Malaria parasites undergo a rapid and extensive metamorphosis after invasion of the host erythrocyte

Abstract

Within the human host, the symptoms of malaria are caused by the replication of malaria parasites within erythrocytes. Growth inside the erythrocyte exposes the parasites to the normal surveillance of erythrocytes by the host organism, in particular the clearance of erythrocytes in the spleen. Here we show that the malaria parasite Plasmodium falciparum undergoes a rapid, multi-step metamorphosis that transforms the invasive merozoite into an amoeboid-shaped cell within minutes after invading erythrocytes. This transformation involves an increase in the parasite surface area and is mediated by factors already present in the merozoite, including the parasite phospholipid transfer protein PV6. Parasites lacking PV6 do not assume an amoeboid form and instead are spherical and have a smaller surface area than amoeboid forms. Furthermore, erythrocytes infected with P. falciparum parasites lacking PV6 undergo a higher loss of surface area upon infection, which affects the traversal of infected erythrocytes through the spleen. This is the first evidence that after invasion, the parasite undergoes a rapid, complex metamorphosis within the host erythrocyte that promotes survival in the host.

Keywords: Plasmodium; Host-pathogen Interaction; Malaria; Membranes.

Figure 1. Amoeboid formation of Plasmodium parasites during the early erythrocytic cycle.

(A) Morphologies of P. falciparum parasites that were scored as amoeboid forms. (B) Timing of amoeboid formation after invasion. Parasites were fixed at the indicated time after removal of ML10 and amoeboid formation was subsequently scored by microscopy. Results shown are the combination of two biological replicates; error bars: +/−SD. Panels on the right are examples of amoeboid forms detected at the indicated time. (C) Live-cell imaging of amoeboid formation. The frame rate was one frame/second, time starting from invasion is indicated in each panel. The arrows indicate the parasite. (D) Timing of amoeboid formation in 19 parasites observed from the moment of invasion by video microscopy; error bars: +/−SEM. (E, F) Amoeboid formation of Plasmodium falciparum parasites after invasion in the presence of either DMSO in cRPMI (DMSO), Hanks’ Balanced Salt Solution (HBSS) or 5 µM cycloheximide (Cyclo) in cRPMI 2 h after removal of ML10 and Plasmodium knowlesi 1–3 h after removal of ML10; error bars: +/−SD. (E) Representative images of live Plasmodium falciparum parasites. The (*) indicates an oddly shaped elongated parasite that was occasionally detected in the cycloheximide-treated samples. (F) Amoeboid formation in the indicated condition. Data represent three biological replicates of a minimum of 100 parasites (P. falciparum) and two biological replicates of Plasmodium knowlesi of at least 50 parasites (two-tailed t-test: ns: not significant; **P < 0.01 (P = 0.0076), error bars: +/−SD. (G) Live-cell fluorescence imaging of Plasmodium knowlesi-infected erythrocytes stained with Bodipy C5-ceramide 2 h after removal of egress inhibitor. All scale bars represent 5 µm. Source data are available online for this figure.

Figure 2. Development of wild-type parasites and parasites lacking PV6 at early time points post-invasion.

(A) Giemsa-stained smears of PV6 DiCre (PfBLD529) parasites treated with DMSO or rapamycin (Rapa). Development of synchronized parasites was examined in the cycle following DMSO or rapamycin treatment at the indicated time after removal of egress inhibitor (see EV1 for outline of experiment). Scale bar represents 5 µm. (B) Size (area) of DMSO and rapamycin-treated PV6-diCre parasites. Data represent three biological replicates with a minimum of 20 parasites per sample. Box represents the interquartile range and the horizontal line within the box represents the median. Whiskers indicate range from minimum to maximum. Significance was determined using Kruskal-Wallis test: ns: not significant; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. For additional statistical analyses and exact p values, please see Appendix Table S1. (C) Live-cell imaging of PV6-DiCre (PfBLD529) parasites treated with DMSO or rapamycin in the previous cycle or with jasplakinolide for 1 h prior to imaging. Parasites were imaged 2 h after removal of ML10. Scale bar represents 5 µm. (D) Quantification of the size (area) of the parasites in the experiment presented in (C). Data are based on measurement of at least 20 rings in each of the three biological replicates performed. Rapa: rapamycin-treated; Jasp: jasplakinolide-treated (Kruskal-Wallis test: ns: not significant; ****P < 0.0001; ns; not significant; error bars: +/−SD). (E) Analysis of the circularity of the parasites in the experiment presented in (C). Rapa: rapamycin-treated; Jasp: jasplakinolide-treated (Kruskal-Wallis test: ns: not significant; ****P < 0.0001; error bars: +/−SD). Source data are available online for this figure.

Figure 3. Amoeboid formation is accompanied by a large increase in surface area and requires PV6.

(A) SBF-SEM slices (sections) and surface renderings of PV6-DiCre (PfBLD529) parasites treated with DMSO and rapamycin 20 min and 2 h after removal of egress inhibitor, illustrating the shape and positioning of the parasite (cyan) within the erythrocyte (red). Indicated in the panels is the distance between the sections. The panel second from right shows a 3D model of the parasite shown in the sections on the left-hand side. On the far right is shown the parasite modelled inside the erythrocyte. Scale bar represents 500 nm. (B) Modelling of schizont and individual merozoite. (C) Surface area of merozoites (Mero) and DMSO-treated (D) and rapamycin-treated (R) parasites at the indicated time after removal of egress inhibitor. Data represent the measurement of at least 15 parasites. The Mann–Whitney U test was performed for statistical analysis: ns-not significant; ***P < 0.001 (P = 0.0004 for comparison DMSO 20 min-DMSO 2 h and P = 0.0001 for comparison DMSO 2h-RAPA 2 h); ****P < 0.0001, Error bars +/− SD. See Fig. EV2 for additional views of these parasites and Fig. EV3 for additional models. (D) Volume of merozoites (Mero) and DMSO-treated (D) and rapamycin-treated (R) parasites at the indicated time after removal of egress inhibitor. Data represent the measurement of at least 15 parasites. The Mann–Whitney U test was performed for statistical analysis: ns-not significant; **P < 0.01 ((P = 0.0034 for comparison DMSO 20 min-DMSO 2 h and P = 0.0051 for comparison DMSO 2h-RAPA 2 h); error bars +/− SD. See Fig. EV2 for additional views of these parasites and Fig. EV3 for additional models. Source data are available online for this figure.

Figure 4. Connections between limbs of the amoeboid-shaped parasites are very narrow and can separate the nucleus into lobes.

(A) Consecutive SBF-SEM sections, showing three limbs of a parasite and the thin connections between them. Numbers in upper left-hand corner of the panels indicate distance between the sections, N indicates the nucleus and the scale bar represents 500 nm. (B) Three-dimensional models of infected erythrocytes (red) 2 h after removal of egress inhibitor, showing the nucleus (dark blue) in a wild-type (DMSO) parasite and a parasite lacking PV6 (RAPA); parasites are shown in cyan. Scale bar represents 500 nm. (C) Live-cell imaging of an infected erythrocyte stained with ceramide and the DNA dye SiR-Hoechst. Infected erythrocytes were imaged 2 h after the removal of egress inhibitor to reveal the ceramide and DNA staining. Also shown is the cell imaged using DIC. The imperfect alignment of the ceramide and SiR-Hoechst staining is the result of the movement of the parasite during imaging. Scale bar represents 5 µm. Source data are available online for this figure.

Figure 5. Phenotype of P. falciparum parasites lacking PV6.

(A) SBF-SEM sections from DMSO-treated (upper panel) and rapamycin-treated (lower panel) PV6 DiCre parasite culture, showing the accumulation of membrane whorls around parasites lacking PV6. Arrows indicate some of the lipid whorls. See Fig. EV4A for additional images. (B) Live-cell fluorescence imaging of parasites expressing an mNeonGreen-HA₃-PV6 fusion (PfBLD717) one hour after removal of ML10. Membranes were labelled using Bodipy C5-ceramide (orange). (C) Immunofluorescence staining of DMSO-treated and rapamycin-treated PV6-DiCre parasites 6 h after removal of ML10 using anti-RESA (green) and anti-EXP2 (red) antibodies. DNA was stained with Hoechst 33342 (blue) and the erythrocyte plasma membrane was visualized with Wheat Germ Agglutinin (WGA)-Alexa Fluor 647 (white). In all panels, the scale bar represents 5 µm. Source data are available online for this figure.

Figure 6. Size and functional alteration of erythrocytes infected with parasites lacking PV6.

(A) Surface area of uninfected erythrocytes (uRBCs) and erythrocytes infected with PV6-diCre parasites (iRBCs) 30 min after removal of egress inhibitor. The surface area of uRBCs and iRBCs in culture of wild-type (DMSO-treated) parasites and parasites lacking PV6 (Rapamycin-treated) was determined using imaging flow cytometry (Appendix Fig. S4). (B) Quantification of the surface area loss of uninfected (uRBC) and infected (iRBC) erythrocytes from cultures of DMSO-treated and rapamycin-treated parasites 30 min after removal of ML10, representing four experiments. Data represent four biological replicates (T-test; **P ˂ 0.01 (P = 0.0078), error bars: +/− SEM). (C) Retention of erythrocytes infected with wild-type (DMSO-treated – D) parasites and parasites lacking PV6 (Rapamycin-treated – R) on a microsphiltration column. Data represent 4 biological replicates with a minimum of 6 technical replicates. Significance was determined using Mann–Whitney test; ***P ˂ 0.0001 (P = 0.0001), error bars: +/−SD. (D) Retention of erythrocytes infected with PV6-DiCre parasites treated with DMSO or rapamycin in an ex vivo human spleen. Blood containing erythrocytes infected with DMSO-treated (solid black line) and rapamycin-treated parasites (dashed black line) approximately 6 h after the removal of ML10 was continuously perfused through the spleen and the parasitemia was determined at the indicated times. Interpolation curves are shown in grey lines (R² = 0.84 for DMSO and R² = 0.90 for Rapa samples). Source data are available online for this figure.

Figure EV1. Phenotypic analyses of parasites lacking PV6.

(A) Schematic representation of the protocol used to analyse the phenotype of the P. falciparum PV6 DiCre parasites. Synchronized parasites were treated with a minimum of 10 nM rapamycin (or an equivalent volume of DMSO) for 1 h at the trophozoite stage (~30 h post-invasion) in the first cycle. Close to the end of the first cycle, an egress inhibitor (either Compound 2 or ML10) was added to arrest the parasites at a very late schizont stage. When most of the parasites appeared arrested, the egress inhibitor was removed to initiate a round of synchronized invasion, starting cycle 2, allowing for observation of the phenotype of PV6 over time. (B) Live-cell imaging of amoeboid formation. The frame rate was set at one frame/second, time starting from invasion is indicated in each panel. The arrows indicate the parasite. Scale bar represents 5 µm.

Figure EV2. Additional three-dimensional views of the models presented in Fig. 3.

Three-dimensional models from SBF-SEM data of the parasites presented in Fig. 3, illustrating the shape of the parasite (cyan) and its positioning within the erythrocyte (red) and additional SEM-SBF sections. The scale bars represent 500 nm.

Figure EV3. Three-dimensional models of PfBLD529 parasites treated with DMSO or rapamycin.

(A) These models were obtained in the same experiment as those presented in Fig. 3 and Fig. EV2. The invasion of these parasites was carefully synchronized using an egress inhibitor to allow the development of the parasites to be followed over time. Time indicated refers to the time after removal of egress inhibitor. The scale bars represent 500 nm. (B) Three-dimensional models from SBF-SEM data of DMSO-treated and rapamycin-treated parasites for which invasion was not synchronized. The parasites had been synchronized at the start of cycle 1 (Fig. EV1) and were allowed to progress to the next cycle without further synchronization. The scale bars represent 500 nm. (C) Surface area of the parasites shown in panel B; D-DMSO, R-rapamycin. Error bars +/− SD. Data represent the measurement of at least 7 parasites. The Mann–Whitney U test was performed for statistical analysis (***P < 0.001, P = 0.0002). (D) Volume of the parasites shown in (B). Error bars +/− SD. Data represent the measurement of at least 7 parasites. The Mann–Whitney U test was performed for statistical analysis. No significance difference was detected between the wild-type parasites and the parasites lacking PV6.

Figure EV4. Nuclear morphology in wild-type amoeboid-shaped parasites and parasites lacking PV6.

(A) Consecutive SBF-SEM sections, showing the thin connections between limbs of two different parasites. Numbers in the upper left-hand corner of the panels indicate distance between the sections and the scale bar represents 500 nm. (B) Three-dimensional models from SBF-SEM data of erythrocytes infected with DMSO-treated and rapamycin-treated parasites 2 h after removal of egress inhibitor. Erythrocytes (red), parasite (cyan) and the nucleus (blue) are highlighted. The scale bar represents 500 nm. (C) Live-cell fluorescence imaging of infected erythrocytes labelled with C5-Bodipy-ceramide (orange) and SiR-Hoechst (red) 2 h after removal of egress inhibitor. Note that in the merged imaged, the parasite and the SiR-Hoechst do not overlap perfectly owing to the movement of the parasite during the acquisition of the images. The scale bar represents 5 µm. (D) Three-dimensional model showing the nucleus (blue) in wild-type parasites (magenta) 20 min after removal of egress inhibitor. Scale bar represents 500 nm.

Figure EV5. Additional SBF-SEM images of wild-type parasites and parasites lacking PV6.

(A) Individual SBF-SEM sections of different erythrocytes infected with wild-type PfBLD529 parasites (left; DMSO) and PfBLD529 parasites lacking PV6 (right; Rapa). In each case, the left-hand panel shows a close view of the parasite and the right-hand panel shows the entire infected erythrocyte. Note the accumulation of membranous whorls next to the parasites lacking PV6. The close views of the parasites in the top row are also shown in Fig. 5. Scale bars represent 500 nm. (B) Analysis of export of RESA in erythrocytes infected with DMSO-treated and rapamycin-treated parasites. Samples prepared for IFA were imaged as described in the Methods section. The amount of RESA exported to the erythrocyte was measured for 18 parasites obtained from two independent experiments. Error bars +/− SD. The Mann–Whitney U test was performed for statistical analysis (***P < 0.001 (P = 0.0005)).

Figure EV6. Analysis of samples used for microsphiltration and ex vivo spleen perfusion.

(A) Retention of PV6-diCre parasites cultures treated with DMSO, rapamycin or jasplakinolide. Data represent 2 biological replicates with a minimum of 6 technical replicates. Error bars: +/− SD. The Mann–Whitney U test was performed for statistical analysis: ns, not significant; ****P < 0.0001. (B) Retention of untreated erythocytes (RBCs) and fixed erythrocytes (rigid RBCs) on the columns used for the microsphiltation analysis of erythrocyte infected with wild-type parasites or parasites lacking PV6. Data represents 1 biological replicate with 6 technical replicates. (C) Physiological state of ex vivo spleen during the course of the perfusion with infected erythrocytes. The pH and levels of bicarbonate, lactate and glucose were measured periodically to ensure the proper functioning and viability of the spleen. (D) Gating strategy used for cytometric analysis of blood passaged through ex vivo spleen.