Functional Characterization of Plasmodium falciparum Surface-Related Antigen as a Potential Blood-Stage Vaccine Target

We have identified and functionally characterized a novel Plasmodium falciparum surface-related antigen (PfSRA) as a potential multistage vaccine candidate. The antigen is localized on both merozoites and gametocytes with high anti-PfSRA growth inhibition assay activity in laboratory strains and clinical isolates.

Malaria is a deadly infectious disease that affects inhabitants of the tropics and subtropical regions of the world and accounts for approximately 212 million cases and 429 000 deaths annually [1]. The clinical manifestation of the disease begins from the blood stage of the infection, during which the parasite invades human erythrocytes [2]. Plasmodium falciparum erythrocyte invasion is a complicated process that involves an array of receptor-ligand interactions [3] and/or protein-protein interactions [4][5][6][7][8] that occur at the parasite-host cell interface and facilitate the recruitment of the parasite's invasion machinery.
Thus, recent identification of P. falciparum surface proteins that are accessible to both humoral and cellular immune systems is a major advancement toward vaccine development against malaria [9][10][11][12]. This has given great impetus to the idea of a multiantigen vaccine as an intervention strategy against blood-stage malaria. Therefore, the selection and prioritization of candidate antigens are critical aspects of this strategy. Although several blood-stage antigens have been extensively studied, few have demonstrated the desired qualities for a vaccine candidate. One of the remarkable observations from the P. falciparum genome sequencing project was that nearly 60% of the parasite' s genes lacked sequence similarity to genes from other known organisms, and thus these genes have remained hypothetical with no defined functional roles [13]. Subsequently, the availability of more comprehensive genomic, proteomic, and transcriptomic datasets from both humans and Plasmodium has paved the way for further characterization of these hypothetical proteins using informatics-based approaches. This is required for successful identification of a repertoire of novel P. falciparum merozoite antigens that could be explored as targets for a rational vaccine design [14].
A detailed understanding of the functional roles of P. falciparum novel merozoite antigens, their localization, and their fate during invasion is critical to the identification of targets of host immunity and prioritization of merozoite antigens for inclusion in blood-stage malaria vaccines.
Herein, we have identified a novel P. falciparum protein (PlasmoDB ID: PF3D7_1431400/PF14_0293) that we have named P. falciparum surface-related antigen (PfSRA), based on its dual subcellular localization on both merozoites and gametocytes. Native PfSRA is proteolytically processed into multiple fragments in parasite culture supernatant, and the 32-kDa fragment of PfSRA exhibits erythrocyte-binding activity. More important, antibodies against PfSRA potently inhibited merozoite invasion of erythrocytes by both sialic acid-dependent and sialic acid-independent parasite strains. The data also demonstrated that PfSRA is a target for naturally acquired immune responses in humans.

Screening for New Blood-Stage Vaccine Candidates
To identify new blood-stage proteins as potential vaccine candidates, a systematic screening procedure was implemented. This included analysis of temporal gene expression relative to other well-characterized invasion-type genes and in silico interrogation of protein structural features. Once all of these selection criteria were ascertained, sequence alignment analysis was done to evaluate the conservation level of the target gene across the different Plasmodium species orthologs. Finally, we scanned the entire protein sequence using a current state-of-the-art online threading program to identify coiled-coil regions [15].

Peptide Synthesis and Immunogenicity Studies
Three synthetic peptides corresponding to the immunogenic epitopes were synthesized by GeneScript on the basis that they harbor coiled-coil signatures corresponding to the conserved regions in PfSRA orthologs.

Erythrocyte-Binding Assay
Plasmodium falciparum strains 3D7 and W2mef were maintained in culture as described previously [16]. Schizonts were purified using Percoll-alanine gradient centrifugation [20], followed by saponin lysis, and the recovered parasite pellets were further lysed in sodium dodecyl sulfate-polyacrylamide gel electrophoresis sample buffer. Parasite culture supernatant or ringstage invasion supernatant were used as the source for native PfSRA, which was detected by immunoblotting using primary α-rabbit antibodies to the 3 peptides (α-PfSRA P1, α-PfSRA P2, and α-PfSRA P3). The immunoblots were developed using goat α-rabbit horseradish peroxidase-conjugated, secondary antibody, and enhanced chemiluminescence reagents (Thermo Scientific). Erythrocyte-binding assay (EBA) was performed using parasite culture supernatant as described previously [21].

Permeabilized and Nonpermeabilized Immunofluorescence Assays
Immunofluorescence microscopy was carried out using P. falciparum (3D7)-infected erythrocytes smeared onto glass slides and fixed in prechilled methanol for 30 minutes. Fixed erythrocytes were permeabilized using 0.1 % Triton X-100 formulated in phosphate-buffered saline. Nonpermeabilized, liquid immunofluorescence assay (IFA) was carried out as described previously [22]. After the washing step for both IFA conditions, the slides were blocked for 1 hour in PBS containing 3% bovine serum albumin. Slides were probed with primary and secondary antibodies for the respective antigens and mounted in vectashield (Vector Laboratories Inc) supplemented with 4′,6-diamidino-2-phenylindole for staining the nucleus. Fluorescence microscopy was performed on an Olympus fluorescence microscope (BX41). Images captured were processed using the open access Fiji-Image J software (National Institutes of Health).

Antibody Internalization Assay
To test for the internalization of α-PfSRA P3 antibodies, α-PfSRA P3 and α-PfMSP-1 19 polyclonal antibodies were incubated with tightly synchronized segmenting schizonts that were allowed to rupture and release merozoites. The viability of the released merozoites was assessed at the different stages (early, mid, and late) of an invading merozoite. For each stage, smears were prepared, fixed, and examined by immunofluorescence microscopy as described above.

Growth Inhibition Assays
Growth inhibition assays (GIAs) were performed as described previously [23], and parasitemia levels were determined using flow cytometry on a BD FORTESSA X-20 with Flo J software. Erythrocyte invasion inhibitory effects of the α-PfSRA antibodies were estimated by comparison of percentage of invasion of controls with test assay.

Identification of Plasmodium falciparum Surface-Related Antigen as a Potential Malaria Vaccine Candidate
In a systematic screen of uncharacterized P. falciparum proteins for potential blood-stage vaccine candidates, we performed data-mining analysis for genes with peak mRNA expression levels in late schizogony using data from P. falciparum transcriptome studies [24,25] and another study on the prediction of PfSUB-1 protease specificity [26]. The details of all analyses have been described in the Supplementary Material (Supplementary Figure 1A and 1B). Overall, PfSRA emerged as the top hit with both signal peptide and a predicted glycosylphosphatidylinositol attachment site.
Furthermore, we generated sequence alignments with other orthologs of PfSRA and showed that the C-terminus of PfSRA had 5 positionally conserved cysteine residues across the different Plasmodium species orthologs (Supplementary Figure 2). Additional predictive analysis from PSI-Pred revealed that the PfSRA protein sequence harbored coiled-coil signatures (Supplementary Figure 3). This signature forms stable structures that elicit functional antibodies and was considered as a basis for the design of 3 PfSRA chemically synthesized peptides used for antibody generation.

Specific Antibodies Using Synthetic Peptides
Despite several optimization procedures for expression in Escherichia coli, attempts to express recombinant PfSRA protein were unsuccessful. However, we designed 3 peptides for synthesis that corresponded to the conserved regions of PfSRA in other Plasmodium species orthologs ( Figure 1A). These PfSRA-derived peptides (PfSRA P1, PfSRA P2, and PfSRA P3) include coiled-coil signatures (Supplementary Figure 3). Immunization of rabbits with the 3 synthetic peptides (PfSRA P1, PfSRA P2, and PfSRA P3) by Genescript resulted in high titers of PfSRA peptide-specific antibodies, which were used in subsequent experiments.
To identify the fragment(s) of PfSRA that possess erythrocyte-binding activity, we performed erythrocyte binding assays. The 32-kDa fragment of PfSRA was clearly detected in the eluate, suggesting that it bound erythrocytes. Under higher exposure, the 113-kDa full-length PfSRA and the 58-kDa processed fragment were also detectable (Supplementary Figure 3), suggesting a weaker binding relative to the 32-kDa fragment. Of interest, the binding was sensitive to neuraminidase and trypsin treatments but resistant to chymotrypsin treatment ( Figure 1F). As a control, recombinant EBA-175 (R217) bound erythrocytes with the same sensitivity to enzyme treatments ( Figure 1F).

Surface-Related Antigen in Asexual Stage
Stage-specific expression analysis in P. falciparum asexual stages by IFAs showed that all α-PfSRA peptide antibodies labeled the surface membranes of merozoites in intact schizonts and released merozoites, respectively (Figure 2A-C). However, α-PfSRA antibodies did not label developing ring-stage parasites, which served as a useful internal control for subsequent experiments in asexual stages.

Surface-Related Antigen During Erythrocyte Invasion
To determine the fate of native PfSRA during invasion, invading merozoites at different time-points (early, mid, late, and postinvasion) were colabeled with α-PfSRA P3 and α-PfMSP1 19 antibodies. We observed labeling of both α-PfSRA P3 and α-PfMSP1 19 antibodies in all of the time points of invasion ( Figure 3C and 3D) suggesting that a fragment or the unprocessed forms of PfSRA are carried into erythrocytes during invasion. As a control, late-stage parasites and internalized merozoites were colabeled with α-PfSRA P3 and the gametocyte surface marker α-Pfs48/45 antibodies. Whereas α-PfSRA P3 antibody labeled internalized merozoite, no labeling was observed with α-Pfs48/45 antibody ( Figure 3B).

Antigen by Naturally-Acquired Antibodies
Consistent with the ELISA data, we also established by immuno-dot blotting that plasma samples from malaria-infected children recognized PfSRA P1 ( Figure 5A). Thus, α-PfSRA P3specific human antibodies against the C-terminus of the protein were purified from a pooled sample of plasma from Kintampo children that showed immunoreactivity ( Figure 4D). Consistent with the data from rabbit α-PfSRA antibodies ( Figures 2E and  3A), human α-PfSRA P3 immuno-affinity purified antibodies stained schizonts and released merozoites that colocalized with α-PfMSP1 19 antibodies on the merozoite surface ( Figure 5B).

Surface-Related Antigen in Gametocytes
We performed stage-specific expression analysis in gametocytes (stage III-V) by microscopy and observed that α-PfSRA P3 antibodies specifically labeled the membranes of gametocytes, whereas the control antibody (PfMSP1 19 ) did not label gametocytes ( Figure 6A)  (stage II-V) with the respective α-PfSRA peptide antibodies and the gametocyte surface marker α-Pfs48/45 antibody and showed close colocalization that appeared to be stage-dependent ( Figure 6B-D). Furthermore, PfSRA expression in gametocytes appeared not to be sex-specific, with α-PfSRA P1 antibodies labeling both male and female gametocytes and α-tubulin antibodies labeling only male gametocyte (Supplementary Figure 5).

Related Antigen Peptide-Induced Antibodies
We evaluated the invasion inhibitory activity of α-PfSRA antibodies against P. falciparum 3D7 and W2mef and showed that all PfSRA peptide antibodies exhibited 70%-80% inhibition at 750 μg/mL, whereas the preimmune control did not show any inhibition (Supplementary Figure 6A and 6B). Because all 3 PfSRA peptide antibodies were inhibitory at higher concentrations (100-750 μg/mL) (Supplementary Figure 6A and 6B), we have shown that, whereas the preimmune sera and a control immunoglobulin G (IgG) did not inhibit parasite invasion at lower concentrations (25-75 μg/mL), α-PfSRA antibodies exhibited a concentration-dependent inhibition of parasite invasion ( Figure 7A). Of all 3 of the PfSRA peptide antibodies tested, only α-PfSRA P1 antibodies showed 60% inhibition of parasite invasion at 75 μg/mL. As a control, anti-Basigin antibodies showed 75% inhibition of parasite invasion at 10 μg/mL ( Figure 7A).  To exclude the possibility that serum contaminants might be interfering with parasite invasion, we generated antibodies against a shorter construct of R1-peptide using the same purification procedures for α-PfSRA peptide-specific antibodies, and no inhibition of parasite invasion was observed ( Figure 7A). Because all 3 α-PfSRA antibodies were inhibitory at 100-750 μg/mL, we tested combinations of α-PfSRA peptide antibodies against 3D7 and observed that this strategy did not greatly impact on parasite invasion inhibition ( Figure 7B). Furthermore, we performed invasion inhibition assays with 2 Ghanaian clinical isolates (MISA010 and MISA011) and observed similar patterns of parasite invasion inhibition for α-PfSRA peptide antibodies ( Figure 7C and 7D).

DISCUSSION
The focus of this work was the identification and functional characterization of a potential blood-stage malaria vaccine candidate (PfSRA). Using Bioinformatics and data-mining analysis of published P. falciparum transcriptomes and proteomes [24,25], we identified PfSRA to be reminiscent of a surface protein, defined by the presence of a signal peptide and a predicted glycosylphosphatidylinositol attachment signal [28,29]. Indeed, PfSRA has been shown to have PfSUB-1 cleavage sites that were assigned using the Prediction of Protease Specificity analysis platform [25]. The presence of these proteolytic cleavage sites in PfSRA may allude to processing events occurring prior to invasion of erythrocytes. Besides, proteolytic processing in the malaria parasite has been shown to be relevant for cascades of interaction occurring at the parasite-host cell interface [21,30]. Furthermore, because orthologs are good candidates for multispecies vaccines, we generated sequence alignments with other PfSRA orthologs, which showed that the C-terminus of PfSRA has 5 positionally conserved cysteine residues across the different species orthologs. Our attempt to express the recombinant PfSRA that possesses 67% unstructured regions was unsuccessful. Similarly, others have attempted the expression of the mature recombinant PfSRA in HEK293E cells using a codon-optimized gene (Geneart) and have also been unsuccessful [28]. Considering that PfSRA is an asparagine-rich protein with numerous potential N-linked glycosylation sites, we did not attempt expression in insect cells because the system produces proteins with more complex N-linked glycosylation [31] that may not present the relevant sugar epitopes of native glycosylation [32]. Therefore, the challenges associated with the expression of recombinant PfSRA necessitated the design of chemically synthetized peptides for an epitope-based vaccine strategy.
Plasmodium falciparum surface-related antigen harbors coiled-coil domains that are known to be less polymorphic [33,34], and they form stable structures that elicit functional antibodies that block relevant domains in many organisms [35][36][37]. Interestingly, these domains have been evaluated as potential targets for peptide-based vaccines [38][39][40]. The 3 PfSRA peptide antibodies from rabbits specifically detected breakdown products of native PfSRA in ring-stage invasion supernatant or parasite culture supernatant and schizont lysates. This indicated  that PfSRA synthetic peptides are antigenic mimics of the native parasite protein. Our data showed that PfSRA is postsynthetically processed by cleavage into parasite culture supernatant, and this is consistent with the PfSUB-1 cleavage sites it harbors. The merozoite surface is remodeled by a series of proteolytic processing events; the physiological relevance of these events in the malaria parasite remains poorly described. However, there are reports suggesting that proteolytic processing may result in activation, structural rearrangement, or acquisition of other new functional properties of native parasite proteins [41].
We have described the fate and shedding pattern of native PfSRA by temporal immunofluorescence imaging using α-PfSRA and α-PfMSP1 19 antibodies that bound the merozoite surface and were internalized during erythrocyte invasion. Consistent with this observation is a previous report that α-PfMSP1 19 antibodies were carried into invaded erythrocytes,  disrupted intra-erythrocytic development, and inhibited erythrocyte invasion [42][43][44]. Although the molecular mechanism underlying the internalization of antibodies remains debatable, it was suggested that the tight junction between the merozoite and the erythrocyte might consist of transient interactions that allow the passage of antibodies or surface proteins [19]. Because all rabb3it α-PfSRA peptide antibodies recognized different PfSRA polypeptide fragments in ring-stage invasion supernatant or parasite culture supernatant and schizont lysates at varying thresholds, it was expedient to investigate whether processing could influence differential subcellular localization of PfSRA in the parasite. Interestingly, all 3 rabbit α-PfSRA peptide antibodies showed circumferential association on the merozoite surface at the timing of schizont rupture and merozoite release. Similarly, we performed colabeling in IFAs with gametocytes (stage II-V) using all 3 rabbit α-PfSRA peptide antibodies with the gametocyte surface marker Pfs48/45. A clear, punctate rim-fluorescence pattern was observed for all 3 rabbit α-PfSRA peptide antibodies that appeared to colocalize with Pfs48/45 in a stage-dependent manner based on the colocalization coefficient. Therefore, the consistency in the staining patterns of all 3 rabbit α-PfSRA peptide antibodies in both asexual and gametocyte stages suggested that proteolytic processing of PfSRA does not cause changes in the subcellular localization of the protein. Although the distribution of proteins in male or female gametocytes could be linked to functional divergence between the sexes [45], we observed the expression of PfSRA in both male and female gametocytes. Consistent with this observation is an existing report on Plasmodium berghei gametocyte egress and sporozoite traversal protein (PbGEST) expression in both sexes [46].
Also, it was imperative to determine whether PfSRA in released merozoites was accessible to humoral immune surveillance during the short period of erythrocyte invasion. Our serological screens, buttressed by immuno-dot blot assays, with plasma samples from malaria-infected children residing at different endemic sites showed differences in total IgG recognition frequencies for PfSRA peptides. This could be linked to varying transmission intensity rates as reported for samples collected from different endemic sites in previous studies [33,47]. In most cases, the reactivity of all 3 PfSRA synthetic peptides was low, and the likely explanation for this could be the hindered accessibility of PfSRA in the native context. However, our data from IFAs showed that the immuno-affinity purified, human α-PfSRA peptide antibodies labeled native PfSRA, which suggests that malaria-infected populations have naturally acquired antibodies against PfSRA.
Generally, several receptors have been characterized based on their sensitivities to different enzyme treatments. Notably, neuraminidase removes sialic acids from glycophorins, trypsin cleaves peptide backbones of several receptors (glycophorin A, glycophorin C, and complement receptor 1), and chymotrypsin cleaves glycophorin B and complement receptor 1, among others [48]. We have shown that the 32-kDa-processed fragment of native PfSRA binds normal human erythrocytes, but the molecular identity of the receptor remains unknown. However, we have classified the putative receptor for PfSRA as sialic acid-dependent on the basis of its binding specificity, which is sensitive to treatments with both neuraminidase and trypsin but resistant to chymotrypsin, a binding phenotype that fits the description of the receptor glycophorin C. The enzyme sensitivity profile of PfSRA binding to erythrocytes is similar to that observed for PfEBA-140 (region II), the parasite ligand for glycophorin C [49]. Additional investigations are required to determine whether PfSRA also interacts with glycophorin C, possibly via a different binding site.
Our data revealed that PfSRA peptides induce functional antibodies that inhibited P. falciparum erythrocyte invasion of both laboratory strains and clinical isolates. The observed invasion inhibitory activity of rabbit α-PfSRA peptide antibodies could be attributed to indirect effects of antibody binding to the merozoite surface or direct inhibition of proteolytic processing events. The demonstration that PfSRA synthetic peptides induced erythrocyte invasion inhibitory antibodies and the successful purification of a limited amount of PfSRA-specific human antibodies from patient plasma suggested that the synthetic peptides possessed structural integrity or conformation that mimics the native PfSRA.
In summary, this study has provided relevant new information regarding the proteolytic processing of PfSRA that supports the idea of targeting these cleavage events for development of antimalarial therapies. The expression of PfSRA in late stages of gametocytes is an unprecedented opportunity that should be explored for potential transmission-blocking vaccines. Also, PfSRA-specific immune responses triggered in natural infections may inform the inclusion of PfSRA as a candidate for epitope-based, blood-stage malaria vaccine development.