A Profound Membrane Reorganization Defines Susceptibility of Plasmodium falciparum Infected Red Blood Cells to Lysis by Granulysin and Perforin

Abstract

Malaria remains one of the most serious health problems in developing countries. The causative agent of malaria, Plasmodium spp., have a complex life cycle involving multiple developmental stages as well as different morphological, biochemical and metabolic requirements. We recently found that γδ T cells control parasite growth using pore-forming proteins to deliver their cytotoxic proteases, the granzymes, into blood residing parasites. Here, we follow up on the molecular mechanisms of parasite growth inhibition by human pore-forming proteins. We confirm that Plasmodium falciparum infection efficiently depletes the red blood cells of cholesterol, which renders the parasite surrounding membranes susceptible to lysis by prokaryotic membrane disrupting proteins, such as lymphocytic granulysin or the human cathelicidin LL-37. Interestingly, not the cholesterol depletion but rather the simultaneous exposure of phosphatidylserine, a negatively charged phospholipid, triggers resistance of late stage parasitized red blood cells towards the eukaryotic pore forming protein perforin. Overall, by revealing the molecular events we establish here a pathogen-host interaction that involves host cell membrane remodeling that defines the susceptibility towards cytolytic molecules.

Keywords: blood-stage malaria; cholesterol; granulysin; perforin; phosphatidylserine (PS); plasma membrane; pore forming proteins (PFPs).

Figure 1

Plasmodium falciparum infections render parasitized RBCs susceptible to GNLY while they become simultaneously resistant to PFN lysis. (A, B), non-synchronized iRBCs cultures were treated with Alexa Fluor 488-labeled GzmB (0.4 μM) in combination with GNLY (0.4 μM) or with PFN (500 mU/μl) for 20 minutes before fixation and assessment by confocal microscopy. Nuclei were stained with Hoechst and are indicated in blue. Late stage parasites are indicated with white arrows and early stage parasites with gray arrows. Scale bars are 10μm in (A) and 2μm in (B) MACS-purified (late stages) and flow-through cells (uninfected and early stages) were treated with indicated concentrations of GNLY and PFN for 15 minutes at 37°C before centrifugation to separate the cellular pellet from the supernatant. The supernatants were subjected to optical densities measurement at a wavelength of 405 nm to assess hemoglobin release (C) and to immunoblot analysis to monitor release of plasmodial LDH and RBC GAPDH (D).

Figure 2

Cholesterol is depleted of late-stage iRBCs. Non-synchronized iRBCs were treated with Alexa Fluor 488-labeled streptolysin O (SLO-AF488; 2 U/μl) for 60 minutes on ice before fixation and assessment by confocal microscopy. Nuclei were stained with Hoechst. Scale bars are 10μm. Four representative images from three independent experiments are shown in (A). The white arrows indicate the late stage iRBC (not stained with the SLO-AF488) and the gray arrows show the early stage iRBC (stained with the SLO-AF488). The quantification (n>40 for each stage form in every independent experiment) of the three independent experiments is presented in (B). (C) iRBCs were MACS purified to enrich for late stages before filipin staining and fluorescence measurement by plate reader and compared to MACS flow-through cells (early stages and uninfected). Averages +/- SEM of three independent experiments are shown. P values of differences, calculated by Student`s t test, are indicated.

Figure 3

Cholesterol depletion increases susceptibility of RBC membranes to GNLY and to PFN lysis. RBCs were pre-treated with 7.5 μM methyl-β-cyclodextrin for 2 hours at 37°C to deplete cholesterol. Depleted cells were then treated with indicated concentrations of GNLY (A), LL-37 (B), PFN (C) or SLO (D) for 15 minutes at 37°C before hemoglobin release was assessed by plate reader at 405 nm. Averages +/- SEM of specific lysis in three independent experiments are shown. P values of differences between treatment conditions, calculated by Student`s t test, are indicated.

Figure 4

Plasmodium falciparum infection triggers exposure of phosphatidylserine in late stage iRBCs. Uninfected erythrocytes, mixed stages, 75% Percoll enriched trophozoites and 65% Percoll enriched schizonts were incubated with annexin V-FITC and Draq5 for 15min at RT and subjected to flow cytometry analysis. A representative FACS dot plot of annexin V versus Draq5 staining is shown in (A). Averages and values of single experiments represented by dots in different colors indicating frequencies of Draq5+/Annexin V+ are depicted in (B) and indicating geometric mean of annexin V in Draq5 + are presented in (C). Mixed (D) or MACS-purified late stage iRBCs (E) were stained with Annexin V-FITC for 60 minutes on ice before fixation and analysis by confocal microscopy. Nuclei were stained with Hoechst. Representative images of three independent experiments are shown in (D, E). In (D), scale bars are 10μm. The quantification of the microscopy results and presentation as averages +/- SEM is shown in (F). P values of differences between stages were calculated with the unpaired t-test (B, F) or Mann-Whitney test (C). Independent experiments are represented in different colors and replicates by symbol in (B, C).

Figure 5

Exposure of phosphatidylserine in the outer RBCs plasma membrane leaflet mediates resistance to PFN lysis. RBCs were pre-treated with 1 μM calimycin +/- 2 μM calcium for 3 hours at 37°C, then washed with PBS and stained with Annexin V-FITC for 1 hour on ice for confocal microscopy (A) or treated with indicated concentrations of PFN (D), SLO (E) or GNLY (F) for 15 minutes at 37°C before hemoglobin release assay. Averages +/- SEM of the fluorescence of annexin V stained cells assessed by plate reader is demonstrated in (B). Averages and values of single experiments represented by dots in different colors indicating the frequency of annexin positive cells assessed by flow cytometry are shown in (C). In (A), scale bars are 10μm. P values of significant differences were calculated with unpaired t-test. wo, without.