Plasmodium falciparum Glutamic Acid-Rich Protein-Independent Polyclonal Antibodies Inhibit Malaria Parasite Growth in Human Erythrocytes

Abstract

Plasmodium falciparum glutamic acid-rich protein (PfGARP) is a recently characterized cell surface antigen encoded by Plasmodium falciparum, the causative agent of severe human malaria pathophysiology. Previously, we reported that the human erythrocyte band 3 (SLC4A1) serves as a host receptor for PfGARP. Antibodies against PfGARP did not affect parasite invasion and growth. We surmised that PfGARP may play a role in the rosetting and adhesion of malaria. Another study reported that antibodies targeting PfGARP exhibit potent inhibition of parasite growth. This inhibition occurred without the presence of any immune or complement components, suggesting the activation of an inherent density-dependent regulatory system. Here, we used polyclonal antibodies against PfGARP and a monoclonal antibody mAb7899 to demonstrate that anti-PfGARP polyclonal antibodies, but not mAb7899, exerted potent inhibition of parasite growth in infected erythrocytes independent of PfGARP. These findings suggest that an unknown malaria protein(s) is the target of growth arrest by polyclonal antibodies raised against PfGARP.

PfGARP is an erythrocyte surface antigen encoded by Plasmodium falciparum, the causative agent of severe human malaria pathophysiology. We provide evidence for a mechanism that targets growth arrest of intraerythrocytic parasites by polyclonal antibodies against PfGARP via a novel PfGARP-independent pathway.

Malaria is a major parasitic disease that afflicts a significant fraction of the world's population [1]. The World Health Organization estimated 247 million cases of malaria infection worldwide in 2021 resulting in approximately 627 000 deaths due to complications of malaria, mainly in children in sub-Saharan Africa. This dramatic increase in mortality underscores the need for improved mechanistic understanding of the disease, ultimately contributing to the development of new treatment strategies. Plasmodium falciparum is the most prevalent malaria parasite in sub-Saharan Africa, accounting for 99% of malaria cases and deaths. A unique feature of P. falciparum malaria is the adhesion of infected red blood cells (iRBCs) to host endothelial cells [2]. Previously we identified an exported protein called P. falciparum glutamic acid-rich protein (PfGARP) as a potential contributor to cytoadhesion [3]. PfGARP (PF3D7_0113000) is expressed during the trophozoite and schizont stages of P. falciparum development consistent with knob formation and the cytoadhesion phenomenon [4, 5]. Unlike the var gene family including PfEMP1, there is only 1 PfGARP gene on chromosome 1, of which no known paralogs exist in the parasite genome. PfGARP consists of 26% glutamic acid residues [6] and is only present in other human Plasmodium species except P. falciparum, and primate parasites including Plasmodium reichenowi and Plasmodium gaboni [7–9]. These features are consistent with the presence of antibodies against PfGARP in plasma of children resistant to malaria [10]. Moreover, PfGARP gene and protein expression were increased in parasites isolated from Tanzanian children with malaria [10]. Therefore, a possibility exists for therapeutic targeting of PfGARP as a potential vaccine candidate.

A previous study has shown that the N-terminal lysine repeats of PfGARP localize PfGARP to the RBC membrane [11]. This evidence is consistent with our previous observation that human RBC band 3 anion exchanger functions as a potential extracellular host receptor for PfGARP [3]. Recently, Raj et al reported that antibodies targeting the C-terminal segment of PfGARP exhibit potent inhibition of trophozoite-infected RBCs by inducing programmed cell death of iRBCs [12]. Because our polyclonal and monoclonal antibodies raised against a specific segment of PfGARP (amino acids [aa] 370–444) failed to show any growth inhibitory activity of iRBCs [3], we compared the inhibitory activity of monoclonal and polyclonal antibodies raised against recombinant PfGARP (aa 410–673) under identical conditions as reported [12]. Our results demonstrate that the monoclonal antibody termed mAb7899 [12] does not exert any inhibitory activity, and polyclonal antibodies raised against PfGARP (aa 410–673) inhibit parasite growth in both wild-type and PfGARP knockout parasites. These findings suggest a mechanism for growth arrest of parasites by a PfGARP-independent mechanism with implications for novel therapeutics and multisubunit vaccine against malaria.

METHODS

Parasite Culture

P. falciparum wild-type 3D7 (Malaria Research and Reference Reagent Resource Center, Manassas, VA) and PfGARP knockout parasites (Jeffrey D. Dvorin, Harvard Medical School) were cultured in vitro in complete malaria medium (CMM) containing RPMI-1640 supplemented with 0.5% AlbuMax II, 25 mM HEPES, 50 mg/L hypoxanthine, and 50 mg/L gentamicin in a 37°C incubator maintained with a gas mixture of 5% CO₂, 3% O₂, and balanced by N₂. Blood smears were fixed with 100% methanol for 30 seconds, stained by Wright Giemsa, and parasitemia was quantified by microscopy (60× magnification, oil immersion). To harvest synchronized parasites and culture supernatant, iRBCs infected with mature-stage parasites from 3D7 and PfGARP knockout (KO) cultures were purified using a Miltenyi Biotec LS magnetic column (magnetic-activated cell sorting [MACS]).

Purification of His-GARP-410 and Generation of Polyclonal Antibodies

The Escherichia coli codon-optimized open reading frame of PfGARP encoding aa 410–673 (GenScript) was subcloned into the pET30a vector using the BamHI and EcoRI restriction sites, and protein expression was induced with Isopropyl ß-D-1-thiogalactopyranoside (IPTG) overnight at 18°C. Protein purification was performed through a 2-step process using a nickel-nitrilotriacetic acid (Ni-NTA) matrix and Mono-Q column on an Äkta fast protein liquid chromatography (FPLC) system. Purified His-GARP410 was emulsified in an equal volume of TiterMax Gold adjuvant, and 50 µg antigen was injected intraperitoneally at 2-week intervals for a total of 4 doses in female BALB/cJ mice.

Generation of mAb7899 Monoclonal Antibody

The mAb7899 was produced commercially by Absolute Antibody, United Kingdom. This monoclonal antibody was designed based on the DNA sequences of heavy- and light-chain variable regions of mAb7899 published earlier [12]. For consistency, we obtained the monoclonal antibody used in this study from the same company that synthesized the mAb7899 monoclonal antibody used in the previous study; all conditions for production were identical [12]. All conditions were identical for the production of mAb7899. The quality control certificate analysis by Absolute Antibody showed a single band of mAb7899 under nonreducing conditions with endotoxin level of <0.069 EU/mg as determined by Limulus amebocyte lysate (LAL) chromogenic endotoxin assay. The properties of mAb7899 are as follows: isotype (mouse lgG1), molecular weight (146 355.94 Da), and extinction coefficient (239 070 M⁻¹ cm⁻¹). Mouse mAb7899 was purified by affinity chromatography in 20 mM histidine solution, pH 6.0, and filtered by 0.22 μm with a final concentration of 1.0 mg/mL (determined by Absolute Antibody).

Immunoblotting and Immunofluorescence Assays

For isolation of mature-stage parasites, cultures were passed through a magnetic column (Miltenyi Biotec/MACS) and resuspended in distilled water (1:25. v/v) followed by centrifugation at 3000g. The schizont pellets were resuspended and lysed in sodium dodecyl sulfate (SDS) sample buffer with reducing agent and clarified at 3000g for 5 minutes. For radioimmunoprecipitation assay buffer (RIPA) extraction of the schizonts pellet, parasites were resuspended in RIPA buffer (150 mM sodium chloride, 50 mM Tris-HCl pH 8.0, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS) on ice for 10 minutes and clarified at 3000g for 5 minutes. Alternatively, parasite pellets were prepared by lysis with 0.15% saponin in phosphate-buffered saline (PBS) on ice for 10 minutes, followed by washing with PBS before solubilization with SDS-polyacrylamide gel electrophoresis (SDS-PAGE) sample buffer. Proteins were analyzed by 12% or 4%–20% gradient SDS-PAGE and immunoblotting. The mAb7899 was diluted 1:1000 in the blocking buffer and incubated for 2 hours at 37°C before developing with an anti-mouse-horseradish peroxidase immunoglobulin G (IgG). The His-GARP-410 polyclonal serum was diluted at 1:100, 1:200, or 1:500 and incubated overnight at 4°C. Secondary antibodies (Cell Signaling Technology DyLight anti-mouse 800 and anti-rabbit 680) were diluted to 1:25,000 or 1:30,000 in blocking buffer for 45 minutes in the dark at room temperature and signal was detected using an Odyssey CLx. An anti-Pfaldolase rabbit polyclonal antibody diluted at 1:1000 was used to normalize loading of parasite proteins. The V5-tagged PfGARP 3D7 parasite line was cultured to 200 mL at 2% hematocrit and 4% parasitemia and tested by immunoblotting with an anti-V5 mAb (Bio-Rad SV5-Pk1). Mature-stage parasites were MACS purified and lysed in RIPA buffer with protease inhibitors (Pierce Tablets catalog No. PIA32955; Thermo Scientific).

For immunofluorescence assay (IFA), thin blood smears were air dried and fixed with ice-cold methanol. Smears were blocked with 3% bovine serum albumin in PBS for 30 minutes and washed twice with PBST (0.05% Tween-20). Anti-glycophorin A (1:200 dilution), mAb7899 (1:1000 dilution), or anti-PfGARP antisera (1:25–1:200) were diluted in blocking buffer and incubated at room temperature for 1 hour. Smears were washed 3 times in PBST and incubated with anti-mouse IgG conjugated with Alexa Fluor 488 (1:1000) for signal detection. Labeled smears were mounted with ProLong Diamond Antifade mountant with 4′,6-diamidino-2-phenylindole (DAPI). Cells were imaged using a Nikon Eclipse TE2000-E microscope with an exposure of 300 msec.

Parasite Growth Inhibition Assays

Parasites (wild-type 3D7 and PfGARP knockout) were synchronized to the ring stage through 5% sorbitol treatment. Ring-stage parasites were plated at a final parasitemia of 0.4%–1% in a final hematocrit of 2%–4% in a total volume of 100 µL in microtiter wells. Nonimmune and immune mouse sera were heat inactivated at 56°C for 30 minutes and precleared with human RBCs for 1 hour before use in the assay. The mAb7899 monoclonal antibody was used under conditions as reported earlier [12]. Parasite measurements were performed in triplicate over 3 independent biological experiments. Anti-GARP410 serum was diluted 1:10 for use in the inhibition assay. Due to limited quantity, inhibition assays with mouse anti-GARP410 antisera were performed in 2 independent experiments using technical triplicates. After 48 hours, thin blood smears were prepared and stained with Wright-Giemsa. Human RBCs infected with ring-stage parasites were quantified with a total of 500 RBCs per slide by at least 2 blinded microscopists (CJS and RMK) independently and the results from the triplicate wells were averaged. Statistical analysis comparing treatment groups to control was performed using a 1-way ANOVA. Generation of polyclonal antibodies in mice was approved by the Tufts University Institutional Animal Care & Use Committee (IACUC). Human erythrocytes obtained from a blood bank are not subject to approval for in vitro studies.

RESULTS

The observations reported by Raj et al [12] were based on the development of 2 key antibodies against PfGARP, including polyclonal antibodies generated in mice and a recombinant monoclonal antibody termed mAb7899. Both antibodies recognized an approximately 100-kDa band, presumed to be native PfGARP, by immunoblotting and exerted potent inhibition of P. falciparum growth in human erythrocytes [12]. No growth inhibition of parasites was observed in PfGARP null (KO) parasites consistent with their model that recognition of native PfGARP on the surface of iRBCs is required to trigger programmed cell death [12]. To compare the conditions between our published findings [3] demonstrating no inhibitory effects of polyclonal and monoclonal antibodies [3] and the potent growth inhibition activity of antibodies published earlier [12], we first generated an identical mouse polyclonal antibody raised against the same amino acids (410–673) of PfGARP [12]. In addition, the previous study [12] generated a monoclonal antibody, termed mAb7899, commercially (Absolute Antibody) and published the DNA sequences of the heavy- and light-chain variable regions [12]. For consistency, we obtained the identical mAb7899 monoclonal antibody synthesized by the same commercial supplier that used purification conditions as reported in the previous study [12].

The polyclonal antibodies we generated in mice against PfGARP (His-GARP410; aa 410–673) served as an internal positive control (Figure 1A, red). Because the polyclonal antibodies generated by Raj et al [12] were not available due to their limited quantity, we constructed an identical PfGARP plasmid to express recombinant PfGARP (aa 410–673). Using tandem affinity and ion-exchange chromatography (FPLC), a highly purified recombinant carboxyl segment of PfGARP (termed His-GARP410 or GARP410) was used for immunization of mice using TiterMax gold adjuvant (Figure 1B). The genetic background of mice and the adjuvant used in our experiments were identical to those used in the previous study [12]. Immunoblotting identified 3 of 4 immunized mice exhibiting a robust immune response, whereas non- and preimmune sera showed no reactivity (Figure 1C). Serum samples from the final bleeds detected recombinant His-GARP410 protein (Figure 1D, right lane) as well as an independent construct termed MBP-GARP-L (Figure 1D, left lane) showing sequence overlap with the construct used by Raj et al [12] (Figure 1A, blue and red). Importantly, in contrast to the previous study [12], our polyclonal sera failed to detect any approximately 100-kDa band in 2 freshly prepared 3D7 and FCR3 stocks of wild-type parasite lysates even at dilutions as low as 1:100 (Figure 1E and 1F, asterisks). An anti-Pfaldolase rabbit polyclonal antibody was used to confirm adequate loading of the parasite proteins (Figure 1E and 1F, lower panels). A potential loss of PfGARP gene expression in the parasite strains was ruled out by independent on-bead mass spectrometry analysis of immunoprecipitated proteins at Harvard Meical School proteomics core facility.

Figure 1.

Generation of PfGARP constructs and characterization of antisera in mice. A, Schematic of full-length PfGARP (black), and MBP-GARP-L (blue) and His-GARP410 (red) constructs. Relevant size of each construct is given and monoclonal antibody epitope of mAb7899 is marked with a black box in GARP410 construct. B, Coomassie staining of fast protein liquid chromatography-purified His-GARP410 by chromatography through Ni-NTA and Mono-Q columns. C, Dot immunoblotting with the final bleeds of mouse sera using His-GARP410 Ag immobilized on nitrocellulose. D, Immunoblotting with the most reactive polyclonal antibody against His-GARP410 serum (1:2000 dilution) using MBP-GARP-L and His-GARP410 recombinant proteins. E, Immunoblotting with anti-His-GARP410 serum (1:100 dilution) using 3D7 wild-type saponin-lysed parasite lysate. F, Immunoblotting with anti-His-GARP410 serum (1:100 dilution) using FCR3 wild-type saponin-lysed parasite material. Asterisks denote the location where PfGARP signal should be detected. The polyclonal antibody detected the recombinant His-GARP410 protein in both E and F. Anti-Pfaldolase antibody was used as a loading control in both E and F. Representative data of at least 3 independent experiments. G, Immunoblotting with mAb7899 (1:1000 dilution) using 3D7 wild-type saponin-lysis parasite lysate. An asterisk marks the location of putative approximately 100-kDa band of PfGARP (left). An internal positive control with MBP-GARP-L was included on the same membrane (right). Anti-Pfaldolase was used as a loading control (lower). Identical results were obtained using mAb7899 kindly shared by the authors of previous study [12] and mAb7899 obtained by us directly from Absolute Antibody. Representative of at least 3 independent experiments. H, Immunoblotting with anti-His-GARP410 polyclonal serum (1:200 dilution) on water-lysis of magnetic-activated cell sorting-purified schizonts separated on a 4%–20% SDS-PAGE gradient gel. Anti-Pfaldolase (red) was used as loading control. Representative of at least 3 independent experiments. I, Immunoblotting of V5-PfGARP construct by anti-V5 mAb (left). As expected, V5-PfGARP protein was detected by anti-V5 mAb and no signal was detected in uRBC ghosts. The same V5-PfGARP construct was blotted with anti-GARP410 polyclonal antibody (1:100 dilution; middle). No signal from V5-PfGARP was detected by anti-GARP410 polyclonal antibody. Nonspecific spectrin staining confirmed the protein loading of uRBC ghosts. The same amount of parasite lysate was blotted with anti-Pfaldolase antibody to confirm parasite protein loading (right). J, Immunofluorescence analysis of iRBCs (magnification 100X, oil immersion) with preimmune and anti-His-GARP410 polyclonal serum (1:25 dilution) overlayed with DAPI (blue) signal. Representative images from at least 2 independent experiments. Abbreviations: Ag, antigen; DAPI, 4′,6-diamidino-2-phenylindole; iRBC, infected red blood cells; mAb, monoclonal antibody; PfGARP, Plasmodium falciparum glutamic acid-rich protein; SDS-PAGE, sodium dodecyl sulfate polyacrylamide gel electrophoresis; uRBC, uninfected red blood cells; MBP, maltose binding protein.

To investigate whether endogenous PfGARP can be detected by mAb7899, parasite lysates were prepared using either saponin lysis or hypotonic lysis with water of magnetically (MACS) purified schizonts following the published protocol [12]. Using both saponin and water-lysed MACS schizont lysates, our multiple attempts failed to detect the approximately 100-kDa band of PfGARP with mAb7899 under these conditions (Figure 1G, asterisk). As expected, a positive control of recombinant MBP-GARP-L protein (Figure 1A, blue) was detected by mAb7899 on the same nitrocellulose membrane (Figure 1G, right lane). An anti-Pfaldolase rabbit polyclonal antibody was used to confirm loading of the parasite proteins (Figure 1G, lower panel). To resolve this discrepancy, we contacted the authors of the previous study [12] who kindly shared a very limited amount of their mAb7899 (20 µL) to confirm the presence of endogenous PfGARP by immunoblotting. Our attempts failed to detect any endogenous PfGARP by their mAb7899 along with positive controls shown in Figure 1G. Because Raj et al [12] detected the endogenous approximately 100-kDa band of PfGARP with their polyclonal antibody using the LiCOR detection system, we tested our PfGARP polyclonal antibody raised against His-GARP410 (Figure 1A, red) using a fluorescent secondary antibody suitable for the LiCOR system (Figure 1H). We used water-lysed MACS-purified schizonts lysate to detect PfGARP with the polyclonal antibody. No endogenous approximately 100-kDa band of PfGARP was detected using the LiCOR system (Figure 1H). A nonspecific approximately 250-kDa doublet was identified as host spectrin using antibodies against human erythrocyte membrane spectrin (Figure 1H). Anti-mouse and anti-rabbit secondary antibodies alone showed no cross-reactive bands (data not shown). Based on these results, we conclude that polyclonal antibodies raised against His-GARP protein (Figure 1A, red) and mAb7899 monoclonal antibody (Figure 1G) do not recognize the endogenous approximately 100-kDa PfGARP in the parasite lysates.

To further investigate the lack of detection of endogenous PfGARP by multiple antibodies, we utilized a V5-tagged PfGARP construct kindly provided by the Dvorin laboratory (Harvard Medical School). This is the same V5-tagged PfGARP construct that was used in the previous study [12]. We confirmed the expression of V5-PfGARP by anhydrotetracycline induction using a V5-mAb (Figure 1I, left panel, right lane). The same parasite lysate was probed with the polyclonal antibody raised against the C-terminus GARP410 (Figure 1A, red). No V5-PfGARP signal was detected by immunoblotting (Figure 1I, right panel, right lane). Nonspecific spectrin staining confirmed the loading of RBC membrane proteins (Figure 1I, right panel, left lane). Because several immunoblotting approaches failed to detect any signal of endogenous and V5-tagged PfGARP, we performed IFA analysis to detect endogenous PfGARP using the anti-PfGARP antibodies. To rule out any potential concern of antibody abundance in the polyclonal serum, we performed IFA at higher antibody concentrations (1:25 dilution). A diffuse iRBC-specific intracellular signal was detected by the preimmune serum (Figure 1J, upper panel). Again, as compared to preimmune serum, no immunoreactive PfGARP signal was detected at the cell periphery by IFA (Figure 1J, lower panel). The lack of PfGARP at the periphery of infected erythrocytes is inconsistent with the intense staining of PfGARP as reported in the previous study [12]. Together, these results demonstrate that neither endogenous and full length recombinant V5-tagged PfGARP nor surface-anchored PfGARP are detectable by IFA using the anti-PfGARP antibodies under these conditions.

A striking observation reported in the previous study [12] showed that both polyclonal and monoclonal antibodies against recombinant C-terminus of PfGARP (Figure 1A, red) exerted potent growth inhibition of intraerythrocytic parasites in a PfGARP-dependent manner [12]. No effect of these antibodies on parasite growth was observed in a PfGARP KO parasite line [12]. Because we could not detect any endogenous PfGARP by multiple anti-PfGARP antibodies (Figure 1), we first measured the growth inhibition activity of anti-PfGARP polyclonal antibodies in both wild-type and PfGARP KO parasites. The PfGARP KO parasite line was generously shared by the Dvorin laboratory (Harvard Medical School). This is the same parasite PfGARP KO strain that was used in the previous study [12]. First, we measured the antibody titer of immunized mice and selected the highest potency titer serum (Figure 2A) for the growth inhibition experiments. Unexpectedly, we observed inhibition of parasite growth in both wild-type and PfGARP KO parasites (Figure 2B and 2C). The growth delay in the polyclonal antibody-treated parasites resulted in the accumulation of higher numbers of schizonts with very few to no ring-stage parasites (Figure 2D). These results demonstrate that the polyclonal antibodies raised against recombinant GARP410 exert potent parasite growth inhibitory activity in both wild-type and PfGARP KO parasites.

Figure 2.

Malaria parasite GIAs using anti-PfGARP polyclonal antibodies. A, Immune response of serially diluted anti–His-GARP410 polyclonal antibodies generated in mice was measured by ELISA using recombinant His-GARP410 antigen as substrate. Experiments were performed in duplicate. B, GIAs were performed in both wild-type 3D7 and PfGARP KO parasites using serum (1:10 dilution) from each immunized mouse shown in A. All experiments were performed as technical triplicates. C, Combined data from B were analyzed with a 1-way ANOVA. The error bars represent mean ± standard deviations of 3 biologically independent replicates. D, Representative images from the PfGARP KO treated parasites from B are shown. Abbreviations: CMM, complete malaria medium; ELISA, enzyme-linked immunosorbent assay; GIA, growth inhibition assays; KO, knockout; ns, not significant; PfGARP, Plasmodium falciparum glutamic acid-rich protein; RBC, red blood cell; WT, wild type. ***= p <0.001; ****= p <0.0001.

Because the polyclonal antibodies produced in mice are generally of limited quantity, the highly potent and abundant mAb7899 monoclonal antibody would be ideal to validate parasite growth inhibition as reported earlier [12]. To investigate this critical issue, we contacted the original commercial supplier of mAb7899 [12] and requested an independent production of recombinant mAb7899 under identical conditions as described [12]. Parasite lysates from magnetically purified schizonts of wild-type 3D7 and PfGARP KO parasites were prepared by lysis with RIPA as well as SDS sample buffers along with the purified GARP410 protein (Ponceau S staining, SDS lysis, Figure 3A; RIPA lysis data not shown). Multiplex fluorescence-based immunoblotting was used to detect the reactivity of affinity-purified mAb7899 along with anti-Pfaldolase as loading control. While mAb7899 detected recombinant GARP-410 protein under both RIPA lysis (Figure 3B, left panel, green band) and SDS lysis (Figure 3B, right panel, green band) conditions, no approximately 100-kDa band representing endogenous PfGARP was detected in wild-type parasite lysates under both lysis conditions (Figure 3B). These results further confirm that endogenous PfGARP is not detectable by mAb7899 in the wild-type parasite lysates. Next, we tested the growth inhibitory activity of mAb7899 under conditions as described earlier [12]. No inhibition of parasite growth was observed by mAb7899 at 200 μg/mL in both wild-type parasites (Figure 3C) and PfGARP KO parasites (Figure 3D). Furthermore, titration of mAb7899 up to 300 μg/mL had no effect on both wild-type and PfGARP KO parasites (Figure 3E). The representative images of parasites treated with CMM and mAb7899 are shown (Figure 3F–H). Together, these findings demonstrate that mAb7899 does not exert any growth inhibitory effect on either wild-type or PfGARP KO parasites (Figure 3).

Figure 3.

Characterization of mAb7899 monoclonal antibody. This antibody was generated by Absolute Antibody, for Tufts University, Boston, and is the same antibody that was used in the previous study [12]. A, Ponceau S-stained image of purified His-GARP410 protein and SDS sample buffer-lysed wild-type 3D7 and PfGARP KO parasite lysates. B, Multiplex fluorescence immunoblots of RIPA-lysed (left panel) and SDS-lysed (right panel) wild-type 3D7 and PfGARP KO parasites. Blots were probed with mAb7899 and rabbit anti-Pfaldolase antibodies. Signals were detected using conjugated anti-mouse 800 and anti-rabbit 680 fluorescent secondary antibodies. Imaging was performed with the Odyssey CLx with autoexposure settings. C, GIA with mAb7899 at 200 µg/mL using wild-type 3D7 malaria parasites. GIAs were performed in triplicates averaging 3 independent biological experiments. The error bars represent mean ± standard deviations of 3 biologically independent replicates. D, GIA with mAb7899 at 200 µg/mL in both wild-type 3D7 and PfGARP KO parasites. The antimalarial drug pyrimethamine was used as control. GIAs were performed in duplicate averaging 2 independent biological experiments. The error bars represent mean ± standard deviations of 2 biologically independent replicates. E, GIA titration with mAb7899 at 100 µg/mL, 200 µg/mL, and 300 µg/mL in both wild-type 3D7 and PfGARP KO parasites. GIAs were performed in duplicate averaging 2 independent biological experiments. The error bars represent mean ± standard deviations of 2 biologically independent replicates. F, Representative Wright-Giemsa smears of PfGARP KO parasites treated with CMM. G, Wild-type D7 parasites treated with mAb7899. H, PfGARP KO parasites treated with mAb7899 (magnification 60×, oil immersion). Abbreviations: CMM, complete malaria medium; GIA, growth inhibition assay; KO, knockout; ns, not significant; PfGARP, Plasmodium falciparum glutamic acid-rich protein; RBC, red blood cell; SDS, sodium dodecyl sulfate; WB, Western blot. ***= p <0.001; **** = p <0.0001.

Finally, we evaluated the erythrocyte membrane reactivity of mAb7899 monoclonal antibody in infected erythrocytes by immunofluorescence microscopy. The erythrocyte membrane localization was confirmed using an anti-glycophorin-A monoclonal antibody (anti-GYPA) (Figure 4A). No membrane reactivity with mAb7899 was detected in mature parasite-infected erythrocytes (Figure 4B). These results provide further support to our model that mAb7899 neither recognizes endogenous PfGARP in the wild-type P. falciparum-infected erythrocytes nor inhibits intraerythrocytic growth under conditions as described in the previous study [12].

Figure 4.

IFA of glycophorin A and mAb7899 reactivity in infected erythrocytes. A, IFA of uninfected RBCs and wild-type 3D7 parasite-infected RBCs labeled with GYPA and conjugated with Alexa Fluor 488 (magnification 100×, oil immersion). B, IFA of uninfected RBCs and wild-type 3D7 parasite-infected RBCs labeled with mAb7899 and conjugated with Alexa Fluor 488 (magnification 100×, oil immersion). Parasite nuclei were labelled with DAPI counterstain (data not shown). RBCs were exposed for 300 milliseconds. Representative of 5 independent experiments. Abbreviations: DAPI, 4′,6-diamidino-2-phenylindole; GYPA, aPanti-glycophorin A; IFA, immunofluorescence analysis; RBC, red blood cell; uRBC, uninfected RBC.

DISCUSSION

It is puzzling as to why the antibodies directed against the recombinant PfGARP carboxyl terminal segment failed to detect any endogenous PfGARP under conditions reported in the previous study [12]. Raj et al [12] utilized protein, mRNA, or DNA vaccination as means of inducing an immune response. In principle, there could be a difference in the folding and posttranslational modifications of polypeptides produced in the eukaryotic system. However, PfGARP expressed in bacteria was detected as an approximately 100-kDa band by immunoblotting using polyclonal antibodies that were affinity purified using recombinant PfGARP bound to Protein A beads (Figure 1K in Raj et al [12]). Therefore, it appears that antibodies generated in mice against recombinant C-terminal segment of PfGARP can recognize denatured full-length PfGARP by immunoblotting of lysates from iRBCs extracted in the RIPA buffer [12]. The detection of approximately 100-kDa PfGARP in the wild-type 3D7 parasite lysate and its absence in PfGARP KO parasites (Figure 4B in Raj et al [12]) by immunoblotting using anti-PfGARP polyclonal antibodies is inconsistent with the data reported in this study. Of note, the original identification of PfGARP [6] also failed to identify endogenous PfGARP polypeptides in both asexual and sexual blood stages using similar immunoblotting techniques [6].

Given that the C-terminus of PfGARP has an unusually high content of glutamic acid residues [6], we speculate that the growth inhibitory activity of polyclonal antibodies against recombinant GARP410 may be targeting an uncharacterized parasite antigen expressed on the surface of iRBCs [4, 5, 10, 13]. There are many poorly characterized malaria parasite proteins encoding highly repetitive glutamic acid-rich sequences that could be recognized by the anti-PfGARP polyclonal antibodies thus clarifying the mechanistic basis of parasite growth inhibition [12]. In contrast, the monoclonal antibody mAb7899 recognizes a linear epitope in PfGARP (aa 443–459; VKNVIEDEDKDGVEIIN) [12]. It is therefore unlikely that mAb7899 recognizes a glutamic acid-rich epitope in the P. falciparum genome. This lack of recognition of an epitope targeted by the anti-PfGARP polyclonal antibodies could be one explanation why mAb7899 failed to inhibit parasite growth in both wild-type and PfGARP KO parasites (Figure 2). In any case, even if the mAb7899 can recognize endogenous PfGARP by immunoprecipitation and mass spectrometry, its biological function remains to be ascertained because it does not inhibit parasite growth as reported earlier [12].

The basis of our inability to detect endogenous PfGARP and inhibition of parasite growth by mAb7899 remains unclear at this stage. In fact, a resolution of this discrepancy may also begin to clarify another similar observation reported earlier by Raj et al [14]. They demonstrated that polyclonal and monoclonal antibodies raised against malaria parasite surface antigen termed PfSEA-1 inhibit parasite egress by suppressing intraerythrocytic parasite replication [14]. A subsequent study by Simon Draper's laboratory [15] at the University of Oxford could not replicate these findings using the similar polyclonal antibodies against PfSEA-1. Because PfSEA-1 is also an important malaria vaccine candidate [14], the availability of monoclonal antibodies developed by Raj et al [14] for PfSEA-1 and PfGARP [12] will help to independently validate this important contradiction thus contributing to the development of an effective multisubunit vaccine against malaria [16]. A pressing need now is to identify the protein target(s) in P. falciparum-infected human erythrocytes recognized by anti-GARP410 polyclonal antibodies that could serve as a potential therapeutic and vaccine target against severe malaria.

Notes

Acknowledgments. We are grateful to Dr Jeffrey Dvorin of Harvard Medical School for kindly sharing the PfGARP knockout and V5-tagged PfGARP parasite lines. We are thankful to Drs Zhiyan Cao and Hongyan Wang for sharing the pET-30a plasmid. We thank Christian Rosa Birriel for help with the confocal microscopy, and Dr Andrew Levin for feedback during the course of this study. Finally, we are grateful to Donna-Marie Mironchuk for her many contributions to the administrative organization of the project, proofreading, and improvements of figures.

Author contributions. A. C. and C. S. conceived and designed the study. C. S. generated recombinant proteins and performed the bulk of the experiments. T. H. generated polyclonal antibodies against PfGARP. R. K. performed experiments using mAb7899. C. S. wrote the initial draft of the manuscript. A. C. assembled and edited the subsequent versions of the manuscript and guided the project from its inception to completion. All authors approved the final manuscript.

Data availability. All data generated or analyzed during this study are available within the article.

Availability of materials. Unique materials are available under reasonable request for nonoverlapping studies. We contacted the same commercial supplier that generated the original mAb7899 for Brown University [12]. The mAb7899 monoclonal antibody generated for Tufts University recognizes the same recombinant C-terminal segment of PfGARP (aa 410–673) used by the previous study [12]. Tufts University has permission to distribute mAb7899 for academic studies. Investigators are welcome to contact Absolute Antibody, UK (info@absoluteantibody.com) directly for mAb7899. Alternatively, we are willing to share our limited stock of mAb7899 and recombinant His-GARP410 protein to serve as a positive control for academic research studies.

Financial support. This work was supported by the National Heart, Lung, and Blood Institute (NIH grant number RO1-HL060961 to A.C.).