ABSTRACT
Blood-stage malaria vaccines that target single Plasmodium falciparum antigens involved in erythrocyte invasion have not induced optimal protection in field trials. Blood-stage malaria vaccine development has faced two major hurdles, antigenic polymorphisms and molecular redundancy, which have led to an inability to demonstrate potent, strain-transcending, invasion-inhibitory antibodies. Vaccines that target multiple invasion-related parasite proteins may inhibit erythrocyte invasion more efficiently. Our approach is to develop a receptor-blocking blood-stage vaccine against P. falciparum that targets the erythrocyte binding domains of multiple parasite adhesins, blocking their interaction with their receptors and thus inhibiting erythrocyte invasion. However, with numerous invasion ligands, the challenge is to identify combinations that elicit potent strain-transcending invasion inhibition. We evaluated the invasion-inhibitory activities of 20 different triple combinations of antibodies mixed in vitro against a diverse set of six key merozoite ligands, including the novel ligands P. falciparum apical asparagine-rich protein (PfAARP), EBA-175 (PfF2), P. falciparum reticulocyte binding-like homologous protein 1 (PfRH1), PfRH2, PfRH4, and Plasmodium thrombospondin apical merozoite protein (PTRAMP), which are localized in different apical organelles and are translocated to the merozoite surface at different time points during invasion. They bind erythrocytes with different specificities and are thus involved in distinct invasion pathways. The antibody combination of EBA-175 (PfF2), PfRH2, and PfAARP produced the most efficacious strain-transcending inhibition of erythrocyte invasion against diverse P. falciparum clones. This potent antigen combination was selected for coimmunization as a mixture that induced balanced antibody responses against each antigen and inhibited erythrocyte invasion efficiently. We have thus demonstrated a novel two-step screening approach to identify a potent antigen combination that elicits strong strain-transcending invasion inhibition, supporting its development as a receptor-blocking malaria vaccine.
INTRODUCTION
Malaria is a leading public health threat, with almost 3 billion people at risk of contracting the disease. Plasmodium falciparum, the causative pathogen of the most severe form of malaria, accounts for 1.2 million deaths annually (1). An effective blood-stage malaria vaccine would be a great asset in controlling and eliminating the disease, which is caused primarily by blood-stage parasites. However, blood-stage malaria vaccine development against P. falciparum has been hindered by the enormous complexity of the parasite, widespread antigenic polymorphisms, and inadequate knowledge of host-parasite interactions as well as naturally acquired immunity.
Global efforts to develop blood-stage malaria vaccines against P. falciparum have focused on a few antigens tested mostly individually in clinical trials (2). Unfortunately, the leading blood-stage vaccine candidates merozoite surface protein 1 (MSP-1) and apical membrane antigen 1 (AMA-1), while being essential for the parasite, have elicited very limited protection in field trials (3, 4), which is attributed to their extensive polymorphisms that enable immune escape (5). Antibodies that impair P. falciparum erythrocyte invasion are one of the effector mechanisms known to mediate immunity against blood-stage malaria parasites. A significant association of invasion inhibition measured in vitro with a reduced risk of malaria has been reported (6, 7), and thus, in vitro invasion-inhibitory activity appears to be a useful surrogate marker to predict the efficacy of antibodies induced by a blood-stage vaccine. Therefore, it is crucial to identify and validate novel, efficacious P. falciparum blood-stage targets that elicit strain-transcending invasion-inhibitory antibodies.
Apart from MSP-1 and AMA-1, there are not many essential parasite ligands involved in P. falciparum erythrocyte invasion. Two families of P. falciparum erythrocyte binding proteins, EBA (erythrocyte binding antigens) and PfRH (P. falciparum reticulocyte binding-like homologous proteins), have been identified as major determinants of erythrocyte invasion (8–10). However, due to redundancy, EBA or PfRH proteins (with the exception of PfRH5) are not essential for the parasite, and their antibodies individually do not block invasion in a strain-transcending manner (8–14).
Recent reports have demonstrated that the PfRH5-basigin interaction is crucial for erythrocyte invasion (13) and that PfRH5 antibodies exhibit strain-transcending invasion inhibition (14). PfRH5 antibodies were generated by using the adenovirus-modified vaccinia virus Ankara (AdHu5-MVA) prime-boost regime, which is known to produce strong immune responses (14). The production of full-length native PfRH5 as a recombinant protein that exhibits erythrocyte binding activity and elicits invasion-inhibitory antibodies has proven challenging (15). Polymorphisms in PfRH5 affect its erythrocyte binding specificity for both human and Aotus erythrocytes (11). Thus, while PfRH5 undoubtedly appears to be a very promising target, its vaccine potential still requires further validation.
P. falciparum has the ability to switch its invasion phenotype (16, 17) and generate polymorphisms to enable immune escape. Thus, the targeting of single antigens is unlikely to be effective for blood-stage malaria vaccines. Analogous to the antimalaria combinatorial drugs administered to prevent the onset of drug resistance (18), a combination vaccine approach that targets multiple antigens may be more effective in limiting the parasite's ability to escape host immunity. Therefore, our approach for developing a receptor-blocking blood-stage malaria vaccine against P. falciparum is based on the targeting of the functional erythrocyte binding domains of key merozoite ligands involved in erythrocyte invasion, which would simultaneously block diverse invasion pathways and produce significant invasion inhibition.
Recently, a few reports have demonstrated that targeting combinations of merozoite antigens (including EBA and PfRH) yielded potent invasion inhibition (15, 19, 20). However, those studies did not demonstrate invasion-inhibitory efficacy against heterologous P. falciparum clones, which exhibit different invasion phenotypes. The demonstration of potent strain-transcending invasion inhibition against multiple P. falciparum clones is a major prerequisite for taking any candidate antigen into vaccine development. Another important challenge is to identify potent antigen combinations from the large and expanding repertoire of merozoite ligands that are involved in erythrocyte invasion. To identify potent antigen combinations, it will not be feasible to coimmunize all possible antigen mixtures and assay their invasion-inhibitory activity. Therefore, in the current study, we demonstrate a two-step screening approach that allowed us to identify a triple-antigen combination from a pool of six merozoite proteins that elicited potent strain-transcending neutralizing antibodies.
Our candidate antigens include three members of the PfRH family of proteins (PfRH1, PfRH2, and PfRH4), EBA-175 (PfF2), P. falciparum apical asparagine-rich protein (PfAARP), and Plasmodium thrombospondin apical merozoite protein (PTRAMP), which bind erythrocytes with different specificities and have been shown to be major determinants of different invasion pathways (Fig. 1) (8–10, 21–38). The erythrocyte binding domains of these adhesins have been elucidated (24, 27, 29, 31, 32). PTRAMP contains an adhesive thrombospondin repeat (TSR) domain that has been implicated in playing a conserved role in erythrocyte invasion (33, 34).
Subcellular localizations of the six P. falciparum merozoite proteins and their erythrocyte binding characteristics. A diagram of a merozoite invading an erythrocyte is depicted, highlighting the localizations of the six antigens in the different apical organelles (rhoptries and micronemes). The erythrocyte binding characteristics of the six antigens are tabulated with respect to the identity of their erythrocyte receptor, the type of the receptor with respect to its sialic acid content, and sensitivity to enzymes such as trypsin. Our antigens are localized in different apical compartments and are believed to be released at different time points during invasion and in addition bind erythrocytes with different specificities, implying an involvement in different invasion pathways. Localization data for PfRH1 and PfRH4 are based only on confocal immunofluorescence microscopy, and no immunoelectron microscopy data are reported for these two proteins. S. No., serial number.
Our antigen portfolio represents merozoite adhesins that are localized in different apical organelles (rhoptries and micronemes) (Fig. 1), which release their contents in a sequential manner at different time points during invasion (39). PfRH1, PfRH2, and PfAARP are localized in the neck of the rhoptries (23, 29, 32), whereas PTRAMP is located in the bulb of the rhoptries (F. A. Siddiqui and C. E. Chitnis, unpublished data). PfEBA-175 and PfRH4 are localized in micronemes (26, 27, 40, 41). Furthermore, these antigens bind different erythrocyte receptors and mediate different invasion pathways, which are defined by their dependence on sialic acids and sensitivity to enzymes such as trypsin (Fig. 1) (21–29, 31, 32). Therefore, our antigen portfolio represents a combinatorial diversity involving different parasite ligands mediating distinct invasion pathways and probably even different steps of erythrocyte invasion.
In the present study, we have raised specific antibodies against the recombinant receptor binding domains of these six merozoite antigens and have systematically evaluated their invasion-inhibitory potential in different triple combinations. Through this first step of screening, we identified a potent antibody combination against PfRH2, PfEBA-175, and PfAARP that exhibited maximum inhibition in a strain-transcending manner. In the next step, mice were coimmunized with this antigen combination, and the antibodies raised against the antigen mixture were also found to exhibit potent invasion inhibition, consistent with the results from the first screen. Our study has not only established a proof of principle for the development of a receptor-blocking, blood-stage vaccine that targets solely erythrocyte binding domains of key merozoite ligands but also validated a two-step screening approach for the downselection of potent antigen combinations for vaccine development.
(This work was previously presented at the XI Symposium on Vectors and Vector Borne Diseases, Jabalpur, Madhya Pradesh, India, 15 to 17 October 2011, and the 2012 Molecular Approaches to Malaria [MAM] Meeting, Lorne, Victoria, Australia, 19 to 23 February 2012.)
MATERIALS AND METHODS
Ethics statement.The animal studies described below were approved by the International Centre for Genetic Engineering and Biotechnology (ICGEB) Institutional Animal Ethics Committee (IAEC) (reference no. MAL-51), according to the guidelines of the Department of Biotechnology, Government of India.
P. falciparum parasites and human erythrocytes.P. falciparum clones used in this study were 3D7 (42), Dd2 (43), HB3 (44), MCamp (45), and 7G8 (46). These parasite clones were kindly provided by Louis Miller (NIH). The parasites were cultured in vitro according to methods described previously by Trager and Jensen (47). The parasites were grown in a 2% suspension of human O+ erythrocytes and RPMI 1640 medium supplemented with 0.5% Albumax (Gibco, Life Technologies, Grand Island, NY), 24 mM HEPES, 360 μM hypoxanthine, 24 mM sodium bicarbonate, and 10 μg/ml gentamicin at 37°C with 5% CO2–5% O2–90% N2. Normal human erythrocytes (O+) were obtained from the Rotary Blood Bank in New Delhi, India. The serum and leukocytes were removed, and the erythrocytes were washed with modified RPMI 1640 medium, as mentioned above, with no supplement of Albumax (mRPMI).
Enzymatic treatment of erythrocytes.Enzymatic treatments of the erythrocytes were performed as described previously (24, 27, 37). For trypsin-chymotrypsin, 109 erythrocytes in 10 ml mRPMI were incubated with l-1-tosylamide-2-phenylethyl chloromethyl ketone (TPCK)-treated trypsin (Sigma-Aldrich) or Nα-p-tosyl-l-lysine chloromethyl ketone (TLCK)-treated chymotrypsin (Sigma-Aldrich) at a final concentration of 1 mg/ml with rocking at 37°C for 1 h and then washed with mRPMI. The erythrocytes were then treated with 1 mg/ml of trypsin-chymotrypsin inhibitor (Sigma-Aldrich) at room temperature for 15 min. Cells were washed with mRPMI and stored at 4°C for a maximum of 24 h prior to the assay. For neuraminidase, 2.5 × 109 erythrocytes in 5 ml of mRPMI (pH 6.7) were incubated with 0.037 U of Vibrio cholerae neuraminidase (Roche Diagnostics) at 37°C for 1 h with rocking and then washed twice with mRPMI.
Expression of the receptor binding domains as recombinant proteins.The respective gene sequences encoding the receptor binding domains of the different merozoite antigens were cloned in the following T7 promoter-based pET expression vectors, obtained from Novagen (EMD Millipore), as described previously: rPfRH240 (pET-24b) (24), rPfRH430 (pET-11a) (27), rPfAARP20-107 (pET-28a) (32), and EBA-175 (rPfF2) (pET-28a) (35). The rPfRH430-pET11a plasmid was kindly provided by Louis Miller (27). For rPfRH140, an Escherichia coli codon-optimized synthetic gene encoding amino acids 500 to 833 of the receptor binding domain of PfRH1 (29) was obtained from GeneArt (Life Technologies) and cloned into pET-24b using the NdeI and XhoI restriction enzymes. Similarly, a codon-optimized synthetic gene encoding the ectodomain of PTRAMP (PFL0870w) (amino acid residues 26 to 307) was cloned into pET28a. The recombinant proteins rPfRH140, rPfRH240, rPfRH430, rPfAARP20-107, rPfF2, and rPTRAMP26-307 were expressed in E. coli with a C-terminal hexa-His tag and purified as described previously (24, 27, 29, 32, 35). Briefly, E. coli BL21(DE3) cells (Novagen, San Diego, CA) were transformed with the expression plasmids and cultured in Luria broth at 37°C. Protein expression was induced with 1 mM isopropyl-β-d-thiogalactopyranoside (IPTG) at a culture optical density at 600 nm (OD600) of ∼0.6. Cells were grown for 4 h after induction, harvested by centrifugation at 3,000 × g, and lysed by sonication. All proteins were found in inclusion bodies, with the exception of rPfAARP20-107 and rPTRAMP, which were expressed in a soluble form. The inclusion bodies were washed and solubilized in 8 M guanidine-HCl. The proteins were purified from solubilized inclusion bodies by metal affinity chromatography using a Ni-NTA (nitrilotriacetic acid) resin (Qiagen), refolded, and further purified by ion-exchange chromatography (GE Healthcare, Sweden), as described previously (24, 27, 29, 35). rPfAARP and rPTRAMP were expressed as soluble proteins and required no in vitro refolding. The expressed proteins obtained after bacterial lysis were purified first by metal affinity chromatography using Ni-NTA resin (Qiagen) and then by anion-exchange chromatography using Q-Sepharose resin (GE Healthcare, Sweden), followed by gel filtration (GE Healthcare, Sweden). The recombinant proteins were characterized by SDS-PAGE and Coomassie staining.
Antibodies and IgG purification.Rabbits were immunized intramuscularly with 100 μg of each recombinant protein emulsified with complete Freund's adjuvant (CFA) (Sigma, St. Louis, MO) on day 0, followed by two boosts emulsified with incomplete Freund's adjuvant (IFA) (Sigma, St. Louis, MO) on days 28 and 56. The sera were collected on day 42 and from terminal bleeds on day 70. Sera were tested for antibody titers and the specific recognition of each recombinant protein by an enzyme-linked immunosorbent assay (ELISA).
For coimmunogenicity experiments, a group of six mice (BALB/c) was immunized on day 0 with the antigen mixture formulated with complete Freund's adjuvant followed by two boosts emulsified with incomplete Freund's adjuvant on days 28 and 56. In parallel, each of the three individual antigens was also immunized separately in groups of 6 mice. Terminal bleeds were collected on day 70. Sera were tested for antibody titers and the specific recognition of each recombinant protein by ELISA.
Total IgG was purified from rabbit and pooled mouse sera by using a protein G affinity column (GE Healthcare, Uppsala, Sweden), in accordance with the manufacturer's instructions. The purified IgGs were dialyzed with mRPMI and used in invasion inhibition assays.
ELISAs.Antisera were tested for their recognition of the corresponding recombinant proteins by an ELISA. Briefly, 96-well plates were coated overnight with 0.2 μg per well of the recombinant protein and further incubated with 5% skimmed milk in phosphate-buffered saline (PBS) for blocking at 37°C for 2 h. Serial dilutions of the primary sera (1:1,000-fold onwards) were prepared and incubated in the respective wells. The ELISA plate was subjected to stringent washing with PBS containing 0.05% Tween 20 and finally with PBS alone. Thereafter, a 1:10,000 dilution of the horseradish peroxidase-conjugated secondary antibody (Sigma, St. Louis, MO) was added to each well and incubated for 60 min at 37°C. The enzymatic reaction was developed by the addition of o-phenylenediamine dihydrochloride (OPD) and hydrogen peroxide. The reaction was terminated by the addition of sulfuric acid, and the OD492 was recorded by using an ELISA microplate reader (Molecular Devices). Preimmune (prebleed [PB]) sera were used at similar dilutions as a control.
Erythrocyte binding assays.Erythrocyte binding assays (EBAs) were performed as described previously (24, 27) Soluble parasite proteins were obtained from P. falciparum 3D7 culture supernatants, as described previously (24, 27). Culture supernatants (500 μl) or recombinant protein (0.5 μg) was incubated with the different enzyme-treated human erythrocytes (100 μl) at 37°C for 1 h. After incubation, the suspension was centrifuged through dibutyl phthalate (Sigma, St. Louis, MO). The supernatant and oil were removed by aspiration. Bound parasite proteins were eluted from the erythrocytes with 1.5 M NaCl. The eluate fractions were analyzed for the presence of the proteins of interest using specific antibodies in immunoblots, as described previously (24, 27, 37).
Invasion inhibition assay.Invasion inhibition assays were done as described previously (24). Briefly, the parasites were first synchronized by the purification of schizont-stage parasites on a Percoll gradient, followed by 2 to 3 rounds of treatment of the ring-stage parasites with sorbitol. Schizont-stage parasites at an initial parasitemia level of 0.3% at 2% hematocrit were incubated with purified IgG and incubated for one cycle of parasite growth (40 h postinvasion). The parasite-infected erythrocytes were stained with ethidium bromide dye and measured by a fluorescence-activated cell sorter (FACS)-based assay, as described previously (24). Invasion inhibition was calculated with respect to purified preimmune IgG as well as immune IgG generated against a nonrelated peptide (KESRAKKFQRKHITNTRDVD) (from human pancreatic RNase) formulated with the same adjuvant (CFA/IFA) used for raising the other antibodies. P values were calculated by using the Student t test.
RESULTS
Expression of the receptor binding domains as recombinant proteins and generation of specific antibodies.The erythrocyte receptor binding domains of PfRH4 (rPfRH430), PfRH1 (rPfRH140), PfRH2 (rPfRH240), PfAARP (rPfAARP20-107), and EBA-175 (PfF2) were well characterized in previous reports (24, 27, 29, 32, 35) (Fig. 2). We have successfully expressed these binding domains along with the ectodomain of PTRAMP as recombinant proteins (see Fig. S1 in the supplemental material). Consistent with data from previous reports, the erythrocyte binding specificities of all recombinant proteins produced against the receptor binding domains matched the specificities of their respective parasite proteins, thus confirming that the recombinant proteins were functionally folded similarly to their native counterparts (see Fig. S2 in the supplemental material). Antisera were raised against the individual purified recombinant proteins formulated with complete and incomplete Freund's adjuvant in rabbits. Endpoint titers of around 1:320,000 were observed by ELISAs at an OD492 against all antigens (data not shown), which were consistent with data from previous reports (24, 27, 29, 32, 35).
Functional domains of the six P. falciparum merozoite proteins expressed as recombinant proteins. Shown is a schematic diagram of the six merozoite proteins involved in erythrocyte invasion, and the erythrocyte binding domains against which the recombinant proteins have been produced are highlighted (gray).
The antibodies specifically detected the native proteins from parasite culture supernatants of the expected molecular masses in immunoblots (see Fig. S2 in the supplemental material). While the functional specificities of antibodies against PfRH4, PfAARP, and PfEBA-175 in inhibiting the erythrocyte binding of the native parasite proteins were reported previously (27, 32, 37), no such data were reported for PfRH1 and PfRH2. Purified IgG against both rPfRH140 and rPfRH240 also inhibited the erythrocyte binding of both recombinant and native PfRH1 and PfRH2, respectively, in a specific dose-dependent manner (see Fig. S3 in the supplemental material).
PfRH antibody combinations produce an additive inhibition of erythrocyte invasion by P. falciparum.The invasion-inhibitory activities of different combinations of purified total IgG against each of the three PfRH proteins were assessed over one cycle against P. falciparum clone 3D7 (24, 27), which invades erythrocytes using both sialic acid-dependent and -independent pathways (Fig. 3). The purified total IgGs were tested individually (2.5 to 10.0 mg/ml) as well as in double-antibody (2.5 mg/ml and 5 mg/ml each) and triple-antibody (3.3 mg/ml each) combinations (Fig. 3). The maximum total IgG concentration tested was 10 mg/ml, as this concentration is close to the physiological concentration of IgG in human sera (48). Individual PfRH antibodies exhibited dose-dependent invasion inhibition, confirming a specific effect (Fig. 3). PfRH240 IgG exhibited a maximum inhibition of 54% (10 mg/ml) (Fig. 3).
Erythrocyte invasion-inhibitory activity of PfRH antibodies in combination against P. falciparum clone 3D7. Total IgGs purified from rabbit sera raised against the receptor binding domains of RH1, RH2, and RH4 were tested for their invasion-inhibitory activities individually (2.5 to 10 mg/ml) and in combination. Combinations of two RH IgGs were assessed at two concentrations (2.5 and 2.5 mg/ml, and 5.0 and 5.0 mg/ml), and combinations of three RH IgGs were tested at 3.3 mg/ml each. AMA-1 IgG (5 mg/ml) was used as a positive control. The negative control is the average of the inhibition of purified IgG from preimmune rabbit sera and a control rabbit immunized with a nonrelated peptide. Three independent assays were performed in duplicate. The error bars show the standard errors of the means.
PfRH IgG combinations at 2.5 mg/ml each did not produce any significant increase in invasion inhibition; however, the combinations at 5.0 mg/ml each produced an additive effect (Fig. 3). Individually, the three PfRH IgGs blocked invasion by 17 to 29% at 5 mg/ml (Fig. 3). Among the double combinations, the combination of RH2 and RH4 (RH2+RH4) IgGs inhibited invasion by 54%, while the inhibition of RH1+RH2 and RH1+RH4 was 30 to 37% (Fig. 3). The most potent additive inhibition was observed with a combination of three antibodies (RH1+RH2+RH4, at 3.3 mg/ml each) that produced 66% inhibition, compared to the low-level inhibition (<12%) exhibited by each individual IgG at 3.3 mg/ml (Fig. 3).
Invasion-inhibitory activities of combinations of antibodies against the PfRH proteins PfF2, PfAARP, and PTRAMP.To further assess whether the inhibition obtained by targeting PfRH proteins could be augmented by the inclusion of other key merozoite target antigens, we tested the invasion-inhibitory activities of all 20 possible triple-antibody combinations of purified total IgG (3.3 mg/ml each) against a pool of 6 antigens (RH1, RH2, RH4, PfF2, AARP, and PTRAMP). Invasion inhibition was assayed against P. falciparum clones 3D7 and Dd2 (Fig. 4). Against 3D7, individual IgGs against each of the 6 antigens produced 8 to 25% inhibition, with AARP IgG being the most potent (25%) at 3.3 mg/ml (Fig. 4A). Different combinations of three IgGs displayed a potent inhibition of erythrocyte invasion (Fig. 4A), with the maximum inhibition (79%) being elicited by PfF2+RH2+AARP IgGs, which is higher than the 66% observed with RH1+RH2+RH4 IgGs (Fig. 4A).
Invasion-inhibitory activities of antibody combinations against P. falciparum. Total IgGs purified from rabbit sera against the six proteins (PfRH1, PfRH2, PfRH4, PfF2, PfAARP, and PTRAMP) were assayed individually at a concentration of 3.3 mg/ml and in combinations of three IgGs (3.3 plus 3.3 plus 3.3 mg/ml) against sialic acid-independent clone 3D7 (A) and sialic acid-dependent clone Dd2 (B). Three independent assays were performed in duplicate. The error bars show the standard errors of the means. P values were calculated by using the Student t test.
Against the sialic acid-dependent clone Dd2, AARP IgG inhibited erythrocyte invasion with the same efficiency as that observed with 3D7 (Fig. 4B). Total IgG against PfF2 and PfRH1, which are sialic acid binding proteins, exhibited higher-level inhibition against Dd2 than against 3D7 (Fig. 4). This is consistent with the fact that Dd2 is known to express higher levels of PfRH1 and utilize it for invasion (22). Conversely, PfRH2 and PfRH4 IgGs exhibited poor inhibition against Dd2, again consistent with Dd2 expressing low levels of these two proteins (16, 17, 23). Similar to 3D7, some triple-IgG combinations displayed potent invasion inhibition against Dd2 (Fig. 4B). PfF2+RH1+AARP yielded the maximum invasion inhibition (75%) against Dd2 (Fig. 4B), consistent with the three antigens being involved in sialic acid-dependent invasion. However, this combination yielded only 48% inhibition against clone 3D7 (Fig. 4A). The most effective combination against 3D7, PfF2+RH2+AARP, exhibited the second highest level of inhibition (68%) against Dd2 (Fig. 4) and was thus considered to be most efficacious against both parasite clones.
The efficacy of PfF2+RH2+AARP was further analyzed against three other diverse P. falciparum clones. In addition to 3D7 and Dd2, we tested the inhibition of invasion by sialic acid-independent clones (7G8 and HB3) and a sialic acid-dependent clone (MCamp). PfF2+RH2+AARP IgGs inhibited the invasion of all five clones with an invasion inhibition efficiency ranging between 67 and 79% (Fig. 5). Two other IgG combinations (RH1+RH2+RH4 and RH1+RH2+PfF2) were also tested for the inhibition of invasion by multiple clones. The invasion-inhibitory efficiencies of these combinations were not similar for all clones (Fig. 5), with broad inhibition efficiencies of between 36 and 67% (Fig. 5). Thus, the PfF2+RH2+AARP antibody combination was identified from our first step of screening to exhibit the most potent, strain-transcending, invasion-inhibitory activity.
Strain-transcending activities of three triple-antibody combinations against five diverse P. falciparum clones. The invasion inhibitions of three antibody combinations (PfF2+RH2+AARP, RH1+RH2+RH4, and RH1+RH2+PfF2) were assayed with five P. falciparum clones: 3D7, 7G8, HB3, Dd2, and MCamp (MC). 3D7, 7G8, and HB3 are sialic acid-independent clones; Dd2, MCamp, are sialic acid-dependent clones. Three independent assays were performed in duplicate. The error bars show the standard errors of the means.
Immunogenicity studies of antigen combinations.After evaluating antibody combinations that were mixed in vitro for their invasion-inhibitory activities, we wanted to test whether the most potent combination identified would elicit similar invasion-inhibitory antibodies when coimmunized together as a single formulation. In this next step, the most potent antigen combination (PfF2+RH2+AARP) and another combination (RH1+RH2+RH4) as a control were used to immunize mice (BALB/c). The individual antigens in each combination were also used to immunize mice separately. All antigens were formulated with the adjuvant CFA/IFA. The ELISA results (OD492) showed that the antibody titers (endpoint, 1:320,000) against each protein immunized individually were not significantly altered when immunized as a mixture with the two other antigens (Fig. 6). The immunogenicity curves for PfF2, PfAARP, PfRH1, and PfRH4 were identical and overlapping whether the antigens were immunized alone or in their respective combinations (Fig. 6). PfRH2 was the common antigen in both antigen mixtures, and only its antibody titers were observed to be marginally lower when immunized as an antigen mixture than with its individual immunization (Fig. 6). Thus, our recombinant antigens were immunogenic and did not elicit any significant immune interference when immunized in combination.
Immunogenicity of the antigens in mice when used alone and in combination. Antigen combinations (PfF2+RH2+AARP and RH1+RH2+RH4) as well as the individual corresponding antigens were used to immunize mice. Day 70 sera from the immunized animals were probed against the respective antigens to determine their immunogenicity. (A) Immunogenicity of antibodies against PfF2+RH2+AARP (purple) compared with that of the sera raised against the individual antigens (green). (B) Immunogenicity of antibodies against RH1+RH2+RH4 sera (purple) compared with that of antibodies against sera raised against the individual corresponding antigens (green). For ELISAs, each individual antigen was coated separately onto 96-well plates for both the coimmunized as well as the individual antigens. Preimmune or prebleed (PB) sera were used as controls. The data points represent average values for the six mice included in each group. Two independent experiments were done in duplicate. The error bars represent the standard errors of the means.
Invasion-inhibitory activity of antibodies raised against the coimmunized antigen mixtures.Consistent with the invasion inhibition observed with the antibody combinations mixed in vitro, the antibodies raised against the antigen mixtures were highly potent and equally efficient in inhibiting erythrocyte invasion (Fig. 7). Against the 3D7 clone, antibodies against the PfF2+RH2+AARP antigen mixture displayed 69% and 85% inhibitions at concentrations of 5 and 10 mg/ml, respectively (Fig. 7A). PfF2 IgG, RH2 IgG, and AARP IgG individually exhibited 23%, 52%, and 56% inhibitions, respectively, at 10 mg/ml (Fig. 7A). Thus, the PfF2+RH2+AARP formulation induced antibodies that were much more potent than the individual antibodies at the same IgG concentrations, clearly suggesting an additive inhibitory effect. The invasion inhibition exhibited by all antibodies was observed to be dose dependent (Fig. 7). A similar trend was observed with P. falciparum clone Dd2, which was inhibited by the antibodies raised against the PfF2+RH2+AARP mixture by 62% and 80% at 5 and 10 mg/ml, respectively (Fig. 7B). Consistent with the results from the first screen, PfF2 IgG inhibited Dd2 more potently than 3D7, and similarly, PfRH2 IgG inhibited Dd2 with a lesser efficiency than 3D7 (Fig. 7). The antibodies raised against the RH1+RH2+RH4 mixture inhibited the invasion of 3D7 with lower efficiencies of 58% and 69% at 5 and 10 mg/ml, respectively (Fig. 7A). On the other hand, these antibodies exhibited a much lower inhibitory activity (42% at 10 mg/ml) with the Dd2 clone, which expresses lower levels of PfRH2 and PfRH4 (Fig. 7B). These results were consistent with the inhibition observed for the same antibody combinations in the first screen.
Invasion-inhibitory activities of antibodies raised against the two coimmunized antigen formulations. Total IgGs from mouse sera raised against the immunogens (RH1, RH2, RH4, PfF2, AARP, RH1+RH2+RH4, and PfF2+RH2+AARP) were evaluated for their invasion-inhibitory activities (at concentrations of 1, 3.3, 5, and 10 mg/ml) against sialic acid-independent clone 3D7 (A) and sialic acid-dependent clone Dd2 (B). Two independent assays were performed in duplicate. The error bars show the standard errors of the means.
Antibodies raised against both the antigen mixtures were assayed for their invasion-inhibitory activities against three more P. falciparum clones (7G8, HB3, and MCamp) that exhibit phenotypic variation in erythrocyte invasion (see Fig. S4 in the supplemental material). The antibodies against the PfF2+RH2+AARP antigen mixture displayed strain-transcending inhibition efficiencies of around 80 to 87% against five diverse P. falciparum clones, consistent with those observed when antibodies against these antigens were tested in combination (Fig. 8; see also Fig. S4 in the supplemental material). The invasion inhibition observed with the antibodies against the RH1+RH2+RH4 antigen mixture against the five clones varied over a broader range of 40 to 70%, similar to that observed for the antibody combination (Fig. 8; see also Fig. S4 in the supplemental material). Thus, our results demonstrated that the invasion-inhibitory activities of antibody combinations mixed in vitro are comparable to the invasion-inhibitory activities of antibodies raised against formulations containing a mixture of antigens.
Strain-transcending activities of mouse antibodies raised against the coimmunized triple-antigen combinations. The strain-transcending neutralizing activities of purified total IgGs raised against the coimmunized PfF2+RH2+AARP and RH1+RH2+RH4 antigen formulations were evaluated against three sialic acid-independent (3D7, 7G8, and HB3) and two sialic acid-dependent (Dd2 and MCamp) clones at a concentration of 10 mg/ml. Two independent assays were performed in duplicate. The error bars show the standard errors of the means.
DISCUSSION
An innovative approach to produce potent invasion inhibition of P. falciparum that was recently reported is to simultaneously target a combination of key merozoite antigens involved in erythrocyte invasion (15, 19, 20). However, due to the high order of molecular redundancy exhibited by P. falciparum (8–10), this strategy poses the enormous challenge of identifying efficacious antigen combinations from the large pool of merozoite ligands that could elicit potent, strain-transcending, invasion-inhibitory antibodies. Logically, this would involve large-scale immunizations of a number of antigen mixtures and analysis of their invasion-inhibitory activities, a task that will be very time-consuming considering the huge number of merozoite ligands and the combinations needed to cover them.
Here, we have used a systematic, two step screening process that appears to circumvent the problem of coimmunizing many antigen combinations and has enabled us to identify a potent antigen combination from a pool of six parasite ligands (PfRH1, PfRH2, PfRH4, PfAARP, EBA-175, and PTRAMP) that elicits strong, strain-transcending, invasion-inhibitory antibodies.
The PfRH family includes five functional members (PfRH1, PfRH2a, PfRH2b, PfRH4, and PfRH5) that bind erythrocytes with different specificities (8–11, 21–29). PfRH2a and PfRH2b share 2,700 identical amino acids in their ectodomains (23–25), which comprise their receptor binding domain (rPfRH240) (24). Thus, both PfRH2a and PfRH2b are considered identical for the purposes of this study and are referred to here as PfRH2. In our present study, where we aimed at establishing a proof of principle for a subunit combination, receptor-blocking, blood-stage malaria vaccine, we have focused only on PfRH1, PfRH2, and PfRH4. PfRH1 binds a sialic acid-containing erythrocyte receptor (21, 22), whereas PfRH2 and PfRH4 bind erythrocytes in a sialic acid-independent manner (24, 25, 27, 28). Most importantly, the differential expression of PfRH proteins has been found to be responsible for determining the invasion phenotypes of diverse P. falciparum strains (16, 17, 22, 23, 30). Such a dependence of the invasion specificity on the expression levels of a parasite ligand highlights the importance of the PfRH family, making them our primary target antigens.
EBA-175 is one of the most well-characterized parasite ligands involved in erythrocyte invasion (8–10, 31, 35–38), and its minimal receptor binding domain has been mapped to a cysteine-rich region, known as PfF2, that binds sialic acid residues of the erythrocyte receptor glycophorin A (31, 35). The EBA-175/glycophorin A pathway mediates erythrocyte invasion in a wide range of P. falciparum strains that exhibit either sialic acid-dependent or -independent pathways (36, 37). EBA-175 polymorphisms have no effect on its binding specificity (37, 38), and EBA-175 antibodies raised against the receptor binding domain blocked the invasion of heterologous parasite strains (37). These data support the inclusion of EBA-175 in a blood-stage receptor-blocking malaria vaccine.
The novel P. falciparum apical asparagine-rich protein (PfAARP) was previously reported by our group to be a conserved merozoite adhesin involved in erythrocyte invasion that binds erythrocytes in a sialic acid-dependent manner. PfAARP antibodies exhibited potent invasion inhibition (32), which made it an obvious choice to be included in our study. PTRAMP contains the adhesive thrombospondin repeat (TSR) domain and has been shown to exhibit specific erythrocyte binding activity (Siddiqui and Chitnis, unpublished).
These six antigens are localized in different apical organelles (rhoptries and micronemes) and are translocated to the merozoite surface at different time points during invasion (39). Therefore, with our antigen portfolio, we have been able to evaluate a substantial combinatorial diversity involving multiple ligand-receptor interactions, distinct invasion pathways, and probably even different steps of invasion for their potential as targets for the blockade of erythrocyte invasion by P. falciparum.
Our approach of targeting the functional binding domains of multiple parasite ligands involved in erythrocyte invasion represents a new strategy for the development of a receptor-blocking, blood-stage malaria vaccine. We successfully produced the recombinant erythrocyte binding domains of all these proteins and raised specific antibodies against them.
In the first step, we tested the invasion inhibition efficiencies of all 20 possible triple-antibody combinations from our pool of six antibodies. We restricted our antibody combinations to a maximum of three antigens, as this appeared practical from the later standpoint of vaccine production. Among all 20 possible antibody combinations, PfF2+RH2+AARP clearly exhibited the strongest inhibition against five well-characterized P. falciparum clones that broadly represent the various invasion phenotypes reported for parasites from different parts of the world in a strain-transcending manner.
The inhibitory activities of the antibody combinations in our assay reflected the differences in the expression and utilization of the corresponding parasite ligands among the different P. falciparum clones, implying that the invasion inhibitions exhibited by the antibodies were specific. Furthermore, a number of antibody combinations involving PTRAMP did not yield strong inhibition, substantiating that the additive inhibition observed with the antibody combinations in our assays was specific.
To confirm whether the potent invasion inhibition observed with the antibodies combined in vitro would hold when the antigens are immunized together in a multiantigen vaccine, we undertook the next step of coimmunizing the three antigens in a single formulation in mice. We observed that the antigen mixtures did not produce any immune interference, as the antibody titers against each antigen were similar whether immunized alone or in combination. The potent invasion-inhibitory activities of the antibodies raised against the triple antigen formulation PfF2+RH2+AARP were consistent with those observed with the antibodies combined in vitro in terms of both efficiency and strain-transcending activity. Thus, with our screening approach, we have demonstrated that a potent antigen combination that yields strong strain-transcending invasion-inhibitory antibodies can be identified without immunizing all possible antigen combinations. This work has strong implications for vaccine design against other pathogens, such as HIV, that also involve the evaluation of neutralizing antibodies.
Previous studies reported high-level growth inhibition (67 to 80%) only against clone 3D7 over two cycles using antibody combinations against EBA175/PfRH2/PfRH4 (15) and EBA175/PfRH2/PfRH4/PfRipr (19). In those reports, the functional minimal receptor binding domains were not targeted, and instead, the immune response was directed against regions of these proteins, EBA-175 regions III to V, RH2a/b, and RH4.9 (15, 19, 20), which are different than our target domains and whose functional role is not yet defined. The mechanisms by which these antibodies block invasion remain unknown. Most importantly, the strain-transcending ability of these antibodies to block the invasion of a number of heterologous parasite clones was not reported (15, 19, 20).
Previous studies on AMA-1 have demonstrated potent invasion inhibition only against homologous parasite clones (49). The inability of AMA-1 antibodies to block the erythrocyte invasion of heterologous P. falciparum clones has been verified in vitro (49). The lack of cross-strain neutralization was a major impediment for the efficacy of AMA-1 in field trials (5). Therefore, the strain-transcending activity of antibodies against parasite proteins is a critical attribute that must be demonstrated to realize the significance of the blood-stage antigens as potential vaccine candidates.
Our approach is different from those of previous reports in that we have targeted solely functional erythrocyte binding domains of multiple parasite proteins involved in invasion and have identified a potent antigen combination that elicits strong, strain-transcending, parasite-neutralizing antibody responses. Our data provide strong support for the development of the leading combination PfF2+RH2+AARP as a receptor-blocking blood-stage vaccine for P. falciparum malaria. Achieving high-level invasion-inhibitory activity against conserved functional domains of parasite antigens involved in invasion may be the key to the development of an effective blood-stage malaria vaccine.
ACKNOWLEDGMENTS
We are grateful to Louis Miller (NIH) for providing the P. falciparum clones used in the study and the rPfRH430 expression plasmid. We deeply appreciate the technical assistance of Alka Galav, D. B. Chandramouli, Rakesh Kumar Singh, and Ashok Das from the ICGEB animal facility in performing the animal experiments.
D.G. and C.E.C. are recipients of a Ramalingaswami fellowship and a TATA innovation fellowship, respectively, from the Department of Biotechnology (DBT), Government of India. D.G. is also the recipient of a Grand Challenges Exploration grant from the Bill and Melinda Gates Foundation. This work was supported by the Bill and Melinda Gates Foundation through the Grand Challenges Explorations Initiative (grant GCE OPP1007027 to D.G.); the DBT, Government of India, through the Ramalingaswami fellowship program (grant BT/HRD/35/02/14/2008 to D.G.), the rapid grant scheme for young investigators (grant BT/PR13376/GBD/27/260/2009 to D.G.), a program support grant (grant BT/01/CEIB/11/V/01 to D.G., C.E.C., and V.S.C.), the Vaccine Grant Challenges Program (grant ND/DBT/12/040 to D.G., C.E.C., and V.S.C.), and the Indo-Australia Biotechnology Fund (grant BT/PR12677/ICD/55/14/2009 to C.E.C.); and the European Commission through and EVIMalar grant (project no. 242095 to C.E.C.). A.K.P. is a recipient of a postdoctoral research associateship of the DBT; K.S.R., T.S., S.G., and K.R.M. are recipients of senior research fellowships of the Council of Scientific and Industrial Research, Government of India; and H.S. and F.A.S. are recipients of junior and senior research fellowships of DBT, respectively. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
All authors report no potential conflicts of interest.
FOOTNOTES
- Received 9 October 2012.
- Returned for modification 6 November 2012.
- Accepted 19 November 2012.
- Accepted manuscript posted online 26 November 2012.
Supplemental material for this article may be found at http://dx.doi.org/10.1128/IAI.01107-12.
- Copyright © 2013, American Society for Microbiology. All Rights Reserved.
