INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

Plasmodium falciparum causes the most virulent kind of malaria in humans and is almost exclusively responsible for all malaria-related deaths in the world. Several parasite antigens from the asexual erythrocytic stages, such as merozoite surface protein 1 (MSP-1), MSP-2, the apical membrane antigen 1, etc., which are targets of the potentially protective immune responses, are now being developed as candidates for vaccines (reviewed in reference 22). However, a major problem in developing an effective vaccine is the high degree of genetic diversity and antigenic variation found in the target antigens (5, 9, 27, 29, 35, 40). This problem is further aggravated by the fact that in several cases, these variant regions constitute immunodominant determinants with the potential to divert immune responses from critical epitopes and/or obstruct maturation of high-affinity antibodies to these epitopes (2, 3, 15). These critical epitopes might represent structures involved in some important processes, such as the merozoite invasion, which is a crucial event in the life cycle of the parasite and are, hence, rather conserved. For example, MSP-1 of P. falciparum has several blocks which display a high degree of polymorphism among various strains of the parasite (reviewed in reference 29). However, its C-terminal region, termed MSP-1₁₉, with its epidermal growth factor-like domains, is essentially conserved even across the species, and it is this region that has been shown to be critically implicated in the merozoite invasion of erythrocytes (6, 7, 20). Similarly, a highly conserved region II motif present in the circumsporozoite protein of all Plasmodium species sequenced so far seems to play an essential role in the sporozoite invasion of hepatocytes (11, 30). In fact, we (12), as well as others (36), have shown that immunization with synthetic peptides modeling highly conserved regions of the P. falciparum antigens can even protect mice against live challenge with the murine malaria parasites, viz., P. berghei or P. yoelii. Such conserved portions of malarial proteins are currently the subject of active investigation as putative vaccine molecules.

The antigen which contains repeats of acidic and basic residues (ABRA) of P. falciparum (41) seems to be another such highly conserved molecule. It is a 101-kDa protein located on the surface of merozoites as well as in the parasitophorous vacuole within the infected erythrocytes (14, 46). Significantly, this protein is also present in the immune clusters of merozoites which are formed at the time of rupture of mature schizonts in the presence of immune serum. Formation of such clusters prevents the dispersal of merozoites, resulting in a marked decrease in parasitemia, which is considered an indicator of protective immunity (17, 28). Furthermore, ABRA is almost fully conserved among various laboratory isolates of P. falciparum (46), possesses a chymotrypsin-like activity (33), and has a partial protein sequence homology with an extracellular cysteine protease of another protozoan, Trichomonas vaginalis (16). All these findings about ABRA seem to indicate its potential role in a protease-mediated process(es), such as merozoite invasion of erythrocytes, which is a critical event in the life cycle of the parasite. Because of its location on the merozoite surface, its presence in the immune clusters of merozoites, its highly conserved sequence, and its reported protease activity, ABRA represents an attractive molecule for development as a vaccine candidate.

In the present study, we have attempted to delineate the putative epitopic sequences of ABRA by using a battery of eight synthetic peptides based on its most hydrophilic regions and its repeat sequences. We found that these sequences represented target epitopes for the serum immunoglobulin G (IgG) antibodies in a large proportion of humans recovering from P. falciparum infection. They also stimulated the peripheral blood mononuclear cells from convalescing patients from an area of endemicity. Our results indicated that five out of six sequences from the nonrepetitive regions and one of the two repeat sequences elicited a boostable, IgG-type antibody response in rabbits immunized with the carrier-free peptides. We also found that antibodies against two of the peptides, both nonrepetitive, exerted a strong inhibitory effect on the merozoite invasion of erythrocytes, indicating the importance of these constructs in inducing a potentially protective antibody response.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Parasite. The FID-3 isolate of P. falciparum was maintained in continuous culture essentially according to the methods described by Trager and Jensen (42), and a detergent-soluble extract of the parasite proteins was prepared as described previously (38) for use as the antigen in the enzyme-linked immunosorbent assay (ELISA) or the immunoblotting assay. A schizont-rich preparation of the parasites was also used to obtain genomic DNA by phenol-chloroform extraction and ethanol precipitation following standard procedures. The quality and yield of genomic DNA was ascertained by agarose gel electrophoresis; this DNA was used to obtain the full-length ABRA-encoding gene and its various fragments by amplification using PCR as described below.

Synthetic peptides. Analysis of the predicted amino acid sequence of ABRA (46) according to the Chou-Fasman algorithm revealed six stretches, ranging from 12 to 18 residues, having a hydrophilicity score of 40% or more. These sequences are as follows (amino acid numbers are according to the numbering system of Weber and colleagues [46]): AB-1, ¹⁹NIISCNKNDKNQ³⁰; AB-2, ⁹⁹ANNSANNGKKNNAEE¹¹³; AB-3, ³⁹⁵YKAYVSYKKRKAQEK⁴⁰⁹; AB-4, ⁴⁴⁸LKNKIFPKKKEDNQAVDT⁴⁶⁵; AB-5, ⁵¹⁸VPPTQSKKKNKNET⁵³¹; and AB-6, ⁶³⁹ENDVLNQETEEEMEK⁶⁵³. In addition, two more constructs, representing the repetitive sequences in ABRA, were also synthesized: AB-7, TNDEEDTNDEEDTNDEED, and AB-8, KEEKE EKEEKEEKEKEKE. The procedures employed for synthesis, purification, and characterization of the synthetic constructs AB-1 to AB-8 were essentially the same as those described in our earlier work (24, 37, 38). Briefly, peptides were synthesized by stepwise solid-phase synthesis in an automated peptide synthesizer (model 430A; Applied Biosystems, Foster City, Calif.) and purified by gel filtration followed by reverse-phase high-performance liquid chromatography. The purity and authenticity of the synthetic peptides were ascertained by reverse-phase analytical high-performance liquid chromatography and amino acid analysis, respectively.

Recombinant ABRA constructs. The full-length ABRA gene (but lacking an N-terminal putative signal sequence) and its three major fragments were amplified from the parasite genomic DNA by PCR and cloned in Escherichia coli by standard molecular biology protocols. Briefly, the ABRA gene encoding amino acids (aa) 23 to 743, i.e., the full-length protein except for the putative signal sequence, was amplified by PCR using the P. falciparum genomic DNA as the template. A forward primer (5'-CGGGATCCCGATGAACATG-3') representing the N-terminal region and incorporating a BamHI restriction site and a reverse primer (5'-AACCCAAGCTTATTTTGATTCTTCAG-3') representing the C terminus and incorporating a HindIII restriction site were used for this amplification reaction. The amplified DNA fragment (~2.1 kb) was cloned into pGEMT vector (Promega Corporation, Madison, Wis.), and the nucleotide sequence of the cloned gene was partially determined by the dideoxynucleotide chain termination method.

View larger version (17K): [in this window] [in a new window]   FIG. 1.   Schematic representation of P. falciparum ABRA and its three fragments obtained by PCR amplification of the parasite genomic DNA and expressed as recombinant proteins as described in Materials and Methods. The small, numbered horizontal bars indicate the positions of the sequences chosen for synthetic peptides AB-1 to AB-8 as described in Materials and Methods. ss, signal sequence.

Human samples. Serum samples were collected from eight normal, healthy individuals who had no known past history of malaria and were negative for malaria by slide examination at the time of drawing blood (24). With the P. falciparum lysate used as the capture antigen, these sera (each diluted 1/200) yielded an average OD₄₉₀ of 0.35 with a standard deviation of 0.05 (0.35 ± 0.05) by ELISA. With the synthetic constructs in the same assay, these sera gave average OD₄₉₀s of 0.10 ± 0.02, 0.15 ± 0.02, 0.13 ± 0.03, 0.19 ± 0.05, 0.10 ± 0.02, 0.14 ± 0.04, 0.10 ± 0.01, and 0.2 ± 0.04 with AB-1 to AB-8, respectively.

Lymphocyte proliferation assay. Peripheral blood samples were collected from human volunteers who were living in an area of endemicity, had suffered from confirmed P. falciparum malaria infection several months prior to the study, and had been cured with the standard chloroquine regimen. They were malaria negative by slide examination at the time of sample collection and gave informed consent to participate in the study. Of 18 subjects approached, only 11 agreed to give a blood sample for this part of the study. The peripheral blood mononuclear cells (PBMC) were separated by centrifuging each of the blood samples on a density gradient (Histopaque-1077; Sigma Chemical Co., St. Louis, Mo.). The lymphocyte proliferation assay was set up in the 96-well tissue culture plates (catalog no. 3595; Costar Scientific Corp., Cambridge, Mass.) with PBMC cultured in RPM 1640 medium supplemented with 25 mM HEPES, 0.2% sodium bicarbonate, 50 µM -mercaptoethanol, 1.0 mM pyruvic acid, and 10% pooled human serum (AB/Rh⁺ group). Seven of eight ABRA constructs were used at three doses each, i.e., 10, 1.0, and 0.1 µg/well, with each well containing the cell suspension in a total volume of 200 µl. Each dose was tested in triplicate wells; concanavalin A, at a previously determined optimal concentration of 1 µg/well, was used as a nonspecific, polyclonal mitogen. On day 3 (concanavalin A) or day 5 (peptide), cultures were pulsed with [methyl-³H]thymidine (0.5 µCi/well; Amersham International plc, Little Chalfont, Buckinghamshire, England) for 6 h. The cells were then harvested onto glass fiber filters by using the PHD cell harvester (Cambridge Technology, Inc., Watertown, Mass.), and the ³H incorporation was determined by -emission liquid scintillation spectroscopy. The results were expressed as stimulation indexes (SIs); the SI represents the ratio of counts per minute obtained in the presence of the peptide to those obtained in the absence of the peptide.

Animals and their immunization. Animals used in this study were procured from the Small Animal Facility of the National Institute of Immunology, New Delhi, India. Animals were housed, fed, and used in the experiments following guidelines set forth in the National Institutes of Health manual Guide for the Care and Use of Laboratory Animals (30a).

ELISA. Sera were tested for the presence of antibodies in an ELISA, using carrier-free peptides or parasite lysate as the capture antigen. Procedures employed for the preparation of the parasite lysate (FID-3 isolate of P. falciparum) and for performing the assay were essentially as described previously (24, 38). Briefly, wells of flat-bottom Immulon-2 plates (Dynatech Laboratories Inc., Chantilly, Va.) were coated with the previously determined optimal concentration of capture antigen (carrier-free synthetic peptides or parasite lysate); the uncovered reactive sites were blocked with 5% milk powder solution in PBS. The antigen-coated wells were then sequentially incubated with appropriate dilutions of the first antibody followed by optimally diluted, enzyme-labeled secondary antibody (horseradish peroxidase-labeled anti-human or anti-rabbit IgG), with thorough washing of plates in between the incubations. The enzyme reaction was developed with o-phenylenediamine dihydrochloride as the chromogen and hydrogen peroxide as the substrate. After stopping the reaction with sulfuric acid, the OD₄₉₀ of the reaction product in the wells was recorded by using a Microplate Reader (Molecular Devices, Palo Alto, Calif.). In an ELISA using parasite lysate as the capture antigen, P. falciparum patient sera giving an OD₄₉₀ of 0.2 or more than the average mean OD₄₉₀ obtained with the normal sera (i.e., an OD of 0.55) were defined as positive sera. We realize that this arbitrary cutoff OD value of 0.55 is rather high, but it ensures stringent specificity against background noise in the sera from a region of endemicity like India. The same criterion of the difference between OD values (OD) being 0.2 was applied to determine the specific positivity of clinical sera against the individual ABRA peptides. In the end point titrations, the last dilution of a test serum yielding an OD₄₉₀ twice or more than twice that obtained with the respective preimmune serum (diluted 1/100) was taken as the end point titer.

Immunoblotting. The reactivity of the rabbit antipeptide sera with the parasite protein(s) was further ascertained by immunoblotting. The whole parasite lysate was fractionated on a 10% gel by SDS-PAGE under reducing conditions and transferred onto a nitrocellulose membrane following standard procedures. After the uncovered reactive sites of the nitrocellulose membrane were blocked by saturation with 5% nonfat milk powder solution in PBS overnight, the membrane was probed with various preimmune and immune rabbit sera by using a Mini Protean II Multi-Screen apparatus (Bio-Rad Laboratories). Total lysates of bacteria expressing the recombinant fragments of ABRA, viz., AB-N, AB-M, and AB-C, were also similarly fractionated and probed with the respective region-specific rabbit antibody, i.e., anti-AB-1, anti-AB-3, and anti-AB-8 antisera, respectively. The parasite proteins and the recombinant ABRA fragments in the bacterial lysates, separated by SDS-PAGE and transferred onto nitrocellulose paper, were incubated first with rabbit antipeptide sera and then with the horseradish peroxidase-labeled anti-rabbit IgG antibodies. The final enzyme reaction was developed with H₂O₂ as the substrate and 4-chloro-1-naphthol as the chromogen.

Immunofluorescence assay. Sera from rabbits immunized with ABRA constructs were tested for their reactivity with the authentic parasite protein(s) in the immunofluorescence assay as well (38). All rabbit sera were preadsorbed with fresh human erythrocytes so as to get rid of any heterophile antibody possibly present in these sera (38). Multispot antigen slides were made from a parasite-infected erythrocyte suspension prepared from an asynchronous culture of P. falciparum (strain FID-3). The antigen spots, air dried and fixed with the acetone-methanol (9:1, vol/vol) mixture, were sequentially incubated with serial dilutions of the test sera and the optimally diluted fluorescein isothiocyanate-labeled anti-rabbit IgG solution. The slides were finally mounted in the buffered glycerol containing p-phenylenediamine dihydrochloride (1 mg/ml) as the antifading reagent and examined under a fluorescence microscope (Wild Leitz GmbH, Wetzlar, Germany), alternately in visible and UV light, to see specific binding of the antibody to the parasite.

Merozoite invasion inhibition assay. The in vitro cultures of the FID-3 strain of P. falciparum were synchronized at the ring stage by two treatments with 5% sorbitol solution (25) and incubated further for about 30 h, so that at the time of setting up of the assay, nearly 90% of the parasites were >4N segmenters. The cultures were grown in RPMI 1640 medium supplemented with 10% human serum plus 5% normal (preimmune) or immune rabbit serum and incubated in a candle jar at 37°C for 20 h. Additional controls included culture wells with no rabbit serum, wells with rabbit anti-P-8_(MSP-1) serum (negative control [38]), and wells containing rabbit anti-18-mer_{(conserved TRAP motif)} total immune IgG at a concentration known to cause 50% inhibition of the merozoite invasion in this assay (38). In each experiment was included a parallel set of culture wells with appropriate rabbit sera, monitored every 2 to 3 h by microscopy for any possible toxic effect of these sera on the parasite or parasitized erythrocytes. At the end of the assay, smears were drawn from aliquots taken from each well, stained with Giemsa stain, and examined under a microscope by two researchers; only the ring-infected cells were counted as parasitized cells for calculating percent parasitemia (number of parasitized erythrocytes out of a total of 100 erythrocytes); at least 10,000 cells were counted to determine the level of parasitemia in each smear. In a subsequent experiment, various concentrations of the purified IgG fractions isolated from preimmune and immune rabbit sera were also incorporated in the test system.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Immunogenicity of synthetic ABRA peptides. We synthesized six nonrepetitive sequence peptides based on their high hydrophilicity score, potentially facilitating their accessibility to the immune system and, in addition, two more constructs modeling the repeat regions of ABRA. In an ELISA for measuring levels of circulating IgG antibodies, we found that each of these eight constructs bound appreciable amounts of antiparasite antibodies in the sera from a large proportion of human subjects recovering from natural P. falciparum infection (Fig. 2). Of 50 such sera that we tested, 33 (66%) yielded a positive ELISA reaction with one or more ABRA constructs; of these 33 positive samples, we found that only 3 reacted with all eight of the constructs, 6 reacted with seven constructs, 5 reacted with six constructs, 3 reacted with five constructs, 2 each reacted with four, three, and two constructs, and 10 reacted with only one of the eight constructs. Synthetic construct AB-3 was by far the most frequently recognized epitopic sequence, since as many as 50% of the clinical samples bound to this peptide, and as a group, the positive samples also yielded the highest average OD₄₉₀ value (0.56 ± 0.41) in the peptide ELISA, while AB-1, which was recognized by 36% of these sera in the same assay yielded the lowest average OD₄₉₀ value (0.31 ± 0.09). We found four samples to be negative for both parasite- and peptide-specific antibodies in this assay. Two other samples which were seronegative for the parasite antigen yielded a positive ELISA reaction with one (AB-3) and two peptides (AB-2 and AB-6), respectively.

View larger version (32K): [in this window] [in a new window]   FIG. 2.   Distribution profile of antibody levels obtained by ELISA using sera from 50 clinical malaria patients. Each serum was diluted 1/200 and tested in duplicate against each of the eight ABRA constructs. The OD490 value was obtained by subtracting the OD490 value given by a normal human serum pool from that given by the respective clinical serum. Each data point represents a mean of duplicate values. A OD490 value of 0.2 or more was defined as positive. The solid horizontal line in each column represents the mean OD490 value. Pf Ag, P. falciparum antigen.

View larger version (20K): [in this window] [in a new window]   FIG. 3.   Time course of IgG-type antibody responses generated in rabbits as monitored by an ELISA using homologous peptides as the capturing antigens. Rabbits immunized with the carrier-free ABRA peptides received boosters at week 4 (solid arrow); animals immunized with AB-2, AB-6, and AB-7 received boosters one more time at week 6 (broken arrow). Each serum was tested at a single dilution of 1/200 in duplicate wells.

Cross-reactivity of antipeptide sera with parasite protein(s) and recombinant ABRA constructs. Furthermore, rabbit antibodies generated against synthetic peptides also cross-reacted with the native parasite protein in three different assays. Results of an ELISA using parasite lysate as the capture antigen are presented in Fig. 4. Although the general pattern of the time course of antibody reactivity with the parasite lysate antigen was comparable to that obtained with the peptides used as capture antigens (Fig. 3), the level of reactivity with the parasite antigen was predictably low but well within the range expected of antibodies raised with short, carrier-free peptide immunogens (38). A careful analysis of data obtained with the preimmune (week 0) and test (week 8) sera revealed that in the hierarchy of responses obtained with the peptide ELISA (Fig. 3), AB-6, with an OD of test serum/OD of preimmune serum ratio (T/P ratio) of 44.5, ranked at the top, followed by AB-4 (T/P ratio, 39.2), AB-8, AB-1, AB-5, and AB-3; as mentioned earlier, the remaining two peptides, AB-2 and AB-7, induced barely detectable levels of antibody response, their T/P ratios being 3.25 and 2.75, respectively. With the parasite antigen (Fig. 4), on the other hand, the highest T/P ratio (15.8) was obtained with AB-3 followed by AB-1, AB-4, AB-6, and AB-5; the T/P ratios obtained for the remaining sera were indicative of marginal levels of antibody as detected in this assay. The most notable feature of the results obtained in this assay was the poor cross-reactivity of anti-AB-8 antibodies with the parasite antigen (T/P ratio, 1.9), in stark contrast to its very high peptide-specific reactivity (T/P ratio, 27.71). In the immunoblot assay, all six of the high-titer, antipeptide sera cross-reacted with the parasite protein(s); thus, the rabbit antisera, raised against AB-1, AB-3, AB-4, AB-5, AB-6, and AB-8, all reacted with a parasite protein of the expected size, i.e., ~101 kDa (Fig. 5, lanes 9, 11 to 14, and 16, respectively); furthermore, sera against AB-1, AB-3, AB-4, and AB-5 also recognized some additional protein bands of lower molecular weight in the parasite lysate; we observed a rather weak recognition of several parasite proteins, including one at ~101 kDa, with the anti-AB-7 serum, which had shown virtually no antibody titer in ELISAs with the peptide or the parasite lysate. As expected, rabbit anti-AB-2 serum recognized no parasite protein in the immunoblot assay; none of the rabbit preimmune sera reacted with any parasite protein (Fig. 5, lanes 1 to 8).

View larger version (18K): [in this window] [in a new window]   FIG. 4.   Cross-reactivity of rabbit antipeptide antibodies with the parasite protein(s) as monitored in an ELISA using parasite lysate prepared from the asexual blood stages of P. falciparum as the capturing antigen. Each serum was tested at a single dilution of 1/200 in duplicate wells. See the legend to Fig. 3 for additional details.

View larger version (100K): [in this window] [in a new window]   FIG. 5.   Cross-reactivity of rabbit antipeptide sera with the parasite protein(s) in an immunoblot assay. The parasite proteins extracted from the asexual blood stages of P. falciparum were separated on an SDS-10% PAGE gel, transferred onto a nitrocellulose membrane, and probed with different antipeptide sera by using a Bio-Rad Mini Protean II Multi-Screen apparatus. Lanes 1 to 8 were probed with the preimmune sera, and lanes 9 to 16 were probed with the respective test sera. Thus, the lanes are for sera as follows: 1 and 9, AB-1 sera; 2 and 10, AB-2 sera; 3 and 11, AB-3 sera; 4 and 12, AB-4 sera; 5 and 13, AB-5 sera; 6 and 14, AB-6 sera; 7 and 15, AB-7 sera; and 8 and 16, AB-8 sera. Apart from the specific protein at approximately 101 kDa (arrow), some other bands were also detected, which may be degradation products of ABRA.

View larger version (41K): [in this window] [in a new window]   FIG. 6.   Three recombinant constructs representing the N-terminal region (AB-N), the middle region (AB-M), and the C-terminal region (AB-C) of ABRA were expressed as recombinant fusion proteins in E. coli under conditions of IPTG induction (lanes I) and immunoblotted with rabbit anti-AB-1 (AB-1), anti-AB-3 (AB-3), and anti-AB-8 (AB-8) sera. ABRA constructs of the expected sizes (arrows) were specifically recognized by the respective sera; no such reactivity was noticed in the control, uninduced bacterial cultures (lanes U).

View larger version (68K): [in this window] [in a new window]   FIG. 7.   Immunofluorescence on air-dried, acetone-fixed monolayers of P. falciparum-infected erythrocytes (arrows) probed with rabbit anti-AB-1 serum. Shown are a brightly fluorescent trophozoite (A and B) and merozoites (C and D) within a mature schizont seen under UV illumination (A and C) and visible light (B and D); uninfected erythrocytes which did not react with antibody are also seen.

Merozoite invasion inhibitory activity of the antipeptide sera. Having established the immunogenicity of ABRA constructs in rabbits directly by immunization and in humans as inferred from our immunological observations of the clinical samples, we considered it logical to ascertain the biological function, if any, of the peptide-specific antibodies. In order to accomplish this, we tested the rabbit anti-peptide sera for activity against the parasite in a growth inhibition assay. We tested sera from only six rabbits that had shown positive antibody responses to six different ABRA constructs, as shown in Fig. 3. The results of the merozoite invasion inhibition assay are shown in Table 2. Surprisingly, the two sera, anti-AB-4 and anti-AB-3, which had yielded the high parasite-specific ELISA titers had only a marginal effect (anti-AB-3) or virtually no effect (anti-AB-4) on the merozoite invasion of the erythrocytes; two other sera (anti-AB-6 and anti-AB-8) caused only about 50% inhibition of invasion, while the remaining two sera (anti-AB-1 and anti-AB-5), which had ELISA titers of only 1/12,800 and 1/25,600, respectively, inhibited the parasite invasion by 94.4 and 77.6%, respectively (Table 2). That this inhibition was, indeed, mediated through antibodies and not any other serum component was further corroborated by the results we obtained with purified IgG fractions from the preimmune and immune rabbit sera in this assay. As apparent from the data presented in Table 3, total immune IgG purified from the sera of rabbits immunized with AB-1 or AB-5 exerted an inhibitory effect on the in vitro merozoite invasion of erythrocytes in a dose-dependent manner. In contrast, a similarly purified IgG fraction from another rabbit, immunized with AB-4 and showing high levels of peptide-specific antibodies (Table 2), had no adverse effect on the merozoite invasion of erythrocytes (Table 3).

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Of several secretory proteins, such as the S antigen, the serine-rich protein, the glycophorin binding protein, etc., which P. falciparum liberates during its asexual erythrocytic cycle, ABRA alone shows virtually no polymorphism (23). In fact, ABRA represents one of the most conserved antigenic proteins of P. falciparum; in a comparative study of three laboratory isolates of the parasite (namely, IMTM-22, FCR-3, and Camp), Chulay and colleagues (14) observed little difference in the size of this protein. Moreover, genes encoding this protein are nearly identical, with only four differences in the ABRA nucleotide sequences from the two isolates FCR-3 and Camp, which otherwise differ significantly in the sequence of the serine-rich, 126-kDa protein (14, 45) as well as in the molecular weights of several other proteins recognized by the growth-inhibitory antibodies (13, 43). At the same time, no protein analogous to ABRA from any other malarial parasite has been described so far, nor has any strong homology between ABRA and any other housekeeping proteins been found in the database search (45). These observations seem to underline the uniqueness of this protein and its potentially significant role in the biology of the parasite. In fact, there is immunological as well as biochemical evidence to suggest the possible involvement of ABRA in the processes of rupture of mature schizonts, release of merozoites, and invasion of fresh erythrocytes. In an elegant study, Nwagwu and colleagues (33) demonstrated chymotrypsin-like proteinase activity associated with the affinity-purified, parasite-derived ABRA protein. Such a proteinase(s) has been implicated in the release of mature merozoites from the parasitized erythrocytes and in the merozoite invasion of fresh erythrocytes (4, 19, 26). ABRA might well be one such proteinase. Thus, it represents an attractive target for chemotherapeutic and immunological intervention. However, there has not been any study on the mapping of epitopic sequences in this protein and their possible immunogenicity and protectivity against the parasite.

We have, therefore, concentrated mainly on the hydrophilic regions of the protein which, we reasoned, would be more accessible for generating an antibody response. In the present study, we have used eight synthetic peptides to delineate putatively protective epitopic sequences in ABRA. Our results with the human sera tend to support our contention, since we found that a majority of sera did contain appreciable levels of antibodies directed against these sequences. It is of interest that the synthetic construct AB-3 yielded the highest ELISA positivity (50%) as well as the highest OD values for serum samples tested. Furthermore, in a lymphocyte proliferation assay with PBMC of human subjects, we found that four of these constructs were recognized as T-helper-cell epitopes as well. Interestingly, AB-1 and AB-3 elicited strong proliferation responses in all 11 subjects tested (Table 1). Although we did not determine the HLA haplotype of these patients, it seems plausible that they represented more than one haplotype, and to that extent, AB-1 and AB-3 appear to be degenerate in their ability to bind to genetically restricted different HLA haplotypes and stimulate T-cell responses. If that indeed is the case, we would have two promiscuous T-helper-cell epitopes available from an asexual blood stage protein of the parasite. We are currently looking into this possibility in greater detail in both mice and humans.

It was encouraging to find that all eight constructs of ABRA we synthesized for the present study indeed represented epitopic targets of antibody responses generated during the natural infections with P. falciparum (Fig. 2). But we realize that the mere presence of circulating antibody in serum as measured in an ELISA provides little information about the possible protective potential of these sequences as immunogens. It seemed to us quite pertinent to assess the immunogenicity of these constructs in an experimental animal model and to ascertain the potential antiparasite activity of the experimentally raised, peptide-specific antibodies.

Interestingly, five of the six nonrepetitive sequences (AB-1 to AB-6) and one of the two repetitive sequences (AB-7 and AB-8), without the use of any carrier protein, induced boostable, IgG-type antibody responses in rabbits (Fig. 3); only AB-2 and AB-7 failed to generate any boostable antibody response. The peptide-specific antibodies cross-reacted with the parasite protein(s) in an ELISA (Fig. 4) and an IFA (Fig. 7). In an immunoblot also, rabbit anti-AB-5 serum recognized a protein band at about 101 kDa and two more bands with lower molecular masses, which could be the autoproteolytic products of ABRA or cross-reactive epitopes present in other proteins; a similar pattern of bands was seen with the rabbit sera against other immunogenic ABRA constructs (Fig. 5). To an extent defined by these results, the synthetic constructs seem to mimic the portions of native protein faithfully enough to induce antibodies which recognized the authentic parasite protein in three different immunoassays. More significantly, we found that antibodies to four synthetic constructs also exerted antiparasite activity, causing 40 to 90% inhibition of the in vitro merozoite invasion of erythrocytes with as little as 5% antiserum incorporated in the culture system (Table 2). Although it remains far from being fully established how antibodies exert their influence upon the intracellular parasite, several workers have used this assay as a suitable in vitro correlate of potentially protective antibody (13, 34, 39, 43, 44). In the case of ABRA, monoclonal antibodies have been demonstrated to agglutinate freshly released merozoites, thus preventing reinvasion of erythrocytes (14). Rabbit antibodies which we raised against synthetic peptides modeling parts of ABRA may also function in a similar way. However, it is also likely that antibodies to ABRA or its parts might be working by blocking its protease activity, which, along with other proteases, is possibly involved in the secondary processing of MSP-1, a process shown to be critical for producing invasive merozoites (19).

Notwithstanding the apparent immunodominance of AB-3 in humans, as evident from the serological results (Fig. 2), rabbit anti-AB-3 serum, though yielding a reasonably high ELISA titer, was found to be poor in its direct antiparasite activity; it caused only ~30% inhibition of the parasite in vitro; similarly, AB-4-specific rabbit antibodies had virtually no direct inhibitory effect on parasite growth. However, these negative results do not rule out the possibility of their growth-inhibitory potential in cooperation with monocytes. In fact, a number of studies investigating the importance and relevance of this mechanism of parasite clearance have provided convincing evidence for such a phenomenon occurring in malaria (8, 10, 17, 18).

Synthetic peptides such as those evaluated in the present study provide a promising alternative to conventional vaccines or those being produced by recombinant DNA technology. A number of studies have demonstrated the feasibility of using such constructs as potentially protective immunogens (1, 12). In the first-ever human clinical trials of a synthetic peptide vaccine, a polymeric, multicomponent malaria vaccine, SPf66, has been found to be safe and immunogenic in both adults and children residing in widely distant geographical regions with different transmission rates (1, 31, 32), although its efficacy has become a subject of controversy and contradiction (31, 32). At the same time, results of these trials have underlined the need to continuously search for other molecules and for better ways to generate a protective immune response. Our present study, which delineates a couple of putatively protective epitopes in ABRA, represents only the first essential step in our ongoing efforts to evaluate this protein as a possible vaccine candidate. An important question would relate to the immunogenicity of these peptide constructs in the context of different major histocompatibility complex haplotypes on the one hand and various adjuvant formulations, other than Freund's, on the other hand. Studies have indicated that adjuvants could play a critical role in determining the magnitude and specificity of immune responses to a particular epitope(s) (21, 41a). We already have preliminary data to indicate that AB-1 and AB-5 are recognized in the context of several different major histocompatibility complex haplotypes of mice, although the titers of antibodies were not very high (unpublished observations). Using different adjuvant formulations, we are now in the process of establishing the repertoire of humoral responses to ABRA in terms of IgG subclasses of antibodies induced by various ABRA constructs, the pattern of their affinity maturation for the antigen, and their qualitative or functional features, such as their growth-inhibitory or merozoite invasion-inhibitory potential.