Purification and production of Plasmodium falciparum zygotes from in vitro culture using magnetic column and Percoll density gradient

Cell culture media and supplements were obtained from Corning Mediatech (Manassas, VA, USA). O⁺ human RBCs and human serum were obtained from Valley Biomedical (Winchester, VA, USA). Schneider’s media, Giemsa, DAPI, and Percoll were obtained from Sigma Aldrich (St. Louis, MO, USA). A MACS separator was obtained from Miltenyi (Bergisch Gladbach, Germany). Anti-Pfs25, and anti-Pfs28 antibodies were obtained from BEI Resources (Manassas, VA, USA). Alexa Fluor-conjugated anti-mouse IgG antibodies were obtained from Thermo Fisher Scientific [14].

The GFP-expressing strain, P. falciparum NF54HT-GFP-luc [14] was maintained at 5% haematocrit in RPMI 1640 Media supplemented with 10% human serum, 2 mM l-glutamine, 25 mM HEPES, and 50 µM hypoxanthine at 37 °C under 5% CO₂, 5% O₂, 90% N₂. Culture medium was replaced daily and new RBCs were added when the parasitaemia reached 3% [14, 15].

Gametocyte cultures were initiated from asexual parasites at 2% parasitaemia and 5% haematocrit. Cultures were maintained under the same conditions as described above for approximately 14 days with daily media changes without adding fresh RBCs. Exflagellation was studied from days 12 to 14 by counting exflagellation centers in 20 random and defined regions on a haemocytometer. The gametocyte culture was considered to be mature when more than one exflagellation event occurred in each region [16‐18].

Mature gametocytes were obtained by centrifugation at 800×g for 10 min (9 min acceleration, 3 min brake) and resuspended in zygote media consisting of RPMI 1640, Schneider’s, and Waymouth’s media in equal portion with 20% fetal bovine serum (FBS), 4% human RBC lysate, 0.04% NaHCO₃ (Sigma-Aldrich, MO), and 0.25% trehalose (Sigma-Aldrich, MO) at pH 7.4. Gametocytes were induced in zygote media for 6 h at 27 ± 2 °C to promote zygote transformation as described in the Fig. 1a legend [16].

To induce ookinete transformation, purified parasites from the three interfaces were washed in PBS, transferred to a 12-well plate, and incubated in 200 µL of ookinete medium (identical to zygote medium) at 27 ± 2 °C. Ookinete transformation was studied over a time course of 24 h. Specific parasite morphology was studied at 0, 8, 16, and 24 h post-incubation. Fifty µl of culture was collected from each well for immunofluorescent staining and smeared on a standard microscope slide [18].

RBCs, gametocytes, macrogametes, and zygotes were collected from 10 ml of zygote culture and counted under a haemocytometer. Ookinetes were obtained from 200 µl of ookinete culture. Different stages were identified according to their shape and size. RBCs are cells with round and reddish color, Gametocytes are elongated, and ookinetes are elongated banana shaped. Zygotes are oval shaped, while macrogametes are circular and transparent. Two experiments were performed and the representative data was shown. Data is presented in tables as total cells counted.

Before and after each step of purification, 10 µl of culture was collected and smeared on a slide. Each slide was fixed with methanol and stained with 10% Giemsa solution in water for 10 min before imaging.

Samples collected after each purification step during the ookinete- induction time course were fixed on glass slides using 4% paraformaldehyde in PBS for 15 min. After washing with PBS, the slides were blocked in blocking buffer (3% BSA in PBS) for 1h and incubated with primary antibodies against Pfs25 (MRA-28, 1:300) or Pfs28 (MRA-17, 1:300) for 1h at RT. Slides were then washed three times and blocked again as described above. Next, a secondary antibody (Alexa Fluor 594, Goat Anti-Rat IgG, 1:1000) was added and incubated for 45 min at RT before mounting with slowfade-antifade mounting medium containing DAPI (Thermo Fisher Science, MA) [18].

Zygotes were purified as described above through two MACS columns and two Percoll density gradients. Macrogametes were collected from the flow-through of the two MACS columns and further purified with a Nycodenz gradient [20]. After purification, both zygotes and macrogametes were fixed with 4% para-formaldehyde in PBS for 15 min and permeabilized for 5 min with 0.1% TritonX-100 in PBS. After being washed with PBS, test samples were incubated with Vybrant DyeCycle Orange DNA stain (Thermo Fisher Scientific Inc., MA) at a concentration of 10 uM for 30 min at 37 °C and washed again. Samples were resuspended in 2% BSA/2 mM EDTA in PBS. Before analysis, samples were filtered into polystyrene tubes through mesh cell-strainer caps. Flow cytometry for both macrogametes and zygotes was performed on either a BD LSR II or BD FACSymphony flow cytometer (BD Biosciences, CA). Analysis was conducted using FlowJo software (FlowJo, version 10.6.1).

Gametocytes were incubated in zygote media for 15 or 30 min and smeared on glass slides. Parasites were fixed with either methanol or paraformaldehyde as describe above. Methanol-fixed slides were stained with Giemsa. Paraformaldehyde-fixed slides were stained with anti-Pfs25 as described above. The nuclear diameter of zygotes and macrogametes was measured by using a micrometer on Giemsa-stained slides with a light microscope or by using measurement software on DAPI-stained slides with a confocal microscope [21].

Giemsa-stained images were acquired with a light microscope (Nikon Eclipse 50i). Images were viewed at 40x, 60x, or 100x magnification with a 10x eyepiece. All other images were taken using a DeltaVision Elite High Resolution Microscope (GE Healthcare Life Science, PA) designed for fluorescence imaging and analysed using DeltaVision software (SoftWoRx software, version 6.5.2).

Zygotes and macrogametes are similar in size. Most previously published purification methods were not able to differentiate them [11]. To distinguish zygotes from macrogametes, body and nuclear size were compared through Giemsa and IFA staining. Zygotes were purified by 2 MACS and 2 Percoll methods and macrogametes were purified from the flow-through of the two MACS columns and further purified with a Nycodenz gradient [20]. Fixed parasites were stained with an anti-Pfs25 antibody and DAPI (for nuclear staining) (Fig. 2a). The zygotes were verified by Pfs28 staining (Fig. 2b). While body diameter is comparable between zygotes and macrogametes, zygotes display significantly larger nuclei (Table 1), in agreement with previous observations [19]. The difference in nuclear diameter served as a metric to distinguish zygotes from macrogametes during the purification experiments. The DNA content of the zygote is more than that of the macrogametes (Additional file 1: Figure S1).

Gametocytes matured after 6 h of incubation in the zygote medium and transformed into zygotes. At this point, the parasite population consisted of zygotes, along with a large number of uninfected RBCs, untransformed gametocytes, and macrogametes (Table 2). Two sequential MACS columns were used for zygote purification. After the first MACS column, 93.8% of gametocytes, 78.1% of macrogametes, and 99.9% of uninfected RBCs had been removed. After the second column, 98.4% of gametocytes, 94.3% of macrogametes, and 100% percent of uninfected RBCs were removed compared to the starting parasite culture (Table 2). Thus, two sequential magnetic columns efficiently removed uninfected RBCs and a significant proportion of earlier parasite stages (gametocytes and macrogametes). Zygotes were enriched after using the first MACS column (30.9 fold) and were further enriched after using a second MACS column (37.7 fold). Further purification methods were tested to identify a protocol capable of yielding pure, highly enriched zygotes.

A simpler method of zygote purification using an Accudenz (previously known as Nycodenz) gradient after MACS column purification was tested. A gradient of 60%, 35%, and 25% Accudenz [19] was centrifuged at low speed (1000xg) for 30 min. Centrifugation produced two bands; zygotes were only present in the upper band. However, the upper band was also contaminated with gametocytes and uninfected RBCs (upper band: 40.7% zygotes). Only cell debris and haemozoin pigment were found in the lower band (Additional file 2, 3: Figure S2a and Table S1). Another purification method was tested in which no MACS column was used and parasites were layered onto a gradient consisting of 60% (bottom layer), 35% (middle layer), and 25% (top layer) Percoll solutions. After centrifugation at 1000xg for 30 min, three bands were formed, but the entire gradient was smeared with lysed RBCs (Additional file 2: Figure S2b), likely due to RBCs bursting during centrifugation. Moreover, all three bands were contaminated with gametocytes (Additional file 4: Table S2). Thus the use of a single MACS column with Accudenz and a single Percoll gradient did not improve on zygote purification compared to two sequential MACS columns.

Since neither of these methods proved effective at removing earlier P. falciparum stages (gametocytes and macrogametes), two sequential discontinuous Percoll gradients were used after the two MACS columns. For the first Percoll gradient, 60% (bottom) and 40% (top) layers were used. Centrifugation gave rise to two separate bands containing parasites (Fig. 1b). The upper band consisted of a higher number of zygotes (1.7 x 10⁷) compared to the lower band (4.8x10⁶). To further purify the zygotes, the parasites in the upper band were added to a second Percoll gradient: 60% (bottom layer), 35% (middle layer), and 25% (top layer). Centrifugation yielded three separate bands containing parasites (Fig. 1b). The purest population of zygotes was present in the upper band (45.5-fold enrichment; 100% zygotes, compared to 2.2% zygotes before purification). The middle and lower bands contained fewer zygotes and were contaminated with RBCs, gametocytes, and macrogametes along with cell debris and haemozoin pigment. (Figure 3 and Table 2).

While the use of two MACS columns and two Percoll gradient centrifugation steps provides a pure population of zygotes, it also causes a moderate loss of transformed zygotes due to multiple steps of purification. To attempt to minimize the loss of zygotes during purification, the use of a single MACS column and a single Percoll density gradient (60%, 35%, and 25%) at a lower centrifugation speed (1000xg) and longer spinning time (30 min) was tested. After centrifugation, three bands consisting of parasites were present, and zygotes were present in all bands. However, there was greater contamination from other parasites stages (the final zygote percentage was 88.9%; Additional file 2, 5: Figure S2c and Table S3). Thus, the method “2 MACS 2 Percoll” is the preferred method to obtain the purest population of zygotes. The experiments of “2 MACS 2 Percoll” were performed in duplicates. Best results were presented in the text and similar results were obtained from the other experiments (Additional file 6: Table S4).

Parasites separated via the 2 MACS 2 Percoll protocol were subjected to zygote-to-ookinete transformation in ookinete medium for an additional 24 h at 27 °C. A larger percentage of zygotes from the upper band transformed into ookinetes compared to the middle and lower bands (Table 3). The percent of transformation, indicated as P trans (z-o), was calculated as the ratio of the number of transformed ookinetes to the number of zygotes. Zygotes purified by 1 MACS 1 Accudenz and 1 MACS 1 Percoll methods were also incubated in ookinete media. The percentages of transformed ookinetes from these two methods were much lower compared to those measured after the 2 MACS 2 Percoll protocol (Additional file 7: Table S5). Purified zygotes were incubated in ookinete medium and sampled over a 24-h time course. Giemsa staining showed that purified zygotes from all three bands were able to transform into retorts and then ookinetes (Fig. 4a). Zygote, retort, and mature ookinete morphology was studied by IFA using Pfs28, a surface marker of these stages [10]. Early stage retorts with a small protrusion, as well as late stage retorts were found at 8 and 16 h of the transformation. Expression of Pfs28 confirmed the presence of zygotes in the purified samples. Transformed ookinetes were morphologically similar to mosquito-derived ookinetes and expressed Pfs28 (Fig. 4b).

Classical zygote purification methods utilize either wheat germ agglutinin (WGA) to gather RBCs away from parasites [13, 22] or lysis of RBCs with ammonium chloride [23]. WGA does not remove RBCs effectively due to differential binding of the WGA to young and old RBCs [24]. Ammonium chloride solution treatment also does not lyse RBCs completely. It needs further purification to remove the residual RBCs.

Other attempts have been made to isolate Plasmodium macrogametes and zygotes, including the use of metrizamide (a non-ionic radiopaque density gradient medium for the centrifugation of biological particles, now unavailable) to separate macrogametes and zygotes prepared from in vitro cultures: gametocytes were layered in a discontinuous gradient (70% and 10%) and centrifuged. Recovered parasites were further purified using a WGA-Sepharose 6 MB column and 67% of macrogametocytes were recovered [11]. A separate study to isolate Plasmodium gallinaceum zygotes used a Hypaque/Ficoll gradient method [13]. The interface material was collected and zygotes, macrogametes, white blood cells, and RBCs were found. Infected and uninfected RBCs were removed by several rounds of WGA agglutination processing followed by column purification through glass wool. The final sample contained zygotes (90% pure) with contaminating macrogametes.

A similar study using the NF54 strain (wild type) of P. falciparum employed an Acccudenz gradient to purify macrogametes and zygotes. Briefly, gametocytes were transformed into gametes and zygotes. A density gradient of Accudenz (16% lower, 11% middle, and 6% top layers) was made and the transformed material was loaded onto the gradient and centrifuged. Macrogametes and zygotes were harvested from the 6–11% interface. The percent purity was not determined and the samples were contaminated with RBCs [12]. The method described here (2 MACS 2 Percoll) improves on these earlier procedures, which were less effective in producing a pure zygote population.

The 2 MACS 2 Percoll method provides advantages over other methods because Percoll (a relatively low-viscosity colloidal silica that is regularly used in purification of various cells types) efficiently separates zygotes without causing physical or functional damage [25]. Removal of RBCs with a MACS column is a simple and quick method and has been used previously to separate other Plasmodium stages [13, 23, 26]. It was found that zygotes bind to the column and other cell types were removed from the columns during the washing steps.

The methodology described here produces large quantities of highly purified zygotes that could be used to identify new markers for further study. These potentially important zygote antigens could be targeted individually or in combination with known sexual stage antigens [27‐30]. Additionally, these markers could be used to assess the quality of an in vitro system for the mass production of a genetically attenuated parasite (GAP) vaccine [31]. For example, sexual stage antigens could serve as developmental biomarkers and function as quality control reagents for culturing systems.