PfCap380 as a marker for Plasmodium falciparum oocyst development in vivo and in vitro

The luciferase expressing strain P. falciparum NF54HT-GFP-luc [14] was maintained in complete media (CM), which is Roswell Park Memorial Institute (RPMI) 1640 Media, containing 25 mM HEPES, 2 mM l-glutamine, and 50 μM hypoxanthine (Mediatech, VA) plus 10% human serum (Valley Biomedical, VA). The strain was sub-cultured in CM with 5% O⁺ human erythrocytes (Valley Biomedical, VA) and gametocytes seeded using previously established methods [15‐18]. Mature sexual stage gametocytes were induced by allowing continuous growth of cultures without the addition of fresh red blood cells (RBC), as previously described [19]. Briefly, asexual cultures were inoculated at 2% trophozoite parasitaemia and 5% haematocrit. Media was changed daily and thin blood smears were observed beginning on day 9 to determine maturation of gametocytes. Mature gametocytes (day 12–14 post-induction) were fed to 7-day-old Anopheles stephensi mosquitoes to initiate infection. The mature gametocyte culture was diluted 1:1.8 with fresh whole blood and then diluted 1:1 with human serum. Mosquitoes were allowed to feed for 20 min and were then incubated at 27 °C and 75% humidity and given 8% dextrose with 0.05% para-aminobenzoic acid (Sigma-Aldrich, MO). To check midgut infection of P. falciparum, mosquitoes were dissected between days 2–9 after gametocyte infection.

Gametocyte culture was seeded and matured as described above. Media was changed daily and tri-gas (90% N₂, 5% CO₂, 5% O₂) added to the flask for ~ 90 s. Emergence of male gametes (exflagellation) was measured between days 12–16 and the culture was considered mature when ≥ 1 exflagellation event per field was observed in ~ 4000 RBC. Once matured, gametocytes were placed in ookinete media [20] comprised of RPMI 1640, Schneider’s (Sigma-Aldrich, MO), and Waymouth’s (Mediatech, VA) medias in a 1:1:1 ratio along with 20% fetal bovine serum (FBS), 4% human RBC lysate, 0.04% NaHCO₃ (Sigma-Aldrich, MO) 0.25% trehalose (Sigma-Aldrich, MO), adjusted to pH 7.4. The culture was incubated for 30 h in ookinete media at 27 °C with shaking at 15 RPM.

Methods to purify Plasmodium ookinetes by magnetic column purification have been previously described [21]. For the studies here, mature ookinetes were incubated for 30 h and condensed into a volume of 5 mL, and then purified using LS MACS columns (Miltenyi Biotec, Bergisch Gladbach, Germany) with an attached 24 or 25-gauge flow restrictor (Strategic Applications, IL). For magnetic isolation, columns were mounted on a powerful magnet of 0.5 Tesla magnetic force (QuadroMACS Separation system; Miltenyi Biotec, Bergisch Gladbach, Germany). Prior to purification, columns were washed with RPMI. Then the culture was passed over the column while attached to the magnet. The column was washed with 10 mL of RPMI and the bound ookinetes were released by removing the column from the magnetic field and eluting in 5 mL of RPMI in a sterile collection tube. Ookinetes were washed with RPMI and resuspended in oocyst medium. Purity of the ookinetes was determined by counting ookinetes using a haemocytometer. A second round of purification using magnetic columns was performed in the same manner to obtain highly purified ookinetes. Viability of the ookinetes was determined by the Trypan blue (Amresco, OH) exclusion method. Twenty thousand purified ookinetes were seeded into individual wells of 8-well-chamber slides (Electron Microscopy Sciences, PA) that were pre-coated overnight at 4 °C with 25 μg/mL laminin (Corning, NY), 25 μg/mL laminin/entactin (Corning, NY), and/or 50 μg/mL mouse collagen IV (Corning, NY). BSA (1% BSA in PBS) was used as a negative control. Ookinetes were seeded into the pre-coated wells with oocyst medium (RPMI 1640 and Schneider’s medias in a 1:1 ratio along with 15% FBS, 0.04% w/v NaHCO₃, 0.25% w/v trehalose, 50 μg/mL hypoxanthine (Sigma-Aldrich, MO), 10 mM HEPES (Sigma-Aldrich, MO), and 10× lipoprotein cholesterol solution [12]. Other supplements including 4% human RBC lysate, 0.015% silkworm haemolymph, and 0.001% haemin chloride (Sigma-Aldrich, MO) were added to the medium. Oocyst medium was changed on the fifth day after plating ookinetes.

PfCap380 protein sequences from P. falciparum (PFC0905c), P. vivax (PV095215), P. berghei (PB000071.00.0 and PB300510.00.0), P. yoelii (PY00597), and P. gallinaceum (PGAL8A_00395900.1) were retrieved from PlasmoDB.org. Amino acid sequence alignment was performed using ClustalW [22].

Anti-PfCap380 rabbit polyclonal antibodies (antisera) were raised by GenScript (GenScript, NJ). Briefly a gene fragment of PfCap380 corresponding to amino acids 1954–2068 (115 amino acids) was cloned into the pET-30a expression vector and expressed in BL21 (Thermo Fisher Scientific, MA) bacteria. The expressed protein contained a His-tag and was purified. The correct protein size was confirmed using SDS-PAGE, and the protein gel was stained and detained using the eSTAIN L1 staining kit (GenScript, NJ). Also, Western blot analysis was performed to identify the immunogen by presence of the His-tag. SDS-PAGE was performed and the gel was transferred to a PVDF membrane, that was washed in PBS—Tween 20 three times for 5 min each. The membrane was incubated with primary antibody (mouse-anti-His; GenScript, NJ) in milk for 30 min and washed as before. The membrane was blocked in milk and incubated with the secondary antibody (anti-mouse IgG alkaline phosphatase or HRP) for 45 min. The membrane was washed as before and then exposed to identify the immunogen, and the correct size was confirmed. Once the correct protein size and presence of the His-tag was confirmed, the purified protein immunogen was used with Freund’s adjuvant to immunize rabbits and obtain polyclonal antibodies against PfCap380.

All images were taken using 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). Images were taken at 40×, 60×, or 100× magnification with a 10× eyepiece. Merging of separate colour channels was performed using Image J [26].

Eight-well-chamber slides were pre-coated over night with laminin, collagen type IV, and entactin, as described. BSA was used as a negative control. After coating, excess material was removed and washed with sterile PBS. Fixed numbers of ookinetes (20,000 per well) were seeded and allowed to transform into early oocysts after 2, 3, and 6 days in culture. At the end of 2, 3, or 6 days completing the developmental experiment, the wells were washed, fixed and stained as described before with anti-PfCap380 antisera. For each condition, a total of 20 fields (containing ~ 700 oocysts) were counted at 40× magnification. For each experiment, average values of oocyst binding per field were determined and then normalized to BSA. Each group was compared to BSA and statistical significance was determined using Student’s T tests.

To quantify DNA from oocysts, ookinetes were seeded into 6-well-plates, and oocysts were transformed and grown as described above using laminin, entactin, and collagen IV as the basal lamina. On days 2, 3, and 6, oocysts were collected in PBS to count the number of live oocysts. Oocysts were resuspended in 400 μL of lysis buffer (0.2% Triton in Tris–EDTA buffer). Total double stranded DNA (dsDNA) from 12,500 live oocysts was measured per sample using the Quant-iT PicoGreen dsDNA Assay Kit according to the manufacturer’s protocol. (P7589, Thermo Fisher Scientific, MA). To each oocyst sample, 100 μL of PicoGreen solution was added. Then, oocysts were gently shaken and incubated for 5 min before measuring fluorescence using a microplate reader (SpectraMax M2, Molecular Devices, CA) at wavelengths 480 for excitation and 520 for emission. Total oocyst DNA concentration was calculated using a known DNA standard.

To identify a stage-specific oocyst marker for oocyst developmental studies, a thorough literature search was performed. The only known oocyst-specific protein to date is an oocyst capsule protein, Cap380, which was originally identified in P. berghei (PbCap380) from a subtraction library of genes enriched for expression during the oocyst stage [27]. Cap380 is a unique protein and all Plasmodium species have Cap380 orthologues. When comparing the Cap380 amino acid sequence across several Plasmodium species, the N-terminal half of the protein shows greater similarity than the C-terminal half [13]. In order to select the PfCap380 peptide immunogen, the region of P. falciparum that is homologous to the antigenic region of P. berghei was analyzed [13]. From these amino acids in P. falciparum (1954–2283), a shorter peptide fragment (1954–2068) within the larger region was selected since the expressed peptide was predicted to exhibit high stability. Next, this region of PfCap380 was compared to the orthologues for P. berghei, P. yoelii, P. vivax, and P. gallinaceum and it was found that PfCap380 shares greatest amino acid similarity with P. gallinaceum followed by P. vivax, P. berghei and P. yoelii (see Additional file 1: Figure S1A, B). The general dissimilarity of amino acids between P. falciparum and P. berghei in this region explains why the anti-PbCap380 antibody that was previously generated does not cross react with P. falciparum (unpublished observations). PfCap380 was selected as an immunogen for antibody design since it is expressed in the oocyst stage and exhibits stage-specific expression [13].

While a previous study in P. berghei generated an antibody against PbCap380, the anti-PbCap380 antibody does not cross-react with P. falciparum infected midgut oocysts, prohibiting usability in P. falciparum oocyst developmental studies [13], Unpublished observations]. Therefore, antisera against PfCap380 was raised using a similar strategy that was used for the anti-PbCap380 antibody [13]. The first 115 amino acids of PfCap380 (1954–2068) homologous to the peptide fragment of PbCap380 that was used for immunization was selected. The expressed His-tagged PfCap380 peptide fragment showed the predicted size in SDS-PAGE analysis, ~ 15 kDa (See Additional file 2: Figure S2A, performed by GenScript). Western blot analysis using an anti-His antibody also confirmed expression of the same size protein fragment (See Additional file 2: Figure S2B, performed by GenScript). The PfCap380 peptide immunogen was then used to raise polyclonal rabbit antibodies by using one prime immunization followed by two boosting immunizations. In sum, an immunogen corresponding to PfCap380 amino acids 1954–2068 was used to immunize rabbits and obtain polyclonal antisera for P. falciparum oocyst developmental studies.

Towards developing a culturing system for production of P. falciparum mosquito stages, including oocysts, in vivo oocyst development was first studied to serve as a standard. Oocysts were identified in mosquito midguts by following GFP expression using a fluorescent reporter strain, P. falciparum NF54HT-GFP-luc [14]. Oocysts were generally round or oval in shape and grew in size during the course of the developmental study (from 2 to 9 days after mosquito infection). The measured oocyst diameter was ~ 7 μm on days 2–3, ~ 20 μm on day 6, and ~ 30–40 μm on day 9. To study in vivo oocyst development of P. falciparum infected mosquito midgut oocysts, the expression of PfCap380 was monitored by immunofluorescence assays (IFA) using anti-PfCap380 antisera. The expression of PfCap380 was found to begin on day 2 as a small circular or oval pattern surrounding the oocysts and was present at all time points (days 2–9) (Figs. 1a and 2). No signal was observed for negative controls using only secondary antibodies (See Additional file 3: Figure S3A). These results demonstrate that midgut P. falciparum oocysts express Cap380, and that Cap380 can be recognized by the anti-PfCap380 antisera.

Circumsporozoite protein (CSP) is another oocyst marker, but is expressed in other stages than the oocyst since it is also expressed in SPZ. Previous studies using fluorescence and immunoelectron microscopy demonstrate that CSP is localized on the oocyst plasma membrane [13, 28‐30] directly below the oocyst capsule where Cap380 is localized [13]. Separate P. falciparum developmental studies on in vivo oocyst expression of CSP using western blot and ELISA methods found CSP to be expressed on days 8–10 after mosquito infection [31, 32], while immunoelectron microscopy of P. berghei oocysts demonstrate low but detectable CSP expression in early oocysts [28]. IFA were performed on P. falciparum mosquito-derived oocysts to study CSP and compare the utility of CSP to Cap380 as an oocyst marker. In contrast to the PfCap380 expression patterns, CSP was expressed later and in a different outer layer of the oocyst. CSP signal was not found on day 2 as it was for Cap380. Instead, CSP expression was very low but detectable on day 3 and was higher on days 6 and 9 (Figs. 1b, 2). This high-resolution fluorescence microscopy study revealed multi-nucleated oocyst DNA below the oocyst membrane and capsule using DAPI staining (Fig. 2c–e). The data also suggest that CSP appears to localize below PfCap380 (Fig. 2e) but higher resolution microscopy (immunoelectron microscopy) would be required to be absolutely certain the two proteins are located separately. Other studies demonstrate that CSP localizes to the plasma membrane [28‐30] directly below the oocyst capsule, where PbCap380 is localized [13]. The present studies demonstrate the utility of PfCap380 as a marker and the anti-PfCap380 antisera to study in vivo oocyst development from days 2–9 after mosquito infection.

A goal of this work is to study P. falciparum mosquito stage development in vitro to ultimately manufacture SPZ for malaria vaccines. In vitro oocysts were produced by culturing P. falciparum gametocytes to maturity, then promoting the parasites to transform through the stages—zygote, ookinete, and oocyst—as outlined in Fig. 3. Zygotes and ookinetes were produced by making minor modifications to an existing protocol [20, 33]. Once ookinetes formed, magnetic column purification was used to remove red blood cells [21] and obtain highly purified ookinetes. For Plasmodium development in the mosquito, ookinetes bind to the mosquito midgut basal lamina, and this interaction is thought to trigger the transformation of ookinetes into early oocysts [34]. Considering this knowledge, basal lamina components were provided along with nutrient rich media (oocyst media) to mature ookinetes, supporting transformation into early oocysts. Then, oocysts were studied over a 6-day developmental time course.

To select basal lamina components that would support ookinete binding and transformation into oocysts, Plasmodium development in mosquitoes was considered as well as the previously published P. falciparum culturing system. The study by Warburg and Schneider utilized Matrigel as a basal lamina, which is a gel secreted by Engelbreth-Holm-Swarm mouse sarcoma cells and contains proteins including laminin, entactin, and collagen IV [12]. Matrigel also contains growth factors in undefined amounts and would be problematic for use in manufacturing human vaccines due to lack of complete characterization of unknown, tumor derived factors. Therefore, this study was aimed to identify the minimal basal lamina requirements for in vitro oocyst production and focused on the protein components of Matrigel. Laminin and collagen IV are known components of the mosquito basal lamina and are considered important for ookinete binding and parasite development [34‐38]. Laminin, a glycoprotein, incorporates into the developing oocyst [39] while collagen IV is a structural sheet-forming protein of the basal lamina. Entactin is a glycoprotein that connects laminin and collagen IV. The tested basal lamina components include laminin alone, collagen IV alone, and a combination of laminin, entactin and collagen IV (L/E/C). The number of oocysts per field was counted and shown as fold change compared to BSA (negative control) (Fig. 4). On day 2, fold change values were as follows, laminin—0.53, collagen IV—4.0, L/E/C—4.4; on day 3, laminin—0.4, collagen IV—3.4, L/E/C—3.8; on day 6, laminin—0.4, collagen IV—3.0, L/E/C—3.8. The binding study of P. falciparum to basal lamina components revealed that collagen IV alone or in combination with laminin and entactin provided the greatest binding compared to controls (BSA) or laminin alone.

High transformation rates for gametocytes to ookinetes (24 ± 0.02%) and ookinetes to early oocysts (85 ± 0.04%) was observed using the L/E/C basal lamina (See Additional file 4: Figure S4). Within 24 h of in vitro development, ookinetes round up to form early oocysts, attach to the reconstituted basal lamina, and begin growing. The oocyst media consists of components described by Warburg and Schneider [12], but without the use of Drosophila S2 cells. Novel supplements to the oocyst media include human RBC lysate, silkworm hemolymph, and hemin chloride. After screening several factors, the novel supplements were found to support oocyst transformation and viability. Early oocysts grew in size from day 2 to 6 reaching the maximum size ~ 5–7 μm; however, most of the oocysts were developmentally arrested, but still viable (positive GFP signal) for ~ 7 more days. In sum, the oocyst culturing system promotes high ookinete to early oocyst transformation by using collagen as a basal lamina component and novel oocyst media.

To study the development of in vitro produced oocysts, the expression and localization of PfCap380 and CSP was studied by IFA on days 2, 3, and 6 after seeding ookinetes (Fig. 5). Similar to mosquito-derived oocysts, in vitro oocysts expressed PfCap380 from day 2 and maintained the expression pattern up to day 6, (Fig. 5a). The majority of oocysts (> 90%) exhibited a similar expression pattern to the images shown (Fig. 5). On day 6, most oocysts showed strong PfCap380 staining as compared to the negative controls (See Additional file 3: Figure S3B). For CSP, expression began at day 2 and continued through day 6. However, in most oocysts, only a portion of the oocyst circumference showed CSP staining (Fig. 5b).

To test whether PfCap380 is expressed only in the oocyst stage or also in other mosquito stages, IFA were performed on gametocytes and ookinetes using anti-PfCap380 and co-labeled using antibodies against gametocyte markers (anti-Pfs230 and anti-Pfs48/45) and an ookinete marker (anti-chitinase). PfCap380 is not expressed in gametocytes or ookinetes; although, the background staining for PfCap380 was greater for ookinetes than for gametocytes (Fig. 6). To confirm this was background or non-specific staining rather than specific staining, the anti-PfCap380 antisera was directly labelled with a fluorophore and under these conditions, anti-PfCap380 does not bind ookinetes (See Additional file 5: Figure S5). An earlier study of PbCap380 expression also showed that Cap380 expression is restricted to the oocyst stage [13]. Since the IFA data showed binding of the PfCap380 antisera only to oocyst stages as previously reported for P. berghei, additional control experiments with the antisera were not performed (i.e., SDS-page or Western blot analysis on parasite extracts or mosquitoes). All gametocytes express Pfs230, while only a subset of gametocytes express Pfs48/45, and ookinetes express chitinase (Fig. 6). The observed expression patterns for gametocyte and ookinete markers are also in agreement with previous studies [40, 41]. The IFA studies on in vitro produced P. falciparum support oocyst expression of PfCap380 and utility of anti-PfCap380 antisera on in vitro produced oocysts.

A previous study using the P. berghei in vitro culturing system demonstrates that oocyst endomitosis occurs by showing that DAPI foci increase over a 15-day time course [10]. To determine whether genome replication occurs in the present in vitro culturing system, total double-stranded DNA (dsDNA) was measured using PicoGreen, which binds specifically to dsDNA. The average concentration values (ng/mL) were as follows for day 2–16.3; day 3–19.4, and day 6–23.9 (See Additional file 6: Figure S6). These data indicate that DNA replication occurs during the six-day time course of the in vitro oocyst culture.

The oocyst capsule begins forming soon after the transition from ookinetes to early oocysts and increases in size to accommodate several thousands of growing SPZ. The actual mechanism of how the oocyst capsule forms and grows is not completely understood, yet it is believed that the capsule is an ordered and dynamic structure that facilitates transport of nutrients into the oocyst and exit of metabolites [42]. The capsule is not a lipid bilayer membrane, and is presumed to consist of several layers of proteins and carbohydrate moieties. Studies support a role for the capsule in protecting developing SPZ from the mosquito immune system [13, 43].

The localization of Cap380 begins on the surface of early oocysts and incorporates into the oocyst capsule [13]. Previous microarray studies on P. berghei showed that PbCap380 is expressed in oocysts but not in gametocytes, SPZ, nor in blood stage parasites [44]. In the present studies, Cap380 localizes to the capsule of P. falciparum oocysts, and is not expressed in other sexual stages (gametocytes or ookinetes) (Fig. 6, See Additional file 5: Figure S5). The microscopy study presented suggests that the P. falciparum oocyst capsule consists of Cap380, and CSP appears to underlie Cap380 on the oocyst plasma membrane (Fig. 2). While the exact function of Cap380 is not yet known, knockout studies of PbCap380 indicate the importance of the protein for oocyst survival and SPZ production. In these studies, normal numbers of early oocysts formed, but further development was blocked and oocysts were eliminated without production of SPZ [13].

The anti-PfCap380 antisera presented here binds to the midgut oocyst capsule from day 2 and through day 9 (Fig. 2). While CSP can be used to study oocyst development, strong CSP expression is not seen until 6 days after mosquito infection, and CSP is not a stage-specific marker as it is also strongly expressed in SPZ. To study oocyst development in the infected mosquito midgut, anti-PfCap380 antisera serves as a powerful stage-specific reagent to monitor oocyst growth and development as early as 2 days after infection with gametocytes.

Ookinetes bind to the mosquito midgut basal lamina, and this interaction is thought to trigger the transformation of ookinetes into early oocysts [34]. In the present study, basal lamina components were tested for binding affinity of early oocysts and, therefore, ability to support transformation into early oocysts. The results demonstrate the need for basal lamina components, especially collagen IV, since collagen IV alone or together with laminin and entactin support early oocyst transformation when compared to controls or laminin alone (Fig. 4). A high rate of ookinete to oocyst transformation (85%) was observed. The only other published study on in vitro production of P. falciparum oocysts did not describe the oocyst transformation rate [12]. The high rate of transformation in the present culturing system could be due to high viability of ookinetes (> 95%) or possibly to culturing in the absence of the mosquito immune response, which counteracts parasite survival.

Early in vitro P. falciparum oocysts produced in the present culturing system exhibit the following characteristics: (1) Transformation of ookinetes into oocysts (i.e., rounding up into oocysts) within 24 h; and (2) Development to the diameter of ~ 5–7 μm. These results are consistent with other early oocyst Plasmodium culturing systems [9, 10, 45]. The anti-PfCap380 antisera detects early oocysts from day 2 and the staining pattern is distributed on the entire surface of the oocyst capsule (Fig. 5a). CSP expression occurs later as consistently strong expression was observed starting on day 6 (Fig. 5b). The in vitro development of oocysts arrests after day 6, and DNA replication occurs as a significant but slow increase in DNA concentration was observed. These results can be explained by nutritional deficiency in the oocyst media or lack of proper signaling molecules for further oocyst maturation. Additional studies on an improved, more robust in vitro culturing system will be necessary for production of mature SPZ-producing oocysts.