In-silico Leishmania Target Selectivity of Antiparasitic Terpenoids

Neglected Tropical Diseases (NTDs), like leishmaniasis, are major causes of mortality in resource-limited countries. The mortality associated with these diseases is largely due to fragile healthcare systems, lack of access to medicines, and resistance by the parasites to the few available drugs. Many antiparasitic plant-derived isoprenoids have been reported, and many of them have good in vitro activity against various forms of Leishmania spp. In this work, potential Leishmania biochemical targets of antiparasitic isoprenoids were studied in silico. Antiparasitic monoterpenoids selectively docked to L. infantum nicotinamidase, L. major uridine diphosphate-glucose pyrophosphorylase and methionyl t-RNA synthetase. The two protein targets selectively targeted by germacranolide sesquiterpenoids were L. major methionyl t-RNA synthetase and dihydroorotate dehydrogenase. Diterpenoids generally favored docking to L. mexicana glycerol-3-phosphate dehydrogenase. Limonoids also showed some selectivity for L. mexicana glycerol-3-phosphate dehydrogenase and L. major dihydroorotate dehydrogenase while withanolides docked more selectively with L. major uridine diphosphate-glucose pyrophosphorylase. The selectivity of the different classes of antiparasitic compounds for the protein targets considered in this work can be explored in fragment- and/or structure-based drug design towards the development of leads for new antileishmanial drugs.


Introduction
Several closely-related protozoan parasites in the genus Leishmania are etiological agents for a number of clinical forms of leishmaniasis. These clinical forms of leishmaniasis are characterized as either cutaneous, diffuse cutaneous, disseminated cutaneous, mucocutaneous, visceral or post-kala-azar dermal. Species causing this protozoan disease have been reported in several tropical and Neotropical regions including Africa, the Americas, Eastern Europe, central and western Asia, the Indian subcontinent as well as in Australia. There are over 350 million people at risk of infection in Leishmania-endemic regions. There are several treatment options for leishmaniasis, although the effectiveness of the available drugs depends on which clinical form is being treated, and also on the specific geographical location. For a review of the current treatment options please see reference [1]. There remains a need for better chemotherapy for cutaneous, visceral and post-kala-azar dermal leishmaniasis, as well as, Leishmania-HIV co-infection.
Current chemotherapy of visceral and cutanoeus leishmaniasis includes miltefosine, a compound that has been demonstrated to inhibit P13K/Akt signaling pathway, and fluconazole, a sterol 14α-demethylase inhibitor. In addition to currently targeted Leishmania proteins, several other proteins have also been identified, or suggested as potential drug targets in Leishmania [2][3][4][5]. Most of these targets include enzymes that are critical to the metabolism of glucose, sterols, nucleotides and glycosylphosphatidylinositol, as well as enzymes important for the maintenance of trypanothione and polyamine levels. Many of these proteins have been shown to be important to the survival of the parasites. Other targets include cyclin-dependent-and mitogen-activated protein kinases, topoisomerases and cathepsin-like proteases.
Numerous phytochemical agents have exhibited either in vitro or in vivo antileishmanial activity [5,[53][54][55][56][57][58][59][60][61]. While the activities of many of these compounds are notable, what is generally unknown are the biochemical targets of these agents. In this study, we have carried out a molecular docking analysis of known antiparasitic plant-derived isoprenoids with established drug targets with known structures available from the Protein Data Bank.

Monoterpenoid Docking
The structures of the monoterpenoids examined in this study are shown in Figure 1, while the corresponding docking energies are summarized in Tables 1-3. The overall strongest docking monoterpenoid ligands were the acyclic geranial, geraniol, and neral, probably owing to their flexibility. These ligands, however, did not show docking selectivity to any of the Leishmania protein targets, but rather docked strongly to most of the proteins investigated. The protein targets that showed predominantly strong docking by monoterpenoids were L. major uridine diphosphate-glucose pyrophosphorylase (LmajUGPase), L. major methionyl t-RNA synthetase (LmajMetRS), and L. infantum nicotinamidase (LinfPnC1). Geranial had a docking energy of −76.9 kJ/mol with LmajUGPase, comparable in docking energy with several other proteins. Both enantiomers of piperitone showed significantly stronger docking to Lmaj UGPase (−68.0 kJ/mol) than the other targets, suggesting selectivity for that protein. Geranial was also the strongest docking ligand with LmajMetRS (−76.8 kJ/mol), but perilla alcohol (−73.6 kJ/mol) was selective for that protein target. Carvone, piperitone, and α-thujone showed significantly selective docking to LinfPnC1 (docking energies less than −73 kJ/mol). Interestingly, although the monoterpenoids showed a docking propensity for LinfPnC1, higher terpenoids (sesquiterpenoids, diterpenoids, and triterpenoids) showed very little inclination to dock to this protein, generally with positive docking energies (see below).
Monoterpenoids represents a very small percentage of terpene-derived compounds that have been reported to have antileishmanial activity, and the docking energies of monoterpenoids were generally weaker than those obtained for limonoids, withanolides, triterpenoids, steroids and diterpenoids with these same targets (see below). Their docking energies were much higher than the energies obtained for the co-crystallized ligands of those protein targets. The higher docking energies of these compounds correlate with their small size (and molecular weight), and the minimal intermolecular interactions they are able to have with the protein targets. So, comparatively, it appears that monoterpenoids will not be prime leads for structure-based antileishmanial drug discovery. However, they may be useful in fragment-based drug discovery [62,63]. Additionally, several terpene-derived compounds are used in topical formulations. Therefore, those monoterpenoids that have antileishmanial activity and no reported toxicity at physiologically relevant concentration/dosage should be evaluated as possible components of topical polytherapy for leishmaniasis.

Sesquiterpenoid Docking
Sesquiterpenoids examined in this work are shown in  docking energies of sesquiterpenoids are summarized in Tables 4-15. The germacranolide sesquiterpenoids exhibited the overall strongest docking energies toward the Leishmania protein targets, with 16,17-dihydrobrachycalyxolide the strongest-docking germacranolide. This ligand showed docking selectivity toward LmajMetRS (docking energy = −152.9 kJ/mol) and L. mexicana phosphor-mannomutase (LmexPMM) (docking energy = −136.3 kJ/mol). The two proteins most selectively targeted by the germacranolides in terms of docking energies were LmajMetRS and L. major dihydroorotate dehydrogenase (LmajDHODH). Although most germacranolides did not dock with LinfPnC1, tatridin A did show docking selectivity for this protein target. Based on molecular weight, the strongest-docking germacranolide was 4α,5-epoxy-8-epi-inunolide, and this ligand showed docking selectivity toward both LmajMet RS and LmajDHODH.              The guaianolide with the strongest docking energy was diguaiaperfolin, probably owing to its dimeric structure and larger molecular weight (716.77 amu). This ligand did show notable docking (−154.5 kJ/mol) with LmajDHODH as well as with LmajUGPase (docking energy = −151.8 kJ/mol). 8-[4-Hydroxy-5-(5-hydroxytigloyloxy)tigloyl]santamarin was the strongest-docking eudesmanolide, and this ligand showed docking selectivity to LmajMetRS (docking energy = −153.1 kJ/mol). The proteins most strongly targeted by both the guaianolides and the eudesmaolides were also LmajMetRS and LmajDHODH. Interestingly, the miscellaneous sesquiterpenoids preferentially targeted L. major pteridine reductase 1 (LmajPTR1), and plagiochilin A showed notable selectivity (docking energy = −132.2 kJ/mol) for this protein.
Several electrophilic sesquiterpenoids have exhibited antiprotozoal activity [5] and many of these showed docking selectivity to LmajDHODH. The active site of this protein has some potential nucleophilic residues, namely Ser 69, Ser 196, and Cys 131. Suitably oriented electrophilic ligands could form covalent bonds with these nucleophiles and thus inhibit the enzyme. Thus, for example, the germacranolide tatridin A docked preferentially to LmajDHODH, and the lowest-energy docked pose oriented the electrophilic carbon of the α-methylene lactone moiety close to the sulfur atom of Cys 131 (see Figure 6). Similarly, the lowest-energy docked orientation of 11-epi-ivaxillin places one of the epoxide groups near to the sulfur atom of Cys 131 (Figure 7). Conversely, the lowest-energy docked pose of neurolenin B is such that the electrophilic carbon of the α-methylene lactone group of the ligand is near the hydroxyl group of Ser 69 ( Figure 8). The ligand with the lowest docking energy to LmajDHODH was the guaianolide dimer diguaiaperfolin (−154.5 kJ/mol). The lowest-energy pose for this ligand placed the cyclopentenone moiety near the sulfur atom of Cys 131 ( Figure 9).

Triterpenoid Docking
Triterpenoids, including limonoids, withanolides, quassinoids, and steroids are shown in Figures 20-24       Grandifotane showed selective docking to LmajDHODH with a docking energy of −154.4 kJ/mol. The lowest-energy pose of the ligand placed the compound at the binding site of the co-crystallized ligand, 5-nitroorotic acid ( Figure 25). The furan ring of the ligand is sandwiched between the riboflavin monophosphate cofactor and Cys 131. There are hydrogen-bonding interactions between the docked ligand and residues Asn 199, Asn 68, Ser 69, and Ser 130. Grandifotane, isolated from Khaya grandifoliola [65], has apparently not been reported to be antiparasitic. The bark and seeds of K. grandifoliola, however, have shown significant antimalarial activity [66]. 6-O-Acetylswietenolide, also isolated from K. grandifoliola, has shown antiplasmodial activity [67], and this compound showed preferential docking to LmexGPDH. The lowest-energy docking pose of 6-O-acetylswietenolide with LmexGPDH ( Figure 26) lies in a cavity surrounded by Arg 274, Phe 26, Lys 125, Phe 156, Val 298, Lys 210, and Ala 157. Lys 125 and Lys 210 have been identified as critical to catalytic activity of this enzyme [15]. With the exception of lanosterol, the steroid ligands showed preferential docking to LmajMetRS. The lowest-energy poses for these steroids show them all occupying the methionyl adenylate binding site ( Figure 27). The tetracyclic steroidal structures all occupy the same position in a hydrophobic pocket surrounded by Asp 486, Trp 515, Lys 522, His 228, Gly 227, His 513, and Tyr 218, and hydrogen-bonded by way of the 3-hydroxyl group of the steroid to Trp 516 and Gly 514. Clerosterol has shown in vitro antileishmanial activity [68], while saringosterol, stigmasterol, and 24-hydroperoxy-24-vinylcholesterol, in addition to clerosterol, have shown in vitro antitrypanosomal activity [69]. -Sitosterol has shown modest antitrypanosomal activity [70].               Not surprisingly, all of the steroids and many of the triterpenoids examined in this study showed significant docking preference for L. infantum sterol 14α-demethylase (LinfCYP51). This had been noted previously with Trypanosoma brucei sterol 14α-demethylase [71]. In particular, 24-hydroperoxy-24-vinylcholesterol (docking energy = −121.5 kJ/mol) and 24,25-epoxywithanolide D (docking energy = −128.1 kJ/mol) were strongly docking with LinfCYP51. The lowest-energy pose of 24-hydroperoxy-24,25-vinylcholesterol with LinfCYP51 places the hydroperoxy group of the ligand adjacent to the heme Fe ( Figure 28); this ligand, then, can presumably oxidize the Fe and render the enzyme inactive.   An examination of docking energies with respect to ligand molecular size suggests that for terpenoid ligands there is a threshold where larger size does not correspond to stronger binding to the protein target. A plot of molecular weights of representative terpenoids (monoterpenoids, germacranolide sesquiterpenoids, labdane diterpenoids, and triterpenoids) and docking energies to three different protein targets (LmajMetRS, LmexGPDH, and LdonCyp) ( Figure 29) shows that strongest docking energies are terpenoids with molecular weights around 360-430 amu. residues unchanged. The resulting models were refined with conjugate gradient minimization with no experimental energy terms used in the crystallographic and NMR System (CNS) program suite [97]. The resulting detailed model was refined with conjugate gradient minimization with no experimental energy terms used. All atoms of the molecules were unrestrained and were minimized for 500 steps with a continuous dielectric constant of one. Figure 29. Plots of docking energies vs. molecular weights for representative isoprenoids with three Leishmania protein targets.

Conclusions
Numerous antiparasitic plant-derived natural products have been identified but the molecular target(s) of most of these compounds remain unknown. This gap in knowledge impedes further characterization and optimization of the antiparasitic activity of many of these compounds. In this molecular docking study, we have identified molecular targets in Leishmania that preferentially interact with certain classes of antiparasitic isoprenoids from plants. Consequently, Leishmania proteins that have structural motifs similar to those identified in this work may be explored as potential drug targets by antileishmanial drug discovery programs. It is important to point out that: (a) there are likely additional Leishmania proteins or other biochemical targets that have not yet been identified; (b) some of the antiparasitic terpenoids examined in the study may have poor bioavailability due to limited solubility, membrane permeability, hydrolysis, or other metabolic transformations; (c) the ligands may also target homologous isozymes in humans. Therefore, pharmacokinetic and pharmacodymanic studies as well as structure-based design and optimization studies are needed to resolve issues of bioavailability and selectivity. In summary, this in-silico molecular docking study has provided evidence for what classes and structural types of terpenoids may be targeting certain Leishmania protein targets and could provide the framework for synthetic modification of antiparasitic terpenoids, de novo synthesis of structural designs, and further phytochemical investigations.