Evidence for Globally Shared, Cross-Reacting Polymorphic Epitopes in the Pregnancy-Associated Malaria Vaccine Candidate VAR2CSA

Parasite and mammalian cell lines. P. falciparum parasites were grown in O+ erythrocytes and human serum. Clinical isolate Pl0711 was isolated from peripheral blood samples of a pregnant woman with placental malaria (17) and selected on CSA prior to analysis. Laboratory isolates FCR3, 7G8, HB3, and 3D7 were initially selected on the human placental BeWo cell line (45) or CSA and subsequently maintained by panning on 200 μg/ml bovine CSA (Fluka Biochemika) coating plastic petri dishes. The FCR3 and IT4 parasite lines are isogenic due to historical contamination (31) and are therefore considered to have the same var2CSA gene. Chinese Hamster Ovary K1 cells (CHO-K1) and CSA-deficient CHO-745 derivative lines were cultured in F-12 Kaign's medium (Gibco). CHO-745 cell lines transfected with CD36 (CHO-745-CD36) (10) or VAR2CSA DBL domains (19) were maintained under selection by G418 at 500 μg/ml (Invitrogen). CHO-745 (CSA-negative), CHO-745-CD36, and CHO-K1 (CSA-positive) cells were used to confirm the binding phenotype of parasites in this study.

Design of DBL synthetic genes. var2csa synthetic genes were designed with optimized codons for human expression so they could be shuttled between different plasmid vectors for human, rabbit, and yeast cell expression. Potential N-glycosylation sites were removed from synthetic genes by converting asparagine to glutamine or by replacing asparagine with an amino acid from another VAR2CSA allele.

Recombinant protein expression in P. pastoris.Synthetic genes encoding FCR3 VAR2CSA DBL domains (accession no. AY372123) DBL1 (residues 58 to 383), DBL2 (residues 536 to 858), DBL3 (residues 1221 to 1541), DBL4 (residues 1594 to 1888), DBL5 (residues 2003 to 2270), and DBL6 (residues 2322 to 2590) were cloned into the pPIC9K vector containing an N-terminal α-factor secretion signal (Invitrogen). Synthetic genes contained a His₆ tag on the C terminus. Constructs were digested with SacI and transformed into electrocompetent P. pastoris strain GS115. The transformation results in DNA insertion at the AOX1 locus, generating a His⁺ Mut⁺ phenotype. Yeast clones were grown in buffered complex medium (1% yeast extract, 2% peptone, 1% yeast nitrogen base, 1 M potassium phosphate buffer, pH 6.0) plus 2% glycerol as a carbon source. Starting from an overnight culture, yeast cultures were grown for 3 days at 20°C, with shaking at 250 rpm. After each 24-h interval, recombinant protein expression was induced by adding methanol to a 0.5% final concentration. Recombinant proteins were harvested 72 h postinduction. Proteins were purified by using nickel-nitrilotriacetic acid-agarose (Qiagen) on an Econo-Pac chromatography column (Bio-Rad) and eluted by a step gradient from 30 mM to 450 mM imidazole. Since FCR3-DBL3 was produced at lower levels, this domain was also cotransformed with the plasmid pPICZαA-PpPDI, which overexpresses P. pastoris protein disulfide isomerase (PDI), to increase recombinant protein production, and fermented in 5-liter bioreactors as described previously (40). Recombinant protein identity was confirmed by mass spectrometry.

Plasma samples.The human plasma samples used in these studies were collected from East African donors, under protocols approved by the relevant ethics review committees. Study participants provided written, informed consent before donating samples and included adult men and multigravid women from Kenya (28) as well as from Tanzania (2).

Recombinant protein characterization. P. pastoris recombinant proteins were analyzed in 4 to 20% sodium dodecyl sulfate-polyacrylamide gel electrophoresis gels under reduced or nonreduced conditions (Bio-Rad). Gels were stained with Coomassie blue or transferred to a nitrocellulose membrane and detected by Western blot analysis using anti-His tag antibodies (Invitrogen). DBL1, -3, -4, and -6 recombinant proteins were tested by enzyme-linked immunosorbent assay against immune and nonimmune sera according to published methods (32).

Construction of VR1051/DBL plasmids for DNA immunization.The six synthetic genes used for P. pastoris protein production (minus the His₆ tag) were cloned into the VR1051 vector (34). VR1051 contains a tissue plasminogen activator leader sequence to direct recombinant proteins into the secretory system and expresses proteins fused to the P2P30 universal helper T-cell epitope. In addition, we cloned 3D7-DBL2x (residues 542 to 853; accession no. XM_001350379) and A4-CIDR1 (residues 402 to 846; accession no. L42244). Plasmid DNA was prepared by using the EndoFree plasmid DNA purification kit (Qiagen). Inserts were confirmed by DNA sequencing. All constructs contained less than 10 endotoxin U/ml (Pyrosate kit PSD10; Associates of Cape Cod, Inc.).

Immunizations.Immunizations were performed at R&R Rabbitry according to animal immunization guidelines. In brief, rabbits received 100 or 500 μg of recombinant protein in complete Freund's adjuvant for first immunizations and half as much protein in five subsequent immunizations (incomplete Freund's adjuvant). For DNA immunization, rabbits were immunized intramuscularly six to 10 times by injection of 500 μg of DNA with Vaxfectin adjuvant (DNA-to-vaxfectin molar ratio of 2:1) according to published methodologies (20).

Real-time PCR.Quantitative real-time PCR was performed on RNA harvested ∼10 to 12 h postinvasion as described previously (44). Published cross-reactive primers to the VAR2CSA DBL4 domain (33, 42) or gene-specific primers to the two HB3 var2CSA genes were used. Primer sequences for HB3 allele A were 5′ TGACGAAAATTTATTTAAAGC-3′ (forward) and 5′ TTCACACTTGTTTTCATCTC-3′ (reverse); for HB3 allele B, the primer sequences were 5′ TCAGCAAAATTCATCTGATG-3′ (forward) and 5′ CAATTGATCTGTACCTGTA-3′ (reverse). The DBL4 and HB3 primer sets were optimized to the same conditions as the 56 IT4 var primers and housekeeping controls (44). Results are shown as the difference in expression relative to the housekeeping P. falciparum adenylosuccinate lyase (Asl) (PFB0295w) (44).

Flow cytometry analysis.For flow cytometry analysis, a confluent monolayer of CHO-745 cells was lifted by phosphate-buffered saline (PBS) containing 4 to 8 mM EDTA and 1 million CHO-745 cells were incubated with 1/25 dilutions of rabbit anti-VAR2CSA sera. The monoclonal antibody (MAb) 179 (6 μg/milliliter; Affymax Research Institute) and 12CA5 mouse ascites fluid (1/250) (MAb HA.11; BAbCo) were used to monitor the expression of DBL recombinant proteins on CHO-745 transfectants (19). Primary antibodies were detected by using Alexa Fluor 488-conjugated goat anti-rabbit IgG (1/500; Molecular Probes) and Alexa Fluor 488-conjugated goat anti-mouse IgG (1/1,000; Molecular Probes).

Mature-stage IEs were washed and resuspended in a 100-μl volume of PBS-1% bovine serum albumin containing 1/25 dilutions of rabbit sera preadsorbed on uninfected erythrocytes. As controls, mouse monoclonal BC6 was used to determine A4var PfEMP1 expression in the A4ultra culture (36) and pools of immune plasma from males or multigravid women (1/10 dilutions) were used. Bound antibodies were detected with fluorescein isothiocyanate-conjugated goat F(ab′)2 anti-human IgG (Beckman, PN IM1627) or Alexa Fluor 488-conjugated goat anti-rabbit IgG. Infected erythrocytes were detected using ethidium bromide (Sigma) at 0.02 mg/ml. Samples were analyzed in a Beckman Coulter Epics XL-MCL (Tree Star, Inc.). Preimmune sera were used to set a gate to determine the percentage of CHO-745 cell transfectants or infected erythrocytes reactive with immune sera and to calculate the difference in median fluorescence intensity (MFI) between preimmune and immune sera. Data were converted to normalized MFI by subtracting the MFI value of preimmune serum from the MFI value of immune serum.

Antibody inhibition of IE binding to CHO cells.FCR3-CSA IEs were enriched by 0.7% porcine gelatin flotation and resuspended at 1 × 10⁷ IEs/ml in RPMI medium (pH 7.2). Infected erythrocytes were preincubated for 1 h at 37°C in a 1/25 final dilution of preimmune or immune serum. After washing, IE-antibody complexes were added to CHO-K1 cells (CSA positive) and incubated for 1 h at 37°C in RPMI 1640 binding medium containing a 1/25 dilution of rabbit antibodies. The percentage of IE binding inhibition in the presence of immune serum was calculated relative to preimmune serum.

Protease treatment of IE surface antigen.For protease treatments, IEs were washed three times in PBS (pH 7.2) and incubated with trypsin-N-tosyl-l-phenylalanine chloromethyl ketone (Sigma) or chymotrypsin-Nα-p-tosyl-l-lysine chloromethyl ketone (Sigma) at 1 mg/ml or 0.1 mg/ml in PBS for 15 or 30 min at 37°C. Control samples were incubated with PBS alone. Digestion was stopped by the addition of 1 mg/ml soybean trypsin/chymotrypsin inhibitor (Sigma) for 5 min or culture medium containing 10% human serum. Protease-treated IEs were tested by flow cytometry for reactivity with rabbit antibodies as described above.

Expression of P. pastoris recombinant proteins.To determine conditions for producing individual DBL immunogens, the six DBL domains of the IT4/FCR3 form of VAR2CSA were expressed as secreted His-tagged recombinant proteins in the methylotropic yeast P. pastoris (Fig. 1A). Using this system, recombinant DBL1 (rDBL1), rDBL4, and rDBL6 were produced at approximately 1 to 4 mg/liter of yeast culture, while rDBL3 protein was produced at lower levels. However, rDBL3 protein production could be increased to 0.5 mg/liter broth by cotransformation with a plasmid that overexpresses P. pastoris PDI (40) and by controlled fermentation conditions. In contrast, we were unable to produce appreciable amounts of rDBL2 and rDBL5 proteins, although these same construct boundaries were previously used for expressing proteins at the surfaces of CHO-745 cells (19).

• Open in new tab
• Download powerpoint

FIG. 1.

Production of FCR3-VAR2CSA recombinant proteins using P. pastoris. (A) Schematic diagram of FCR3-VAR2CSA domain organization. Potential CSA-binding domains are black (19). DBL amino acid boundaries are labeled, and the positions of the mutated N-glycosylation sites are indicated by circles (black circles for glutamine replacement and gray circles for conservative amino acid replacements). At the bottom, 5 μg of rDBL1, rDBL4, and rDBL6 and 3 μg of rDBL3 were loaded on a 4 to 20% gradient Tris-HCl polyacrylamide gel electrophoresis gel under nonreduced and reduced conditions and stained with Coomassie blue stain (B) or detected using anti-His tag antibodies (C). Molecular mass standards are shown outside the panels. (D) Enzyme-linked immunosorbent assay showing gender-specific recognition of P. pastoris rDBL proteins by human IgG. Three groups of human plasma were tested: 30 P. falciparum exposed women (W), 12 P. falciparum exposed men (M), and 5 nonimmune adults (Non Im). Median (central bar), central 50% (boxes), and central 90% (whiskers) levels of human plasma IgG reactivity with recombinant proteins. Statistics were determined by the Kruskal-Wallis test. Each of the four groups had an overall significant P value, and pairwise comparisons within each group achieving significance are shown (*, P < 0.05; **, P < 0.01; ***, P < 0.001). OD, optical density.

Under sodium dodecyl sulfate-polyacrylamide gel electrophoresis reducing conditions, all four proteins migrated at the expected molecular weights (Fig. 1B and C). However, rDBL3 and rDBL1 contained minor, lower-molecular-weight products. These minor products are likely nicked proteins, which can occur due to the limited lysis of P. pastoris and subsequent release of proteases (11, 40). Furthermore, in contrast to the other proteins, the majority of rDBL6 migrated as a dimer or larger forms under nonreducing conditions (Fig. 1B and C). DBL domains contain several conserved intradomain disulfide bonds (35, 38); however, whether a disulfide interaction normally occurs between DBL6 domains in the native VAR2CSA protein is not known. All four P. pastoris recombinant proteins were recognized in a gender-specific manner by plasma from Tanzanian women with histories of malaria exposure during pregnancy compared to plasma samples from malaria-exposed men from the same area of endemicity or to nonimmune samples (Fig. 1D).

Antibodies to VAR2CSA react with homologous FCR3-CSA-infected erythrocytes.Rabbits were immunized with the four recombinant proteins or the six individual IT4/FCR3 VAR2CSA DBL domains cloned into a plasmid DNA vector (VR1051) (3). In addition, rabbits were immunized with plasmid DNA encoding the 3D7-DBL2 VAR2CSA and the A4var-CIDR1 domain from a nonplacental-type PfEMP1 protein. A4var was selected as a control because a MAb specific to the A4var PfEMP1 protein (MAb BC6) can be used to select IEs expressing this specific PfEMP1 variant from the IT4/FCR3 parasite genotype (36). The immunization protocols are summarized in Table 1.

Antibody titers, as determined by end-point dilution, ranged between 10⁴ and 10⁶ for both protein and DNA immunizations and peaked after four immunizations (data not shown). To assess antibody cross-reactivity between different VAR2CSA domains and between different allelic forms of a given domain, each serum sample was screened against the DBL1, DBL2, and DBL3 domains in IT4-VAR2CSA (homologous domains) and all six DBL domains from 3D7-VAR2CSA (heterologous domains) expressed at the surface of CHO-745 cells (19) (Table 2). The six DBL domains of IT4/FCR3 and 3D7 have 85%, 76%, 81%, 80%, 85%, and 64% identity, respectively. In general, antisera had strong reactivity to both homologous and heterologous forms of the immunizing domain but limited cross-reactivity to other VAR2CSA domains (Table 2). There were only a few examples of partial cross-reactivity between different VAR2CSA domains; IT-rDBL3 antiserum (rabbit 1) reacted with both DBL1 and DBL3 domains, IT-rDBL4 (rabbit 1) reacted with DBL4, DBL6, and DBL2β, and IT-rDBL6 reacted with DBL6 and DBL1. Cross-reactivities between VAR2CSA domains were confined to rabbits in the protein immunization group. Therefore, antisera were generally highly specific to the type of domain used for immunization and VAR2CSA recombinant proteins contain cross-reactive epitopes.

To investigate whether VAR2CSA antisera could recognize native protein, rabbit sera were tested by flow cytometry on FCR3-CSA-infected erythrocytes (homologous parasites) and A4ultra-infected erythrocytes. Using gene-specific primers to the 56 known var genes in the IT4/FCR3 parasite genome (44), we found that var2CSA was the predominant var gene transcribed in FCR3-CSA (greater than 19:1 ratio relative to housekeeping gene P. falciparum Asl) and A4var was the predominant var gene in the A4ultra culture (∼6:1 ratio) (Fig. 2). Conversely, var2CSA and A4var were present at a ratio much less than 0.1:1 when the gene was not “on” (Fig. 2). As expected, FCR3-CSA-infected erythrocytes bound CSA and were preferentially recognized by pooled sera from multigravid women, while A4ultra-infected erythrocytes did not bind CSA and were recognized equally well by pooled sera from women or males (Fig. 3 and data not shown).

• Open in new tab
• Download powerpoint

FIG. 2.

Transcriptional analysis of var genes in CSA-binding parasite lines. var gene transcription in FCR3-CSA and A4ultra parasite lines was compared using the full set of 56 IT4 var primers. The FCR3-CSA parasite line predominantly transcribes var2csa. The A4ultra parasite line predominantly transcribes A4var and expresses AFBR6 var at lower levels. Gene expression is compared to that of the housekeeping gene Plasmodium falciparum adenylsuccinate lyase. For the insets, var2csa transcription was separately analyzed using two primers specific to the VAR2CSA DBL4 domain (DBL4e var2csa and 79a PFL0030c) or with allele-specific primers to HB3 var2CSA alleles A and B. STS, seryl-tRNA synthetase (PFD7_0073); Arg, arginyl-tRNA synthetase (PFL0900c); Glut, glutaminyl-tRNA synthetase (PF13_0170).

• Open in new tab
• Download powerpoint

FIG. 3.

Reactivity of VAR2CSA antisera with FCR3-CSA- and A4ultra-infected erythrocytes. Sera from DNA immunizations (one rabbit) or protein immunizations (two rabbits) were tested against homologous (FCR3-CSA) or heterologous (A4ultra) parasite isolates by flow cytometry. The results show the means and standard deviations (error bars) of the normalized MFI and percentage of infected erythrocytes reactive (% IE) from four distinct experiments for FCR3-CSA and in triplicate for A4ultra. The histogram provides an example of how the values for normalized MFI and the percentage of IEs reactive were calculated for serum raised against pDNA-A4-CIDR1 reacting with A4ultra. The percentage of IE indicates the shift or percentage of IEs that label with immune sera above the level of preimmune sera. Normalized MFI is the difference in surface fluorescence intensity between preimmune and immune sera. Pooled sera from multigravid (multig.) Kenyan women or males were used as controls for gender-specific recognition of CSA-binding parasite lines. Immune serum reactivities below the pDNA VR1051 vector alone were considered negative.

Rabbit antibodies to the three VAR2CSA recombinant proteins (rDBL1, rDBL3, and rDBL6) and four plasmid DNAs (DBL1, DBL3, DBL5, and DBL6) reacted with homologous FCR3-CSA-infected erythrocytes (Fig. 3). Rabbit sera were considered negative if they produced a signal lower than that of plasmid DNA vector alone. In general, positive sera recognized more than 50% of IEs and the reactivities of anti-DBL1 and anti-DBL6 sera were similar to or higher than that of pooled multigravid sera. Conversely, VAR2CSA antisera did not recognize A4ultra-infected erythrocytes, although this parasite line was recognized equally well by anti-A4var CIDR1 sera and the A4var MAb, BC6 (Fig. 3).

In contrast to the case for other domains, antibodies to DBL2 and DBL4 reacted poorly or not at all with infected erythrocytes. Previous studies have shown that DBL4 reactivity can be revealed by treating infected erythrocytes with protease conditions that have been shown to partially cleave VAR2CSA (2, 6). To determine the effect of protease treatment, FCR3-CSA-infected erythrocytes were treated with trypsin or chymotrypsin. Both treatments led to a dose- and time-dependent increase in DBL4 reactivity and a partial loss of DBL1 and DBL3 reactivity (Table 3). Of interest, DBL4 is the most conserved VAR2CSA domain (8). Taken together, these observations suggest that the DBL4 immunogen may not be folded properly or that this domain may be less accessible to antibodies in the native protein.

None of the sera were able to efficiently block IE binding to CSA, although we obtained a partial binding inhibition, approximately 24% of FCR3-CSA-infected erythrocytes with the rDBL1x protein immune serum and also approximately 20% with the pDNA-DBL1x immune serum (data not shown). However, there was no CSA binding inhibition when sera were pooled from DBL1-DBL3 or DBL4-DBL6 (DNA immunization) or from DBL1 and DBL3 (protein immunization) (data not shown). Taken together, antibodies generated to individual VAR2CSA DBL domains were able to recognize native PfEMP1 protein, but unable to block infected erythrocyte binding.

VAR2CSA antibodies partially cross-react on geographically diverse CSA-binding infected erythrocytes.To study the antigenic relationship of different CSA-binding variants, rabbit sera were tested against four different CSA-binding parasite lines: 3D7-CSA (isolated from Amsterdam; origin ambiguous), HB3-CSA (isolated from Central America), 7G8-CSA (isolated from South America), and Pl0711 (a clinical isolate from a pregnant woman in Africa). These four parasite isolates all appear to have origins distinct from those of the FCR3 parasite, whose exact provenance is unknown due to historical contamination but which groups with southeast Asian isolates by DNA genotyping (26). Whereas FCR3, 3D7, HB3, and 7G8 parasites have been cloned in vitro and represent single-parasite genotypes (7, 29), the Pl0711 has only recently been adapted to in vitro cultivation and is likely to contain more than one parasite genotype (17).

Based upon quantitative reverse transcription-PCR, the three long-term adapted parasite isolates all transcribed var2CSA at ratios greater than 3:1, while var2csa was transcribed at an approximately 1:1 ratio in Pl0711 compared to the P. falciparum Asl housekeeping control gene (Fig. 2). Of interest, HB3 has two different var2CSA alleles (alleles A and B) and the HB3-CSA culture appeared to be a mixture of both variants. Consistent with transcriptional data, all four CSA-binding lines were recognized in a gender-specific manner by endemic sera, although the proportion of IEs reacting with pooled maternal antibodies differed among isolates and the MFI for all four CSA-binding lines was less than that for FCR3-CSA (Fig. 4), which may indicate weaker variant antigen surface expression.

• Open in new tab
• Download powerpoint

FIG. 4.

Cross-reactivity of VAR2CSA antisera with geographically diverse CSA-binding isolates. Antisera to the FCR3 VAR2CSA recombinant proteins were tested against four other different CSA-binding lines (3D7, 7G8, HB3, and P10711 clinical) from different continental regions. Sera were screened from all the plasmid DNA immunizations and from one of the two rabbits that were most reactive to FCR3CSA after protein immunization. Data were analyzed at least in duplicate as described in the legend for Fig. 3. Pooled sera from multigravid (multig.) Kenyan women or males were used as controls for gender-specific recognition of CSA-binding parasite lines. Immune serum reactivities below the pDNA VR1051 vector alone were considered negative. Error bars indicate standard deviations.

VAR2CSA immune sera cross-reacted with these parasites in different patterns, and none of the antisera recognized all four isolates (Fig. 4), indicating that vaccine sera did not target conserved epitopes exposed in the native protein. Instead, antibody cross-reactivity appeared to be due to shared polymorphic epitopes. For instance, DBL3 antibodies cross-reacted with 7G8-CSA and HB3-CSA lines and were very weakly cross-reactive reactive with P10711, but did not label 3D7-CSA above the negative control. This pattern of reactivity corresponded with molecular modeling and sequence analysis showing that FCR3 is identical to 7G8 at polymorphic segments 1, 3, and 5 and shares identity to HB3 alleles at segments 3 and 5, but has no polymorphic segments in common with 3D7 (Fig. 5). The presence or absence of common polymorphic epitopes also correlated with the antibody cross-reactivity to the DBL1 and DBL5 domains (data not shown). Therefore, common and pangeographically shared polymorphic epitopes were found in multiple VAR2CSA DBL domains.

• Open in new tab
• Download powerpoint

FIG. 5.

Amino acid conservation and diversity within the VAR2CSA DBL3 domain. The DBL3 sequences from the four CSA-selected parasite lines were aligned. The top shows the predicted DBL3 structure modeled on EBA-175 (38) and the location of five predicted polymorphic loops in color. As determined previously by molecular modeling and segmentation analysis (8), the sequence polymorphism at each loop is restricted and globally overlapping (colored sequence segments). At the bottom, polymorphic residues that are common between IT4 and other VAR2CSA alleles are indicated with a white font and are outlined in black. The locations of semiconserved B, D, F, and H blocks are indicated. The DBL structure and polymorphic segments for all but the 7G8 sequence are reprinted from reference 8 with permission of the publisher. Asterisks indicate conserved residues.