Lead optimisation of dehydroemetine as an antimalarial option through drug repositioning and molecular modelling

Malaria is a life-threatening infectious disease caused by a parasitic protozoan belonging to the genus Plasmodium. Resistance has started to emerge for the currently recommended artemisinin combination therapies used to treat malaria. There is a great need for new anti-malarial drugs, but traditional drug discovery timelines take approximately 10-17 years. Drug repositioning offers a very effective alternative to de novo drug design and could minimise the long timescales involved in the process of bringing a drug to the market. Work at the University of Salford identified the previously widely-used anti-amoebic drug emetine dihydrochloride as a potent inhibitor of Plasmodium falciparum K1 strain parasites (IC50: 47 + 2.17 nM). Emetine is not currently in use due to its emetic and cardiotoxic effects.The cryo-EM structure (3J7A) depicting the emetine binding site on the 40S subunit of the 80S P. falciparum ribosome has recently been published as a potential target binding site for emetine, enabling rational design of safer synthetic analogues. This study focuses on using molecular modelling tools to predict activities of the two isomers of dehydroemetine, a synthetic analogue of emetine developed in the 1960s by Roche. Studies conducted in the mid-20th century found dehydroemetine to be less toxic than emetine; however, its use as an anti-amoebic was replaced due to the introduction of the safer drug metronidazole, although dehydroemetine still remains as a WHO recommended drug for use in amoebiasis in metronidazole treatment failure. Using MOE and Autodock vina docking software, the activities of two isomers of dehydroemetine; (-)-R,S-dehydroemetine and (-)-S,S-dehydroisoemetine were predicted. Docking studies predicted a better binding conformation for (-)-R,S-dehydroemetine compared to (-)-S,S-dehydroisoemetine. The two diastereomers of dehydroemetine were synthesised and tested against the multi-drug resistant K1 strain of Plasmodium falciparum. It was found that of the two diastereomers, (-)-R,S-dehydroemetine (IC50: 69.58 + 2.62 nM) shared almost similar potency with emetine (47 + 2.17 nM), whereas (-)-S,S-dehydroisoemetine (IC50: 1.85 + 0.2 µM) was much less active as predicted through modelling studies. Stage specificity studies of both compounds showed more activity against the trophozoite erythrocytic stages of the parasite. The IC50 speed assays carried out to verify onset of antimalarial activity in emetine and the dehydroemetine diastereomers, found that although the compounds started activity within 24 hours, the IC50 values were achieved only at 48 hours. Cell cytotoxicity assays showed both compounds to be potent against HepG2 hepatic cancer cell lines seeded at 5000 cells/well with low selectivity indices (SI = ~3 and ~1 for (-)-R,S-dehydroemetine and (-)-S,S-dehydroisoemetine). To achieve further dose reduction to minimise toxicity, CalcuSyn-based drug interactivity assays were optimised and validated on known synergistic antimalarial drug combinations (atovaquone/proguanil). Drug interaction analysis performed using the validated CalcuSyn method found the antimalarial drug atovaquone to be synergistic with (-)-R,S-dehydroemetine at IC50, IC75 and IC90 values (Combination indices (CI): 0.88-0.89). Proguanil displayed moderate synergy (CI: 0.67), additivity (CI: 1.04) and antagonism (CI: 1.6) with (-)-R,S-dehydroemetine at IC50, IC75 and IC90 combinatory doses, respectively. Combination studies with artemether (CI: 1.6) and doxycycline (CI: 1.5) were found to be antagonistic. Emetine and (-)-R,S-dehydroemetine were found to display gametocidal activity against both male and female gametes with no cross-resistance observed on 3D7A, Dd2 and W2 strains. Further experiments were conducted to investigate the possibility of a multimodal mechanism of action for emetine and its analogues. ATP proliferation experiments on parasite-infected erythrocytes showed emetine and (-)-R,S-dehydroemetine caused a reduction in cellular ATP concentrations. Emetine and (-)-R,S-dehydroemetine also resulted in the disruption of the parasite mitochondrial membrane potential in infected erythrocytes in a manner similar to the slow-acting antimalarial atovaquone. The results support multimodal mechanisms of action in addition to inhibition of protein translation.Experiments were also conducted to determine the cellular basis of cardiotoxicity previously reported in emetine and dehydroemetine therapy for amoebiasis. hERG K+ channel blocking assays showed that emetine and (-)-R,S-dehydroemetine did not result in inhibition (Selectivity indices > 285), indicating a differential mechanism of action for cardiotoxicity. To explore this further, experiments were done to evaluate the effect of emetine and (-)-R,S-dehydroemetine on calcium regulatory mechanisms on sheep ventricular myocytes. Both emetine and (-)-R,S-dehydroemetine displayed potentiation of ryanodine receptors, lowering the threshold for generation of calcium waves with consequent increased susceptibility for arrhythmias. (-)-R,S-dehydroemetine however, also displayed negative inotropic effects in comparison to emetine, thus making the compound potentially less arrhythmogenic and supporting early clinical observations reported in the 1960s during anti-amoebic therapy