INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

Several Plasmodium falciparum proteins play a role in merozoite invasion of erythrocytes (2, 15). Among these, the proteins that participate in the sequence of events leading to invasion include MSP-1, which possibly mediates initial contact between merozoites and erythrocytes, and EBA-175, a micronemal protein, which binds to erythrocytes and may be involved in junction formation.

EBA-175 may bind to erythrocytes via two mechanisms: an initial binding, which is dependent on sialic acid, and a secondary binding, which is not dependent on sialic acid. A conserved region of 42 aa of EBA-175, EBA-peptide 4(1062-1103), has been implicated in the binding to the erythrocyte (16), although it is not essential for the initial sialic acid-dependent binding (17). We have synthesized peptides from this putative erythrocyte binding region of EBA-175 and used them for identification of the minimum peptide sequence mediating attachment to erythrocytes. This peptide binding is not dependent on sialic acid. We also report that the erythrocyte binding sequence is recognized by IgG antibodies of children with acute malaria but not by IgG antibodies of children with asymptomatic infections nor by IgG antibodies of adults living in regions of malaria transmission. Antibodies to EBA(aa1076-96) can be induced in mice by immunization.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Abbreviations used in this paper: aa, amino acids; EBA, erythrocyte binding antigen; ELISA, enzyme-linked immunosorbent assay; Fmoc, fluorenylmethoxycarbonyl; HOBt, hydroxybenzotriazol; HPLC, high-pressure liquid chromatography; Ig, immunoglobulin; MBHA, methylbenzhydrylamine; MSP-1, merozoite surface protein 1; NMM, N-methylmorpholine; NMP, N-methylpyrrolidone; OD, optical density; PBS, phosphate-buffered saline; PPD, purified protein derivative; SD, standard deviation; TBTU, 2-(1H-benzotriazol-1-yl)-1,1,3,3-tetramethyluronium tetrafluoroborate.

Synthetic peptides. The sequences of the synthetic peptides synthesized are as follows. PfMSP-1 peptide has the sequence YSLFQKEKMVL, a sequence included in the malaria vaccine SPf66. The EBA-175 peptides contain overlapping amino acid sequences from the EBA-175 region 1062 to 1104: EBA(aa1062-86) peptide, SNNEYKVNEREDERTLTKEYEDIVL; EBA(aa1076-96) peptide, TLTKEYEDIVLKSHMNRESDD; EBA(aa1086-1104) peptide, LKSHMNRESDDGELYDENS. The following truncated EBA(aa1076-96) peptide variants were produced: EBA(aa1077-96), LTKEYEDIVLKSHMNRESDD; EBA(aa1078-96), TK EYEDIVLKSHMNRESDD; EBA(aa1079-96), KEYEDIVLKSHMNRESDD; EBA(aa1080-96), EYEDIVLKSHMNRESDD; EBA(aa1081-96), YEDIVLKSHMNRESDD; EBA(aa1082-96), EDIVLKSHMNRESDD; EBA(aa1083-96), DIVLKSHMNRESDD; EBA(aa1084-96), IVLKSHMNRESDD; EBA(aa1085-96), VLKSHMNRESDD; EBA(aa1086-96), LKSHMNRESDD; and EBA(aa1087-96), KSHMNRESDD. The EBA-175 peptides covered a sequence reported to be involved in erythrocyte binding, while the PfMSP-1 peptide is included in the SPf66 vaccine (14) and has been reported to bind to erythrocytes (1).

Peptide binding to glycophorin in ELISAs. Glycophorin A, glycophorin B, and desialylated glycophorin A were purified as previously described (11, 20). Another batch of glycophorin A was purchased from Sigma (St. Louis, Mo.). Glycophorin preparations (10 µg/ml) were coated in 100 mM NaHCO₃ (pH 9.6) on Maxisorp microtiter plates (Nunc, Roskilde, Denmark). All coatings were performed overnight at 4°C. The wells were washed four times in 0.5 M NaCl-3 mM KCl-1 mM KH₂PO₄-8 mM Na₂HPO₄-1% Triton X-100. This washing procedure was done after each of the following incubation steps: (i) biotinylated peptides in twofold dilution series (stock 2 mg/ml) diluted in incubation buffer (washing buffer plus 15 mM bovine serum albumin [pH 7.2]) were incubated overnight at 37°C; (ii) 100 µl of streptavidin peroxidase (DAKO) diluted in incubation buffer was added per well at room temperature for 1 h. Enzyme activities were quantitated after addition to each well of 100 µl of 0.67-mg/ml 1,2-phenyldiamine hydrochloride (DAKO) dissolved in 100 mM citric acid-phosphate buffer (pH 5.0) containing 0.015% (vol/vol) H₂O₂. The reactions were stopped by adding 50 µl of 2.5 M H₂SO₄ per well, and the OD values were measured in an ELISA scanner at 495 nm. All tests were done in duplicate.

Flow cytometric analysis of peptide binding to erythrocytes. Erythrocytes were washed three times in 10 volumes of PBS each time. Biotinylated peptides were resuspended in PBS and incubated in 50 µl (1 volume) with 10⁶ erythrocytes per reaction for 1 h at room temperature and then washed three times in 20 volumes of PBS. Quantum red-conjugated streptavidin (Sigma, St. Louis, Mo.) and Ig (Miles Inc., Elkhart, Ind.) at 2 mg/ml as blocker was added, and the mixture was incubated for 15 min at room temperature. The samples were analyzed by flow cytometry after being washed three times with 20 volumes of PBS.

Effect on parasite growth in vitro. P. falciparum isolate 3D7 was kept in continuous cultures as described by Jepsen and Andersen (6), with RPMI 1640 supplemented with 21 mM sodium bicarbonate, 25 mM HEPES buffer and 10% human serum. The parasites were grown in 4% (vol/vol) group 0 positive human erythrocytes.

Immunization of mice. EBA(aa1076-96) peptide (1.2 mg) was conjugated to 1 mg of PPD of mycobacteria (5 mol of peptide per mol of PPD) with equal volumes of 0.2% glutaraldehyde diluted in 0.1 M phosphate buffer (pH 7.5).

Human sera. Sera were collected from three different regions of endemic malaria infection. (i) Sera were collected in 1984 from 15- to 67-year-old donors from villages in a region of holoendemic infection in Irian Jaya, Indonesia. At the time of serum collection, the principal infections were with P. falciparum and P. malariae. Splenomegaly was common in the study population. (ii) Sera were collected in 1984 during the medium transmission season from 19- to 28-year old soldiers in a region of hyperendemic infection in Juba at the White Nile river in Sudan. The principal infection was with P. falciparum. (iii) Sera were collected in 1986 from 15- to 80-year-old donors in a region of holoendemic infection in Enugu near Nyssuka, Nigeria. The principal infection was with P. falciparum.

Antibody reactivity with native proteins and synthetic peptides in ELISA. Peptides conjugated to ovalbumin (1 µg/ml) (to test mice sera) or incorporated into immunostimulating complexes (1 µg/ml) (to test human sera) were coated in 100 mM NaHCO₃ (pH 9.6) on Maxisorp microtiter plates (Nunc) overnight at 4°C. Washing, incubation, substrate diluents, and OD measurements were as described for the glycophorin solid-phase enzyme assay. The washing procedure was done after each of the following incubation steps: (i) mouse or human sera made to 1% (vol/vol) in incubation buffer were incubated for 1 h at room temperature; (ii) 100 µl of peroxidase-conjugated rabbit anti-human IgG or biotinylated rabbit anti-mouse IgG antibodies (Amersham) per well diluted in incubation buffer was incubated for 1 h at room temperature, and (iii) 100 µl of streptavidin-conjugated peroxidase (DAKO) diluted in incubation buffer was added per well at room temperature for 1 h. All tests were done in duplicate.

Statistical methods. The data obtained with serum samples collected from Tanzanian children were analyzed by the Mann-Whitney rank sum test for intergroup comparisons because of the skewed data distributions. All P values less than 0.05 were considered significant. The calculations were performed with Sigmastat (Jandel Scientific, San Rafael, Calif.) software.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Evaluation of synthetic peptides. Peptide purities were ascertained through HPLC purification and within-peak evaluation with diode array detection. The peaks were >95% pure, and all had good purity factors. Mass spectrometry confirmed the predicted molecular masses within 3 Da.

EBA peptide binding to erythrocytes. Biotinylated EBA(aa1076-96) peptide showed a strong concentration-dependent binding to glycophorin A immobilized on polystyrene plates compared to biotinylated peptides EBA(aa1086-1104) and EBA(aa1062-86) as well as the PfMSP-1 peptide (Fig. 1a). Similar binding of biotinylated EBA(1076-96) peptide to human erythrocyte ghosts was detectable in four different experiments (data not shown). Saturation of peptide binding to glycophorin A was obtained with concentrations exceeding 8 µM biotinylated EBA(aa1076-96) peptide. Figure 1b shows that native EBA-175 inhibited the binding of biotinylated EBA(aa1076-96) peptide to glycophorin A; 9% of the peptide binding was not blocked by native EBA-175 and appears to be nonspecific background binding. To test the specificity of the peptide binding to glycophorin A, we also tested peptide binding to desialylated glycophorin A and to glycophorin B immobilized on polystyrene plates in addition to glycophorin A. EBA(aa1076-96) peptide showed similar binding to all the glycophorin preparations tested (data not shown), indicating that peptide binding is not specific for glycophorin A and is not dependent on sialic acid.

View larger version (15K): [in this window] [in a new window]   FIG. 1.   Binding of biotinylated peptides to glycophorin A. (a) Binding of biotinylated peptides to ELISA wells coated with glycophorin A protein. , PfMSP-1 peptide; , EBA(aa1062-86) peptide; , EBA(aa1076-96) peptide; , EBA(aa1086-1104) peptide. Background ODs were less than 200. The mean and SD for four experiments are shown. (b) Blocking of biotinylated EBA(aa1076-96) to polystyrene plate wells coated with glycophorin A by serial dilutions of native EBA-175. The mean and SD for three experiments are shown.

View larger version (19K): [in this window] [in a new window]   FIG. 2.   Binding of biotinylated truncated EBA(aa1076-96) variants to polystyrene wells coated with glycophorin A protein. , EBA(aa1076-96); , EBA(aa1087-96); , EBA(aa1086-96); , EBA(aa1085-96). The mean and SD for three experiments are shown.

EBA peptide binding to erythrocytes in flow cytometry analysis. EBA(aa1076-96) peptide bound to erythrocytes (Fig. 3). Maximum binding was 85% of the erythrocytes at peptide concentrations above 1 mg/ml. The binding decreased to 65% at peptide concentrations of 0.5 mg/ml. EBA(aa1086-1104) peptide did not bind to erythrocytes (data not shown).

View larger version (22K): [in this window] [in a new window]   FIG. 3.   Flow cytometry analysis of peptide binding to erythrocytes. (a) Binding of biotinylated EBA(aa1076-96) (2 mg/ml); (b) binding of control secondary antibody alone.

Inhibition of parasite multiplication in vitro by EBA peptide. EBA(aa1076-96) peptide showed a concentration-dependent blocking of parasite growth in vitro (Fig. 4). At 8 µM EBA(aa1076-96), peptide almost completely blocked parasite multiplication. Peptide EBA(aa1086-1104) showed no blocking of parasite multiplication.

View larger version (18K): [in this window] [in a new window]   FIG. 4.   Blocking of parasite multiplication in vitro by EBA(aa1076-96) () and EBA(aa1086-1104) (). The mean percent inhibition and SD for four independent experiments are shown.

Reactivities of human sera to EBA peptides in ELISA. Serum reactivity with peptide was defined as being positive if the OD of the serum sample was higher than the mean plus 3 times the standard deviation of OD values obtained with 10 Danish controls. When tested at a dilution of 1:100, only a proportion of the sera tested reacted with EBA(aa1076-96) peptide. Of 44 tested Sudanese serum samples, 5 were reactive with the peptide, whereas 6 of 33 Indonesian sera and 2 of 20 Nigerian sera were reactive.

Immunization of mice with EBA peptides. BCG-primed mice immunized with PPD-conjugated EBA(aa1076-96) peptide in the absence of Freund's complete adjuvant produced antibodies against this peptide (Fig. 5). The antibody reactivity increased with each immunization, and intraperitoneal immunization was superior to subcutaneous immunization. Sera from mice immunized three times inhibited EBA(aa1076-96) peptide binding to glycophorin A in the solid-phase assay (Table 1), while sera from the Indonesian, Nigerian, and Sudanese donors did not block peptide binding to glycophorin A whether they were reactive with the EBA peptide or not.

View larger version (16K): [in this window] [in a new window]   FIG. 5.   ELISA for murine IgG reactivities to EBA(aa1076-96). Results of duplicate tests on four mice in each group are shown. , peptide-PPD intraperitoneal immunization; , peptide-PPD subcutaneous immunization; , PPD control intraperitoneal immunization; , PPD control subcutaneous immunization.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

The main finding in this study is the identification of a 12-aa peptide sequence of the malaria vaccine candidate EBA-175 which binds to glycophorin and may be involved in the invasion process of merozoites into erythrocytes. The 12-aa sequence is contained within a 43-aa conserved sequence originally identified as the putative erythrocyte binding region, since rabbit antisera against this region block parasite multiplication and EBA-175 binding to erythrocytes (13, 19).

When merozoites attach to and invade erythrocytes, a sequence of events takes place in which the first step is a lectin-like binding of merozoites to erythrocytes followed by reorientation of merozoites, bringing the apical pole in contact with the erythrocyte and leading to junction formation (7). Two different proteins, EBA-175 and MSP-1, have been implicated in mediating the initial binding of merozoites to the erythrocyte. EBA-175 is a protein, located in micronemes (18) and released into culture supernatants (2), that binds N-acetylneuraminic acid, 2-3-Gal determinants on O-linked carbohydrates of glycophorin A on the erythrocyte membrane (12). EBA-175 may mediate a two-step invasion procedure, an initial lectin-like binding followed by a second, possibly hydrophobic binding, triggering internalization of the merozoite. A cysteine-rich region of EBA, the F2 fragment, mediates the lectin-like binding to glycophorin A (17), while a 65-kDa processing fragment of EBA was reported to bind to an erythrocyte determinant in a sialic acid-independent manner (8). This fragment does not contain the 43-aa sequence, and the relative role of the 65-kDa fragment and the fragment containing the 43-aa sequence in any secondary binding steps remains unknown.

To further characterize the erythrocyte binding sequence, we synthesized three overlapping peptides covering the 43-aa region. We showed that EBA(aa1076-96) bound strongly to erythrocytes and more specifically to glycophorin A when tested in a solid phase binding assay. We also showed that soluble EBA(aa1076-96) bound to intact erythrocytes. The binding of EBA(aa1076-96) to erythrocytes does not appear to be dependent on sialic acid, since the peptide bound to desialylated glycophorin A. We hypothesized that EBA(aa1076-96) is involved in the second step of a two-step binding process which may resemble HIV-1 gp160 binding to lymphocytes, where both gp120 and gp41 processing fragments remain attached to the lymphocytes through different binding sites. Virus entry is facilitated by an envelope-mediated fusion of the viral and target cell membranes. After formation of gp120 binding to CD4 as well as processing of gp160 to gp120 and gp41 fragments, the mobility of the envelope protein is afforded by the noncovalent nature of the gp120-gp41 bond, which may allow efficient exposure of the lymphocyte membrane to the hydrophobic gp41 regions that mediate the fusion process (10). Multiple regions of gp41 are involved in the invasion process by interaction with CD4 and other cellular receptors, as well as being involved in conformational changes of gp41 (21). Likewise, we hypothesized that the initial and specific cystein-rich fragment binding to sialic acid on glycophorin A induces a conformational change in EBA-175 which may expose the EBA(aa1076-96) peptide fragment to subsequent erythrocyte binding. Both the EBA(aa1076-96) peptide-containing fragment and the 65-kDa fragment reported by Kain et al. (8) may participate in the secondary binding to the erythrocyte, which is independent of sialic acid and may be more nonspecific. The two regions play different roles, since an EBA-175 peptide of 42 aa comprising the EBA(aa1076-96) peptide blocks binding of the full-length EBA-175 whereas a peptide from the 65-kDa EBA-175 fragment blocks the binding only of the 65-kDa fragment but not of full-length EBA-175 (8, 19). To further characterize the binding sequence of the EBA(aa1076-96) peptide, we showed that the amino acid sequence, VLKSHMNRESDD, at positions 1085 to 1096 of EBA-175 contained the binding sequence when the binding of 11 truncated peptide variants, EBA(aa1077-96) to EBA(aa1087-96), to glycophorin A immobilized on polystyrene plates was tested. To substantiate the evidence that EBA(aa1076-96) contains the erythrocyte binding site, we tested the ability of EBA(aa1076-96) peptide to block parasite multiplication. The peptide showed a strong parasite-blocking activity. EBA(aa1076-96) peptide may compete with merozoites for binding to erythrocytes.

Antibodies against erythrocyte binding domains of EBA-175 may block the ability of merozoites to invade erythrocytes. Such antibodies may be responsible for the achievement of clinical immunity against malaria. The immune response to EBA-175 among humans living in regions of endemic infection remains poorly characterized. However, lymphocyte proliferation responses to EBA(aa1086-1104) but not to EBA(aa1076-96) among Ghanaian donors have been reported (4). None of the peptides were recognized by IgG antibodies from the Ghanaian donors. In this study, we selected sera from donors living in Indonesia, Nigeria, and Sudan with long exposure to malaria; the sera were highly reactive with a recombinant RAP-1 (5). The majority of these sera had low or negliable IgG reactivity to the EBA(aa1076-96) peptide. However, sera from young Tanzanian children with clinical malaria had high IgG reactivities to EBA(aa1076-96), and their reactivities were higher than the IgG reactivities of sera collected from children with asymptomatic infections. Whether these antibodies play a harmful role needs further investigations. Our data indicate that there is a high antigenic threshold for induction of antibody reactivities against the peptide and that infections do not induce a sustained antibody response against the peptide. We then investigated whether antibodies against the EBA(aa1076-96) peptide could be induced by vaccination of mice. We found that PPD-conjugated peptide induced the formation of antibodies against the peptide and that these murine sera inhibited EBA(aa1076-96) peptide binding to glycophorin A, in contrast to human sera. In conclusion, we have identified a short amino acid sequence of EBA-175 capable of binding to glycophorin A of erythrocytes. The peptide binding does not depend on sialic acid, but it may be involved in a second, relatively nonspecific binding step of EBA-175 during merozoite invasion. The peptide blocks parasite multiplication. Finally, the erythrocyte binding sequence appears to induce only an unstable antibody response during P. falciparum infections of humans, but antibodies against the peptide can be induced by vaccination.