We have developed a model for designing antimalarial drugs based on interference with an essential metabolism developed by Plasmodium during its intraerythrocytic cycle, phospholipid (PL) metabolism. The most promising drug interference is choline transporter blockage, which provides Plasmodium with a supply of precursor for synthesis of phosphatidylcholine (PC), the major PL of infected erythrocytes. Choline entry is a limiting step in this metabolic pathway and occurs by a facilitated-diffusion system involving an asymmetric carrier operating according to a cyclic model. Choline transport in the erythrocytes is not sodium dependent nor stereospecific as demonstrated using stereoisomers of alpha and beta methylcholine. These last two characteristics along with distinct effects of nitrogen substitution on transport rate demonstrate that choline transport in the infected erythrocyte possesses characteristics quite distinct from that of the nervous system. This indicates a possible discrimination between the antimalarial activity (inhibition of choline transport in the infected erythrocyte) and a possible toxic effect through inhibition of choline entry in synaptosomes. Apart from the de novo pathway of choline, PC can be synthesized by N-methylation from phosphatidylethanolamine (PE). There is a de novo pathway for PE biosynthesis from ethanolamine in infected cells but phosphatidylserine (PS) decarboxylation also occurs. In addition, PE can be directly and abundantly synthesized from serine decarboxylation into ethanolamine, a pathway which is absent from the host. The variety of the pathways that exist for the biosynthesis of one given PL led us to investigate whether an equilibrium can occur between all PL metabolic pathways. Indeed, if alternative (compensative) pathway(s) can operate after blockage of the de novo PC biosynthesis pathway this would indicate a potential mechanism for resistance acquisition. Up until now, there is no evidence of such a compensative process occurring in Plasmodium-infected erythrocytes under physiological conditions. Besides, the discovery of a highly parasite-specific pathway (serine decarboxylation and the presence of PS synthase) constitutes a very attractive and promising target, which could be attacked if resistances are built up against choline analogs. Indeed, potential inhibitions of the serine decarboxylase pathway could be very useful in acting instead of, or in surgery with, choline analogs.

The systematic screening of more than 250 molecules against Plasmodium falciparum in vitro has previously shown that interfering with phospholipid metabolism is lethal to the malaria parasite. These compounds act by impairing choline transport in infected erythrocytes, resulting in phosphatidylcholine de novo biosynthesis inhibition. A thorough study was carried out with the leader compound G25, whose in vitro IC50 is 0.6 nM. It was very specific to mature parasites (trophozoïtes) as determined in vitro with P. falciparum and in vivo with P. chabaudi -infected mice. This specificity corresponds to the most intense phase of phospholipid biosynthesis activity during the parasite cycle, thus corroborating the mechanism of action. The in vivo antimalarial activity (ED50) against P. chabaudi was 0.03 mg/kg, and a similar sensitivity was obtained with P. vinckei petteri, when the drug was intraperitoneally administered in a 4 day suppressive test. In contrast, P. berghei was revealed as less sensitive (3- to 20-fold, depending on the P. berghei-strain). This difference in activity could result either from the degree of synchronism of every strain, their invasion preference for mature or immature red blood cells or from an intrinsically lower sensitivity of the P. berghei strain to G25. Irrespective of the mode of administration, G25 had the same therapeutic index (lethal dose 50 (LD50)/ED50) but the dose to obtain antimalarial activity after oral treatment was 100-fold higher than after intraperitoneal (or subcutaneous) administration. This must be related to the low intestinal absorption of these kind of compounds. G25 succeeded to completely inhibiting parasitemia as high as 11.2% without any decrease in its therapeutic index when administered subcutaneously twice a day for at least 8 consecutive days to P. chabaudi -infected-rodent model. Transition to human preclinical investigations now requires a synthesis of molecules which would permit oral absorption.