If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
PfEBA175 has an important role in the invasion of human erythrocytes by Plasmodium falciparum and is therefore considered a high priority blood-stage malaria vaccine candidate. PfEBA175 mediates adhesion to erythrocytes through binding of the Duffy-binding-like (DBL) domains in its extracellular domain to Neu5Acα2–3Gal displayed on the O-linked glycans of glycophorin-A (GYPA). Because of the difficulties in expressing active full-length (FL) P. falciparum proteins in a recombinant form, previous analyses of the PfEBA175-GYPA interaction have largely focused on the DBL domains alone, and therefore they have not been performed in the context of the native protein sequence. Here, we express the entire ectodomain of PfEBA175 (PfEBA175 FL) in soluble form, allowing us to compare the biochemical and immunological properties with a fragment containing only the tandem DBL domains (“region II,” PfEBA175 RII). Recombinant PfEBA175 FL bound human erythrocytes in a trypsin and neuraminidase-sensitive manner and recognized Neu5Acα2–3Gal-containing glycans, confirming its biochemical activity. A quantitative binding analysis showed that PfEBA175 FL interacted with native GYPA with a KD ∼0.26 μm and is capable of self-association. By comparison, the RII fragment alone bound GYPA with a lower affinity demonstrating that regions outside of the DBL domains are important for interactions with GYPA; antibodies directed to these other regions also contributed to the inhibition of parasite invasion. These data demonstrate the importance of PfEBA175 regions other than the DBL domains in the interaction with GYPA and merit their inclusion in an EBA175-based vaccine.
Background: The GYPA-PfEBA175 interaction is important for erythrocyte invasion by the malaria parasite.
Results: The entire ectodomain of EBA175 interacted with GYPA with different biochemical parameters to the previously determined GYPA-binding fragment containing two DBL domains.
Conclusion: Regions outside of the tandem DBL domains contribute to GYPA binding by EBA175.
Significance: These findings may assist the design of an EBA175-based malaria vaccine.
). Widespread implementation of control measures within the last decade, including artemisinin combination therapy and the use of insecticide-treated bed nets, has resulted in a significant decrease in the incidence of P. falciparum malaria in endemic countries (
). In the face of emerging resistance to artemisinin in parasites and pyrethroid insecticides in mosquito vectors, however, the need for an effective vaccine for long term control and prevention of malaria remains an important global health objective (
). Vaccines that target the blood stage of the infection are conceptually attractive because the parasite is directly exposed to the host humoral immune system, and it is this stage of the life cycle that is responsible for all the clinical symptoms of malaria (
). The blood stage is initiated when an extracellular form of the parasite called the merozoite recognizes and invades host erythrocytes. Invasion is a complex process involving multiple interactions between host erythrocyte receptors and parasite ligands displayed on the merozoite surface. PfEBA175 was the first P. falciparum invasion ligand identified and interacts with the highly abundant erythrocyte surface sialoglycoprotein, glycophorin-A (GYPA)
). PfEBA175 is considered a leading vaccine candidate because antibodies directed against PfEBA175 are present in malaria-immune individuals, and antibodies raised against recombinantly expressed fragments of PfEBA175 inhibit erythrocyte invasion by P. falciparum in vitro (
). EBA175 is a member of the erythrocyte-binding-like family of Plasmodium proteins, which include the Duffy-binding proteins of Plasmodium vivax and Plasmodium knowlesi (PvDBP and PkDBP) and the P. falciparum paralogs EBA140, EBA181, and EBL1. The members of the erythrocyte-binding-like family share a similar gene structure, and this homology has been used to define six regions, RI–RVI, in their ectodomains (Fig. 1A) (
). Co-crystallization of the RII fragment with a structural analog of this disaccharide, α-2,3-sialyllactose, revealed six putative glycan-binding sites at the dimer interface of the parasite protein, with four of the sites located within two channels that span the dimer and another two in a deep groove accessible only through a cavity at the top of the dimer (
) have proposed a “receptor-induced dimerization model” for the erythrocyte binding of PfEBA175, which postulates monomeric PfEBA175 assembling into a dimer around the dimeric extracellular region of GYPA during invasion. This model is also consistent with recent structural studies investigating the mode of binding of the PvDBP RII to its sulfotyrosine-carrying receptor, DARC.
Although studies of the 68-kDa recombinant PfEBA175 RII have proven highly informative, native PfEBA175 is a much larger protein, being synthesized as a 190-kDa membrane-tethered precursor that is proteolytically cleaved during erythrocyte invasion to release the 175-kDa extracellular region (
), much of the biochemical characterization of the PfEBA175-GYPA interaction has been performed using just the RII fragment.
In the study reported here, we expressed the full-length ectodomain of PfEBA175 (PfEBA175 FL) in a soluble recombinant form using a mammalian expression system. We confirmed that the recombinant protein is biochemically active and exhibited binding properties similar to those of native PfEBA175 isolated from parasite cultures. Using this protein, we investigated the biophysical parameters of the EBA175-GYPA interaction and compared them with those of the RII fragment to demonstrate that regions outside the DBL domains contributed to GYPA binding. These data have important implications for the rational design of an effective PfEBA175-based malaria vaccine.
Recombinant Expression and Purification of Proteins
The sequence encoding the entire extracellular domain of P. falciparum (3D7) EBA175 (PfEBA175 FL) except the signal peptide (amino acids 21–1424) was made by gene synthesis (GeneART). The codons were optimized for expression in human cells, and the potential N-linked glycosylation sites were removed as described (
). The sequence coding for RII of PfEBA175 (amino acids 142–764) was amplified from this construct by PCR. The coding sequences were cloned into pTT3-derived vectors using unique flanking NotI and AscI restriction enzyme sites, between the leader sequence of the mouse variable κ light chain 7–33 (
) and a rat Cd4 (Ig-like domains 3 and 4) tag, followed either by an enzymatically biotinylatable peptide tag, the pentamerization domain of the rat cartilage oligomeric matrix protein (COMP), and ampicillin resistance protein β-lactamase, or a hexa-His tag (
), and excess unconjugated d-biotin was removed by extensive dialysis into HBS. His-tagged proteins were purified from the culture supernatants by affinity chromatography on HisTrap HP columns (GE Healthcare) using an ÄKTAxpress (GE Healthcare) according to the manufacturer's instructions. Size exclusion chromatography (SEC) of nickel-purified proteins was carried out on a Superdex 200 Tricorn 10/600 column (GE Healthcare) in HBS-EP (HBS, 3 mm EDTA, 0.005% v/v Surfactant P20 (GE Healthcare).
Proteins were resolved under reducing conditions, blotted, and detected using horseradish peroxidase (HRP)-conjugated extravidin (Sigma) essentially as described (
). Commercially available glycophorin-A preparations (catalog numbers: G7903 and A9791, Sigma) were biotinylated in vitro by incubation with a 20-fold molar excess of EZ-link sulfo-NHS-biotin (Thermo Scientific) for 30 min at room temperature and dialyzed into HBS prior to its use in the assays. When testing for immunoreactivity, proteins were immobilized with or without prior treatment at 80 °C for 10 min.
Erythrocyte Binding Assays
Biotinylated PfEBA175 and Cd4 (negative control) were multimerized by immobilization on streptavidin-coated Nile Red fluorescent 0.4–0.6-μm microbeads (Spherotech Inc.) by incubation for 45 min at 4 °C (
). The bead arrays were then presented to erythrocytes in flat-bottomed 96-well microtiter plates at a density of ∼3 × 105 cells/well (the ratio of cells/beads = 1:200). After incubation for 1 h at 4 °C with the protein arrays, the cells were washed twice in HBS + 1% BSA (HBS/BSA) and analyzed by flow cytometry. The data were acquired on an LSRII cytometer (BD Biosciences) using the FACS Diva software (BD Biosciences). Nile Red was excited by a blue laser and detected with a 575/26 filter. Forward scatter and side scatter voltages of 430 and 300 V, respectively, and a threshold of 26,100 on forward scatter were applied when analyzing erythrocytes. To test for binding specificity, the erythrocytes were pretreated with either tosylsulfonylphenylalanyl chloromethyl ketone-treated trypsin from bovine pancreas (Sigma) (at 0.25, 0.5, and 1 mg/ml), tosyl-lysyl chloromethyl ketone-treated chymotrypsin from bovine pancreas (Sigma) (at 0.25, 0.5 and 1 mg/ml), or 0.1 milliunit/106 cells of Vibrio cholerae neuraminidase (Sigma), for 1 h at 37 °C. Trypsin- and chymotrypsin-treated cells were washed once, treated with 0.5 mg/ml soybean trypsin-chymotrypsin inhibitor (Sigma) for 10 min at room temperature, and then washed twice more before incubation with PfEBA175-coated beads. Neuraminidase-treated cells were washed twice prior to use in the binding assays.
), was adapted for detecting the interactions of recombinant PfEBA175 proteins, expressed as β-lactamase-tagged pentameric “preys” with biotinylated 'bait’ forms of purified GYPA and synthetic carbohydrate probes (GlycoTech). Briefly, the baits and the preys were normalized to activities that have previously been shown to detect transient interactions (monomeric half-lives less than 0.1 s). The biotinylated baits were immobilized on streptavidin-coated 96-well microtiter plates (NUNC) at concentrations sufficient for the complete saturation of the available binding surface/well. The plates were then washed twice in HBST and blocked with HBS/BSA for 0.5–1 h, before addition of normalized β-lactamase-tagged prey proteins and incubation for 2 h. After washes in HBST and HBS, 125 μg/ml nitrocefin, a β-lactamase substrate, was added to the wells, and its hydrolysis was monitored by absorbance measurements at 485 nm on a Pherastar plus (BMG Laboratories). All steps were performed at room temperature.
Lectin Binding Assay with Purified GYPA
Biotinylated lectins (Vector Laboratories) were immobilized on streptavidin-coated 96-well microtiter plates (NUNC) at 10 μg/ml for 1 h. The plates were then washed twice in HBST and blocked with HBS/BSA for 30 min. The immobilized lectins were next incubated with 0.02 mg/ml purified GYPA (Sigma) for 2 h. After washes in HBST, the plates were incubated with 1 μg/ml of the anti-GYPA mouse monoclonal antibody BRIC256 (Abcam) for 1 h, followed by an alkaline phosphatase-conjugated anti-mouse secondary antibody (Sigma) for 30 min, before the addition of p-nitrophenyl phosphate (Sigma) at 1 mg/ml and measuring the absorbance at 405 nm on a Pherastar Plus (BMG Laboratories). All steps were performed at room temperature.
Surface Plasmon Resonance (SPR)
SPR studies were performed on a BIAcore T100 instrument (GE Healthcare) at 37 °C, using HBS-EP (GE Healthcare) as the running buffer. In each experiment, biotinylated baits were immobilized on streptavidin-coated sensor chips (GE Healthcare), with the negative control bait in the “reference” flow cell and an approximate molar equivalent amount of the “query” baits in the other flow cells. Purified analyte proteins were separated by size exclusion chromatography on a Superdex 200 Tricorn 10/600 column just before their use in the SPR experiments. Increasing concentrations of these proteins were injected over the immobilized baits at 20 μl/min for equilibrium measurements and at 100 μl/min for kinetic measurements. The surfaces were regenerated with a pulse of 5 m NaCl at the end of each injection cycle. Duplicate injections of the same concentration in each experiment were superimposable, demonstrating no loss of activity after surface regeneration. Reference-subtracted sensorgrams were analyzed using the BIAcore evaluation software version 1.1.1 (GE Healthcare). To determine the overall equilibrium binding affinity, binding responses in the steady-state region of the sensorgrams (Req) were plotted against analyte concentration (C) and fitted to the following equation: Req = CRmax/(C + KD), where Rmax is the maximum binding response, and KD is the equilibrium dissociation constant. Kinetic constants were calculated by nonlinear regression fitting to the association and dissociation phases of the sensorgrams. To identify the mechanism of binding, the sensorgrams were globally fitted to three predefined interaction models as follows: simple 1:1 binding (A + B ↔ AB, where A is the soluble analyte and B is the immobilized ligand); conformational change (A + B ↔ AB ↔ AB*); and bivalent analyte (AA + B ↔ AAB; AAB + B ↔ AABB).
Multiangle Light Scattering Measurements (MALS)
Size exclusion chromatography was performed on Superose 6 10/30 (PfEBA175 FL) and Superdex 200 10/30 (PfEBA175 RII) columns (GE Healthcare) equilibrated in HBS (GE Healthcare) at 0.4 ml/min. The column was followed in line by a Dawn Heleos-II light scattering detector (Wyatt Technologies) and an Optilab-Rex refractive index monitor (Wyatt Technologies). Molecular mass calculations were performed using ASTRA 18.104.22.168 (Wyatt Technologies) assuming a dn/dc value of 0.186 ml/g.
P. falciparum Culture and Invasion Assays
The 3D7 and Dd2 strains of P. falciparum were cultured in human O+ erythrocytes at 5% hematocrit in complete medium (RPMI 1640 medium containing 10% human serum), under an atmosphere of 1% O2, 3% CO2, and 96% N2. Invasion assays were performed as described previously (
To raise polyclonal sera against PfEBA175 FL and RII, the purified proteins were injected into rabbits (Cambridge Research Biochemicals, Billingham, UK). The sera were purified on HiTrap Protein G HP columns (GE Healthcare) using an ÄKTAxpress (GE Healthcare) and dialyzed into RPMI 1640 medium (Invitrogen) prior to their use. The anti-PfEBA175 FL and anti-PfEBA175 RII antibodies had similar anti-Cd4 activity, as determined by ELISA. An anti-AMA1 polyclonal (
) have prevented a detailed biochemical investigation of the EBA175-GYPA interaction, with most studies limited to a fragment of EBA175 encompassing the two tandem DBL domains (RII). To investigate the biochemical properties and functional activity of a more physiologically relevant PfEBA175 protein, we took advantage of a strategy utilizing a mammalian expression system and codon-optimized gene sequences that had previously been successful for producing biochemically active Plasmodium proteins (
). Using this method, we expressed the entire full-length ectodomain fragment of PfEBA175 (PfEBA175 FL) from the 3D7 isolate of P. falciparum as a C-terminally tagged soluble fusion protein (Fig. 1A). Western blotting of unpurified culture supernatants confirmed the presence of a protein at the expected size (Fig. 1B). To determine whether the recombinant PfEBA175 FL was correctly folded and biochemically active, its immunoreactivity to two mouse monoclonal antibodies, R217 and R218, raised against the RII fragment of EBA175 expressed in Sf21 insect cells and known to bind to nonlinear heat-labile epitopes within PfEBA175 RII, were tested (
). Immunoreactivity to PfEBA175 FL was observed for both R217 and R218. Importantly, this immunoreactivity was heat-labile suggesting that at least the epitopes recognized by R217 and R218 are correctly folded in the recombinant protein (Fig. 1C). We were also able to demonstrate immunoreactivity to other parts of the EBA175 protein using a mouse monoclonal antibody raised against a yeast-expressed PfEBA175 RVI fragment (
), and so to assess the oligomeric state of PfEBA175 FL, the purified protein was analyzed by SEC. The elution profile of PfEBA175 consisted of two peaks, both containing the protein of interest as demonstrated by denaturing SDS-PAGE. The smaller peak eluted in the void volume of the column, and the size of the major peak was estimated to be ∼500 kDa. A similar mass for PfEBA175 FL was observed by native PAGE, consistent with it being primarily homodimeric in solution (Fig. 1D). To further investigate the oligomeric state of PfEBA175 FL, it was subjected to MALS immediately following SEC. This analysis revealed that the main peak of PfEBA175 FL is predominantly monomeric with a certain degree of self-association (Fig. 1E). The early elution of PfEBA175 FL in SEC and its retarded mobility in native PAGE may therefore be due to a large hydrodynamic shape, possibly due to a high degree of conformational flexibility or an elongated structure.
PfEBA175 FL Binds Human Erythrocytes in a Neuraminidase- and Trypsin-sensitive but Chymotrypsin-resistant Manner
To determine whether recombinant PfEBA175 FL is biochemically active, we first asked whether it could bind human erythrocytes in a manner consistent with its recognition of GYPA. Interactions between cell surface proteins typically have low binding affinities, and multimerized, highly avid binding reagents are often required to facilitate their detection (
). Therefore, to estimate the degree to which the PfEBA175 FL protein could associate with human erythrocyte surfaces, we first created a highly avid PfEBA175 FL binding reagent by clustering the biotin-tagged monomeric EBA175 protein around fluorescent streptavidin-coated beads. The EBA175-coated beads were then presented to human erythrocytes, and the extent to which they bound was quantified by fluorescence-activated cell sorting. We observed that PfEBA175 FL-coated beads were clearly able to bind to human erythrocytes relative to the negative control (Fig. 2, A–C). The binding of EBA175 to erythrocytes is known to be sensitive to the treatment of the cells with the enzymes trypsin and neuraminidase but is insensitive to chymotrypsin (
). Therefore, to assess the specificity of the binding of PfEBA175 FL-coated beads, erythrocytes were pretreated with these enzymes. Treatment with trypsin reduced the binding of PfEBA175 FL to erythrocytes in a dose-dependent manner (Fig. 2A), whereas a similar treatment with chymotrypsin had no significant effect, even at the highest concentration (Fig. 2B). Pretreatment of the erythrocytes with V. cholerae neuraminidase, which preferentially cleaves α2–3-linked sialic acids of O-linked tetrasaccharides, was sufficient to prevent all EBA175 binding (Fig. 2C).
PfEBA175 FL Binds Human Native GYPA and Neu5Acα2–3Gal-containing Glycans
To determine whether recombinant PfEBA175 FL bound native GYPA directly, we used a preparation of GYPA extracted from human erythrocytes. First, we characterized the glycan profile of the native GYPA preparation using three lectins with known carbohydrate-recognition specificities, which revealed a predominance of Neu5Acα2–3Gal with a much smaller amount of Neu5Acα2–6-linked glycans (Fig. 2D). An asialylated form of GYPA included as a control appeared to mainly carry Galβ1–3GalNAc (T antigen), as expected. To detect the direct binding of PfEBA175 FL to native GYPA, we chemically biotinylated GYPA and captured it on streptavidin-coated plates before incubating this with PfEBA175 FL expressed as a highly avid β-lactamase-tagged pentamer (
). Using this approach, pentamerized PfEBA175 FL showed saturable binding to GYPA, but not to its asialylated derivative demonstrating that the recombinant PfEBA175 antigen associates directly with GYPA in a “Neu5Acα2–3Gal”-dependent manner (Fig. 2E).
The expression of the entire ectodomains of EBA175 as a highly avid recombinant protein enabled us to systematically assess the glycan binding properties of EBA175 by screening a large panel of 54 synthetic carbohydrates, primarily selected because they or their close derivatives are known to be present at the surface of human erythrocytes (Fig. 2, F and G, and Table 1). All oligosaccharides containing the Neu5Acα2–3Gal determinant were recognized by PfEBA175 FL (Fig. 2G). Previous work has suggested that the glycan binding properties of PfEBA175 may be responsible for the restriction of P. falciparum to humans because it has been reported that PfEBA175 is unable to bind Neu5Gc (N-glycolylneuraminic acid) (
). Interestingly, we observed clear binding of PfEBA175 FL to the Neu5Gc monosaccharide in the screen, and this was subsequently confirmed by SPR (Fig. 2G). Together, these data demonstrate that the recombinant soluble protein consisting of the entire ectodomain of PfEBA175 is biochemically active and can directly bind native GYPA in a Neu5Acα2–3Gal-dependent manner.
TABLE 1List of the detailed contents of the human erythrocyte synthetic carbohydrate panel
PfEBA175 FL and GYPA Directly Interact with a KD of ∼0.26 μm
The ability to express and purify PfEBA175 FL enabled us to investigate its interaction with native GYPA in detail. We first determined the biophysical properties of the PfEBA175 FL-GYPA interaction using SPR as implemented in a BIAcore instrument. Affinity purified PfEBA175 FL was first resolved by SEC, and fractions encompassing the main peak were pooled and serial dilutions injected over GYPA immobilized on a sensor chip. Binding of PfEBA175 FL to native GYPA was observed and quantified once equilibrium had been reached to derive an equilibrium dissociation constant (KD) of ∼0.26 μm (Fig. 3A). Although this interaction is relatively weak as expected, it is still ∼4-fold stronger than the two other P. falciparum ligand-erythrocyte receptor interactions characterized in our laboratory, PfRH5-Basigin and PfMTRAP-Semaphorin 7A, each of which has a KD of ∼1 μm (
). To characterize the mechanistic details of the interaction between GYPA and the entire PfEBA175 ectodomain, we performed a kinetic analysis using PfEBA175 FL. The association and dissociation kinetic rate constants, ka and kd, were determined using nonlinear curve fitting to a set of reference-subtracted sensorgrams using the initial binding and wash-out phases. Both the ka (8.6 ± 0.2 × 105m−1 s−1) and kd (0.09682 ± 0.00008 s−1) values were typical for a relatively weak protein-protein interaction. To gain a better mechanistic understanding of the interaction, the binding data were fitted to three models (Fig. 3B). The fits to the conformational change and bivalent analyte models were similar and better than that of the simple 1:1 binding model (Fig. 3B). To test whether a conformational change was involved in the binding of PfEBA175 FL to GYPA, we injected PfEBA175 FL until saturation was achieved for a range of different contact times. We observed that variations in contact time did not influence the dissociation phase, suggesting that the interaction of PfEBA175 FL with GYPA does not involve a slow, temporally resolvable conformational change that stabilizes the complex (Fig. 3C). Using SEC-MALS, we established that PfEBA175 FL was primarily monomeric in solution albeit with some self-association. To test the relevance of the bivalent analyte model for the PfEBA175-GYPA interaction, we analyzed the propensity of PfEBA175 FL to interact with itself by SPR. We quantified the observed homophilic binding using the dissociation phase of the interaction to avoid the confounding problem of analyte self-association, which would lead to an underestimate of the affinity in equilibrium analyses (Fig. 3D). PfEBA175 FL self-associated with a kd value (∼0.04 s−1), consistent with a role for EBA175 dimerization in its interaction with GYPA.
PfEBA175 RII Interacts with GYPA with a 10-Fold Lower Affinity than PfEBA175 FL
The fragment of EBA175 containing the tandem DBL domains (RII) is known to bind GYPA (
). To compare the binding properties of the RII fragment with the full-length PfEBA175, we also expressed RII as a soluble, recombinant form using HEK293 cells. A biochemical characterization of the PfEBA175 RII protein revealed that it was expressed at the expected size, was recognized by the conformation-sensitive monoclonal antibodies R217 and R218, and bound human erythrocytes in a neuraminidase-sensitive manner suggesting that it is correctly folded (Fig. 4, A–C). PfEBA175 RII eluted as a monodisperse peak at ∼125 kDa when analyzed by SEC consistent with it adopting a primarily monomeric form in solution (Fig. 4D). Further analysis by MALS confirmed that PfEBA175 RII, similar to PfEBA175 FL, was primarily monomeric, but it showed some propensity to self-associate (Fig. 4E).
We used a pentamerized enzyme-tagged PfEBA175 RII protein to show that it could bind GYPA, although in comparison with the normalized PfEBA175 FL control, the interaction detected using this fragment was more sensitive to dilution of the protein, suggesting it had a lower affinity for GYPA (Fig. 5A). Using SPR, we quantified this binding and found that PfEBA175 RII bound GYPA with a KD of ∼2 μm, ∼10-fold weaker in comparison with the PfEBA175 FL-GYPA interaction (Fig. 5B). Consistent with this, a comparative kinetic analysis showed that the weaker interaction strength of PfEBA175 RII was due to a faster dissociation rate, ruling out the possibility that a significant fraction of the PfEBA175 RII protein was functionally inactive (Fig. 5C). One reason for the weaker binding of the RII fragment to GYPA could be due to a reduced ability to self-associate. To examine this, we compared the binding affinity of the RII fragment and the full-length ectodomain to PfEBA175 FL and found that the former interacted with an ∼70-fold weaker binding affinity (Fig. 5D). It has been previously shown that both the GYPA peptide backbone and sialic acid are required for EBA175 binding (
). To determine whether these binding properties could be distinguished and/or attributed to different regions of EBA175 protein, we directly compared the binding of both PfEBA175 FL and PfEBA175 RII analytes to GYPA and the oligosaccharide Neu5Acα2–3Gal. Binding of PfEBA175 RII to GYPA and Neu5Acα2–3Gal was indistinguishable from one another, whereas PfEBA175 FL bound GYPA with ∼2-fold higher affinity than for the glycan alone (Fig. 5E). These results suggest that PfEBA175 RII interacts with the glycan moieties of GYPA but that the extracellular regions of EBA175 outside of RII also contact the polypeptide backbone of GYPA, contributing further binding energy to the interaction.
Comparison of PfEBA175 FL and PfEBA175 RII as Vaccine Candidates
Given the important role of EBA175 in erythrocyte invasion and the early success in expressing region II as an active recombinant protein, this fragment has been advanced as a potential malaria vaccine (
). Because we found that regions of EBA175 outside of the tandem DBL domains influenced its ability to interact with GYPA, we asked whether PfEBA175 FL would be able to elicit a more potent invasion-blocking antibody response than the RII fragment alone. To address this, rabbits were immunized with an equal mass of both proteins to raise polyclonal antisera, which were subsequently purified and tested for their relative ability to inhibit erythrocyte invasion by P. falciparum in vitro. We used two different laboratory strains of P. falciparum that differed in their sensitivity to invade neuraminidase-treated erythrocytes as follows: Dd2, a strain that is dependent on sialic acid, and 3D7, which can invade through a sialic acid-independent route. The anti-PfEBA175 FL and anti-PfEBA175 RII sera, when tested at the same mass per volume (mg/ml) quantities, showed similar efficacy in inhibiting erythrocyte invasion by the two strains of P. falciparum (Fig. 6A). To assess whether antibodies targeting domains other than RII contributed to the inhibition of EBA175-dependent invasion, we first normalized the immunoreactivity of both antisera to RII, which revealed that anti-PfEBA175 FL contained ∼5-fold less anti-RII antibodies than the anti-PfEBA175 RII serum (Fig. 6B). After normalizing for anti-RII immunoreactivity, we observed that antibodies elicited against PfEBA175 FL were able to inhibit parasite invasion more potently than those raised against the RII fragment alone, suggesting that antibodies directed against extracellular regions of PfEBA175 outside of RII contribute to the ability to inhibit invasion (Fig. 6C).
P. falciparum EBA175 has long been considered an attractive anti-malarial vaccine target because of its important role in erythrocyte invasion mediated through its interactions with GYPA expressed on the surface of host erythrocytes. The technical difficulties associated with expressing Plasmodium proteins recombinantly have meant that most biochemical and vaccine research has relied on expressing subfragments of the EBA175 ectodomain, most commonly the tandem DBL domains known as region II. In this study, we successfully expressed the entire full-length ectodomain of PfEBA175 as a functionally active soluble recombinant protein that enabled us to perform a detailed biochemical analysis of its interaction with native GYPA, and we directly compared this with the region II subfragment.
One interesting finding from a systematic interaction screen against a panel of glycans was that the full-length PfEBA175 bound Neu5Acα2–3Gal-containing glycans as expected but also interacted with Neu5Gc. These results are consistent with the observations of Orlandi et al. (
) who reported that the binding of native PfEBA175 to erythrocytes was potently inhibited by Neu5Gc and oligosaccharides containing Neu5Acα2–3Gal. Neu5Gc is not present on human erythrocytes, due to the absence of the enzyme cytidine monophosphate-N-acetylneuraminic acid hydroxylase, which is required for the conversion of Neu5Ac to Neu5Gc; however, it is the predominant form of sialic acid on the erythrocytes of other apes (
). This difference in sialic acid composition has been proposed to be responsible for the restriction of P. falciparum and the related parasite Plasmodium reichenowi to their respective human and chimpanzee hosts. Martin et al. (
) recombinantly expressed PfEBA175 RII and P. reichenowi EBA175 RII (PrEBA175 RII) on the surface of monkey COS cells and observed binding only to human and chimpanzee erythrocytes, respectively, leading to the proposal that PfEBA175 RII recognizes Neu5Ac-carrying GYPA, whereas PrEBA175 RII binds Neu5Gc-containing GYPA. The findings in our study do not support this hypothesis. Further investigation of the contribution of the EBA175-GYPA interaction to the restriction of Plasmodium spp. parasites to their natural hosts is therefore warranted, and this will be facilitated by the ability to express the entire ectodomain of EBA175 in an active form. This is of topical interest due to the recent discovery of new Plasmodium species that infect gorillas and that are the closest known relatives of P. falciparum (
). Our quantitative binding analysis did not find any evidence to support a conformational change in EBA175 (at least a relatively slow one that could be detected using our SPR method) but did clearly show that the full-length ectodomain of EBA175 bound native GYPA with an affinity one order of magnitude higher than the region II fragment alone suggesting that extracellular regions of PfEBA175 outside of RII do play some role in the interaction with GYPA.
The region II fragment of EBA175 crystallized as an anti-parallel dimer with the putative glycan-binding sites being formed at the dimer interface, leading to the suggestion that GYPA, which forms a dimeric complex at the erythrocyte surface, induces EBA175 dimerization upon binding (
). The SPR data we obtained for the interaction of PfEBA175 with GYPA fitted better to a bivalent analyte model than to a simple 1:1 binding model, which is consistent with the dimerization of EBA175 playing a role in its interaction with GYPA. Although we showed that both PfEBA175 RII and PfEBA175 FL were primarily monomeric in solution, both were capable of self-association, and the full-length ectodomain of EBA175 could bind itself with ∼70-fold higher affinity than the RII fragment. Extracellular regions outside of RII may therefore facilitate the interaction with GYPA by promoting homodimerization of EBA175.
Our binding analysis also revealed that soluble PfEBA175 FL has a 2-fold lower affinity for Neu5Acα2–3Gal than for GYPA, but PfEBA175 RII bound both with similar affinities; therefore, RII probably interacts primarily with the glycan moieties of GYPA, whereas PfEBA175 FL forms contacts with both the oligosaccharides and the polypeptide backbone. We attempted to further investigate this by expressing GYPA as a recombinant protein so that it could be purified in large quantities and biochemically manipulated. Although we were able to express and purify the ectodomain of GYPA in a soluble form (rGYPA) and show that it was antigenically active by monoclonal antibody binding, it was unable to bind PfEBA175 FL (data not shown). We attributed the inability of rGYPA to bind PfEBA175 to under-sialylation because glycan profiling of rGYPA using a panel of lectins showed significantly lower levels of sialylation relative to native GYPA. Despite increasing the level of rGYPA sialylation by co-transfecting our GYPA expression construct with plasmids encoding an α-2,3-sialyltransferase and/or CMP-sialic acid transporter, this increase was not sufficient to confer binding to PfEBA175 (data not shown). In addition, we were unable to detect any binding to native GYPA using a fragment consisting of regions III to VI of EBA175 by AVEXIS (data not shown); among other possibilities, the interaction of EBA175 with the GYPA polypeptide backbone could therefore be dependent on the binding of RII to the glycan moieties on the receptor.
In conclusion, the 10-fold higher binding affinity of PfEBA175 FL for GYPA in comparison with the PfEBA175 RII fragment is likely due to the participation of the extracellular regions outside of RII in the homodimerization of EBA175 and the formation of additional contacts with GYPA.
The ability to express the entire ectodomain of P. falciparum EBA175 as a biochemically active recombinant protein and the finding that regions of EBA175 outside of the tandem DBL domains are important for GYPA binding suggested that PfEBA175 FL could be a better vaccine candidate than PfEBA175 RII alone. Consistent with these observations, when normalized for immunoreactivity to RII, the antisera to PfEBA175 FL was ∼5-fold more potent than the antisera to PfEBA175 RII. This suggests that antibodies directed against extracellular regions of PfEBA175 outside of RII also contribute to inhibiting erythrocyte invasion. We envisage that the findings reported here will contribute to a more complete understanding of the molecular basis of erythrocyte invasion by P. falciparum and eventually lead to the development of an effective vaccine against this infectious disease.
We thank Leyla Bustamante and Michel Theron for support in performing and analyzing P. falciparum invasion assays, Laura Romanelli for technical assistance with GYPA expression trials, and Susan Lea for MALS expertise.