Opsonic phagocytosis of Plasmodium falciparummerozoites: mechanism in human immunity and a correlate of protection against malaria

Two different longitudinal cohort studies, referred to as Chonyi and Ngerenya, were included to address different aspects of the acquisition and boosting of antibodies, their relationship to immunity, and the impact of malaria transmission rates on antibodies. Details of the study area and population are published [22]. Malaria transmission occurs in two seasonal peaks, with average annual entomological inoculation rates (EIRs) of 20 to 53 (Chonyi), and 10 (Ngerenya) infective bites/person/year [23, 24]. Briefly, participants were recruited during cross-sectional surveys in October 2000 (Chonyi, n = 536) and October 2002 (Ngerenya, (n = 295)), at the start of a malaria transmission season. A single blood sample was collected at recruitment and participants were subsequently monitored actively each week for six months to detect clinical episodes of malaria. Regular screening for parasitemia was not performed during the follow up visits, but was measured only when participants reported any symptoms suggestive of malaria. Local age-specific criteria defining clinical episodes of malaria were: fever (>37.5°C) plus any parasitemia for children less than one year old, and fever plus a parasitemia >2,500/μL for participants older than one year [22]. In Ngerenya, children also underwent active surveillance for malaria in the six months prior to sample collection. In this report, data are presented for all children from Ngerenya for whom a sample was available (n = 287) and the subset of children from Chonyi who were asymptomatically parasitized (parasite-positive) at the time of sampling (n = 109). For Chonyi, this subset was studied for two main reasons; firstly, although the original cohort was made up of children and adults, 90% of all clinical episodes observed during six months of monitoring occurred in children ≤10 years old; therefore, adults were excluded from the analysis of antibodies in relation to the risk of clinical malaria. Secondly, protective associations have only been observed in the subset of children who were asymptomatically parasitized (parasite-positive) at recruitment and the incidence of malaria in those who were parasite negative at enrollment was low [6, 25‐30]. Therefore, this subset was considered ideal to test the hypothesis that the phagocytosis index was a correlate of protective immunity against clinical episodes of malaria, and comprised children up to ten years of age who were parasite positive at recruitment (n =109). For the initial evaluation of opsonic phagocytosis responses and validation of the assay, and for performing detailed comparisons between opsonic phagocytosis and other measures of malaria immunity, a random selection of samples from Ngerenya children and adults (n = 33) was used for which we had sufficient volumes of sera to perform multiple antibody testing. Total immunoglobulin G (IgG) was also purified from these samples for use in assays. Pooled serum from 20 adult blood donors from the same village was used to affinity-purify antigen-specific antibodies. A reference Malaria Immune Globulin (MIG) reagent (Central Laboratory of the Swiss Red Cross Blood Transfusion Service, Berne Switzerland) [31] was used for validation experiments and as a positive control for the cohort assays. This preparation contains 50 mg/ml of immunoglobulins (98% IgG) purified from a pool of healthy Malawian adult plasma and was originally manufactured to test its potential use as an adjunct therapy to quinine in the treatment of cerebral malaria. Written informed consent was obtained from all study participants or their parents/guardians. Ethical approval was granted by the Kenya National Research Ethics Review Committee (SSC No. 1131).

Merozoites were isolated directly from culture using previously published methods [34, 35]. Briefly, late stage pigmented trophozoites were harvested by magnetic purification on columns and then cultured in medium supplemented with the protease inhibitor trans-epoxysuccinyl-L-leucylamido(4-guanidino) butane (E64) for eight hours to allow maturation to schizonts without rupture. Mature schizonts were collected and passed through a 1.2 μm filter to release and purify merozoites. Free merozoites were stained with ethidium bromide (EtBr) at a final concentration of 1 μg/mL for 30 minutes followed by three washes in RPMI. The cell density was determined using relative counting against CountBright™ Absolute Counting Beads (Invitrogen, Mount Waverly, Victoria, Australia) on a BD FACSCalibur (BD Biosciences, North Ryde, New South Wales, Australia) flow cytometer. The merozoites were then resuspended at 5 × 10⁷ merozoites/mL in RPMI-HEPES and used in assays as described.

Our method was adapted from an established assay for the opsonic phagocytosis of P. falciparum-infected erythrocytes by undifferentiated THP-1 cells [36, 37]. Briefly, freshly cultured THP-1 cells were counted and resuspended at a final concentration of 5 × 10⁵/mL in THP-1 culture medium. Freshly isolated merozoites were transferred into 96-well U-bottomed plates (30 μL/well at 5 × 10⁷ merozoites) that had been pre-coated with fetal calf serum (FCS) (200 μL of FCS, incubated for one hour, followed by a single wash with incomplete RPMI). All antibodies used for opsonization were heat inactivated to exclude any influence of complement. For opsonization, 3.5 μL of test serum was incubated with 30 μL of merozoites (pre-stained with EtBr) for one hour at room temperature in the dark. The plate was washed three times using incomplete RPMI, before resuspension in 150 μL THP-1 culture medium. To obtain three replicates, 50 μL of the opsonized merozoites were co-incubated with 100 uL each of THP-1 cells at 5 × 10⁵ cells/mL in FCS at 37°C for 10 minutes for phagocytosis. Phagocytosis was stopped by the addition of 50 μL cold PBS supplemented with NCS. Plates were immediately washed to remove free or loosely attached merozoites. Three washes were performed using the same buffer at 4°C (centrifugation at 300 × g for four minutes). THP-1 cells were then fixed with 2% paraformaldehyde (PFA) for two hours before analysis by flow cytometry. Several controls were included for each assay: 1) non-opsonized merozoites; 2) merozoites opsonized with non-malaria exposed sera; and 3) merozoites opsonized with pooled highly reactive sera from adults exposed to malaria (MIG). Selected assays had an additional control in which THP-1 cells were pre-incubated with cytochalasin D to inhibit phagocytosis. Flow cytometry was performed in a 96-well format on a BD FACS CantoII (BD Biosciences). In preliminary experiments we established that a ratio of merozoites:THP-1 cells of 10:1 was optimal. The level of phagocytosis was determined by counting the percentage of THP-1 cells that had ingested merozoites and is referred to as the Phagocytosis Index (PI). Results are presented as a relative phagocytosis index (RPI%), with the PI being expressed as a ratio to that of a standard positive control run in each assay. Samples were considered positive for phagocytosis if the RPI exceeded the mean plus three standard deviations of a panel of 10 non-malaria exposed sera from Melbourne blood donors (Melbourne controls).

Isolated human PBMC were resuspended in RPMI-1640 supplemented with 10% FBS at a final concentration of 5 × 10⁶/mL. Freshly isolated merozoites were opsonized and stained with EtBr followed by co-incubation with PBMCs for 10 minutes. The monocyte population was gated on flow-cytometry dot plots using light scatter characteristics and the percentage of EtBr positive monocytes used to determine the phagocytosis index.

Isolated PBMCs were resuspended in RPMI-1640 supplemented with 10% human serum at a concentration of 5 × 10⁶/mL. PBMCs (100 μL) were added to polypropylene tubes and kept on ice. Freshly isolated merozoites were opsonized with either hyper-immune human IgG or serum from malaria-naïve Melbourne donors for one hour before being re-suspended in RPMI-1640 supplemented with 10% human serum at a concentration of 5 × 10⁷/mL. Merozoites (10 μL) were added to the PBMCs and co-incubated at 37°C for six hours. Brefeldin A and Monensin were added to each tube at concentration of 1:1000 and 1:1500, respectively, prior to co-incubation. Cells were then washed with cold fluorescence-activated cell sorting (FACS) buffer after co-incubation and monocytes were labelled by staining with anti-CD14-APC antibodies. Cells were stained with anti-CD69-V450 to determine monocyte activation [38‐40]. The cells were fixed overnight with BD FACS fix buffer (BD Biosciences) then permeabilized with BD Perm/wash buffer. Production of intracellular tumor necrosis factor- α (TNF-α) was detected by staining with anti-TNF-α -PE antibodies. Cells were re-suspended in BD Fix buffer following intracellular staining.

Square glass coverslips (22 mm) were prepared by smearing with a 0.1% solution of polyethyleneimine (PEI) and then dried. Cell samples were incubated on PEI-coated glass coverslips for half an hour. Following incubation, the excess sample was drained, and coverslips with adhered cells were immersed in 2% glutaraldehyde in PBS for one hour. Coverslips were then rinsed three times in PBS for 10 minutes each before being dehydrated in increasing concentrations of ethanol consisting of 10, 20, 40, 60, 80 and 100% ethanol in water for 10 minutes for each step. The coverslips were then dried in a Balzers CPD 030 Critical Point Dryer (Balzers Pfeiffer, Balzers, Liechtenstein) and mounted onto 25-mm aluminum stubs with double-sided carbon tabs and then gold-coated in a Dynavac ‘Xenosput’ magnetron sputter coater (Dynavac, Hingham MA, USA). The cells on coverslips were imaged with the Philips XL30 field emission scanning electron microscope (Philips, Eindhoven, Netherlands) at a voltage of 2 kV.

Merozoites were stained with Hoechst after opsonization with purified IgG from malaria-exposed Kenyan adult sera after which they were co-incubated with THP-1 cells. The THP-1 cells were then resuspended in 2% PFA and mounted on a glass slide. The slides were immediately analyzed by Zeiss Cell Observer (North Ryde, New South Wales, Australia. using a 100× magnification object lens.

Human antibodies against the K1 allelic version of MSP3 [41] and the FC27 allele of MSP2 [42] of P. falciparum were affinity purified from a 50 mL pool of plasma taken from malaria semi-immune adults in Kenya (described under Study Populations, n = 20) by column chromatography (CNBr-activated SepharoseTM 4B, GE Healthcare), according to the manufacturer's instructions, and as previously described [43].

Purified whole merozoites [34, 35] were resuspended in PBS supplemented with a cocktail of protease inhibitors (Roche, Castle Hill NSW, Australia). Whole merozoites were then plated in NUNC Maxisorp™ plates at 100 μL per well and incubated at 37°C for two hours (or overnight at 4°C), followed by six washes with PBS. The plates were then blocked with 200 μL of 1% casein in PBS at 37°C for two hours followed by two washes with PBS. Serum samples were diluted at 1:1000 in PBS and 100 μL of each sample was added to the ELISA plates in duplicate and incubated at 37°C for one hour, followed by six washes in PBS. The plates were further incubated with 100 μL per well of horseradish peroxidase (HRP)-conjugated goat anti-human IgG antibody diluted at 1:2500 in 0.1% PBS-casein at 37°C for one hour followed by six washes with PBS. For determination of IgG1 and IgG3 subclasses, peroxidase conjugated anti-human IgG1/IgG3 antibodies were diluted at 1:1000 in 0.1% PBS-casein. Finally, 100 μL of 2,2'-azino-bis(3-ethylbenzothiazoline-6-sulphonic acid) (ABTS) was added to each well and incubated for 20 minutes for color development. The reaction was stopped by the addition of 100 μL 1% SDS solution. Color development was quantified at 405 nm. Pooled human IgG from malaria exposed Kenyan adults was used at 1:1000 dilution as a positive control and individual sera from malaria non-exposed Melbourne adults were used at 1:1000 dilution as negative controls.

Data were analyzed using Prism 5 (GraphPad Software, Inc) and Stata 11 (StataCorp). For the Ngerenya cohort, the presence/absence of phagocytosis was analyzed in relation to single and multiple clinical episodes of malaria using the modified Poisson regression model [44]. Subgroup analyses were performed for children who had recent exposure to malaria (positive malaria slide) in the six months before samples were collected. In Chonyi, >94% of children were positive for phagocytosis, precluding the use of the presence/absence of phagocytosis to relate to malaria risk. However, the levels of phagocytosis were nearly normally distributed allowing us to define three levels of phagocytosis by tertiles: high, medium and low. Risk of malaria episodes was analyzed for each tertile using standard survival analysis techniques [17]. All analyses included age as a potential confounder.

We developed and validated an assay to measure opsonic phagocytosis using the human monocyte THP-1 cell line and purified intact merozoites isolated using recently established methods [34, 35]; we refer to these as opsonic phagocytosis assays (OPA). We demonstrated that antibody-mediated opsonic phagocytosis: (1) was specific to the IgG fraction of malaria-exposed sera in a dose-dependent manner; (2) was inhibited by pre-treatment of THP-1 cells with cytochalasin D, which is a known inhibitor of macrophage and monocyte phagocytosis; (3) was comparably quantified using either THP-1 cells or freshly isolated human PBMCs; and (4) robustly measured the internalization of merozoites into phago-lysosomes, as demonstrated by staining merozoites with the pH-dependent dye pHrodo™, which only becomes fluorescent in the acid environment of phago-lysosomes (Figure 1A-D). We found equivalent levels of phagocytosis using merozoites stained with pHrodo™, compared to those stained with EtBr, indicating that our assay robustly quantified internalized merozoites and was not confounded by surface bound merozoites. Opsonic phagocytosis was optimally detected at 10 minutes [see Additional file 1: Figure S1], and our OPA gave results that were reproducible over a wide range of phagocytic activity [see Additional file 1: Figure S2]; we demonstrated activity using merozoites purified from several different P. falciparum isolates (isolates D10, 3D7, E8B, CS2, W2mef). Opsonic phagocytosis led to activation of monocytes demonstrated by increased intracellular TNF-α production, which is thought to play an important role in parasite clearance and immunity [18], and up-regulated CD69 expression (Figure 2A-B). Scanning electron microscopy captured striking images showing attachment of merozoites to THP-1 cells and the commencement of phagocytosis and internalization (Figure 3A). Internalized merozoites could also be clearly seen using immunofluorescence microscopy (Figure 3B).

Antibody responses were studied in detail in a subset of samples (n = 33) from children and adults to define relationships between opsonic phagocytosis antibodies and other measures of immunity. To understand better the antibodies promoting phagocytosis and the relationship between antibody binding to the surface of merozoites with OPA, we developed an assay to measure antibodies to intact merozoites by ELISA. Activity in OPA was significantly and positively correlated with IgG reactivity against intact purified merozoites, and mediated predominantly by IgG1 and IgG3 subclasses (Figure 4A-B). The GIA is currently the most widely used functional assay for anti-merozoite antibodies, but has not been consistently associated with protection in naturally-acquired or vaccine-induced immunity. We found that opsonic phagocytosis was only weakly correlated with the ability of the same purified IgG to inhibit parasite growth in a standard GIA (Spearman’s rho −0.358, P = 0.041, Figure 4C); similarly, total IgG against whole merozoites was also only weakly correlated with inhibitory activity in GIAs (Spearman’s rho −0.410, P = 0.018, Figure 4D). Others have reported variable correlations (negative, positive and non-significant) between growth inhibitory antibodies and exposure, or antibodies to merozoite surface proteins (MSPs), in some African populations [11, 12, 45], including Kenya, which has raised questions about the GIA as a correlate of human immunity.

In Ngerenya, 48% had antibodies mediating phagocytosis. This proportion rose to 90% among those who had asymptomatic infections at the time of sampling (Figure 5A), suggesting that active infection boosted opsonic antibodies. In Chonyi, where all children had asymptomatic infections, the prevalence of phagocytosis mediating antibodies was comparably high at 94.5% (Figure 5A). In both cohorts, the OPA activity positively correlated with age, reflecting increasing cumulative exposure to P. falciparum; this was statistically significant in the Ngerenya cohort (Figure 5B), but not in Chonyi (Figure 5C). In Ngerenya, the RPI was significantly higher in children who had active P. falciparum infections at sampling compared to aparasitemic children (Figure 5D). Opsonic phagocytosis was also higher for children who had P. falciparum infections in the six months prior to sample collection compared to those who had not been infected (mean RPI 33.4 versus 14.1, P <0.001), again indicating efficient boosting of opsonic phagocytosis. In contrast, growth-inhibitory antibodies were no different among children with infection in the preceding six months than in those without (mean growth inhibition 98.9% versus 98.7%, P = 0.867, Figure 5E), or those with concurrent parasitemia versus uninfected. The RPI was also significantly higher in age-matched children from Chonyi than in those from Ngerenya, consistent with the differences in malaria transmission intensity (Figure 5F). In Ngerenya, children less than six months of age had a higher RPI compared to those within the six-months to one-year age-group, suggesting placentally-transferred maternal opsonic phagocytosis antibodies (mean RPI 18.11 versus 5.22, P = 0.013).

Prior studies showed that antibodies to MSP2 and MSP3 measured by ELISA were the antigen-specific responses most strongly associated with protective immunity in our study population [6]; IgG to MSP2 remained significantly associated with protection in our multivariate analysis described above. We sought to understand the function of these target-specific antibodies, given that they have limited activity in standard GIAs [46‐48]. We affinity-purified human antibodies using recombinant MSP2 and MSP3 from a pool of sera from Kenyan adult residents with extensive malaria exposure. These antigen-specific affinity-purified antibodies showed strong concentration-dependent opsonic phagocytosis activity compared to the unopsonized control (Figure 7), providing the first evidence of a mechanism that may account for this protective association, and identifying two important targets of opsonic phagocytosis.