Complementary medicinal chemistry-driven strategies toward new antitrypanosomal and antileishmanial lead drug candidates

Abstract

Trypanosomiases and Leishmaniases are neglected tropical diseases that affect the less developed countries. For this reason, they did not and still do not have high visibility in Western societies. The name neglected diseases also refers to the fact that they often received little interest at the level of public investment, research and development. The drug discovery scenario, however, is changing dramatically. After a period in which different socioeconomic factors have prevented massive research efforts in this field, such efforts have increased considerably in the very recent years, with significant scientific advancements. In this context, we have embarked on a new drug discovery project devoted to identification of new small molecules for the treatment of trypanosomal and leishmanial diseases. Two complementary approaches have been pursued and are reported here. The first deals with a structure-based drug design, and a privileged structure-guided synthesis of quinazoline compounds able to modulate trypanothione reductase activity was accomplished. In the second, a combinatorial library, built on a natural product-based strategy, was synthesized. Using whole parasite assays, different quinones have been identified as promising lead compounds. A combination of both approaches to hopefully overcome some of the challenges of anti-trypanosomatid drug discovery has eventually been proposed.

Introduction

At the beginning of the third millennium, protozoan parasitic diseases are still an unsolved public health problem causing morbidity and mortality worldwide. Among the various infections that afflict mankind and livestock, trypanosomatid infections are responsible for more than 20 million cases primarily in tropical and subtropical areas of the globe, and it is estimated that half a billion people are at risk of contracting these diseases (Stuart et al., 2008). In particular, Trypanosoma is responsible for Chagas disease in South America and sleeping sickness (or human African trypanosomiasis, HAT) in Africa (Fevre et al., 2008), whereas Leishmania is responsible for cutaneous and visceral infections, endemic in 88 countries primarily in the Horn of Africa, South Asia and Latin America, with cases also registered in Spain and in the south of Italy (Bern et al., 2008). Because of its main prevalence in the poorest areas of the planet, Trypanosomiases and Leishmaniases lack visibility in the developed countries. For this reason, the World Health Organization (WHO) characterizes them as the most challenging among neglected tropical diseases (NTD). Besides the low visibility, this name is related to the fact that they are often neglected when health agendas and budgets are set, and there has been little in the way of a concerted drug discovery program, both at an academic and an industrial level.

The current drug discovery scenario

The diseases caused by trypanosomatids are very distinctive in terms of pathology, but they share a similar history of strategies for their treatment and control. Despite ongoing attempts to produce a molecular vaccine, present treatment and control rely on chemotherapy (Kinoshita, 2008; Miyahira, 2008; Pinzon-Charry & Good, 2008), but none of the different chemotherapeutic strategies used in the past has been proven consistently successful (Paulino et al., 2005). Many of the anti-trypanosomatid drugs in common use today were discovered at the beginning of the last century, as the tropical pharmacopeia was driven by colonial requirements or, in other cases, first developed for other pharmacological indications. Therefore, although effective during the acute infection stages, they are often nonefficacious against the late stages and produce significant side effects due to their high toxicity. In addition to the deleterious treatments, the number of drug-resistant strains increases sharply. The drugs marketed to treat Chagas disease are nifurtimox (1; Lampit®; Fig. 1) and benznidazole (2; Rochagan®), belonging to the class of nitroaromatic compounds. They show efficacy limited to the disease acute phase and to only some pathogen strains. Moreover, because of the serious side effects, such as anorexia, vomit and diarrhea, nifurtimox was discontinued by Bayer (Paulino et al., 2005). However, recently the use of nifurtimox against trypanosomatid-borne diseases has been considered again, but currently only for the treatment of HAT, and clinical trials are underway to test new drug treatment regimens based on the nifurtimox–eflornithine combination (http://www.dndi.org). As in the former case, the treatment of HAT relies on drugs discovered over 40 years ago, namely suramin (3; Germanin®), a polysulfonated naphthylurea, pentamidine (4; Pentacarinat®), an aromatic diamidine and melarsoprol (5; Mel b, Arsobal®), a melaminophenyl-based organic arsenical. The only exception is eflornithine (6; Ornidyl®), an analogue of the amino acid ornithine, which acts as an inhibitor of the ornithine decarboxylase enzyme, interfering with polyamines' metabolic pathway. The major drawbacks of this drug are the high cost and its effectiveness, which is limited to the Trypanosoma gambiense strain (Fairlamb, 2003; Nok, 2005). Treatment of Leishmaniasis is even more difficult, because of the presence of three different forms and various clinical manifestations (Murray et al., 2005; Croft et al., 2006). Nowadays, there are nearly 25 compounds with antileishmanial effects, but only a few are classified as antileishmanial drugs for humans and most of them are parenteral. First-line treatments are pentavalent antimonials meglumine antimoniate (7, Glucantime®) and sodium stibogluconate (8, Pentostam®), compounds discovered more than 50 years ago, which present severe, unwanted side effects, and whose mechanism of action is still being investigated. Several studies suggest that pentavalent antimony acts as a prodrug that is converted to the active and more toxic trivalent antimony [Sb(III)] (Frezard et al., 2009). A very recent crystallographic study has demonstrated that Sb(III) can then inhibit the enzymatic activity of trypanothione reductase (TR) (Baiocco et al., 2009). The second-line drugs are pentamidine (4) and amphotericin B (9). To reduce renal toxicity, an amphotericin liposomal formulation (Ambisome®) has been successfully developed by the pharmaceutical industry (Golenser & Domb, 2006). However, despite a recent significant reduction in price, this medication remains very expensive for endemic countries. Miltefosine (10, Miltex®), an alkylphosphocholine derivative initially developed as an antineoplastic drug, is the first orally administered therapy for visceral and cutaneous Leishmaniasis. However, it cannot be used during pregnancy because of the potential risk of birth defects, and shows severe gastrointestinal side effects (Berman, 2006, 2008).

In 2005, the genome sequences of the ‘Tritryp’ (Trypanosoma brucei, Trypanosoma cruzi and Leishmania major) (Berriman et al., 2005; El-Sayed et al., 2005; Ivens et al., 2005) and in 2007 those of two other species of Leishmania (Leishmania infantum and Leishmania braziliensis) (Peacock et al., 2007) were completed, eliminating many of the hurdles to performing state-of-the-art drug discovery research on Kinetoplastida (comprising Trypanosoma and Leishmania) parasites. This vast amount of new information has made a more comprehensive and accurate identification and validation of novel drug targets possible. However, new licensed therapies have not yet resulted from genome-dependent experiments, and Trypanosomiases and Leishmaniases, even in the postgenomic era, may remain incurable and often fatal. How can the pharmaceutical community respond to this challenge? Presently, medicinal chemists hold many of the keys to success, but it is mandatory to also exploit in antiparasitic drug discovery the modern strategies already adopted in other more for-profit fields.

In Fig. 2, a current snapshot of the activities in antiparasitic drug discovery is provided. As in other pharmaceutical fields, the drug discovery process, in general referred to as a pipeline, includes several distinct phases from the basic research to clinical trials, and hopefully to the final marketed drug. The advent of modern molecular genetics and high-throughput approaches to chemistry and biology have revolutionized the way in which drug discovery has been conducted in the last 30 years. The ability to produce, through recombinant technology, almost any protein has led to a paradigm shift in which a target-based approach has replaced the former physiology-based approach. In this case, the final goal is to modulate the activity of a particular protein such that a therapeutically beneficial response is induced. This approach has dominated the pharmaceutical research over the last decade, but today other approaches that are not target based are conceivable as well. Use of whole parasite-based assays for screening of molecules selected on the basis of a biological, biochemical or structural rationale has been steadily declining over the years, but is now undergoing a resurgence, mainly due to improvements in assay technologies (Pink et al., 2005; Renslo & McKerrow, 2006). This, in principle, could circumvent some of the obstacles that have emerged for the exploitation of novel, but chemically unvalidated targets identified from genetic or genomic methodologies.

Considering that the two approaches are not mutually exclusive, rather they should be viewed as complementary, in pursuit of novel antitrypanosomal and antileishmanial lead candidates, we have adopted 2 strategies, namely, the structure-based drug design and the nature-inspired library design. Accordingly, the following two paragraphs deal with the two approaches.

Discovery of TR inhibitors through a privileged structure-based approach

As a result of the genome sequencing projects, the possibility of identifying novel drug targets that are of vital importance to the pathogens, but are absent in their hosts, and that are open to selective inhibition without a tendency for the parasite to develop resistance has become a realistic exciting opportunity.

Trypanothione (T[SH]₂) is a low-molecular-mass thiol unique to members of the order Kinetoplastida, which is responsible for maintaining the intracellular reducing environment after parasite aerobic metabolism. T[SH]₂ is regenerated in the thiol state by the enzyme TR (E.C. 1.6.4.8), a homodimeric flavoprotein oxidoreductase. The reduction of trypanothione disulfide (TS₂) to T[SH]₂ catalyzed by TR is homologous to the reduction of glutathione disulfide by glutathione reductase (GR, E.C. 1.8.1.7) in mammals. The similar biological role may in turn be responsible for a high sequence similarity between TR and GR, which display 40% of identity and share three amino acids at the catalytic site, the Cys53, Cys58 and His461′ (from the second subunit), in T. cruzi TR (Krauth-Siegel et al., 2005). However, two major features account for the selectivity of inhibitors against the two enzymes. Because of the presence of an extra protonated amino group in T[SH]₂, TR possesses a negatively charged active site, whereas that of human GR carries an overall positive charge (Krauth-Siegel et al., 2005). Moreover, TR has a larger substrate-binding pocket to fit its bulkier endogenous ligand, T[SH]₂. These peculiarities and the elucidation of the 3D structures of several complexes make TR an ideal candidate for a structure-based drug discovery project. Moreover, the absence of the T[SH]₂ system in mammals, the lack of a functional redundancy within the parasite thiol system, together with the sensitivity of trypanosomes against oxidative stress, render TR as one of the validated drug targets in the search for anti-trypanosomatid drugs (D'Silva & Daunes, 2002; Chibale & Musonda, 2003; Wolf & Dormeyer, 2003; Saravanamuthu et al., 2004; Paulino et al., 2005; Jaeger & Flohe, 2006; Linares et al., 2006; Krauth-Siegel & Comini, 2008; Rivera et al., 2009).

In a search for novel compounds (Cavalli et al., 2009), we reasoned that a scaffold protonated at a physiological pH might be fundamental in conferring TR vs. GR selectivity. However, antiparasitic amidines or guanidines with positively charged groups are in general poorly orally available (Renslo & McKerrow, 2006; Werbovetz, 2006). Therefore, we searched for new chemical entities with superior intrinsic drug-like properties. Privileged structures, as firstly formulated by Evans (1988), are structural motifs common to several drug and lead compounds, and endowed with versatile binding properties. Therefore, in principle, libraries derived from such scaffolds may provide promising compounds in diverse therapeutic areas (Horton et al., 2003; Costantino & Barlocco, 2006). Notably, the likely drug-like properties of privileged structures and substructures might produce more drug-like compound libraries and leads (DeSimone et al., 2004). Properly substituted 2,4-diamino-quinazolines have been labeled as privileged substructures and they are protonated at a physiological pH (Bordner et al., 1988). Numerous quinazoline-based libraries have been developed, and have provided useful ligands for receptor and enzymatic targets (Kamal et al., 2006), also in the field of antiparasitics (Davoll et al., 1972). On these bases, a series of 2-piperazin-1-yl-quinazolin-4-ylamine derivatives (11–17 in Table 1) were, for the first time, designed and tested against TR. Docking of the quinazoline core showed that such a scaffold was able to interact with the TR active site (Fig. 3). Then, different moieties, such as methoxy and dimethylaminoethoxy groups in R¹, and indole, quinone and tetrahydroquinoline moieties in R², were attached to the quinazoline core to provide additional interactions with the enzyme (Fig. 3). The designed compounds were synthesized exploiting straightforward chemical protocols and then tested for their inhibitory activity against T. cruzi TR. All of the synthesized quinazolines proved to inhibit the enzymatic activity, except 11 and 12. The inhibition ranged from 23% to 76%, with the naphthoquinone derivatives 15 and 17 being the most effective inhibitors (Table 1). From these results, we could speculate that the 2,4-diaminoquinazoline scaffold is a good motif for TR recognition. However, the presence of a suitable capping fragment is fundamental for activity. For 14, 15 and 17, the inhibition constants were determined. The naphthoquinone derivatives 15 and 17 inhibited TR, with K_i values of about 10 µM. The degree of inhibition is comparable to that exhibited by other known TR inhibitors, such as acridine or phenothiazine derivatives (Bolognesi & Cavalli, 2006). Compounds 15 and 17 were also tested against human GR. At a concentration of 10 µM, both showed negligible inhibitory activity, further supporting the design rationale. As a subsequent step, the activity of 11–17 against T. cruzi was tested, as well as their cytotoxicity against L6 rat myoblast cells. The results of Table 1 show a good correlation between enzyme inhibition and trypanocidal activity for compound 15, which is the most potent TR inhibitor. On the other hand, compounds 11 and 12, inactive at 100 µM against TR, also interfered with parasite growth, confirming the potential of the quinazoline scaffold for antiparasitic activity. As a drawback, some derivatives showed cytotoxicity toward L6 cells (Table 1). The highest mammalian cytotoxicity was associated with the presence of a naphthoquinone (see 15 and 17), suggesting that toxicity can be tuned down by properly modifying or replacing this moiety. It is also important to highlight that all the designed compounds were active in the whole cell-based assay, and therefore are able to enter cells and reach their intracellular target (Baniecki et al., 2007). However, in view of the design rationale applied, we cannot exclude that other proteins are (also) modulated by our compounds.

In antiparasitic drug discovery, where scientists have been stimulated to find alternative and less costly approaches, the use of validated chemical scaffolds might emerge as a valuable option. New drug candidates must meet stringent requirements, such as being inexpensive to manufacture and distribute and practical to administer (Renslo & McKerrow, 2006). In this scenario, this small library holds promise for further exploration of quinazoline substructure for interaction with TR, with the potential to generate, through iterative medicinal chemistry protocols, preliminary structure–activity relationships and in vivo activities. The aim of reducing the attrition rate, and thus increasing the overall efficiency and productivity in the drug discovery process, might be fulfilled. All these considerations, particularly those concerning the economic aspects, indicate that the exploitation of quinazolines or other privileged structures might have potential for producing new leads against NTD.

Discovery of natural product-inspired antitrypanosomal and antileishmanial compounds

In the second approach, a small focused collection of a combinatorial library of 16 molecules was generated and tested by means of growth assays of parasites. In this case, the molecular target is initially unknown, and a posteriori, by means of the chemical proteomics, the target is tentatively fished out from parasitic cell extracts. We reasoned that a combinatorial chemistry approach would allow the synthesis of large numbers of related molecules more rapidly, more efficiently and more cheaply than classical medicinal chemistry approaches (Link, 2003; Musonda & Chibale, 2004). The chemical economy and effectiveness of a combinatorial discovery process in the context of large pharmaceutical companies cannot be disputed. However, to make combinatorial chemistry cost-effective in synthesizing antiparasitic compounds and tailored to our academic settings, we sought a strategy that could yield small molecules with high hit rates and a concomitant reduced library size. In this respect, natural product-derived and -inspired collections concepts are particularly attractive, because they recognize natural product fragments as evolutionarily selected and biologically prevalidated frameworks to be used for compound collection development (Breinbauer et al., 2002). The likelihood of generating high-quality compounds may be particularly high, if links from such compound classes to the desired biological activity already exist (Lessmann et al., 2007). To meet the criteria outlined above for generating a library of antiparasitic hit compounds, the quinone unit was selected as a suitable template for combinatorial derivatization. Naphthoquinones and other related quinone compounds occur with a high frequency in natural products having activity against Leishmania and Trypanosoma (Kayser et al., 2003). Lapachol (Fig. 4) is a representative example. Building on the 1,4-naphthoquinone and 1,4-anthraquinone natural scaffolds, 16 compounds were synthesized through a parallel approach (see Fig. 4 for the design strategy) by introducing in position 2 aromatic groups mimicking the structural elements of triclosan (Fig. 4), the antimicrobial compound also active against Trypanosoma.

The synthesized library was tested in a medium-throughput whole organism screening, against the blood trypomastigote form of Trypanosoma b. rhodesiense, the intracellular amastigote form of T. cruzi and the axenic amastigote form of Leishmania donovani. All compounds showed good activities against cultured parasites. Notably, the 2-phenoxy-anthracene-1,4-dione derivative 18 reported in Fig. 4 turned out to be the most potent library member, with an IC₅₀ of 50 nM against T. brucei. Many other library entries possessed micromolar IC₅₀ values against T. cruzi and L. donovani. Interestingly, 2-(2,4-difluoro-phenoxy)-anthracene-1,4-dione 19 was slightly more potent than the reference drug miltefosine. The main drawback of this series of compounds was the general cytotoxicity against mammalian cells, with the exception 2-phenoxy-1,4-naphthoquinone (20), which was selected as the library hit. In fact, it showed an IC₅₀ value of 80 nM and a parasite vs. murine cells selectivity index of 74, which is not far from the specifications given by the WHO/TDR for a compound to be considered an anti-trypanosomatid drug candidate (Bolognesi et al., 2008). Proteomic methods are being used for directly isolating and identifying from cellular extracts the protein target that is bound by 20 or, more likely the multiple targets involved in its trypanocidal activity. In fact, a multitarget profile for this compound is easily conceivable, because quinones, like many other natural products, serve plants as potent defense chemicals with an intrinsic pleiotropic mechanism of action (Wink, 2008). During evolution, the constitution of natural products has been modulated so that they usually contain more than one active functional group, allowing them to interact with several molecular targets. Because of their biosynthetic origin, natural products are natively bound to proteins (synthetases). In most cases, natural bioactive compounds have a complex structure and their synthesis involves a range of enzymes, each of which has distinct architectures and molecule-binding cavities, and all of which the molecule under synthesis must be able to interact with. Therefore, the reason why natural products have inherently more potential than synthetic compounds to bind multiple target proteins might indeed be due to this mode of generation (Ji et al., 2009). A comment is required on the possible benefits to identify compounds endowed with a multitarget profile in anti-trypanosomatid drug discovery. As a general rule, chemotherapeutic agents that target single proteins are susceptible to high-level resistance, resulting from a single-step mutation in the target protein. Conversely, drugs that have a low likelihood for the development of endogenous resistance are those that interact with multiple molecular targets, whose structures are determined by multiple genes (Fig. 5). This clearly favors the development of multitarget over single-target compounds also in NTD (Cavalli & Bolognesi, 2009). If this approach is still in its infancy for these infections, for malaria treatment, several multifunctional drug candidates have been developed by academia and industry during the last couple of years (Chibale, 2002; Jenwitheesuk & Samudrala, 2005; Jenwitheesuk et al., 2008; Meunier, 2008), supported by the recent advances in the knowledge of the action and resistance mechanisms involved.

Outlook section

A new drug discovery project aimed at identifying new molecular probes and drug candidates for studying and combating trypanosomatids NTD was initiated a few years ago. We have followed two state-of-the-art medicinal chemical strategies: (1) the structure-based drug design and (2) the nature-inspired library design, generating interesting, although preliminary, results in both of the projects. Adopting the structure-based approach, we have identified novel T. cruzi TR inhibitors, which also discriminate between human GR. The molecules carry privileged scaffolds, which offer the potential to accelerate drug discovery in the field of NTD. Adopting the nature-inspired library design approach, we have synthesized a small focused collection of compounds that displayed remarkable activity against Trypanosoma and Leishmania when tested in a whole parasite assay. At first glance, the two approaches might look antithetical, but they might be combined in an effort to reduce the number of failures and pitfalls generally experienced in the pharmaceutical research. Both have complementary advantages and drawbacks. Classically, target-based drug discovery has enabled researchers to perform what is called ‘rational drug discovery’ and the advent of modern technologies in the drug discovery process. However, it is reductionistic in concept, where the small molecule–organism interplay is reduced to a small molecule–target interplay. Although the plethora of information provided by genomic tools has enabled the identification of potential therapeutic targets on one level, it is clear that sequence information alone cannot resolve the complex patterns of gene activities, which ultimately dictate the physiology and pathology of an organism. The cell might represent an experimental model that preserves the essential elements of a disease network. However, cell-based drug discovery is not a very productive way to search for leads, if carried out in a traditional way. Nevertheless, interesting compounds may be identified through screening of libraries in specialized phenotypic screening systems that mimic specific aspects of a disease (Matter & Keller, 2008). Furthermore, cell-based assays are indispensable tools when searching for multitarget compounds, because they maintain a reasonable experimental efficiency while preserving molecular–pathway interactions (Zimmermann et al., 2007). Certainly, the approach has the benefit that drugs are optimized for biological effects directly in the intact parasite, independent of specific mechanisms.

We believe that we should combine the advantages of target-based drug discovery with an innovative system-based approach, in order to overcome some of the weaknesses of both of the methods. This might take full advantage of the great promise of the genomics era and make new strides in the ongoing effort to develop effective drugs against NTDs.

Acknowledgements

Prof. C. Melchiorre is kindly acknowledged for helpful critical discussion. The University of Bologna, Italy, is gratefully acknowledged for financial support.

References

Author notes