artemisinins was reported in Cambodia in 2006 and has then spread to most of South-East Asia [4,5]. The safety of chemoprophylaxis is also a major concern; for instance, primaquine, atovaquone, and doxycycline are contraindicated in pregnant women and children [6]. All these shortcomings necessitate the discovery and production of new drugs to treat malaria.
In the past 50 years, natural compounds including plant products, have played a major role in drug discovery and have provided value to the pharmaceutical industry [7]. For instance, therapeutics for various infectious diseases, cancer, and other debilitation diseases caused by metabolic disorders have all benefitted from many drug classes that were initially developed based on active compounds from plant sources [8]. Furthermore, quinine and artemisinin, and their synthetic derivatives which are the mainstay of anti-malarial chemotherapy, were also derived from plant sources. In malaria-endemic areas, especially in Africa, many people rely on herbal medicines as the first line of treatment [9]. The common reasons for their preference vary from the cost of standard drugs, availability and accessibility, perceived effectiveness, low side effect, and faith in traditional medicines [10].
Reviews of the antiplasmodial and anti-malarial activities of medicinal plants are needed to drive research into the discovery and production of new anti-malarial drugs. Only a few reviews of the antiplasmodial or anti-malarial activity of medicinal plants have been published in the scientific literature [11][12][13][14][15][16]. These reviews focused only on studies with high antiplasmodial or anti-malarial activity and hardly report on their toxicity. The purpose of this study was to review medicinal plants with moderate to very good antiplasmodial and anti-malarial activities, as well as assess the variation in the activities between different methods. Furthermore, the toxicity of plant species is highlighted.

Methods
The literature was reviewed in search of scientific articles reporting antiplasmodial activities (IC 50 , ED 50 , LD 50 , and parasite suppression rate) of medicinal plants used in Africa to treat malaria. The current study conforms to the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) guidelines [17].

Search strategy and selection criteria
Relevant articles were searched in health-related electronic databases including PubMed, PubMed Central, Google Scholar, and ScienceDirect using the keywords: Traditional herbs or Medicinal plants or Antiplasmodial activity or Antimalarial activity or Herbal medicine or Plasmodium.
The search was limited to studies published in English or containing at least an abstract written in English until May 2020. The titles and abstracts were subsequently examined by two reviewers, independently (parallel method) to identify articles reporting the antiplasmodial activity of medicinal plants. In the case of any discrepancy in their reports, a third reviewer was brought in to resolve the issue. Relevant papers were equally manually cross-checked to identify further references. The following data were extracted from the selected articles by the reviewers: plant species, plant family, place of collection of plant, parts of the plant used, type of study (whether in vitro, in vivo, or human), the extraction solvent used, IC 50 or ED 50 values, parasite suppression rate, isolated compounds, interaction with known malarial drugs (whether synergistic or antagonistic), and toxicity. Articles that did not report antiplasmodial or anti-malarial activity of medicinal plants as well as review articles were excluded. The entire selection process is presented in Fig. 1.
In this study, antiplasmodial activity pertains to studies performed in vitro using different strains of Plasmodium falciparum, meanwhile, anti-malarial activity is reserved for in vivo studies performed using mice and various parasite models (including Plasmodium berghei, Plasmodium yoelii, and Plasmodium chabaudi) and reporting parasite suppression rate.

Categorization of antiplasmodial and anti-malarial activities
For in vitro studies, the antiplasmodial activity of an extract was considered very good if IC 50 < 5 µg/ml, good 5 µg/ml ≤ IC 50 < 10 µg/ml, and moderate 10 µg/ ml ≤ IC 50 < 20 µg/ml [18]. For in vivo studies, the antimalarial activity of an extract is considered very good if the suppression is ≥ 50% at 100 mg/kg body weight/ day, good if the suppression is ≥ 50% at 250 mg/kg body weight/day, and moderate if the suppression is ≥ 50% at 500 mg/kg body weight/day [18]. Antiplasmodial activities of 20 µg/ml and above for in vitro studies and

Results
The PRISMA flowchart (Fig. 1) presents a four-phase study selection process in the present systematic review study. A total of 25,159 titles were identified in the initial search. After the title and abstract screening, 228  articles were retrieved. Of these, a final 200 articles were identified for the review. For this review, the evaluation of the individual plant species was considered as an independent study, so it is common for one article to have more than one study depending on the number of plant species evaluated. In all, there were 722 independent studies. Five hundred and ninety-on (81.9%) of the independent studies were in vitro (Table 1), 90 (12.4%) were in vivo ( Table 2) and 40 (5.5%) were both in vitro and in vivo (Table 3). There was only one human study (clinical trial) conducted so far ( Table 4). The selected research articles were from 31 African countries. Out of the 200 research articles reviewed, most of them were from Nigeria 58 (29.0%), Kenya 24 (12.0%), Ethiopia 13 (6.5%), Cameroon 12 (6.0%), Ivory Coast 11 (5.5%), D.R. Congo 10 (5.0%), and Burkina Faso 7 (3.5%) (Fig. 2). The studies cover the period from 1989 to 2020.

In vitro and in vivo activities of the plants evaluated
Overall, 248 (34.3%) of the studies reported activity that was very good (IC 50 values < 5 µg/ml or suppression rate of ≥ 50% at 100 mg/kg body weight/day), 241 (33.4%) reported good activity and 233 (32.3%) reported moderate activity. For the in vitro studies, a majority 228 (38.6%) reported very good activity; 206 (34.9%) reported good activity and 187 (31.6%) reported moderate activity. Meanwhile for the in vivo studies, a majority 19 (21.1%) reported moderate activity, 16 (17.8%) reported very good activity and 13 (14.4%) reported good activity. For studies reporting both the in vitro and in vivo activity, a majority of 17 (42.5%) reported only moderate activity, 13 (32.5%) studies reported very good activity and 10 (25.0%) reported good activity. Among the plants with very good activity, only one species demonstrated very good activity both in vitro and in vivo (Table 3).
Azadirachta indica and Vernonia amygdalina were the most frequently reported inactive species (Additional file 1: Table S1). Furthermore, Fabaceae, Rubiaceae, Euphorbiaceae, and Asteraceae were the plant families containing the most frequently reported inactive plants. A majority of 95.7% (691/722) of the studies used the crude extract of the plants. The antiplasmodial and/or anti-malarial activity was significantly higher (p = 0.044) in studies using pure compounds compared to those using crude preparations.

Toxicity of plants evaluated for their antiplasmodial and anti-malarial activity
Out of the 198 plants evaluated in toxicity assays, 52 (26.3%) were found to demonstrate some degree of toxicity. The most frequently reported plants with toxicity were Azadirachta indica and Vernonia amygdalina. Plant families harboring the most toxic species were Lamiaceae, Anacardiaceae, Moraceae, Meliaceae, Asteraceae, and Fabaceae. Approximately 33% of the plants tested demonstrated some toxicity in vitro and 26.7% had some degree of toxicity in vivo. Among plants with very good, good, and moderate antiplasmodial activity, 17.8%, 28.3%, and 35.4% had some degree of toxicity, respectively. The leaf was the plant part with the most frequently reported toxicity. Albino mice and Vero E6 cells were the most commonly used assays for the assessment of the toxicity of the plants.

Discussion
Resistance to the frontline anti-malarial drugs is increasing and is now a global concern. With this rising rate of resistance, there is a need to accelerate research into the discovery and development of new anti-malarial drugs. Unfortunately, from this study, it is evident that the progress into the discovery of a new anti-malarial drug in Nd [22] South Africa                          [213] had also arrived at the same conclusion. This review revealed research articles from 31 African countries. Most of the articles were from Nigeria. This is suggestive that Nigeria is leading the podium in research on anti-malarial drug discovery and development, deservedly so, because she is probably the most affected country in the world. It is noteworthy that South Africa which is generally more technologically advanced than Nigeria had very few (8) articles. The African region is the most affected in the world recording the greatest number of cases and malaria attributed deaths. However, the distribution of malaria in Africa is not even, with sub-Saharan Africa harboring disproportionately the greatest number of cases. This is suggestive that research to identify new anti-malarial drugs may be related to the burden of the disease, thus the government policy to control the disease. There is, therefore, the need for policy-driven research into new anti-malarial all across the African region. In this review, IC 50 values of < 20 µg/ml were considered as the cutoff of significant anti-malarial activity. This cutoff is considered the minimum to qualify as a first-pass "hit" in anti-malarial drugs screening [214]. Five hundred and two (502) plant species from 169 families were observed to have moderate to very good anti-malarial activity. The most investigated plant families were Euphorbiaceae, Fabaceae, Rubiaceae, and Annonaceae. However, the plant families containing the most active plants were Apocynaceae, Celestraceae, and Rutaceae. This finding suggests that more emphasis should be given to plants in these families for anti-malarial drug discovery. Besides, the most investigated plant species were Azadirachta indica, Nauclea latifolia, Picralima nitida, and Zanthoxylum chalybeum. Alchornea cordifolia, Flueggea virosa, Crytolepis sanguinolenta, and Zanthoxylum chalybeum were the only plant species with consistently very good antiplasmodial and anti-malarial activities between studies. This is very surprising that no clinical trial using any of these plants has been conducted. Further studies on these plant species should be performed.
This study revealed that overall, a majority of the plants investigated had very good antiplasmodial activity in vitro. That activity decreases as you move to in vivo in most studies, with a majority of plants demonstrating only moderate activity. For example, Gathirwa et al. [146] showed that the activity of Uvaria acuminate decreased from good activity in vitro to inactive in vivo. However, a few studies show that plant activity could also increase from in vitro to in vivo. For example, Ngbolua et al. [211] showed that the activity of Vernonia ambigua increased from in vitro to in vivo analysis. Other examples include studies by Muthaura et al. [20] using Boscia angustifolia, Kweyamba et al. [162] using Commiphora Africana, and Ajaiyeoba et al. [204] using Annona senegalensis. This suggests that plants could still have significant antimalarial activity in vivo although they failed to in vitro. Most investigators usually progress to in vivo studies only when they observe significant antiplasmodial activity in vitro. This may explain the findings of a smaller number of in vivo studies in the current study. The investigation of the anti-malarial activities of plants should continue in vivo despite the dismal performance of the plants in vitro.      The current study revealed substantial inter-study variation in the antiplasmodial activity of several plant species. For example, considerable variation in the antiplasmodial activity was observed for Senna occidentalis, Adansonia digitata, Acanthospermum hispidum, Rotheca myricoides, Anogeissus leocarpus, Annona muricata, Ageratum conyzoides, Albizia coriaria, Ekebergia capensis, Flueggea virosa, Lippia javanica, Maytenus senegalensis, Morinda lucida, Picralima nitida, Trichilia emetica, Vernonia amydalina, and Vernonia colorata. The factors that could have accounted for these differences may include differences in the extraction solvent thus the extraction yield and extracted metabolite. With dichloromethane, mainly the apolar metabolites are extracted. In contrast, with methanol, from polar to moderate apolar metabolites are extracted.
Most (95.7%) of the studies used crude extract for their investigation and rarely the pure compounds (Additional file 1: Table S2 presents a summary of active compounds that have been identified from some of the plants). The finding of a majority of studies in Africa using only the crude extract of plants may be attributed to the absence of the necessary infrastructure to process the plant materials to get the pure compounds. Furthermore, there may be geographical differences in the areas where the plants were collected and this may also affect the activity of the same plant species. For example, despite using the same extraction solvent, the antiplasmodial activity of Acacia nilotica was moderate in South Africa and very good in Sudan. There was also variation between the different assay types. For example, the activities of Vernonia ambigua [211] and Annona senegalensis [204] have been reported to increase from inactive in vitro to very good in vivo. However, a few plant species including Alchornea cordifolia, and Zanthoxylum chalybeum, were observed to be consistently very good between studies. These plant species should be exploited further for their antiplasmodial activity. The activities of the plants were equally observed to increase with the isolation of the active compounds thus reinforcing the need for research into identifying the active compounds of African medicinal plants. The marked difference in the antiplasmodial activity of the crude extract of Artemisia annua and the pure compounds points out the issue that even the compounds which show only low potency and may be discarded from the initial screen for further development may still have active components with therapeutic potential [215]. The strain of the Plasmodium used may also be another factor accounting for the inter-study variation observed; studies using chloroquine-sensitive strains of the parasite like P. falciparum 3D7, D6, NF54 tend to report higher antiplasmodial activity compared to studies using chloroquineresistant strains like P. falciparum W2, Dd5, K1 or D10.
This study revealed that only a few (26.3%) of the plants demonstrated some degree of toxicity. The families hosting the most toxic plant species were Lamiaceae, Anacardiaceae, Moraceae, and Meliaceae. The most toxic plants were Azadirachta indica and Vernonia amygdalina. The former [168] is one of the few plant species that demonstrated very good antiplasmodial activity in some studies. Other plants with high toxicity but very good antiplasmodial/anti-malarial activities include Arenga engleri [25], Celtis integrifolia [52], Ficus platyhylla [50], Gutenbergia cordifolia [21], Helchrysum cymosum [97], Microglossa pyrifolia [92], Opilia celtidifolia [52], Quassia Africana [103], Rumex abyssinicus [92], Clausena anisota [157], Icacina senegalensis [171], Abutilon grandiflorum [200], and Lannea schweinfurthii [205]. The isolation of the active compounds, which has to be done, could eliminate the toxicity, if not all, to a certain degree. For example, Salvia radula crude extract (of aerial parts) has been shown to demonstrate some degree of toxicity, but betulafolientriol oxide isolated from the plant was very active with little or no toxicity against human kidney epithelial cells [120]. There was also considerable variation in the toxicity between the assay types (in vitro or in vivo). As many as 32.8% of the plants demonstrated some level of toxicity in vitro meanwhile 26.7% were toxic in vivo. Since it is customary to evaluate toxicity at the in vitro level and toxic plants are discarded before in vivo evaluation, that may explain while fewer plants were toxic in vivo. Toxicity varied within the same plant species from study to study and could be attributed to differences in the study design as well as differences in the parts of the plants used for testing. From this study, the most toxicity was observed with the leaves. Also, a relationship could be established between toxicity and antiplasmodial activity; as the activity of the plant increases, the toxicity, on the other hand, was observed to decrease. Furthermore, albino mice and Vero E6 cells were the most commonly used assays in the evaluation of toxicity. Unfortunately, the authors could nt make a meaningful relationship between the type of assay and toxicity because of the fewer studies assessing the toxicity of the medicinal plants.
This study, however, is limited in that the analyses may have been compounded by the substantial interstudy variation in the methodologies used by different independent studies for the extraction of plant material, the overall extraction yield, the diversity of extracted metabolites as well as the geographical variations in the different sites used in the plant collection. However, the study has provided important baseline data that may be exploited by researchers in the field for the discovery and development of new anti-malarial drugs.

Conclusion
This study has revealed the slothful progress in the discovery and development of new anti-malarial drugs from African medicinal plants. Despite the encouraging activities demonstrated by the plants in vitro, fewer plants have been evaluated in vivo and just one clinical trial has been conducted so far with Cochlospermum planchonii (Bixaceae). The study also revealed considerable inter-study variation in the antiplasmodial activities of the plants, however, the activity of some plants including Alchornea cordifolia, Azadirachta indica, and Zanthoxylum chalybeum was consistently very good. The study demonstrates a relationship between antiplasmodial activity and toxicity whereby the toxicity of the plants decreases as the antiplasmodial activity increases. Besides, the active compounds were identified in just a handful of the plants. Therefore, there is a need for a policy-driven approach in the discovery and development of new anti-malarial drugs to subvert the rising resistance to the frontline anti-malarial drugs in the world.