Infection of erythrocytes with Plasmodium falciparum stimulates eryptosis, the suicidal death of erythrocytes 12. Eryptosis is characterized by cell membrane scrambling leading to phosphatidylserine exposure at the cell surface 34567. Triggers of cell membrane scrambling include increased cytosolic Ca²⁺ activity 3568 and ceramide 9. Ca²⁺ may enter erythrocytes through Ca²⁺-permeable cation channels, which could be activated by osmotic shock, oxidative stress or energy depletion 8101112. Ca²⁺ further activates Ca²⁺-sensitive K⁺ channels 1314, leading to exit of KCl and osmotically obliged water and thus to cell shrinkage 15. Plasmodium infection imposes oxidative stress onto host erythrocytes, which activates Ca²⁺-permeable cation channels 16 and, thus, fosters cell membrane scrambling and phosphatidylserine exposure at the erythrocyte surface 17. Sustained increase in cytosolic Ca²⁺ similarly stimulates apoptosis of nucleated cells 18. As phosphatidylserine-exposing cells are bound to receptors of macrophages 1920 and phagocytosed 2122, eryptotic cells are rapidly cleared from circulating blood 23.

During malaria, the clearance of infected erythrocytes prior to the development of trophozoites 24 may counteract the development of parasitaemia 25. Along those lines clearance of ring stage Plasmodium-infected erythrocytes is accelerated by sickle-cell trait, beta-thalassaemia-trait, homozygous Hb-C and G6PD-deficiency, genetic conditions associated with a relatively mild course of malaria 72627282930. Moreover, iron deficiency 1 and treatment with lead 2, chlorpromazine 31 and cyclosporine 32 delay the development of parasitaemia and thus foster the survival of Plasmodium berghei-infected mice, presumably at least in part by accelerating erythrocyte death. Erythropoietin, which inhibits the erythrocyte cation channel 33 has similarly been shown to influence the course of malaria 34. Erythropoietin may, however, be effective through mechanisms other than stimulation of eryptosis, which is rather inhibited by the hormone 33.

Azathioprine, a widely used immunosuppressive drug 35363738, has recently been shown to similarly trigger eryptosis 39. The present study explored whether azathioprine accelerates eryptosis of P. falciparum-infected erythrocytes and whether it influences parasitaemia and survival during malaria. Azathioprine (6-mercaptopurine) has previously been shown to inhibit a purine phosphoribosyltransferase of the parasite and thus to interfere with in vitro growth of the parasite 4041. An effect on the survival of infected erythrocytes or in vivo efficacy has, however, not been reported.

To study the in vitro growth of the parasite, P. falciparum-infected erythrocytes were cultured in healthy human erythrocytes and synchronized to ring stage by sorbitol treatment. The initial parasitaemia was 1.3%. Within 48 hours of culture, i.e., after intraerythrocytic amplification, evasion from the host cell, and invasion into new erythrocytes, some 16% of the erythrocytes were infected, while 84% of the erythrocytes remained noninfected (Figure 1A). The percentage of parasitized erythrocytes was decreased by the presence of azathioprine, an effect reaching statistical significance at ≥ 1 μM azathioprine concentration (Figure 1A). Similarly, the intraerythrocytic DNA amplification of the parasite was decreased in the presence of azathioprine, an effect reaching statistical significance at ≥ 1 μM azathioprine concentration (Figure 1B). Together, the data indicate that azathioprine exerts direct effects on the parasite at concentrations ≥ 1 μM.

To explore whether infection of erythrocytes triggers eryptosis, phosphatidylserine-exposing erythrocytes were identified by determination of annexin V-binding in FACS analysis. Prior to infection, the percentage of annexin V-binding erythrocytes was low (1.25 ± 0.20%, n = 6). Infection within 24 hours led to a marked increase in annexin V-binding of both, infected erythrocytes and noninfected bystander cells (Figure 2). The percentage of annexin V-binding was more than double as high in infected than in noninfected erythrocytes (Figure 2), a difference statistically significant both, in the absence and presence of azathioprine. The phosphatidylserine exposure of infected erythrocytes was significantly augmented by azathioprine (Figure 2), an effect observed at 1 μM azathioprine.

Depending on the stage of the parasite development, infection of erythrocytes decreased (early stages; Figure 3A) or increased (late stages; Figure 3B) erythrocyte forward scatter, indicating that early stages initially decreased the host cell volume. Subsequently, during later parasite development, the volume-expanding trophozoites increased the host cell volume. Azathioprine at concentrations of 5 and 10 μM decreased the forward scatter of late stage infected erythrocytes, which was probably due to azathioprine-induced inhibition of intraerythrocytic parasite development (see Figure 1B). In the early stage of infection, however, a statistically significant shrinking effect of azathioprine on infected cells was evident at lower concentrations of azathioprine (≥ 0.1 μM). In summary, these experiments indicate that low concentrations of azathioprine augment eryptosis of the host erythrocyte.

In a last series of experiments, mice were infected with P. berghei to determine the in vivo efficacy of azathioprine treatment. The administration of azathioprine (daily injections of 5 mg/kg b.w. azathioprine subcutaneously) was initiated 8 days after infection. At this time, parasitaemia was less than 5% (Figure 4B). The percentage of infected erythrocytes gradually increased in both, treated and untreated mice. The percentage of parasitized erythrocytes was lower in azathioprine-treated animals than in animals without azathioprine treatment, an effect reaching statistical significance between day 17 and day 20 of infection (Figures 4A and 4B). Accordingly, azathioprine treatment at least transiently decreased parasitaemia (Figure 4A, right panels and Figure 4B).

Azathioprine treatment further affected the survival of P. berghei-infected mice. As illustrated in Figure 4C, all untreated animals died within 22 days after the infection. In contrast, 77% of the azathioprine-treated animals survived the infection for more than 22 days.

The present study unravels a novel effect of azathioprine, i.e. the favorable influence on the course of malaria. Most importantly, azathioprine treatment significantly enhances the percentage of surviving animals after infection with P. berghei. As shown previously, without treatment, the infection of mice with P. berghei is followed by an invariably lethal course of malaria within 22 days 46. In contrast, most of the mice treated with azathioprine survived the infection for 22 days.

Several mechanisms may contribute to the efficacy of azathioprine. In theory, the effect of azathioprine could have been due to its immune-suppressing potency 35363738. However, it is not likely that immunosuppression achieves both, a significant reduction of parasitaemia and a milder course of the disease.

Azathioprine could further affect parasitaemia and host survival by directly affecting the survival and replication of the pathogen or its ability to evade parasitized erythrocytes and to invade noninfected erythrocytes. Indeed, higher concentrations of azathioprine decreased in vitro parasitaemia and DNA/RNA content of parasitized erythrocytes.

The effect of azathioprine could further be secondary to its ability to stimulate suicidal death of erythrocytes 39, an effect, which could contribute to or even account for the blunted parasitaemia and the survival of the infected mice. The drug could be effective by accelerated clearance of infected erythrocytes due to eryptosis. Moreover, the enhanced eryptosis may promote the release of pro-inflammatory cytokines from activated macrophages, thereby resulting in the activation of the hormonal stress response 48.

Phosphatidylserine-exposing erythrocytes are engulfed by macrophages 2122 and thus cleared from circulating blood 23. A wide variety of further endogenous mediators and xenobiotics trigger eryptosis, including haemolysin Kanagawa 49, listeriolysin 50, PGE₂ 51, Bay-5884 52, platelet activating factor 53, chlorpromazine 54, anandamide 55, methylglyoxal 56, paclitaxel 57, curcumin 58 amyloid peptides 59, valinomycin 60, aluminium 61, lead 62, mercury 63 and copper ions 64. Moreover, eryptosis is enhanced in a variety of clinical conditions including iron deficiency 23, sickle-cell anaemia 6566, beta-thalassaemia 7, glucose-6-phosphate dehydrogenase (G6PD)-deficiency 7, phosphate depletion 67, Haemolytic Uremic Syndrome 68, sepsis 69, malaria 25 and Wilson disease 64. Several of those diseases and xenobiotics have already been shown to favorably influence the course of malaria, including sickle-cell trait, beta-thalassaemia-trait, homozygous Hb-C and G6PD-deficiency 72627282930, iron deficiency 1, lead 2, chlorpromazine 31 and cyclosporine 32. Azathioprine may be a particularly attractive substance for the treatment of malaria because it is clinically widely used and thus ample knowledge has been accumulated about its side effects. Nevertheless, further eryptosis-inducing substances may be shown in near future to be effective as antimalarial drugs.

In conclusion, azathioprine accelerates eryptosis of Plasmodium-infected erythrocytes The effect contributes to or even accounts for the favourable effect of azathioprine on parasitaemia and survival of the host during malaria.

ANOVA: Analysis of variance; APC: Allophycocyanin; DNA: Desoxyribonucleic acid; FACS: Fluorescence activated cell sorting; FL: fluorescence channel; G6PD: Glucose 6 phosphate dehydrogenase; HEPES: N-2-hydroxyethylpiperazine-N-2-ethanesulfonic acid; Hb-C: Haemoglobin C; P: Plasmodium; RNA: Ribonucleic acid.

The authors declare that they have no competing interests.

DB performed the in vitro experiments, SK performed the in vivo experiments, CG performed FACS analysis, MF participated in the design of the study and the FACS analysis, evaluated the results and made the illustrations. SMH participated and supervised the in vitro and in vivo experiments, FL designed the study and drafted the manuscript. All authors read and approved the final manuscript.