Novel Endoperoxide-Based Transmission-Blocking Antimalarials with Liver- and Blood-Schizontocidal Activities

Malaria is a potentially life-threatening disease caused by infection with parasites of the genus Plasmodium and transmitted to humans through the bite of infected female Anopheles mosquitoes.¹ Among the five Plasmodium species that commonly infect humans, P. falciparum is responsible for most of the mortality worldwide.² Parasite resistance to antimalarial drugs remains a real and ever-present danger. For this reason, the WHO recommends that P. falciparum malaria should be treated with artemisinin (ART, 1)-based combination therapies (ACT), in which the ART-based component is combined with a second, longer-acting agent.^1,3 Artemisinin and its derivatives are potent blood schizontocides, acting rapidly against parasitic forms that invade erythrocytes and cause disease symptoms.⁴

The ultimate goal of eradicating malaria will benefit greatly from a drug that eliminates all life cycle stages of parasites.⁵ Malaria parasites undergo an asymptomatic, obligatory developmental phase in the liver, which precedes the formation of red blood cell-infective forms.⁶ Thus, the liver stage of infection offers important potential for disease prevention, as intervention at this stage acts before the onset of symptoms, providing a true causal prophylactic strategy.⁷ In addition, P. vivax, the second most prevalent species causing human malaria, and P. ovale infections can generate cryptic parasite forms called hypnozoites that persist in the liver for long periods of time and that, upon reactivation, are responsible for relapses of malaria.⁸ Thus, antiliver stage drugs would also be beneficial for a malaria eradication campaign through elimination of the long-lived hypnozoites of P. vivax and P. ovale in the liver.^8,9 Primaquine (2, PQ, Chart 1) is the only drug currently used for the radical cure of P. vivax and P. ovale malaria and is active against the transient liver forms of all Plasmodium species. Moreover, PQ is also used as a gametocytocidal, i.e., it is active against the blood-circulating sexual forms of the parasite that are transmitted to the mosquito upon a blood meal, and in this way, it is able to block the transmission of infection from the human host to mosquito vectors.^10,11 The liver and sporogonic stages of malaria parasites have remained largely underexploited as antimalarial targets due to the poorly understood biology of these life-cycle stages and the inherent technical difficulties in studying them.^7,11 Only recently have systematic efforts toward the identification of novel liver schizontocidal and transmission blocking scaffolds been reported.^4,12−14

Endoperoxide-based hybrid compounds represent an attractive alternative to ACTs.¹⁵⁻¹⁹ ART contains a 1,2,4-trioxane core that is reductively activated by iron(II) heme, a byproduct of host hemoglobin degradation, to form carbon-centered radicals capable of reacting with heme and proteins.²⁰ An alternative model for the antimalarial mechanism of endoperoxides has been put forward by Haynes and Monti whereby endoperoxides mediate their antimalarial activity through interaction with cofactors. The tetraoxanes reported here are also likely to be capable of oxidizing cofactors such as FADH₂ and differences in activity between trioxanes and tetraoxanes may reflect the different oxidizing capacities of the two heterocycles.^21,22

We recently reported, for the first time, the ability of PQ-ART hybrid molecules, e.g., 3, to impair the liver and erythrocyte stages of Plasmodium, a result that paves the way for the exploitation of this approach for malaria control and eradication.²³ We now extend this hybrid-based strategy to fully synthetic 1,2,4,5-tetraoxanes, a class of peroxides with potent blood schizontocidal activity.²⁴ The aims of this study were (i) to evaluate the efficacy of hybrid compounds 5, 8, 10, and 12 (Scheme 1) against Plasmodium liver and erythrocyte stages and compare their activities to those of their 1,2,4-trioxane counterparts 16 and 18 (Schemes 2 and 3), and (ii) to determine their potential as transmission-blocking agents.

The preparation of hybrid compounds 5, 8, 10, and 12 is outlined in Scheme 1. Compounds 5 and 10, containing an amide linker between the two pharmacophoric moieties, were synthesized by reacting tetraoxanes 4 and 9 with PQ, using TBTU and methyl chloroformate as coupling agents, respectively. Tetraoxanes 4 and 9 as starting materials were prepared via a rapid three-step synthesis that was previously reported.²⁴⁻²⁶

The synthesis of hybrid 8, the amine counterpart of 5, started with the conversion of tetraoxane 4 to the Weinreb amide 6, which was then reduced to the corresponding aldehyde 7 with LiAlH₄.²⁷⁻²⁹ Reductive amination of 7 with PQ and NaBH(AcO)₃ gave compound 5 in moderate yield. Hybrid 12 was synthesized by reductive amination of tetraoxane 11 with PQ and NaBH(AcO)₃.³⁰ The 1,2,4-trioxane-based hybrid 16, the amine counterpart of the previously reported amide 17, was prepared as outlined in Scheme 2.²³ The synthetic pathway started with ART, which was converted to 10β-carboxymethyl-10-desoxy-dihydroartemisinin 13.^31,32 Following the same procedure used for 6, compound 13 was converted to the Weinreb amide 14 and then reduced to the corresponding aldehyde 15 with LiAlH₄. Reductive amination of 15 with PQ in acetic acid gave compound 16 in moderate yield. Finally, hybrid 18, the amide counterpart of 3, was prepared by reacting artelinic acid with PQ and TBTU.

Tetraoxanes 5, 8, 10, and 12 and semisynthetic trioxanes 16 and 18 were first screened for activity against the erythrocyte-stage of chloroquine-resistant W2-strain P. falciparum (Table 1). Compounds 5, 8, 10, and 12 inhibited the growth of parasites with IC₅₀ values ranging from 21 to 45 nM, suggesting that the nature of the linker between the tetraoxane and PQ moieties does not significantly affect antiplasmodial activity. Trioxanes 16 and 18 were slightly more potent than tetraoxanes as antiplasmodial agents, with IC₅₀ values of 9 and 5 nM, respectively (Table 1). These values are of the same order of magnitude as those reported for their counterparts 3, 17, and ART. As expected, PQ showed only modest activity in this assay.

To evaluate the abilities of hybrids 5, 8, 10, 12, 16, and 18 to inhibit infection of liver cells by malaria parasites, the compounds were tested using an in vitro infection model that employs a human hepatoma cell line (Huh7) and the rodent malaria parasite P. berghei.⁷ Parasite load was assessed by bioluminescence measurements following infection with luciferase-expressing parasites, as previously described.³³ The results were compared with those obtained for a tetraoxane lacking the 8-aminoquinoline moiety, the ethyl ester of 4 (Table 1). All hybrids showed high potency against the liver forms of the parasite, with most of the compounds displaying IC₅₀ values in the low to mid nM range. In contrast, the parent tetraoxane showed to be inactive at 10 μM, suggesting that this scaffold is not intrinsically effective against the liver stages of the parasite. Furthermore, hybrids were significantly more potent than PQ and the 1:1 PQ–ART mixture. This is consistent with what has been previously reported for compounds 3 and 17.²³ Additionally, none of the compounds significantly affected Huh7 cell proliferation, indicating that tetraoxane–PQ hybrids are selective and nontoxic antimalarial agents (Table 1).

Since PQ is known to undergo extensive oxidative deamination at the primary amine, the metabolic stability of tetraoxane-based hybrids 5 and 8 was evaluated in rat liver microsomes, at 37 °C. From the results presented in Table 2 it is possible to conclude that tetraoxane 5, containing an amide linker between the two pharmacophoric moieties, displayed a high rate of metabolism in rat liver microsomes and predicted in vivo hepatic extraction ratio, E_H. In contrast, its amine counterpart 8 presented an intermediate E_H values, suggesting that metabolic susceptibility of tetraoxane hybrids are affected by the nature of the linker between the two moieties. Neither primaquine nor its oxidative deamination product, carboxyprimaquine,¹⁰ were detected in the incubation mixtures, suggesting that other metabolic pathways might be operating for 5 and 8. In addition, these hybrids did not degrade when incubated in 80% human plasma for 3 days.

Having determined the activity profile of tetraoxane-based hybrids against the blood- and liver-stage of infection, we then evaluated the in vivo antimalarial efficacy of compound 5 using GFP-expressing P. berghei ANKA-infected male C57Bl/6J mice. This compound was administered i.p. at 30 mg/kg dose once a day for 5 days. Parasitemias were monitored daily by microscopy and flow cytometry. Remarkably, compound 5 completely and irreversibly cleared the parasitemia by day 8 postinfection, and all treated animals survived until the end of the experiment, whereas all control mice succumbed with signs of experimental cerebral malaria at day 6 postinfection (Figure 1), clearly indicating that tetraoxane-based hybrids have curative value. Furthermore, no adverse reactions were observed following administration of 5 at the dosage regimen used in this study.

The potential of tetraoxane hybrids to inhibit the sporogonic cycle of the parasite within the mosquito was also evaluated using an established in vivo infection model consisting of BALB/c mice infected with P. berghei ANKA-GFP and Anopheles gambiae mosquitoes.^34,35 In this model, mice infected with parasitised erythrocytes were treated by a single i.p. injection of each compound at two dose levels (10 and 25 μmol·kg^–1). Two hours after administration, glucose-starved mosquitoes were allowed to feed on the anesthetized mice. The engorged mosquitoes were maintained at 19 ± 1 °C, for 10 days and then collected and dissected for microscopy detection of oocysts in midguts. The criteria used to assess the antimalarial activity of each compound were the percentage of mosquitoes with oocysts and the mean number of oocysts per infected mosquito, when compared to nontreated controls. Inspection of data presented in Table 3 shows that compounds 5, 8, 10, and 12 affected the development of the sporogonic cycle of P. berghei in A. gambiae mosquitoes at the dose levels of 10 and 25 μmol·kg^–1. In particular, compound 5 was superior to PQ, and completely inhibited the appearance of oocysts in mosquito midguts when administered at 25 μmol·kg^–1. Compounds 8, 10, and 12 also markedly decreased the percentage of infected mosquitoes as well as the number of oocysts at the highest dose tested. Although it should be noted that this in vivo model cannot specifically ascribe the drug effect to either gametocytocidal or sporontocidal activity, the results clearly show that hybrids 5, 8, 10, and 12 are effective at interrupting the transmission of malaria parasites to mosquitoes.

In conclusion, a new class of hybrid compounds combining endoperoxides and 8-aminoquinoline pharmacophores was developed to target both blood- and liver-stages of parasites and to block the transmission of infection to mosquito vectors. The in vitro antimalarial profile of these compounds reveals that they display potent inhibitory activity against the blood stage of infection, with IC₅₀ values in the nanomolar range, a level of activity similar to that of artemisinin-based hybrid compounds. Screening against P. berghei liver stage of infection revealed that both the tetraoxane- and trioxane-based series are potent inhibitors of the exoerythrocytic forms of the parasite, with IC₅₀ values in the submicromolar range, being superior to a 1:1 PQ–ART mixture.²³ In vivo studies revealed that compound 5 irreversibly cleared the parasitemia from infected mice, while screening of transmission-blocking activity showed that the tetraoxanes show potent activity in reducing the percentage of infected mosquitoes and the mean number of oocysts per mosquito. These results indicate that these hybrids are excellent starting points to develop agents with the potential to be used in malaria eradication campaigns since they display all the desired antimalarial multistage activities.