A systematic review of pentacyclic triterpenes and their derivatives as chemotherapeutic agents against tropical parasitic diseases

Summary

Parasitic infections are among the leading global public health problems with very high economic and mortality burdens. Unfortunately, the available treatment drugs are beset with side effects and continuous parasite drug resistance is being reported. However, new findings reveal more promising compounds especially of plant origin. Among the promising leads are the pentacyclic triterpenes (PTs) made up of the oleanane, ursane, taraxastane, lupane and hopane types. This paper reviews the literature published from 1985 to date on the in vitro and in vivo anti-parasitic potency of this class of phytochemicals. Of the 191 natural and synthetic PT reported, 85 have shown high anti-parasitic activity against various species belonging to the genera of Plasmodium, Leishmania, Trypanosoma, as well as various genera of Nematoda. Moreover, structural modification especially at carbon 3 (C3) and C27 of the parent backbone of PT has led to improved anti-parasitic activity in some cases and loss of activity in others. The potential of this group of compounds as future alternatives in the treatment of parasitic diseases is discussed. It is hoped that the information presented herein will contribute to the full exploration of this promising group of compounds as possible drugs for parasitic diseases.

INTRODUCTION

Tropical parasitic diseases have been a serious public health problem especially in middle- and low-income countries. These diseases which include malaria, trypanosomiasis, leishmaniasis, schistosomiasis, lymphatic filariasis and onchorcerciasis affect millions of people, resulting in thousands of death annually. The disability-adjusted life year lost estimate for these diseases is very high with a combined annual magnitude of more than 70 million by 2011 (Bhutta et al. 2014; Hotez et al. 2014). At present, there have been reports on the spread of parasitic infections in non-endemic areas which raised more concerns about the feasibility of the global control strategy (Leder et al. 2013). The main obstacles in the control of parasitic diseases are the drugs resistance, toxicity and non-affordability of the available drugs (Buckner et al. 2012). This has prompted a continuous search for safer and more effective treatments especially from natural sources. In this regard, plants have been a prime target for novel therapeutic agents as evident from the large volume of studies being conducted on medicinal plants documented in scientific databases (Rocha et al. 2005; Wright, 2010; Izumi et al. 2011; Ibrahim et al. 2014). Interestingly, considerable success has been recorded with about 65% of all anti-parasitic agents marketed from 1981 to 2010 being originally derived from plant sources and sometimes with synthetic modifications (Newman and Cragg, 2012). This further stimulates research activities on this important area to identify novel bioactive anti-parasitic agents that could potentially be used to combat tropical parasitic diseases. Fortunately, a number of bioactive agents, such as flavonoids, curcuminoids and triterpenoids have shown promising anti-parasitic activities that warrant further drug development studies (Rasoanaivo et al. 2011; Ibrahim et al. 2014).

Pentacyclic triterpenes (PTs) belong to a group of widespread isoprene-derived secondary metabolites collectively known as triterpenes (a sub-class of terpenes). PTs are synthesized mainly by the cyclization of oxidosqualene and squalene and exist in their free form or as components of saponins (glycosides) in many tropical and subtropical plants (Xu et al. 2004; Jäger et al. 2009). The compounds have attracted attention due to their remarkable biological activities. With regard to this, three groups of PT, namely; the oleanane, ursane and lupane groups are considered to be the most important (Dzubak et al. 2006), although other groups such as hopane, taraxastane and friedelane types may also be important. Thus, various derivatives of the biologically important groups of PT are synthesized with the aim of lowering the toxicity and/or increasing the therapeutic activity of the parent compounds. Some of these PT are already registered and/or being marketed in some parts of the world as clinical drugs for the treatment of liver related diseases and diabetes, while others are at various phases of clinical trials (Sheng and Sun, 2011).

Presently, update on the newly discovered PT is a subject of annual review, suggesting an interest to keep track of the advances made in the study of this group of compounds (Dzubak et al. 2006). Moreover, various biological activities, chemistry and therapeutic potencies of the group have been reviewed to highlight the full potencies of this group of compounds. Among the available reviews are the chemistry and metabolic disease-ameliorative effects (Sheng and Sun, 2011), anti-cancer (Laszczyk, 2009), anti-inflammatory (Safayhi and Sailer, 1997), anti-microbial (Wolska et al. 2010), anti-chagasic (da Silva Ferreira et al. 2013b ) and other pharmacological activities (Dzubak et al. 2006). However, a compiled review focusing on the activities of PT against broad range of parasites is lacking. This is despite the potent activities of various members of the group against different parasites as well as the crucial need to develop novel anti-parasitic agents. Hence, a review focusing on the anti-parasitic properties of PT will serve as complementary information in the spectrum of the biological activities of this promising group of phytochemicals.

Available data on plant derived PT investigated for activities against the tropical parasitic infections are reviewed in this paper. This will serve as up-to-date information that could provide direction for future scientific research as well as the future application of this group of compounds as anti-parasitic agents. The article could contribute to the search for effective drugs, which is fundamental in the global fight against parasitic infections.

METHODS

The information presented here is based on PubMed, Medline, SciFinder and Google Scholar search for the PT and their derivatives reported to be tested against parasites of the genera Trypanosoma, Plasmodium, Leishmania, Schistosoma and others, which are considered of importance to tropical countries. Some articles were found through tracking citations in other publications or by accessing the journals’ websites. Various keywords were permutated for the search which include: PT, oleanane, friedelane, ursane, taraxasterane, lupane, hopanes, saponins, anti-parasitic, anti-plasmodial, anti-leishmanial, anti-trypanosomal, anti-filarial, nematicidal and schistomicidal. To the best of our knowledge, all the articles that reported a plant-derived pentacyclic triterpenoid and nortriterpenoids tested against a parasite were included in this paper. Other articles that reported on synthetic modifications of the plant-derived PT were also included to enable full discussion on the structure-activity relationship. In cases where an article contains the name of a compound only, the structures used in this article were obtained from publication series by Hill and Connolly (2015). On the other hand, the names provided in the articles that used nuclear magnetic resonance spectroscopy (NMR) data in validating the structure of the compounds and synthetic compounds are used. Plant names and families were verified at http://www.theplantlist.org database. Overall, the information obtained covered the period; 1985 to the date of submission of this paper.

RESULTS AND DISCUSSION

A total of 112 naturally occurring PT and saponins isolated from 69 plants belonging to 35 families are reported in this paper. Ten of the total number of the compounds are nortriterpenoids of the quinone methide (possessing friedo-oleanane structure) type isolated mainly from five species of the Celastraceae family. Moreover, 62 of the total number of the compounds are the oleanane (including -friedelanes and -saponins), 19 ursane (-saponins), five taraxastane, 15 lupane (-saponins) and one hopane types of PT. These were isolated mostly from the Araliceae, Rubiaceae, Melastomataceae, Compositae and Lamiaceae plant families. Alongside these naturally occurring PT, 79 synthetic PT were also reported, of which 15 are oleanane types, nine are ursane types, one taraxastane type and 54 are lupane types. The structures of all the compounds are provided in Supplementary Fig. 1 (available from http://journals.cambridge.org/PAR).

The anti-parasitic activities of all the PTs were classified as high, moderate or low/no using the criteria suggested by Pink et al. (2005) and Bero et al. (2011) with modifications. Compounds with high potency (in vitro IC₅₀ ⩽ 10 µg mL⁻¹ against protozoa,), moderate potency (in vitro IC₅₀ 10–20 µg mL⁻¹ against protozoa) and low/no activities (IC₅₀ > 20 µg mL⁻¹ against protozoa) are summarized in Supplementary Tables S1, S2 and S3, respectively (available from http://journals.cambridge.org/PAR). Activity of the compounds against other parasites beside protozoa is classified based on the dosage and activity of the standard drug used in the respective studies. Some compounds were tested in in vivo assays which are also summarized in Table 1.

Table 1. In vivo antiparasitic activities of pentacyclic triterpenes

In order to provide a clear view on the anti-parasitic potential of the PT, logical discussions on the activities of the PT against various parasites are made under separate subheadings. The investigated parasites were found to be different species of Plasmodium, Leishmanium and Trypanosoma, as well as various nematodes and Toxoplasma gondii. Finally, some safety and toxicity profiles of the compounds are briefly discussed.

Brief chemistry of PTs

As shown in Fig. 1A–E, the PTs of the quinone methides, oleanane and ursane groups generally have five fused six-membered rings (designated a–e), while the lupane and hopane types have four six-membered and one five-membered rings. The distinguishing feature between the oleanane and ursane types is the localization of the methyl group on the ‘e’ ring, whereas the taraxasteranes differ in the orientation of substituents and the positions of double bonds in the backbone. Also, the lupane and the hopane skeletons differ on the localization of the isopropenyl group on the ‘e’ ring. In plants, all these groups of PT (except the nortriterpenoids quinone methides) commonly originate from cyclization of squalene and oxidosqualene via multiple enzymatic and redox stages involving formation of carbocations (Xu et al. 2004; Vincken et al. 2007). Moreover, in the PT possessing the oleanane and ursane backbone, the C4, C17 and C20 appear to show the highest diversity and unsaturation as well as formation of epoxides, whilst oxygen bridges are formed between the various carbon atoms (Vincken et al. 2007). On the other hand, the C3 and the substituent at C17 have been the primary target for synthetic modification. Finally, the saponins of the various PT are formed via attachment of diverse sugar subunits (ranging from 1 to 8 subunits) to the parent skeleton especially at C3 and C17 and rarely on C4, C16, C20, C21 and C22 (Vincken et al. 2007). Although the physico-chemical properties of saponins as well as the non-glycosylated PT are as diverse as the compounds themselves, the sugar moiety on saponins tend to make them more soluble than the corresponding aglycone (Güçlü-Üstündağ and Mazza, 2007).

Fig. 1. Representative skeletons of the different classes of pentacyclic triterpenes showing the carbon numbers and ring annotations. (A) Quinone methides, (B) oleananes, (C) ursanes, (D) lupanes, and (E) taraxastanes.

Anti-plasmodial activities of PTs

Perhaps the most in vitro active anti-plasmodial plant derived PT belong to the small group of quinone methides. Almost all the compounds belonging to the group isolated from different sources were highly active against both chloroquine sensitive and chloroquine resistant strains of Plasmodium falciparum. The compounds are pristimerin ( 1), isoiguesterol ( 2 ), celastrol ( 3 ), 28-hydroxyisoiguesterin ( 4 ), 17-(methoxycarbonyl)-28-nor-isoiguesterin ( 5 ), 28-nor-isoiguesterin-17-carbaldehyde ( 6 ) and Tingenin B ( 7 ) which all possessed very low IC₅₀ values (<0·5 µg mL⁻¹) against P. falciparum (Supplementary Table S1, available from http://journals.cambridge.org/PAR) (Figueiredo et al. 1998; Maregesi et al. 2010; Ruphin et al. 2013).

The oleanane PT also showed high to low activity against Plasmodia. Epi-Oleanolic acid (OA) ( 11) from Viola verecunda inhibited the growth of the chloroquine sensitive D10 strain of P. falciparum with a very low IC₅₀ of 0·018 µg mL⁻¹ which was close to that of artemisinin (0·015 µg mL⁻¹) (Moon et al. 2007). However, the same compound isolated from Celaenodendron mexicanum had moderate activity against multidrug resistant K1 strain of the parasite (IC₅₀ 12·92 µg mL⁻¹) (Camacho et al. 2000). Another oleanane PT with potent anti-P. falciparum activities is 1-O-[α-L-(rhamnopyranosyl)]-23-acetoimberbic acid 29-methyl ester ( 12 ) from Pittosporum mannii (IC₅₀ 1·2 µg mL⁻¹) (Nyongbela et al. 2013). Furthermore, OA ( 13 ) isolated from different plant species has been shown to possess anti-plasmodial activities with IC₅₀ ranging from 2·1 µg mL⁻¹ against chloroquine sensitive clone D6 (He et al. 2005) to 88 µg mL⁻¹ against multidrug resistant K1 strain of P. falciparum (Steele et al. 1999). Large variation in IC₅₀ values for the same compound often reflects the differences in the parasite strain or sometimes different experimental procedures. Moreover, the variations in the documented anti-plasmodial activities of OA might suggest that strain differences are critical for the anti-plasmodial effects of the oleananes.

Another moderately active anti-plasmodial oleanane PT and the most extensively investigated is maslinic acid (MA) ( 61 ). Incubation of different concentrations of the compound obtained from the fruits of Olea europaea with P. falciparum (at different growth stages) showed that the compound arrests the maturation of the intraerythrocytic parasites from early-ring to schizont stages. The IC₅₀ for the chloroquine sensitive and chloroquine resistant strains of the parasite were 15·13 and 12·29 µg mL⁻¹, respectively (Moneriz et al. 2011a ). The proposed mechanism of the anti-plasmodial activity of oleanane-type PT is via incorporation into the erythrocytes membrane thereby modifying accessibility of the parasites into the cells (Sairafianpour et al. 2003). Indeed, other studies have demonstrated that PT exert their anti-parasitic activities via an interaction with the host cell membranes (Ziegler et al. 2006; Broniatowski et al. 2012).

On the other hand, some oleanane PT such as β-amyrin ( 19 ), arjun glucoside ( 73 ), sericoside ( 74 ), and maytensifolin B ( 22 ) were shown to possess very low or no anti-plasmodial activity (Supplementary Table S3, available from http://journals.cambridge.org/PAR). However, it is also noteworthy that these low active anti-plasmodial oleanane PT totally lack an acid group and/or the C3 hydroxyl or these groups are derivatized/sterically hindered (Cunha et al. 2003). This signifies the role of the polar groups at C27 and C3 in the anti-plasmodial activity of this class of PT.

In the ursane group, ursolic acid (UA) ( 88 ) isolated from Baccharis dracunculifolia had the highest reported in vitro activity against chloroquine sensitive P. falciparum with IC₅₀ of 1 µg mL⁻¹ (da Silva Filho et al. 2009). Additionally, UA from the leaves of Mimusops caffra showed an IC₅₀ of 6·8 µg mL⁻¹ against chloroquine sensitive D10 strain of P. falciparum (Simelane et al. 2013). The activity was boosted by derivatization of the compound to 3β-O-acetylursolic acid ( 89 ) and 3-oxo-ursolic acid ( 99 ) with IC₅₀ of 1·9 and 7·3 µg mL⁻¹ respectively, using the same organism. However, other reports on the anti-plasmodial activity of UA contradict the above findings. For instance, Suksamrarn et al. (2003) and Graziose et al. (2012) reported UA to be inactive against multidrug resistant and chloroquine sensitive strains of P. falciparum respectively. It is pertinent to note that the authors used either different methods or compound dilutions in the anti-plasmodial assay protocol which highlights the need for harmonization of protocols from different laboratories for easier comparison. Other ursanes with potent anti-plasmodial activity are uvaol ( 92 ) and 2α-hydroxy-ursolic acid ( 90 ) isolated from Baccharis dracunculifolia with IC₅₀ of 1·9 and 3 µg mL⁻¹ respectively against a chloroquine resistant K1 strain as well as 3-acetylpomolic acid ( 101 ) (IC₅₀ 2·1 µg mL⁻¹) and pomolic acid (100) (IC₅₀ 3·47 µg mL⁻¹) both isolated from Markhamia tomentosa (da Silva Filho et al. 2009; Tantangmo et al. 2010). Hence, ursane-type PT also provide a promising class of anti-plasmodials for future research.

The lupane-type PT investigated for anti-plasmodial activity include betulinic acid (BA) ( 129 ) isolated from Harungana madagascariensis and Zizyphus vulgaris with IC₅₀ values of 2·33 and 6·3 µg mL⁻¹ against W2 and 3D7 strains of P. falciparum respectively (Lenta et al. 2007; de Sá et al. 2009). A structural analogue of BA isolated from Diospyros quaesita highlights the importance of derivatization of the compound at C3 (true also for OA and UA) for potentiation of anti-plasmodial activity. The 3-caffeate derivative of BA ( 122 ) isolated from the plant was active against both chloroquine sensitive D6 and chloroquine resistant W2 strains of P. falciparum with IC₅₀ of 0·86 and 0·61 µg mL⁻¹ respectively. This activity was enhanced with double acetylation of the caffeate ( 121 ) (IC₅₀ 0·45 and 0·42 µg mL⁻¹, respectively) (Ma et al. 2008). Similar results were obtained with other derivatives such as messagenic acid A ( 123 ) and messagenic acid B ( 124 ) (trans and cis C27 coumaroyl derivatives of BA, respectively) isolated from Gardenia saxatalis which possessed IC₅₀ of 1·5 and 3·8 µg mL⁻¹ respectively against a multidrug resistant strain of P. falciparum while the non-derivatized BA (also OA and UA) was inactive (Suksamrarn et al. 2003). Moreover, cis and trans C3 coumaroyl derivatives of BA ( 138 and 139 ) isolated from Cornus florida were also highly active derivatives against P. falciparum D10 with IC₅₀ of 6·03 and 9·22 µg mL⁻¹ respectively (Graziose et al. 2012). Other naturally occurring lupane-type PT active against the K1 strain of P. falciparum are betulone ( 126 ) and lupenone ( 127 ) with IC₅₀ of 1·32 and 2 µg mL⁻¹, respectively (Gachet et al. 2011). However, synthetic modifications of BA did not lead to profound increase in activity. Ziegler et al. (2004) reported the activities of methyl betulinate ( 125 ), betulinic aldehyde ( 131 ), betulinic acid amide ( 132 ), lupeol ( 128 ) and betulin ( 134 ) which were IC₅₀ of 3·3, 6·2, 6·4, 11·8, and <12 µg mL⁻¹ respectively against P. falciparum. The more interesting finding of the study, however, was that BA, 131 and 134 resulted in a dose-dependent structural change in the membrane of non-parasitized erythrocytes. The compounds consequently prevented entry of P. falciparum merozoites into non-parasitized erythrocytes. These findings demonstrated that lupane-type PT may also restrict parasites’ erythrocyte invasion in vitro via a mechanism that involves modulation of the erythrocytic membrane.

Few PT were investigated for possible in vivo anti-plasmodial activity based on their promising in vitro activities. The only oleanane-type PT that was investigated for in vivo anti-plasmodial activity was MA ( 61 ). Mice were infected with the lethal strain of Plasmodium yoelii and treated with a daily single intraperitoneal dose of 40 mg kg⁻¹ body weight (bw) MA. As found in the in vitro studies, MA demonstrated a static effect on the parasite with accumulated schizonts in the erythrocytes of the infected mice. However, the treated mice consistently maintained lower levels of parasitaemia and remained immunoprotected against further infection with the parasite after 40 days (Moneriz et al. 2011b ). Further analysis of the possible mechanism of action of MA suggested a multi routed mechanism involving the inhibition of a number of proteases necessary for the growth of the parasite. Other binding sites for MA, which include the Plasmodium phospholipase, were putatively proposed in an in silico analysis (Moneriz et al. 2011c ). This remarkable in vivo activity demonstrated by MA calls for similar investigation on other oleanane-type PT especially those with even lower in vitro IC₅₀ than MA such as epi-OA ( 11 ).

On the other hand, 3β-O-acetylursolic acid ( 89 ) was shown to suppress 94·01% of circulating Plasmodium berghei in mice (effective concentration not clear in the report). The compound was also less cytotoxic against HEK293 and HepG2 cell lines (Simelane et al. 2013). Furthermore, taraxasterol acetate ( 115 ) isolated from Pluchea lanceolata at 10 mg kg⁻¹ bw suppressed 52·20% of circulating P. berghei in mice and showed 7 days extension of mean survival time (Mohanty et al. 2013). In another study, in vivo evaluation of betulinic acid revealed that the compound was ineffective in reducing P. berghei in mice even at 250 mg kg⁻¹ bw day⁻¹ (Steele et al. 1999).

One of the greatest limitation on the in vivo activity of PT, especially the less polar among them, is the hydrophobicity. Moreover, another limitation is the high cytotoxicity of some classes. For instance, Pristimerin ( 1 ) isolated from Salacia leptoclada was shown to have a selective index of <1 for P338 leukaemia cell lines (Ruphin et al. 2013), whereas 17-(methoxycarbonyl)-28-nor-isoiguesterin ( 5 ) at 10 mg kg⁻¹ bw was toxic to mice after just one day of administration (Figueiredo et al. 1998). The latter compound which was isolated from S. kraussii although being the most active of all PT against chloroquine-resistant P. falciparum in vitro (IC₅₀ 0·037 µg mL⁻¹), was unable to clear P. berghei in mice treated with 1 and 5 mg kg⁻¹ bw (Figueiredo et al. 1998). Therefore, bioavailability and cytotoxicity should be taken into account when further developing PT as possible anti-plasmodial agents is considered especially if the oral route is to be used. Moreover, these compounds could at least serve as structural backbones for synthesis of less toxic and more efficient compounds.

Anti-trypanosomal activities of PTs

Tingenin B ( 7 ), a quinone methide, is the most active reported PT against Trypanosoma brucei brucei and Trypanosoma cruzi with IC₅₀ < 0·25 µg mL⁻¹ against each of the species. However, as observed with other compounds belonging to the same class, the compound was highly cytotoxic on MCR-5 cells (IC₅₀ 0·45 µg mL⁻¹) (Maregesi et al. 2010). On the other hand, UA ( 88 ) has been reported in many studies to possess anti-trypanosomal activity with low IC₅₀. The compound isolated from Strachynos spynosa possessed an IC₅₀ of 1 µg mL⁻¹ against T. brucei brucei (Hoet et al. 2007). Furthermore, in a study by Abe et al. (2002), UA from Rosmanirus officinalis with an MC₁₀₀ of 40 µg mL⁻¹ was shown to be 86% more effective than the natural trypanocidal compound gossypol (Abe et al. 2002). Other structural analogues of UA were less effective or inactive against trypanosomes. The carboxylic group at C17 appears to be important for the anti-trypanosomal action of UA as evident in lower activities of uvaol ( 92 ) (aldehyde group replacing carboxyl) with an IC₅₀ of 12·3 µg mL⁻¹ and α-amyrin ( 95 ) (methyl group replacing carboxyl) with IC₅₀ of 48 µg mL⁻¹. Likewise, β-amyrin ( 19 ) (IC₅₀ 54·2 µg mL) with CH₃ in place of COOH at C17 of OA (IC₅₀ 2·9 µg mL⁻¹) lost anti-trypanosomal activity against T. brucei brucei (Hoet et al. 2007). The OH at C3 also appears to be equally important in the trypanosomal action of both UA and OA (Cunha et al. 2003; Taketa et al. 2004). This is confirmed by the loss in the activity against T. cruzi of UA in a mixture with OA with addition of acetyl group at C3 ( 14 and 89 , respectively) of both compounds (Cunha et al. 2003). Moreover, oleanonic acid ( 66 ) and 3,11-dioxoolean-12-en-28-onic acid ( 67 ) (IC₅₀ 113·62 and 173·9 µg mL⁻¹ against T. cruzi, respectively) which are similar in structure with OA but with adulterated C3 possessed no activity against the parasite (Cunha et al. 2003; Hoet et al. 2007; Leite et al. 2008). However, replacement of the C3 OH of OA with a polar group in saponin ( 18 ) did not lead to a loss in activity (Taketa et al. 2004). From these findings, it is evident that the presence (and/or property) of the C3 hydroxyl group and C17 COOH group are significant for the trypanocidal activity of the ursane and oleanane-type PT. The role of C3 OH may be, in part, to increase the polarity of the compound because glycosylation of the group in OA with a disaccharide (3-O-[β-D-glucopyranosyl-(1–2)-β-D-galactopyranosyl]) ( 18 ) or addition of potassium in UA ( 93 ) were found to maintain the activity against T. brucei brucei (IC₅₀ 3·05 µg mL⁻¹) and T. cruzi (IC₅₀ 4·26 µg mL⁻¹), respectively (Taketa et al. 2004; Cunha et al. 2006). However, substitution of the neighbouring carbon (C4) to C3 appears to counteract the effect despite the presence of additional polar groups as seen in the loss of activity of both brevicuspisaponin 1 and 2 ( 102 and 103 ) against T. brucei brucei where UA was most potent (Taketa et al. 2004). This suggests that in addition to increasing the polarity of the compounds, the nature and orientation of substituents at positions C3 and C17 may be involved directly in the activities of PT. Furthermore, the double bond between C12 and C13 may also play a role in the activity because friedelanol ( 85 ) lacking any double bond was inactive despite the presence of a C3 OH (da Silva Filho et al. 2004).

The anti-trypanosomal activity of the lupane group appeared in very few reports. Hoet et al. (2007) reported anti-T. brucei brucei activity of betulin ( 134 ), betulinic acid ( 129 ) and lupeol ( 128 ) with IC₅₀ values of 4·0, 14·9 and 19·3 µg mL⁻¹, respectively.

In an in vivo setup, UA, OA and the potassium salt of UA ( 93 ) were potent against the lethal Y strain of T. cruzi in mice treated with daily intraperitoneal dose of 2 mg kg⁻¹ bw. The treatment led to a reduction of parasite load in the infected rats more markedly by UA and the salt (75·7 and 70·4%, respectively) (Supplementary Table S1, available from http://journals.cambridge.org/PAR) (Cunha et al. 2006). In another study, da Silva Ferreira et al. (2010) reported 60 and 40% reduction in Bolivia strain of T. cruzi after treatment of infected rats with UA and OA at doses of 20 mg kg⁻¹ bw day⁻¹ orally. Findings of a later study demonstrated that the effectiveness of UA and OA treatment in T. cruzi infected mice is dependent on the bioavailability of the compound. It was observed that oral administration of the compounds (50 mg kg⁻¹ bw day⁻¹) resulted in 79 and 76% decrease in parasitaemia respectively, while administration of the same concentration via the intraperitoneal route was not effective. Presumably, the intraperitoneal route achieved higher effective concentration of the compounds, which could have modulatory effects on pro and anti-inflammatory cytokines that resulted in an observed immunosuppression (da Silva Ferreira et al. 2013a ). These effects may hence essentially counter the destructive effects of the compounds on the parasites since the immune system at some point of T. cruzi infection participate in parasite clearance (Tarleton, 2007). Hence, at low concentrations (which is achieved via the oral route due to low oral bioavailability of UA and OA or low intraperitoneal dose), the compounds are sufficient to destroy the parasites on their capacity or via other mechanisms.

On a final note, the anti-trypanosomal potential of PT is equally promising. Further research in this area should be directed towards screening more PT (especially the quinone methides) against various species of Trypanosoma. The need to investigate the compounds in animal models is also paramount because the compounds appear to facilitate parasite clearance via stimulation of host mechanisms which cannot be attained in vitro. In this regard, alternating the routes of administration is critical in order to provide a conclusive profile on the full potencies of PT as anti-trypanosomal agents.

Anti-leishmanial activities of PTs

A number of studies have been conducted on the activity of PT against promastigotes and amastigotes of various Leishmania species. A range of saponin glycosides belonging to the oleanane PT isolated from Maesa balansae were very active against Leishmania infantum amastigotes with very low IC₅₀. The most active among them designated maesabalide III ( 25 ) possessed an IC₅₀ of 0·007 µg mL⁻¹. Other maesabalides ( 23, 24, 26, 27 and 28 ) gave IC₅₀ values of 0·014–0·046 µg mL⁻¹ (Germonprez et al. 2005). Oleonolic acid ( 13 ) isolated from Salvia cilicica also possessed activity with IC₅₀ of 0·04 and 0·029 µg mL⁻¹ against promastigotes and amastigote of Leishmania donovani, respectively (Tan et al. 2002). Here also, the C3 OH of OA appears to play a crucial role in the activity as a conformational change tends to decrease the anti-leishmanial activity of the compound. This is because the activity of epi-OA ( 11 ) isolated from Celaendendron maxicanum was hundred-fold lower against the same parasite (IC₅₀ 8·59 µg mL⁻¹) (Camacho et al. 2000). However, acetylation of the C3 OH group of OA may not cause a greater loss in activity. This is evident with acetylation of OA to form 3-OA acetate ( 14 ) which possessed an IC₅₀ 2·49 µg mL⁻¹ against Leishmania amazonensis (Gnoatto et al. 2008). Other oleanane triterpenes with potent activities include hederacolchiside A₁ ( 32 ) (IC₅₀ 0·061 µg mL⁻¹), β-hederin ( 16 ) (IC₅₀ 0·26 µg mL⁻¹) and α-hederin ( 15 ) (IC₅₀ 0·3 µg mL) from two hedera species against amastigotes of Leishmania mexicana (Ridoux et al. 2001; Tantangmo et al. 2010). On the other hand, glycyrrhitinic acid (GRA) ( 33 ), a derivative of β-amyrin ( 19 ) was also potent against L. donovani promastigotes in vitro with an IC₅₀ of 4·6 µg mL⁻¹ (Ukil et al. 2005).

Among the ursanes, UA ( 88 ) isolated from Salvia cilicica appears to be the most active against both promastigotes and amastigotes forms of L. donovani and Leishmania major with low IC₅₀ values of 0·0032–0·042 µg mL⁻¹ (Tan et al. 2002). However, other studies with UA reported much higher IC₅₀ of 2·28 µg mL⁻¹ against L. amazonensis (Torres-Santos et al. 2004), 3·7 µg mL⁻¹ against L. donovani (da Silva Filho et al. 2009) and 4·55 µg mL⁻¹ against Leishmania tarentolae (Graziose et al. 2012). Some structural modification of UA led to reduction in activity as reported for 2α-hydroxy-ursolic acid ( 90 ) and uvaol ( 92 ) (IC₅₀ 19 and 15 µg mL⁻¹ respectively against L. donovani) (da Silva Filho et al. 2009). On the other hand, a bis-(3-aminopropyl) piperazine moiety added to the carboxylic acid of 3β-acetylursolic acid ( 89 ) in compounds 106–108 retained the activity of UA against promastigotes of Leishmania infantum and L. amazonensis (IC₅₀ 6–17 µg mL⁻¹) (Gnoatto et al. 2008). From the above findings, UA appears to be a potent anti-leishmanial agent against multiple species of the parasites. Because the investigated structural modification did not lead to an increase in activity, further modifications of the parent UA may be an experimental strategy for further development of ursane-type PT as anti-leishmanial agents. Other ursane-type PT with promising in vitro anti-leishmanial activity include pomolic acid ( 100 ) and 3-acetyl pomolic acid ( 101 ) from Markhamia tomentosa (IC₅₀ 0·31 µg mL⁻¹ and 3·4 µg mL⁻¹, respectively) against L. donovani and synthetic N-{3-[4-(3-Aminopropyl)piperazinyl]propyl}−3-O-acetylursolamide ( 105 ) (IC₅₀ 3·7 µg mL⁻¹ against L. infantum) (Gnoatto et al. 2008; Tantangmo et al. 2010).

In the lupane group, a few derivatives of BA were active against Leishmania although the parent compound was inactive in multiple studies. Betulinic acid acetate ( 130 ) and trans and cis 3-coumarol derivatives of BA ( 138 and 139 ) isolated from Cornus florida had IC₅₀ values of 0·45, 5·14 and 1·36 µg mL⁻¹ respectively against L. tarentolae (Graziose et al. 2012). Moreover, dihydrobetulinic acid (DHBA) ( 143 ) from Bacopa monniera possessed an IC₅₀ of 2·6 and 4·1 µg mL⁻¹ against L. amazonensis promastigotes and amastigotes, respectively (Chowdhury et al. 2003). Although a number of structural modification of the lupane-type PT led to loss in anti-leishmanial activity (Supplementary Table S3, available from http://journals.cambridge.org/PAR), future research on the group may be targeted towards different synthetic classes of the compounds and species of the parasite.

In an in vivo study, the anti-leishmanial activity of GRA ( 33 ) was further assessed where rats were treated with 50 mg kg⁻¹ bw day⁻¹ (given three times, 5 days interval for 45 days) of the compound. The compound cleared the amastigotes form of the parasite from the liver and spleen of infected animals with a mechanism that involves decrease in the expression of mRNA for anti-inflammatory cytokines [interleukin (IL)-10 and IL-4] and an increase in the level of interferon-γ (IFN-γ) and tumor necrosis factors alpha (TNF-α) (Ukil et al. 2005). This comprehensively resulted in an increased immune response to the infection and clearance of the parasite via an nuclear factor kappa-B (NF-κB)-mediated mechanism. The mechanism through which GRA upregulate NF-κB was further described to involve multiple kinases and phosphatases (Ukil et al. 2011). Indeed, stimulation of the immune system has been deemed a rational strategy for the development of anti-leishmanial drugs (Santos et al. 2008). In a different study, oral and intraperitoneal administration of 10 mg kg⁻¹ bw DHBA to infected golden hamsters caused >90% reduction in parasite load in the spleen and liver of the infected animals. The compound was proposed to exert its effect via a mechanism that involves inhibition of DNA topoisomerases thereby essentially destroying the parasites (Chowdhury et al. 2003, 2011). As observed with other PT, the chemical entities on the C3 and C28 position of the lupeol-type PT is critical for the anti-parasitic activity of the group. Further analysis of the existing members of the group and structural manipulations to enhance activity is recommended.

Anti-nematodal activities of PTs

PTs were also investigated in a number of studies as possible therapeutic agents against lymphatic filariasis, onchocerciasis as well as conditions caused by other parasitic nematodes. Antifilarial activities of oleanane PT were reported against both human lymphatic filaria Brugia malayi and the rodent infective species. Oleanonic acid ( 66 ) and OA ( 13 ) isolated from the stem of Lantana camara were active against B. malayi in vitro with an LC₁₀₀ of 31·25 and 62·50 µg mL⁻¹ respectively (Misra et al. 2007). Glycyrrhetinic acid ( 33 ) and its analogs ( 34–37 ) were also shown to be effective against the adult and microfilarial forms of B. malayi. The acyl derivatives ( 38 and 39 ) were inactive against both growth stages of the parasite, while the others showed IC₅₀ values in the of 0·56–28·63 µg mL⁻¹ range against the microfilariae. However, against the adult worms, only the benzyl amide ( 34 ) and octyl amide ( 35 ) derivatives were active (IC₅₀ 5·95 and 12·04 µg mL⁻¹ respectively) (Kalani et al. 2013).

In animal studies, the OA and oleanonic acid each administered at oral and intraperitoneal doses of 200 and 100 mg kg⁻¹ bw respectively had no effect on the circulating microfilariae of B. malayi in infected mastomys. However, against the adult worms, both compounds showed approximately 56% female worm sterility, although only OA had a filaricidal activity of 18·18% (Misra et al. 2007). Moreover, B. malayi infected jirds were treated with 100 mg kg⁻¹ bw doses of the in vitro active amide derivatives of GRA ( 34 and 35 ). The result showed that only the benzyl amide derivative possessed macrofilaricidal activity (54%) while the other was inactive (Kalani et al. 2013).

The oleanane-type PT, 3-O-acetyl aleuritolic acid ( 75 ) isolated from Discoglypremna caloneura was active against Onchorcerca gutturosa worms. The compound was found to reduce the motility and viability of the worms up to 57·1 and 64·8%, respectively. The reduction in viability was found to be 33·3% more than that of amocarzine and hence compound 75 was considered an interesting compound against filarial infections (Nyasse et al. 2006).

Although the volume of research on the anti-filarial activities of PT is not large, available data suggest them to be potent against different filariid. Hence, future screening of other PT against filariasis may be worthwhile.

Betulin ( 134 ) from Schefflera vinosa as well as OA and UA from Miconia langsdorfii were tested for schistomicidal activities. Among the three compounds, only 134 led to the mortality of the adult worms of Schistosoma mansonii at concentrations of 100 µ m (25% mortality) and 200 µ m (50%) after 120 h of incubation (Cunha et al. 2012). Further research on this subject area should focus on testing newly isolated and available PT on different species of Schistosoma to compliment the library of biological activities of the group as future anti-parasitic agents.

OA isolated from Calendula officinalis was investigated for possible nematocidal activity against the mice intestinal parasite, Heligmosomoides polygyrus. The compound alongside other derivatives exhibited >50% growth inhibition of the larvae incubated with 70 µg mL⁻¹ of the compounds in vitro. The mechanism through which OA and related PT reduces the viability of H. polygyrus was later shown to involve modulation of the pattern of larval antigen glycosylation which appears to lead to a robust increase in cytokine production in mice infected with larvae incubated with the compound (Doligalska et al. 2013). Because anti-filarials act via an immune-mediated mechanisms (Hoerauf et al. 2011), and PTs were shown to modulate the immune system, PTs are logical candidates for in vivo screening as anti-filarial drugs.

MA was also investigated for possible action against the Trichinella, the causative agent of trichinellosis in humans. Against the mammalian infective Trichinella zimbabwensis, the compound orally administered once on 25 dpi or twice on 25 and 32 dpi cleared >90% of the parasite's larvae. This was achieved at a lower dose (2·5 mg kg⁻¹ bw) compared with the anthelmic drug fenbendazole (7·5 mg kg⁻¹ bw) which gave similar efficacy (Mukaratirwa et al. 2016). Hence, MA has shown promising activity against Trichinella and therefore screening of other PT against this parasite will be worthwhile.

Against the plant nematode Meloidogyne incognita, camarinic acid ( 110 ) activity was similar to that of a standard nematicidal drug, furadan, at the same concentration of 1 mg mL⁻¹. The compound which was isolated from Lantana camara led to 100% larval mortality after 24 h exposure (Supplementary Table S1, available from http://journals.cambridge.org/PAR) (Begum et al. 2000). Later studies on this plant showed it to be a repository of PT with varying degrees of nematicidal activities. Lantanilic acid ( 46 ), camaric acid ( 45 ) and OA ( 13 ) from the plant caused 98, 95 and 70% M. incognita larval mortality respectively at 5 mg mL⁻¹ concentration (Qamar et al. 2005). Furthermore, camarinin ( 43 ), lantanolic acid ( 44 ), UA, pomolic acid ( 100 ), lantacin ( 114 ), camarin ( 77 ) and lantoic acid ( 111 ) from the same plant all caused 100% larval mortality at 1 mg mL⁻¹ concentration after 48 h of exposure. Compounds 43, 44 and UA ( 88 ) proved to be comparatively more potent with 90, 10 and 10% larval mortality at 2 µg mL⁻¹ after 72 h exposure (Begum et al. 2008). In a different study with Cordia latifolia, cordinoic acid ( 112 ) isolated from the plant at 5 mg mL⁻¹ concentration led to 100% M. incognita larval mortality after 24 h exposure (Begum et al. 2011). On the other hand, polygalacic acid ( 48 ) and bayogenin ( 49 ) and their saponins 50–60 isolated from Microsechium helleri and Sicyos bulbosus were active against Meloidogyne javanica that also affects plants. Among the compounds, those with a xylose residue attached to the second rhamnose residue at the substituent on C28 ( 50–53 ) were found to be inactive while the others inhibited >74% of the parasite's larvae growth at various concentrations. Moreover, bayogenin which differs from polygalacic acid only in the absence of an OH group at C16 of the latter molecule was the most active together with saponin 58 . Both compounds immobilized 100% of the parasite's larvae at 0·5 µg mL⁻¹ concentration (Hernández-Carlos et al. 2011). From the above findings, it is clear that the activities of PT and their saponin against plant nematodes are promising and warrant further investigation.

Activities of PTs against other parasites

Toxoplasma: Maslinic acid ( 61 ) isolated from Olea europaea inhibited the infectivity of Vero cells by T. gondii tachyzoites with an ID₅₀ of 3·78 µg mL⁻¹ after incubation for 48 h. Moreover, the compound at a concentration of 50 µ m inhibited the motility of 100% of the parasites. The compound was also shown to inhibit key parasite proteases thereby effectively blocking parasite entry into the cells (De Pablos et al. 2010). This dual effect (inhibition of motility and entrance into cells) of MA on T. gondii is interesting as therapeutic approach and hence calls for further screening alongside other PT.

Trichomonas: Only one PT, hederagenin ( 47 ), isolated from Cussonia holstii was investigated for activity against Trichomonas viginalis. The result indicated high in vitro activity with an IC₅₀ of 1·32 µg mL⁻¹ (He et al. 2003). Hence, PT could be suitable candidates for future screening as anti-trichomonas agents.

Toxicity aspects

One of the disadvantages of using PT as therapeutic agents has been known to be associated with high cytotoxicity (Dzubak et al. 2006). However, at low concentrations, some of these PT have proved to be therapeutic (Liu, 2005). Moreover, cytotoxicity studies of some PT, for example MA, reported in vivo safety both in acute and chronic treatments (Sánchez-González et al. 2013). In another study, BA was found to have selective toxicity against cancerous cells but not normal cells (Zuco et al. 2002). Hence, since PT are selective to different cells lines, further toxicity assessments and in vivo safety studies of the most active compounds is warranted.

CONCLUSION AND FUTURE DIRECTIONS

Various research findings from plants of different parts of the world have revealed that PT represent a promising group of phytochemicals with good therapeutic potential against a number of parasitic diseases. However, the studies on the anti-parasitic potential of PT are at preliminary proof of concept stages with only 22 out of the total 191 PT having been investigated in animal models. This underscores the need to re-focus research efforts on in vivo studies of PT against different parasitic infections which may pave the way for further clinical trials and drug development.

On a general note, it is noteworthy that the PT seems to be more promising for future development as anti-malarial agents. This is evident by the propensity of anti-plasmodial studies of PT as well as the potent activities reported for most of the tested PT. However, this does not exclude the possibility of developing therapeutically active PT against other parasites, especially the less studied parasites such as toxoplasma, trichomonas, schistosoma and nematodes.

Another pertinent finding from this review is that quinine methides are the most biologically potent PT with respect to parasitic diseases especially those caused by malaria parasites. Unfortunately however, this class of the compounds also seems to be the most toxic among all the PTs. Thus, studies on quinine methides to target synthetic modifications at various positions of the parent backbone with the aim of minimizing their cytotoxicity, whilst maintaining the anti-parasitic activities should be conducted. In fact, this should be the next step to be taken if research efforts on quinine methides are to be geared along the drug development process.