Isolation of erythrocytes infected with viable early stages of Plasmodium falciparum by flow cytometry



The erythrocytic life cycle of Plasmodium falciparum is highly associated with severe clinical symptoms of malaria that causes hundreds of thousands of death each year. The parasite develops within human erythrocytes leading to the disruption of the infected red blood cell (iRBC) prior to the start of a new cycle of erythrocyte infection. Emerging mechanisms of resistance against antimalarial drugs require improved knowledge about parasite's blood stages to facilitate new alternative antimalarial strategies. For the analysis of young blood stages of Plasmodium at the molecular level, the isolation of ring stages is essential. However, early stages can hardly be separated from both, late stages and non-infected red blood cells using conventional methods. Here, iRBCs were stained with the DNA-binding dyes Vybrant® DyeCycle™ Violet and SYBR® Green I. Subsequently, cells were subjected to flow-cytometric analysis. This enabled the discrimination of early stage iRBCs as well as late-stage iRBCs from non-infected erythrocytes and the properties of the used dyes were evaluated. Moreover, early stage iRBCs were isolated with high purity (>98%) by FACS. Subsequently, development of sorted early stages of the parasite was monitored over time and compared with control cultures. The described flow cytometry method, based on staining with Vybrant DyeCycle Violet, allows the isolation of viable ring stages of the malarial agent P. falciparum, and thereby provides the basis for new, broad-range molecular investigations of the parasite. © 2012 International Society for Advancement of Cytometry

Infection with Plasmodium falciparum leads to malaria which is one of the world's major health burdens. One challenging problem to defeat the disease is the upcoming resistance against antimalarial drugs (1). From 225 million cases of malaria in 2009, the number of deaths is estimated to 781,000 (2). The clinical picture of the disease is the consequence of the erythrocytic life cycle of P. falciparum. After invasion of plasmodial merozoites into erythrocytes, the parasite starts growing to the so-called ring stage. This early stage of the parasite develops to a trophozoites stage (approximately 24 h post-infection). After intraerythrocytic schizogony, the multinucleated late schizont ruptures the host cell releasing new merozoites to start a new cycle of erythrocyte infection (approximately 45 h after invasion). During this intraerythrocytic development of the parasite, numerous structural and biochemical changes within the host cell are induced (3). Accordingly, growth synchronization and isolation of definite stages of P. falciparum-infected red blood cells (iRBCs) are important steps in malaria research. Over the last three decades, several standard procedures for enrichment of parasites or for synchronizing the growth of different blood stages of the parasite have been developed. As a consequence, the asexual parasite life cycle is maintained continuously in culture. The synchronization of early ring stages can be performed by isotonic sorbitol lysis that selectively lyses late-stage iRBCs by osmotic shock (4). Late stages of the parasite can be synchronized either by Percoll® density gradient (5), by magnetic separation (6), or by gelatin sedimentation (7). These standard methods for the synchronization of late-stage iRBCs are based on the morphological changes of the host cell during the parasite's intraerythrocytic life cycle and can also be used to significantly enrich late-stage iRBCs for further analyses. In contrast, early stages of iRBCs lack such significant molecular modifications of the host cell and therefore, can hardly be distinguished and separated from non-infected red blood cells (RBCs). However, for a comprehensive analysis of the complete life cycle of the parasite, it is necessary to develop methods to separate the different viable stages. One powerful tool for cell separation and analysis, flow cytometry (FCM), was already used in the late 1970s for the detection and counting of malaria parasites (8,9). Later, FCM proved to be useful in analyzing the different blood stages of the parasite (10). Although several dyes, e.g. SYTO-16 (11), SYBR® Green I (12–16), YOYO-1 (17–20), acridine orange (21), thiazole orange (22,23), rhodamine DNA dyes (24), and Hoechst 33258 (10) have been already tested (reviewed in Ref. 25), FCM has not become a commonly used method in malaria research. All these dyes have the advantage that, in contrast to non-infected erythrocytes, the parasite-infected erythrocytes contain xenogenic DNA which can be stained and then detected by FCM. However, several disadvantages such as insufficient sensitivity or/and complicated fixation and permeabilization procedures limit the application range of FCM with some of these dyes. In particular, the loss of living parasites restricts their use to different single-cell studies and in assays for drug screening. Here, we report a FCM-based method in which the cell membrane-permeant and highly DNA-selective fluorescent dye Vybrant® DyeCycle™ Violet was used to isolate viable, intraerythrocytic early stages of the parasite. The applicability of this dye has not been studied yet in malaria research and was thus compared with that of the well-established dye SYBR Green I.



Giemsa staining solution was obtained from Merck (Darmstadt, Germany). Gelafundin was procured from Braun (Melsungen, Germany). AlbuMAX II (GIBCO), Vybrant DyeCycleViolet, SYBR Green I and RPMI 1640 were purchased from Invitrogen (Karlsruhe, Germany). BD FACS Accudrop Beads were obtained from BD Biosciences (Heidelberg, Germany). Chemicals not listed were purchased from Sigma (Munich, Germany) or Roth (Darmstadt, Germany).

Cell Culture of Intraerythrocytic P. falciparum

Erythrocyte concentrates (blood group ARh+) were obtained from the Institute of Transfusion Medicine (University Hospital Kiel, Germany). P. falciparum strain FCBR was a kind gift from Prof. Dr. K. Lingelbach (University of Marburg, Germany) and was cultured in continuous culture as described before with modifications (26). Briefly, cultures were maintained in complete medium (RPMI 1640 supplemented with 25 mM HEPES, 20 mM sodium bicarbonate, and 0.5% (w/v) AlbuMAX II) under low oxygen atmosphere (1% O2, 5% CO2, 94% N2, Air Liquide) at 37°C. Cells were incubated at 2.5% hematocrit and growth synchronization was performed by sorbitol lysis (4) and gelafundin sedimentation (27). For sorbitol lysis, cells were centrifuged (570 × g, 5 min, RT) and 10 pellet volumes of prewarmed 5% (w/v) sorbitol were added. After 10 min of incubation at room temperature, the suspension was mixed with one volume of complete medium and centrifuged (570 × g, 5 min, RT). The remaining sediment was transferred to a new culture flask. For gelafundin sedimentation, cells were centrifuged (570 × g, 5 min, RT) and resuspended in 1.4 pellet volume of complete medium. Subsequently, one total volume of prewarmed gelafundin was added and the suspension was incubated at 37°C in a water bath until a phase separation was visible. The upper phase containing the late-stage iRBCs was carefully transferred into a new tube and cells were washed with complete medium prior to parasite propagation. Growth synchronisation procedures were routinely performed at least once a week.

Giemsa Staining and Light Microscopy

Thin air-dried blood smears were prepared on glass slides, fixed with methanol for 30 s, and stained with Giemsa's staining solution for a minimum of 10 min. Parasitemia (percentage of parasite infected-erythrocytes) was determined by counting 1,000 erythrocytes using the Axioskop 40 light microscope equipped with a 100× oil-immersion objective (Zeiss, Jena, Germany). Multiply infected cells were counted as one. Microscopic pictures were taken with an Axio Imager Z1 microscope (63× oil-immersion objective) using AxioVision software (Zeiss).

Cell Staining and Flow Cytometry

Cells from a culture were centrifuged (570 × g, 5 min, RT) and washed once in washing buffer [0.5% (w/v) BSA, 2 mM EDTA in PBS]. Subsequently, up to 1 × 108 cells (1 × 107 cells per ml) were stained with 5 µM Vybrant DyeCycle Violet stain (Invitrogen) in washing buffer. Similarly, cells were stained with SYBR Green I (1:5,000) as described elsewhere (15) at 37°C in the dark for 30 min. Without any washing step, cells were analysed directly by FACSAria flow cytometer using FACSDiva software (BD Biosciences, Heidelberg, Germany). BD FACSAria was equipped with a 488 nm blue laser, a 633 nm red laser, and a 407 nm violet laser. Vybrant DyeCycle Violet stain/DNA complex has fluorescence excitation and emission maxima of 369/437 nm, respectively, and was detected with FL-8 with a bandpass filter of 450/40 nm. SYBR Green I stain/DNA complex has fluorescence excitation and emission maxima of 497/520 and was detected with FL-1 with a bandpass filter of 530/30 nm. Detectors voltages were set as follows, FSC (325 V), SSC (445 V), FL-8 (440 V), and FL-1 (401 V). Drop delay was determined with BD FACS Accudrop Beads as standard procedure before sorting. Cells were sorted under low pressure using a 100 µm nozzle and each sort was reanalysed to determine purity of the samples. Finally, sorted cells were washed twice with complete medium and cultured as mentioned above.


Analysis of P. falciparum iRBCs by FCM Using Vybrant DyeCycle Violet Stain

To analyze samples of P. falciparum-infected erythrocytes and to isolate early stages of iRBCs via FCM, we introduced Vybrant DyeCycle Violet which was already used to analyze DNA content in other living cells and parasites (28,29). This dye is similar in structure to Hoechst dyes and is therefore probably a minor groove binding dye rather than an intercalating dye. First, a non-synchronous culture with 3% parasitemia was used to test the dye as well as the sorting parameters (Fig. 1A). In this experiment, the late-stage iRBCs were isolated from both RBCs and early-stage iRBCs (Figs. 1B and 1C). Re-sort analysis of the cells with low fluorescence showed an enrichment of iRBCs containing early-stage parasites when sorted within G1 (Fig. 1B). Whereas the FSC-A to SSC-A dot plot showed a clear subpopulation of the late-stage iRBCs, the early stages could not be distinguished from RBCs. Late-stage iRBCs were smaller and showed a significant higher level of granulation in comparison with both RBCs and early stage iRBCs (Fig. 1C).

Figure 1.

Analysis of a P. falciparum culture with 3% starting parasitemia. Cells were stained with Vybrant® DyeCycle™ Violet stain and sorted by FACSAria flow cytometer. The histogram and dot plot of FSC-A to SSC-A (A) shows the distribution of 100,000 analyzed cells. Resorts of G1 (B) and of G2 (C) were done with 3,000 cells. In the histogram of resorted G1, an additional peak became detectable that could be assigned to early stages of intraerythrocytic P. falciparum. Compared with G1, cells sorted in G2 showed a clear subpopulation in both, histogram and dot plot. Events not included in gate G1 or G2 are colored red. [Color figure can be viewed in the online issue which is available at]

Early Stages of P. falciparum iRBCs Can be Isolated by FACS

A culture with high parasitemia (17% early stages, 1.5% late stages) was incubated with Vybrant DyeCycle Violet analysed by FACS (Fig. 2). When compared with cultures with low parasitemia (Fig. 1), a remarkably large additional peak was observed (Fig. 2A, G1). Giemsa staining of the sorted cells showed that early-stage iRBCs as well as some RBCs were found in G1 while late-stage iRBCs were present in G2 only (data not shown). G1 was then further divided into three smaller gates (G1a–c) to investigate the best threshold for isolation of early-stage iRBCs (Fig. 3). While G1a contained early-stage iRBCs to more than 65%, the highest percentages were found in G1b (>91%) and in G1c (>98%), which was confirmed by Giemsa staining (Fig. 3C). To compare these results to parasites stained with SYBR Green I, a culture (14.4% early stages, 0.9% late stages) was stained and analysed by FACS (Fig. 2B). The fluorescent signals of SYBR Green I were about 10 times stronger than those of Vybrant DyeCycle Violet, which resulted in a more efficient separation of erythrocytes infected with early stages of the parasite. Moreover, the discrimination between single-infected (first peak in S2) and multiply infected (second peak in S2) erythrocytes was possible. In S1, only non-infected erythrocytes and in S2 only erythrocytes infected with early stages of the parasite were found. In contrast to Vybrant DyeCycle Violet stained parasites, the FSC-A to SSC-A dot plot of SYBR® Green I stained parasites did not show clear subpopulations of early-stage and late-stage iRBCs, because the early-stage iRBCs showed a broader distribution.

Figure 2.

Analysis of a P. falciparum cultures with high starting parasitemia. Cells were stained with Vybrant® DyeCycle™ Violet stain [18.5% parasitemia, (A)] or with SYBR® Green I stain [15.3% parasitemia, (B)] and sorted by FACSAria flow cytometer. Each histogram shows the fluorescence distribution of 100,000 analyzed cells and the corresponding dot plot of FSC-A to SSC-A shows the distribution of these cells according to their size and their level of granulation. In panel (A) G1 (blue) represents a ring-enriched population that can be distinguished from late stages (G2, green) as well as from non-infected red blood cells (RBCs). In panel (B) S1 (yellow) consists solely of RBCs, while S2 (green) represents the ring-enriched and S3 (blue) the late-stage enriched population. Events not included in gates G1, G2, S1, S2, and S3 are colored red. [Color figure can be viewed in the online issue which is available at]

Figure 3.

Isolation of early stages of P. falciparum-infected erythrocytes (iRBCs). Cells were stained with Vybrant® DyeCycle™ Violet stain and sorted by FACSAria flow cytometer. For sorting of early stage parasites, the gate 1 (G1) shown in Figure 2 was divided into three smaller gates (G1a–c). The histogram (A) shows the fluorescence distribution of 100,000 analyzed cells. For each resort of gates G1a–c (B), the fluorescence histogram and corresponding dot plot of forward scatter FSC-A to SSC-A from 5,000 cells are shown. In histograms, the gates G1a–c are visualized by white hachures. Dot plots show characteristics of cells with respect to complete G1 population (blue). Cells that were sorted in G1c were analyzed by Giemsa staining (C). Events not included in gate G1 are colored red. [Color figure can be viewed in the online issue which is available at]

Development of the Parasite is Not Affected by Flow Cytometry Using Vybrant DyeCycle Violet

A culture with high parasitemia (14.5% ring stages, 1% trophozoites) was divided into an untreated control and cells that were stained with Vybrant DyeCycle Violet stain and subjected to FACS. After sorting of cells, cultures from both control and sorted cells were prepared to set the same culture starting point for monitoring the development of the parasites by Giemsa staining. Figure 4 shows images of representative developmental stages of cultured and Giemsa-stained cells at given time points. The control culture showed a higher parasitemia (control: 2.5%; sorted cells: 0.33%, both at 18 h time point) which was caused by remaining schizonts releasing new parasites within the first 6 h of monitoring the cells. In contrast, these schizonts were rejected by FACS. Although the sorted parasites showed a slightly slower development after recultivation, ring-stage parasites were solely present in all cultures after 6 h and the development to late rings/early trophozoites could be observed after 18 h. At this stage, the parasites already started to modify their host cells by establishment of new structures called Maurer's clefts (purple dots indicated by black arrows; Fig. 4) (30). After development of late trophozoites (30 h), schizonts as well as ring stages were found in all cultures, indicating that parasite reinvasion and the beginning of a new parasite life cycle has already taken place (42 h). For comparison, the development of sorted parasites that were stained with SYBR Green I were investigated, using the same time points. After FACS, parasites seemed to be smaller or disintegrated and within some infected erythrocytes undefined structures appeared (Fig. 4). Moreover, during this time course, a development into throphozoite and schizont stages was not detectable. This experiment showed that, in contrast to incubation with SYBR Green I, staining with Vybrant DyeCycle Violet and sorting did not interfere with development and that the parasite is still able to reinfect erythrocytes. In addition, the rate of erythrocyte reinfection was monitored over three cycles in three independent experiments (Fig. 5). A new batch of 3 × 106 sorted and unsorted early-stage parasites was used to monitor the rate of parasitemia in each experiment. The control showed a slightly higher reinfection rate when compared with sorted cells that were stained with Vybrant DyeCycle Violet. On the contrary, parasites stained with SYBR Green I did not show any reinfection of new erythrocytes (Fig. 5).

Figure 4.

Microscopic analysis of the development of unsorted and sorted parasites. P. falciparum cultures were divided into untreated cells as control (upper panel), cells that were stained with Vybrant® DyeCycle™ Violet stain (middle panel), and cells that were stained with SYBR® Green I stain (lower panel). Stained early stages of P. falciparum-infected red blood cells were sorted by flow cytometry and taken back into culture. Subsequently, the development of parasites was monitored by air dried blood smears at indicated time points. Black arrows mark the appearance of Maurer's clefts at 18 h time point. To demonstrate the two different characteristic appearances of the parasite stained with SYBR Green I, two pictures are shown separated by a black line for each time point. [Color figure can be viewed in the online issue which is available at]

Figure 5.

Study of parasite reinfection rate over time of sorted and unsorted parasites. The time curves show the fold change of parasitemia relative to the starting parasitemia of untreated parasites (▴), and parasites that were stained either with Vybrant® DyeCycle™ Violet stain (▪) or with SYBR® Green I (•) and subsequently sorted by flow cytometry. Parasitemia was determined by air-dried and Giemsa-stained blood smears at indicated time points. For each time point, the mean values and standard deviation bars of three independent experiments is shown.


To date several standard procedures for growth synchronization and isolation of late-stage iRBCs have been developed, using the various modifications of the host cell. In contrast to this, the isolation of early stages of viable parasites is still challenging. FCM has the potential to overcome this problem to improve the analysis of molecular biology of the parasite. The fact that only iRBCs contain DNA introduced by the parasite is used to analyze infected blood samples by staining of DNA with intercalating dyes (11–13,17–19,21–23). Unfortunately, some of these dyes have several disadvantages such as low sensitivity and/or the need to permeabilize or fix the cells before staining. Since multinucleated late-stage parasites contain higher amounts of DNA when compared with ring stages, DNA-selective dyes with low fluorescence intensities cannot be used for a reliable detection of early-stage parasites. Moreover, the discrimination between non-infected cells and cells with early-stage parasites is impossible. In addition, the needed permeabilization or fixation procedures are often complicated and cause loss of viability. To overcome these problems, we have chosen Vybrant DyeCycle Violet, which is highly DNA-selective without the need to permeabilize or fix the cells to be studied and compared the results with SYBR Green I stained and sorted parasites. With both dyes, FCM analyzes can be performed directly after staining without any washing procedures. To test the Vybrant DyeCycle Violet stain, we used cultures with low parasitemia and isolated late-stage iRBCs from both, early-stage iRBCs and RBCs. Notably, reanalysis of sorted cells with low fluorescence showed a clear enrichment of early-stage iRBCs that was of substantial interest. The dot plots of Figure 1 and 2 showed that the discrimination between early-stage iRBC and RBC according to size and granulation of the cells is impossible. In contrast, the late-stage iRBCs formed a distinct subpopulation that is in agreement with a report about separation of late-stage iRBCs from uninfected erythrocytes using light-scattering properties only (31,32). In direct comparison with Vybrant DyeCycle Violet, the SYBR Green I staining resulted in much higher fluorescence intensities and is therefore a convenient dye for measuring parasitemia. Even the discrimination between singly and multiply infected erythrocytes was possible as already stated by others (12). After evaluation of sorting parameters with both dyes, early-stage iRBCs with very high purity (>98%) were enriched by FCM. Importantly, the performed life cycle studies using unsorted and sorted cells showed that only the sorted early-stage parasites that were stained with Vybrant DyeCycle Violet stain remained viable. The parasites stained with SYBR Green I were disintegrated and additional structures occurred within the erythrocyte (Fig. 5). Most likely, these undefined structures were the reason for the broader distribution of sorted parasites in the FSC-A to SSC-A dot plot (Fig. 2B).

The sorted viable early-stage parasites are now accessible to further analyses, e.g. proteome analyses as already done by other groups, but with late-stage parasites only (33,34). In particular, the overall protein repertoire between different blood-stages of the parasite can be compared now by various techniques such as quantitative proteomics. These studies may provide new insights into the proteome of early parasites and the identification of proteins presented on the surface of early-stage iRBCs. Notably, membrane proteins are often present in low abundance only and therefore samples of high purity are needed for their identification (35). In addition, the influence of various substances on protein synthesis in early parasites can be investigated as described for late-stage parasites (36,37). Moreover, it is known that the parasite particularly alters the lipid composition of the host cell which is also connected to drug resistance (38,39). The change of lipid composition of erythrocytes infected with different stages of the parasite can now be analyzed and compared with each other. It would be interesting to evaluate which lipids are integrated into the host cell during or directly after invasion of the parasite. For single-cell studies or for parasite cloning, the isolation of individual iRBCs is traditionally accomplished by micromanipulation or by limiting dilution (40,41). However, these standard procedures are time-consuming and imprecise. The FACS-isolated early-stage parasites may be used for different large-scale single-cell studies including drug screening connected to viability assays and sensitivity tests.

In conclusion, this method allows the isolation of viable early blood stages of P. falciparum with high purity. Future application of this method may help to gain additional information about the development of the parasite, which can be used to discover new drugs and therapeutics and to eventually develop novel antimalarial strategies.


The authors thank H. Liessegang, Zoological Institute University of Kiel, and S. Ussat, Institute of Immunology, University of Kiel, for technical assistance.