Trypanocidal Activity of Marine Natural Products

Marine natural products are a diverse, unique collection of compounds with immense therapeutic potential. This has resulted in these molecules being evaluated for a number of different disease indications including the neglected protozoan diseases, human African trypanosomiasis and Chagas disease, for which very few drugs are currently available. This article will review the marine natural products for which activity against the kinetoplastid parasites; Trypanosoma brucei brucei, T.b. rhodesiense and T. cruzi has been reported. As it is important to know the selectivity of a compound when evaluating its trypanocidal activity, this article will only cover molecules which have simultaneously been tested for cytotoxicity against a mammalian cell line. Compounds have been grouped according to their chemical structure and representative examples from each class were selected for detailed discussion.


Introduction
The trypanosomatid diseases human African trypanosomiasis (HAT) and Chagas disease account for over 19,000 deaths and the loss of over 100,000 disability adjusted life years (DALYs) annually [1,2]. The etiological agents of the disease are kinetoplastid parasites of the genus Trypanosoma. Trypanosoma brucei gambiense and Trypanosoma brucei rhodesiense are responsible for HAT, while infection with Trypanosoma cruzi is the causative agent of Chagas disease. Both OPEN ACCESS diseases rely on insect vectors for their transmission; tsetse flies (Glossina spp.) are the vectors for HAT, whereas a number of Triatoma bug species transmit T. cruzi [3,4]. HAT is prevalent throughout 36 sub-Saharan African countries whilst Chagas disease primarily occurs in Southern parts of North America, and South America [5,6].
Initially, inoculation of the parasites into human hosts results in acute disease. In HAT, this is characterized by the presence of the parasites in the vasculature and lymphatic systems. Patients experience fever, nausea, headaches and lymphedema [7]. Without treatment the parasites penetrate the blood brain barrier (BBB) and invade the central nervous system (CNS) initiating chronic or CNS stage disease. CNS stage disease manifests as mental disturbances, anxiety, hallucinations and a characteristic disruption of the sleep-wake cycle [7][8][9][10]. Without treatment the disease is considered fatal [11].
In contrast to HAT, acute Chagas disease is often asymptomatic and as such is not often diagnosed [12]. Approximately one third of infected individuals go on to develop the chronic form of the disease which can remain asymptomatic for 10 to 30 years [12]. The chronic stage can manifest as cardiac or cardiodigestive disorders (megacolon, megaeosphagus), or a combination of these [13]. Chagas related heart disease is one of the major causes of morbidity and mortality in endemic areas [14].
Despite the morbidity and mortality inflicted by HAT and Chagas disease, very few effective drugs are currently available (Figure 1). Acute T.b. gambiense and T.b. rhodesiense infections are treated with pentamidine and suramin, respectively [15]. CNS T.b. rhodesiense infections are treated with melarsoprol, while T.b. gambiense infections are treated with either eflornithine or a nifurtimox/eflornithine combination therapy (NECT) [15]. However, none of these treatments are ideal. Melarsoprol is extremely toxic, resulting in the death of 5% of all patients to whom the drug is administered, and eflornithine has a complicated, protracted administration schedule requiring 56 slow intravenous (i.v.) infusions over 14 days [16,17]. The development of NECT reduced the administration schedule of eflornithine to 14 i.v. infusions over seven days, plus oral nifurtimox every eight hours for 10 days [18,19]. However, NECT is not ideal as parenteral administration is still required and patients must be hospitalized for the duration of treatment. Acute and chronic Chagas diseases are treated with either nifurtimox or benznidazole. Both drugs have lengthy administration schedules requiring bi-or tri-daily administration for 60 to 90 days [20]. Patients frequently experience vomiting, nausea, hepatic intolerance, convulsions and skin disease manifestations [21]. The unpleasant side effects experienced by patients, coupled with administration schedules, result in many patients failing to complete the treatment regimes [22,23]. The paucity of safe, effective and easily administrable drugs for HAT and Chagas disease is partly due to a lack of interest by large pharmaceutical companies. HAT and Chagas disease primarily affect poor, disadvantaged people, with limited access to health care and very little means to pay for drugs. Consequently, there is little incentive for pharmaceutical companies to invest in the research and development of new compounds for these disease indications. It has only been in the last decade, with the establishment of non-for-profit organizations such as the Drugs for Neglected Diseases initiative (DNDi) and the Bill and Melinda Gates Foundation, that substantial investment and progress has been made in drug discovery for HAT and Chagas disease. As a result, one compound, fexinidazole, is now in phase II/III clinical trials for HAT, while a second compound, SCYX-7158, is in phase I clinical trials [24,25]. In addition, during the past five years numerous drug targets have been identified and validated in T.b. brucei which are discussed in detail in a recent review [26]. Promising targets described include, the enzymes S-adenosylmethionine decarboxylase (AdoMetDC) [27,28], N-myristoyltransferase (NMT) [29,30] and trypanothione synthetase-amidase (TrySyn) [31]. For Chagas disease, K777 is currently in pre-clinical trials [32], whilst clinical trials with posaconazole are due for completion in 2013 [33]. Target identification studies have indicated that cysteine protease is the target of K777, thus validating further development of this class of inhibitors. Posaconazole inhibits T. cruzi sterol 14α-demethylase (CYP51) [34], and research continues to identify further inhibitors of this specific target [35][36][37]. Azole antifungals with CYP51 activity have previously entered clinical trials, however, have not demonstrated curative activity [38]. Few validated targets have been identified against T. cruzi and studies to determine new targets will be of benefit for Chagas disease research. Cloning of recombinant proteins based on the identified genome sequence could facilitate this process [39]. The mitochondria and mitochondrial metabolism [40] have been identified as potential sources of new targets for T. cruzi drug discovery research, as well as enzymes involved in pentose phosphate and thymidine synthesis [41].
Non-for profit organizations have highlighted the plight of HAT and Chagas disease patients and have provided the financial resources required for new therapeutics to be identified and developed. However, numerous problems still exist which impede drug development for HAT and Chagas disease. A large proportion of the molecules identified by phenotypic high-throughput screening (HTS) campaigns have undesirable chemical properties and biological characteristics, which makes them unsuitable for further development. Structure activity relationship (SAR) studies are frequently undertaken in order to improve a molecule's physiochemical properties, but this often results in a significant loss of trypanocidal activity. In the last five years, multiple drug targets have been identified in T. brucei spp. and T. cruzi. However, the targets are often inaccessible and it is difficult to develop small molecule inhibitors, which are capable of reaching and interacting with the target. Target-based screening can be utilized to identify potent inhibitors of targets but often the molecules lack trypanocidal activity when subsequently screened against the whole parasite, as they are unable to penetrate the parasites and reach the intracellular target.
The high attrition rate associated with drug discovery and development and the difficulties encountered, means that there still exists a critical need to identify novel compounds for HAT and Chagas disease. Natural products including, marine organisms and metabolites, are one potential source from which unique trypanocidal compounds could be identified.
Natural products are attractive chemical starting points for drug discovery. They have been investigated for a number of different disease indications and biological targets resulting in the identification of both lead molecules and drugs suitable for entry into the drug discovery pipeline. Between 1981 and 2010 natural products and synthetic small molecules either derived from a natural product or based on a natural product, pharmacophore, accounted for over 50% of new chemical entities [42]. Research into the chemistry, pharmacology and therapeutic potential of marine natural products began with the development of self-contained breathing apparatus (SCUBA) in the 1960s and has continued to progress and develop with thousands of compounds now identified [43]. The first marine natural product to be registered by the United States (US) Food and Drugs Administration (FDA) was cytarabine (1β-arabinofuranosylcytosine), a chemotherapeutic agent, in 1969. Since then six other marine natural product based drugs have been approved by the FDA; vidarabine (anti-cancer and anti-viral), ziconotide (an analgesic agent), eribulin mesylate (anti-cancer), brentuximab vedotin (for the treatment of Hodgkin's lymphoma and large cell lymphoma) and the omega-3-ethyl ester preparations, lovaza and vascepa (triglyceride lowering agents). In addition, one further compound, trabectedin (anti-cancer), has been approved by the European Medicines Agency (EMA).
Cytarabine (1β-arabinofuranosylcytosine) and vidarabine (adenine arabinoside) ( Figure 2) are synthetic pyrimidine and purine nucleosides, respectively, developed from nucleosides isolated from the Caribbean sponge Tethya crypta [44,45]. Cytarabine is used for the treatment of acute myeloid and lymphocytic leukemia, while vidarabine was approved in 1976 for the treatment of acute keratoconjunctivitis and recurrent epithelial keratitis caused by Herpes simplex viruses [46][47][48]. The therapeutic effects of cytarabine and vidarabine are thought to arise due to inhibition of DNA polymerase and DNA synthesis [49,50]. Twenty-eight years after the registration of vidarabine, ziconotide, a synthetic equivalent of a peptide originally isolated from the venom of the cone snail Conus magus, was approved by the FDA [51]. The drug is a powerful analgesic due to its ability to selectively and specifically block N-type voltage sensitive calcium channels and is used to manage chronic pain in cancer and AIDS patients [52]. Also in 2004, lovaza, the first drug containing the fish derived omega-3-ethyl fatty acids, eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) was approved for the reduction of triglyceride levels in severe hypertriglyceridemia [53]. This was followed by the registration of vascepa, containing only EPA, in 2012 [54]. Omega-3-ethyl fatty acids are found in all fish species but are most abundant in oily fish, such as salmon, mackerel and herring [55]. The mechanism of action (MOA) for the hypotriglyceridemic effect of omega-3-ethyl fatty acids is not fully understood but has been attributed to the suppression of hepatic lipogenesis, an increase in fatty β-oxidation and down regulation of hepatic nuclear factor-4α (HNF-4α) [56][57][58][59][60]. In 2010, eribulin mesylate, a synthetic macro-cyclic ketone analogue of halichondrin B, a molecule isolated from the marine sponge Halichondria okadai, received FDA approval for the treatment of metastatic breast cancer [61]. Eribulin induces cell death by inhibiting microtubule growth and sequestering tubulin into nonproductive aggregates [62][63][64]. Brentuximab vedotin, a CD30 specific antibody-drug conjugate received FDA approval for the treatment of Hodgkin's lymphoma in 2011. Brentuximab vedotin is composed of monomethylauristatin E (MMAE), a synthetic analogue of the marine natural product dolastatin 10 conjugated with the chimeric anti-CD30 monoclonal antibody, SGN-30 [65]. Dolastatin 10 was originally isolated from the Indian Ocean sea hare Dolabella auricularia in 1987 [66]. MMAE is an anti-tubulin agent which binds to tubulin and prevents microtubule polymerization leading to G 2 -M phase growth arrest and apoptosis [67]. Trabectedin (ecteinascidin) ( Figure 2) has been approved in the European Union (EU) by the EMA. The compound was isolated from the ascidian Ecteinascidia turbinata and is an anti-cancer agent used in the treatment of soft tissue sarcoma and platinum-sensitive ovarian cancer [68]. The MOA of trabectedin is not fully elucidated, however, the compound has been shown to bind to the minor groove of DNA and interact with different binding proteins of the Nucleotide Excision Repair System (NERS) [69][70][71][72]. In addition to the marine natural products which have received regulatory approval and progressed to the market, numerous molecules are currently in clinical development [73].
To date, no marine natural products or derivatives have entered pre-clinical development specifically for trypanosomatid diseases. However, numerous marine natural products which exhibit anti-trypanosomal activity have been reported in the literature.
In this article, the natural products isolated from marine sources for which activity against the protozoan parasites; T.b. brucei, T.b. rhodesiense and T. cruzi has been reported, is reviewed. The majority of the compounds have been identified through phenotypic screening campaigns, which have recently been reviewed in detail [74,75]. It should be noted that although T.b. brucei primarily infects domestic mammals and antelopes and is not the human infective subspecies responsible for HAT, it is frequently used in early drug discovery screening campaigns to identify active compounds [76,77]. Compounds active against T.b. brucei would ultimately be evaluated against the human infective forms of the parasite, T.b. rhodesiense and T.b. gambiense. The bloodstream form of T. brucei spp. is used in phenotypic screening assays, as this is the clinically relevant form of the parasite ( Figure 3A). In T. cruzi infection, the amastigote and the trypomastigote life cycle stages are both found within the human host ( Figure 3B). All lifecycle stages of T. cruzi can be used in assays to evaluate the activity of compounds. However, activity against the amastigote form of the parasite has been deemed to be of primary importance in many assays, with activity against the trypomastigote stage also considered favorable or necessary [78][79][80]. Herein, only activity against the human infective forms, namely amastigotes and trypomastigotes are considered. Many assay formats used in T. cruzi research are based on the method by Buckner et al., whereby compounds are added two hours after addition of T. cruzi β-galactosidase transfected trypomastigotes to host cells [81]. Cells are incubated for seven days before detection of released trypomastigotes via lysis of cells and detection of β-galactosidase activity. This assay may affect both host cell infection and/ or development of amastigotes. The T. cruzi assays discussed in this article are based on this assay format, unless a modification is discussed. In drug discovery screening campaigns for HAT and Chagas disease, a compound is only classed as a "hit", if it has an IC 50 < 10 µM [84][85][86][87]. Compounds with an IC 50 ≥ 10 µM would not be considered suitable for progression along the drug discovery pipeline and would only be used as tools or probes. In this review, the trypanocidal activity of compounds is described according to their IC 50 values and are defined as: IC 50 < 10 µM = promising trypanocidal activity, 10 µM ≤ IC 50 < 20 µM = moderate activity, 20 µM ≤ IC 50 < 30 µM = marginal activity, 30 µM ≤ IC 50 < 40 µM = limited activity, IC 50 ≥ 40 µM = no activity/inactive. When evaluating the activity of compounds against a human pathogen or disease target it is important that the cytotoxicity of the compound is also investigated against a mammalian cell line to allow the selectivity index (SI) of the compound to be determined. The SI for selecting compounds with anti-trypanosomal activity is the ratio of the IC 50 value obtained for mammalian cells divided by the IC 50 against trypanosome species. We have considered herein that an SI < 10 suggests that the compound may be exerting a generally toxic effect. If the SI is ≥10, the compound is considered to have some selective activity against the parasite. However, a significantly greater SI is required in order for molecules to progress along the drug discovery pipeline and eventually into clinical studies.
This article will focus on compounds which have an IC 50 < 40 µM against T.b. brucei, T.b. rhodesiense or T. cruzi and which have also been evaluated against a mammalian cell line. Compounds have been grouped according to their chemical structures into three categories; terpenes, polyketides and xanthones, and alkaloids. Representative examples for each category are discussed in terms of their trypanocidal activity and SI. To allow the activity of compounds to be compared independently of their molecular weight, all literature values have been converted into micromolar concentrations (µM).

Terpenes
The marine sponges Spongia sp. and Ircinia sp. collected from the Turkish coastline of the Aegean Sea yielded a series of linear furanoterpenes and meroterpenes, as well as di-and tri-terpenes all of which were assessed for growth inhibitory activity against a series of protozoan parasites (Figure 4) [88]. 4-hydroxy-3-tetraprenylphenylacetic acid (1) was the most active and selective molecule with an IC 50 value of 1.4 µM against T.b. rhodesiense and a selectivity index (SI) of >150, versus mammalian L6 rat skeletal muscle cells. The related structure heptaprenyl-p-quinol (2) possessing a longer isoprene chain and a hydroquinone terminal unit showed promising activity against T.b. rhodesiense with an IC 50 value of 5.9 µM, however had no selectivity with an almost equivalent IC 50 value of 4.4 µM observed against L6 cells. Demethylfurospongin-4 (3) was selectively active against T.b. rhodesiense with an IC 50 value of 11.8 µM and an SI > 18. The diterpene 11β-acetoxyspongi-12-en-16-one (4) exhibited moderate activity against T.b. rhodesiense with an IC 50 value of 11.5 µM but had no selectivity with an IC 50 of 9.2 µM against L6 cells [88]. A number of trypanocidal molecules with varying degrees of activity have been identified from Agelas sp. marine sponges. The sterol 24-ethyl-cholest-5α-7-en-3-β-ol (5) isolated from the n-hexane extract of the Turkish sponge Agelas oroides showed limited activity against T.b. rhodesiense with an IC 50 value of 34.2 µM [89]. Compound 5 was inactive against both T. cruzi (IC 50 > 72 µM) and L6 cells (IC 50 > 217 µM). These authors used the T. cruzi β-galactosidase assay to estimate compound activity [81].
A series of steroidal saponins characterized by a 2-hydroxycyclopentenone ring D and a glucuronic acid substituent at C-3 isolated from the Caribbean sponge Pandaros acanthifolium have demonstrated wide-ranging biological activity, including inhibition of both T.b. brucei and T. cruzi. Notably, pandaroside G methyl ester (6) had sub-micromolar activity against both T.b. rhodesiense and T. cruzi with IC 50 values of 0.038 and 0.77 µM, respectively [90]. However, the molecule was not specific for T.b. rhodesiense or T. cruzi as it also inhibited mammalian L6 cells with an IC 50 value of 0.22 µM, suggesting the natural product was generally toxic. Related steroidal saponins, the acanthifolisides, were also isolated as minor components from the same sponge collection [91]. Acanthifolioside E (7) showed moderate activity against T. cruzi, with an IC 50
Two cyclic hexapeptides, venturamides A (60) and B (61) were isolated from the Panamanian collection of the marine cyanobacterium Oscillatoria sp. [115] (Figure 10). The two compounds showed moderate activity against T. cruzi with IC 50 values of 14.6 and 15.8 µM, respectively, and mild cytotoxicity to mammalian Vero (monkey kidney epithelial) cells with IC 50 values of 86 and 56 µM, respectively, and thus an SI of < 6. Related cyclic peptides aerucyclamides B (62) and C (63) isolated from the cyanobacterium Microrcystis aeruginosa also displayed anti-trypanosomal activity with IC 50 values of 15.9 and 9.2 µM, respectively, reported for T.b. rhodesiense [116]. Aerucyclamide C had a SI of 12 against L6 cells, whilst the SI of 62 was lower at 8. In a study using natural products as chemical probes to identify the molecular targets of small molecules, two linear peptides, almiramides B (64) and C (65) extracted from a Panamanian collection of the marine cyanobacterium Lyngbya majuscula were found to be low micromolar inhibitors of T.b. brucei with IC 50 values of 6 and 3 µM, respectively [117]. Almiramide C displayed a SI of 11 compared to Vero cells while the SI for almiramide B was slightly lower at 9. Moreover, through a series of target based affinity probes, and fluorescence site localisation imaging studies, the compounds were shown to disrupt glycosome function in the parasite. Glycolysis is an essential pathway in trypanosomatids, and glycosomal enzymes have been identified as a potential drug target in trypanosomes [118].

Conclusions
A large number of structurally diverse marine natural products have been identified with trypanocidal activity. The manadoperoxides isolated from the marine sponge Plakortis cfr. lita are the most promising compounds for HAT. Manadoperoxide B (12) was the most active and selective molecule of the series exhibiting sub-micromolar activity against T.b. rhodesiense whilst highly selective against mammalian cells [94]. This compound was also demonstrated to possess anti-malarial activity, however, it is reported to be more than 700-fold less active against Plasmodium falciparum (D10) than T.b. rhodesiense [119]. As manadoperoxide B has sub-micromolar activity against T.b. rhodesiense and is not cytotoxic, one would anticipate that the physiochemical properties of the molecule, together with the biological activity are being investigated further to ensure the molecule possesses the required characteristics to meet the final target product profile.
The heterocyclic-substituted xanthone analogue 25 isolated from the marine fungus Chaetomium sp. was the most active and selective, marine derived compound for T. cruzi [96]. However, xanthones have been reported to have activity against multiple organisms and disease indications through interacting with a plethora of enzymes and targets [120]. This promiscuous activity may prevent further development of the compounds for Chagas disease.
In the last decade numerous molecules, both natural and synthetic, have been identified with trypanocidal activity. However, only two, have entered pre-clinical development for HAT. Furthermore, despite the identification of new targets and a multitude of in vitro and in vivo studies having been conducted, candidates for Chagas disease have failed to progress to the advanced stages of clinical development. Many of the molecules identified with potent trypanocidal activity, cannot be developed further as they possess unsuitable and undesirable structural and pharmacokinetic properties. This highlights the need to continue to explore other avenues for new chemical entities, whilst reviewing the approaches currently undertaken and the potential reasons for the lack of success. Evaluation of the current in vitro assays used to identify new compounds, in particular the life cycle stage for Chagas disease, is warranted. This is particularly true for the in vivo models where the parasite strain, administration route and duration of the study can impact on the outcomes.
Marine natural products have provided the pharmaceutical industry with many incredibly potent compounds-some developed into therapeutics whilst others providing valuable insights into the biology of disease and desired attributes of the compounds required to ameliorate it. Whilst compounds isolated from this source have yet to progress to pre-clinical development for trypanosomatid diseases, collectively the improvements to the in vitro assays used to identify them, the in vivo models used to evaluate them, and the methodology required for isolating them could change this situation.

Conflicts of Interest
The authors declare no conflicts of interest.