INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Efforts to develop an effective malaria vaccine are concentrated on Plasmodium falciparum, the major cause of malaria morbidity and mortality worldwide (45). Several promising vaccine targets have been identified in each of the major developmental stages of this protozoan parasite (24). These P. falciparum vaccine candidate antigens have been largely identified based on their functional importance and/or as targets of human immune responses. Additional in vivo studies involving the rodent malaria parasites Plasmodium yoelii and Plasmodium chabaudi have facilitated efforts to assess the vaccine potential of several antigens of interest (3, 5, 9, 11, 14-16, 26, 28, 30, 34, 42). Evaluation of the most promising candidate antigens has progressed to human clinical trials, and the initial results are encouraging (18). Even so, it is clear that further efforts to improve overall malaria vaccine efficacy are needed. This includes the characterization of additional parasite antigens and epitopes that induce protective responses upon immunization.

Blood-stage malaria vaccine studies have emphasized the mapping of B-cell epitopes of plasmodial antigens that are targets of protective antibodies. Many immunodominant, linear B-cell epitopes have been identified in association with domains of proteins that contain repeated blocks of amino acids. Many of these determinants are polymorphic and have been suggested to be part of an immune evasion strategy (31, 37). On the other hand, P. falciparum repeat domains, such as those found in the circumsporozoite protein, the ring-infected erythrocyte surface antigen, the serine repeat antigen, and merozoite surface protein 2 (MSP-2) may in fact be targets of protective antibodies (27). A second, very different group of potentially protective B-cell epitopes has been mapped to domains of malaria antigens that are conformationally constrained by multiple disulfide bonds. In P. falciparum, these include the cytoadherence domains of erythrocyte membrane protein 1 (4), the ectodomain of apical membrane antigen 1 (3, 14, 23, 44), the binding domain of the 175-kDa erythrocyte binding antigen (1, 32), and as a group, the epidermal growth factor (EGF)-like domains of MSP-1, MSP-4, MSP-5, and the sexual-stage antigens Pfs25 and Pfs28 (5, 11, 15, 17, 25, 29, 30, 35, 36). Generally these epitopes are less variable between P. falciparum strains, exhibit conservation across species, and appear to be functionally important. As such, they are of considerable interest for malaria vaccine development.

Previously, we reported that immunization with a particulate fraction of blood-stage antigens (pAg) can protect mice against P. yoelii malaria (10). We attempted to utilize the responses of these immunized and protected animals to identify novel, protective blood-stage antigens. One such antigen, pypAg-1, is a novel membrane protein of P. yoelii-infected erythrocytes that was shown to express at least two distinct protective epitopes (8). Using this immunization-based approach to antigen discovery, we now report the identification and characterization of pypAg-2, a second novel antigen of P. yoelii. We present data indicating that pypAg-2 is a glycosylphosphatidylinositol (GPI)-anchored membrane protein of trophozoites and merozoites that lacks repetitive amino acid sequences but contains two C-terminal EGF-like domains. Most importantly, we report that immunization with recombinant pypAg-2 protects mice against lethal P. yoelii blood-stage malaria.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Experimental infections. Male BALB/cByJ mice, 5 to 6 weeks of age, were purchased from The Jackson Laboratory (Bar Harbor, Maine) and housed in the Association for Assessment and Accreditation of Laboratory Animal Care-approved Animal Care Facility of Meharry Medical College. Routine screenings were conducted throughout these studies to ensure that mice remained free of infection with common viral and bacterial pathogens (Assessment Plus Profile; Charles River Laboratories, Wilmington, Mass.). The lethal 17XL strain of P. yoelii was originally obtained from William P. Weidanz (University of Wisconsin, Madison, Wis.) and was maintained as cryopreserved stabilates. Blood-stage infections were initiated by intraperitoneal injection of parasitized erythrocytes obtained from donor mice. The resulting parasitemias were monitored by enumerating parasitized erythrocytes in tail blood thin smears stained with Giemsa.

Immune sera. Sera from five CByB6F1/J mice were obtained 1 week following secondary immunization with pAg but prior to infection. The generation and characterization of these sera and nonimmune adjuvant control sera have been previously reported (10).

Cloning and sequence analysis. A P. yoelii 17XL cDNA library was constructed in the lambda Uni-ZAP XR expression vector (Stratagene, La Jolla, Calif.) using poly(A)⁺ RNA from a mixture of ring, trophozoite, and schizont blood-stage parasites. The library was screened with a pool of sera from mice protected against P. yoelii malaria by pAg immunization. The library construction, screening, and excision of pBSK() phagemid sequences have been described previously (8). Of 26 seroreactive recombinant clones identified, 2 were shown to contain pypag-2 sequences by cross-hybridization studies and partial-sequence analysis.

Production of recombinant pypAg-2. Recombinant pypAg-2 protein (rpAg-2) was produced in Escherichia coli using the pET plasmid vectors and T7 RNA polymerase expression system (43). The complete coding sequence minus the N-terminal signal sequence and hydrophobic C-terminal anchor sequence was PCR amplified from the pypag-2 cDNA clone by using oligonucleotide primers 5'-TTTCATATGGGGAATTGTAATGAAAATGGAAACG-3' (based on nucleotides 307 to 340) and 5'-CATTCATTCCTCGAGAAATGGCTCATGAACAATA-3' (based on nucleotides 1402 to 1435) as 5' and 3' primers, respectively. To facilitate subcloning, NdeI and XhoI restriction sites were incorporated into the 5' and 3' primers, respectively. The amplified fragment was gel purified, digested with NdeI and XhoI, and ligated into NdeI-XhoI-digested pET-15b (Novagen, Madison, Wis.). E. coli strain BL21(DE3)pLysS was used for expression. rpAg-2 contained 20 plasmid-encoded amino acids fused to its N terminus, which included six histidine residues.

Immunoblot analysis. Total P. yoelii 17XL blood-stage antigen and the corresponding pAg fraction were prepared as described previously (10) and solubilized in SDS-PAGE sample buffer with 2.5% SDS and 5% 2-mercaptoethanol. Total blood-stage antigen (10 µg/lane), the pAg fraction (10 µg/lane), and purified rpAg-2 (1 µg/lane) were separated by SDS-PAGE on 10% polyacrylamide gels under reducing conditions by the method of Laemmli (33) and analyzed by immunoblotting (8). Control and pAg immunization mouse sera were assayed at a dilution of 1:100. Preimmune and immune rabbit sera were diluted 1:500. Bound antibody was detected with ¹²⁵I-protein A (ICN Biomedicals, Irvine, Calif.).

Immunoprecipitation studies. Blood from P. yoelii 17XL-infected mice (25 to 30% parasitemia) was collected and washed three times in medium. Erythrocytes were resuspended in RPMI 1640 supplemented with 2 mM L-glutamine, 2.5 mg of glucose per ml, 20 mM morpholinepropanesulfonic acid (MOPS) (pH 7.4), 0.1 mM hypoxanthine, and 50 µg of gentamicin per ml (0.2 ml [packed-cell volume] per ml of medium). [³H]palmitic acid (30 to 60 Ci/mmol; NEN Life Science Products Inc.) was added to 100 µCi/ml, and the parasitized cells were incubated at 37°C in 7% CO₂ for 4 h. Palmitic acid is incorporated into the GPI anchors of several protozoan surface proteins, including those of P. falciparum MSP-1 and MSP-2 (20, 21). Following incubation, cells were collected by centrifugation at 1,800 × g for 10 min and solubilized on ice in 10 volumes of buffer containing 20 mM Tris-HCl (pH 7.8), 50 mM NaCl, 5 mM EDTA, 1% Triton X-100, 0.5% deoxycholate, 1 mM phenylmethylsulfonyl fluoride, and 1 mM iodoacetic acid. Insoluble material was removed by centrifugation at 100,000 × g for 1 h.

Indirect-immunofluorescence assay. The indirect-immunofluorescence assay to localize the expression of pypAg-2 was conducted as previously reported (8). Thin smears of P. yoelii 17XL-infected erythrocytes were air dried and fixed in acetone-methanol. Preimmune and immune rabbit sera were assayed at a dilution of 1:200 in PBS. The secondary antibody utilized was fluorescein isothiocyanate (FITC)-conjugated goat anti-rabbit immunoglobulin G (IgG) (Sigma Chemical Co., St. Louis, Mo.) diluted 1:160 in PBS. Stained cells were washed, mounted with Fluoromount-G (Southern Biotechnology Associates, Inc., Birmingham, Ala.), and visualized by fluorescence microscopy. For assays of merozoites, an enriched preparation of P. yoelii 17XL schizonts was prepared by centrifugation of parasitized blood through a 15 to 30% metrizamide step gradient (19).

Immunizations. Groups of five BALB/cByJ mice were immunized subcutaneously with 25 µg of rpAg-2 in adjuvant, adjuvant alone, or PBS alone. Following the primary immunization, the animals were boosted twice at 3-week intervals with the same dose of antigen. In the first experiment, RAS (50 µg of monophosphoryl lipid A plus 50 µg of synthetic trehalose dicorynomycolate [Ribi ImmunoChem Research Inc., Hamilton, Mont.]) was utilized as the adjuvant for all three immunizations. In the second experiment, rpAg-2 was emulsified in CFA (1:1 emulsion [Life Technologies]) for the primary immunization. For the two remaining boosts, CFA was replaced with recombinant mouse interleukin-12 (IL-12) (1 µg/dose) (R & D Systems Inc., Minneapolis, Minn.). Ten days following the third immunization, mice were infected intraperitoneally with 10⁵ P. yoelii 17XL parasitized erythrocytes and blood parasitemias were monitored. The statistical significance of differences in mean parasitemia between immunized and control groups was calculated by analysis of variance using the StatMost statistical analysis software package (DataMost Corp., Salt Lake City, Utah).

Nucleotide sequence accession number. The sequence reported in this paper has been deposited in the GenBank database under accession number AY005132.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

pypag-2 encodes a novel P. yoelii blood-stage antigen. The pypag-2 gene of P. yoelii was identified during the screening of a cDNA expression library with sera from mice immunized with a protective fraction of P. yoelii blood-stage antigens (8). The complete coding and noncoding strands of pypag-2 were determined from the largest of the pypag-2 cDNA clones isolated (Fig. 1A and B). This cDNA contains 2,103 bp with a single open reading frame from nucleotides 253 through 1482. The encoded protein is 409 amino acids in length, with a predicted molecular mass of 46.8 kDa and isoelectric point of 4.81. An N-terminal signal sequence (amino acids 1 to 22) and a C-terminal hydrophobic anchor sequence (amino acids 389 to 409) are predicted. Unlike many characterized plasmodial antigens, blocks of repeated amino acids are lacking in the predicted pypAg-2 protein. Of particular interest, pypAg-2 contains 18 cysteine residues, with 14 of these found within the C-terminal 160 amino acids. Data bank searches revealed significant similarities between this C-terminal cysteine-rich domain of pypAg-2 and the C terminus of P. falciparum MSP-1 (Fig. 1C). The two sequences have approximately 33% identity over 115 amino acids with alignment of 12 cysteine residues. These 12 cysteine residues mediate the folding of the C terminus of MSP-1 into two EGF-like domains (39). Outside of this domain, pypAg-2 has little similarity to MSP-1 or other known protein sequences.

View larger version (69K): [in this window] [in a new window]   FIG. 1.   Sequence analysis of pypag-2. (A) Diagramatic representation of the pypag-2 cDNA sequence. (B) DNA and deduced amino acid sequences of the 2,103-bp insert of pypag-2. Cysteine residues are indicated in bold. The predicted N-terminal signal sequence and C-terminal hydrophobic anchor sequence are doubly underlined. (C) Amino acid alignment of the C-terminal cysteine-rich domains of pypAg-2 and P. falciparum MSP-1 (accession number X02919). Identical residues are indicated by letter, and conservative and neutral substitutions are indicated by colons. The 12 cysteine residues of P. falciparum MSP-1 that form the two EGF-like domains are indicated in bold and highlighted by asterisks.

Identification of native pypAg-2. A 45-kDa pypAg-2 recombinant protein was expressed in E. coli, purified by nickel-chelate affinity chromatography, and refolded in the presence of glutathione to facilitate disulfide bond formation. rpAg-2 contained the complete coding sequence of pypAg-2 minus the hydrophobic N-terminal signal and C-terminal anchor sequences. Figure 2 shows a Coomassie blue-stained polyacrylamide gel of purified rpAg-2. Immunoblot analysis confirmed the strong reactivity of rpAg-2 with the pAg immunization serum pool that was initially used to screen the P. yoelii expression library (Fig. 3A, lane 2). Control sera from mice immunized with adjuvant alone were nonreactive with purified rpAg-2 (lane 1).

View larger version (57K): [in this window] [in a new window]   FIG. 2.   Expression and purification of rpAg-2. A Coomassie blue-stained SDS-polyacrylamide gel containing E. coli lysates of uninduced (lane 1) and induced (lane 2) cells expressing rpAg-2 and purified rpAg-2 alone (2.5 µg, lane 3) is shown. Molecular mass markers are indicated.

View larger version (44K): [in this window] [in a new window]   FIG. 3.   Immunoblot analysis of recombinant and native pypAg-2. (A) Immunoblot of purified rpAg-2 probed with sera from mice immunized with the protective pAg preparation of P. yoelii blood-stage antigens (lanes 2) or with sera from adjuvant control mice (lane 1). (B) Immunoblot of P. yoelii total blood-stage antigen (lanes 1 and 3) or protective pAg fraction (lanes 2 and 4) probed with preimmune rabbit serum (lanes 1 and 2) or rabbit antiserum raised against rpAg-2 (lanes 3 and 4). Molecular mass markers are indicated.

pypAg-2 is a GPI-anchored membrane protein. The prediction of signal and anchor sequences in pypAg-2, combined with the similarity to MSP-1, suggested that pypAg-2 might also be a GPI-anchored membrane protein. To test for membrane association, P. yoelii blood-stage antigens were solubilized in Triton X-114 and partitioned into aqueous and detergent phases. Phase-separated proteins were then assayed by immunoblotting using the polyclonal rabbit serum raised against rpAg-2. As shown in Fig. 4A, pypAg-2 fractionated predominantly in the Triton X-114 detergent phase (lane 4), with only a small amount detected in the aqueous phase (lane 3). These data support the conclusion that pypAg-2 is in fact membrane associated. To determine if the membrane association of pypAg-2 was mediated by a GPI anchor, the rpAg-2 rabbit antiserum was used to immunoprecipitate P. yoelii 17XL blood-stage proteins metabolically labeled with [³H]palmitic acid. As shown in Fig. 4B, pypAg-2 labeled with [³H]palmitic acid was immunoprecipitated with the rpAg-2 antiserum (lane 2) but not with preimmune rabbit serum (lane 1). Combined, these data indicate that pypAg-2 is a GPI-anchored membrane protein.

View larger version (34K): [in this window] [in a new window]   FIG. 4.   Membrane association of pypAg-2. (A) Immunoblot of Triton X-114 aqueous-phase (lanes 1 and 3) and detergent-phase (lanes 2 and 4) proteins of P. yoelii blood-stage parasites probed with preimmune rabbit serum (lanes 1 and 2) or rabbit antiserum raised against rpAg-2 (lanes 3 and 4). (B) Immunoprecipitation of P. yoelii 17XL blood-stage antigens metabolically labeled with [3H]palmitic acid using preimmune rabbit serum (lanes 1) or rabbit antiserum raised against rpAg-2 (lanes 2). Molecular mass markers are indicated.

pypAg-2 is expressed on trophozoite and merozoite membranes. The membrane association of pypAg-2 was further investigated by indirect immunofluorescence. Thin films of P. yoelii-parasitized erythrocytes were prepared, air dried, fixed, and reacted with the polyclonal rabbit antiserum raised against rpAg-2. Bound antibody was detected with a FITC-conjugated goat anti-rabbit IgG antibody. The expression of pypAg-2 was localized both to the trophozoite membrane (Fig. 5B) and to the membrane of merozoites prior to schizont rupture (Fig. 5C). No fluorescence of P. yoelii-parasitized erythrocytes incubated with preimmune rabbit serum was apparent (Fig. 5A).

View larger version (93K): [in this window] [in a new window]   FIG. 5.   Immunofluorescence staining of fixed smears of erythrocytes infected with P. yoelii. Thin blood smears were incubated with preimmune rabbit serum (A) or rabbit antiserum raised against rpAg-2 (B and C) followed by a FITC-conjugated secondary antibody. The same fields of cells were examined by phase-contrast (left panels) and fluorescence (right panels) microscopy.

Immunization with rpAg-2 protects against lethal P. yoelii malaria. To determine if pypAg-2 is a target of protective immune responses, mice were immunized with rpAg-2 in adjuvant and challenged with lethal P. yoelii 17XL. In the first experiment, mice were immunized with rpAg-2 formulated with RAS. As shown in Fig. 6A, rpAg-2 immunization protected mice against a lethal challenge infection. Uncontrolled parasitemia developed in animals immunized with RAS or PBS alone. As the parasitemia progressed to a high level in control mice, they became moribund, and all were euthanized by day 10 postinfection. In contrast, rpAg-2-immunized mice controlled their infection during this 10-day period, with blood parasitemia not exceeding 10%. Parasitemia peaked in rpAg-2 mice between days 10 and 14, ranging from 4.7 to 34.5% (mean, 19.19% ± 13.08%). All rpAg-2 immunized mice resolved in their infection by day 22 of infection.

View larger version (13K): [in this window] [in a new window]   FIG. 6.   Protection against lethal P. yoelii malaria induced by immunization with rpAg-2. (A) Groups of five BALB/cByJ mice were immunized with rpAg-2 () with RAS as adjuvant. Five control animals were immunized with RAS alone () or PBS (). (B) Groups of five BALB/cByJ mice were immunized with rpAg-2 () with CFA and IL-12 as the adjuvant. Five control animals were immunized with CFA and IL-12 alone (). Animals were boosted twice at 3-week intervals and challenged with 105 P. yoelii 17XL-parasitized erythrocytes. The resulting parasitemias were monitored by enumerating parasitized RBCs in tail blood thin smears stained with Giemsa and are expressed as the mean and standard deviation. Differences in mean parasitemia between immunized and control mice were statistically significant on days 4 to 10 (P < 0.05 by analysis of variance).

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

In our previous studies, immunization of mice with a particulate fraction of blood-stage antigens protected against P. yoelii malaria (10). Although reduced in complexity, the immunizing antigen preparation contained a mixture of blood-stage proteins. Consequently, we expected that multiple targets of the protective response existed. Further studies showed that the pAg-induced protection was B-cell dependent and associated with the production of antibodies that recognized six to eight P. yoelii antigens. As such, the number of antigens contributing to the protective response appeared to be limited. Using the P. yoelii pAg immunization sera, we have attempted to define those protective targets.

We previously characterized pypAg-1 as one P. yoelii antigen that contributed to the pAg-induced protection. This antigen is expressed by P. yoelii trophozoites, exported to the erythrocyte membrane, and eventually secreted from infected erythrocytes (7, 8). In addition to a putative transmembrane domain, the N-terminal portion of pypAg-1 contains a domain unusually rich in tryptophan residues. This tryptophan-rich domain is conserved in pypAg-3, another P. yoelii antigen with a similar pattern of expression that was also identified with the pAg immunization serum (7). The C terminus of pypAg-1 contains a 155-amino-acid, highly charged repeat domain. Immunization with either the N-terminal or C-terminal domains of pypAg-1 induced a four- to sevenfold reduction in P. yoelii 17X blood-stage parasitemia (8). While significant, this was several-fold less than the maximal level of protection induced by crude pAg immunization, suggesting that additional components were involved. One of these components appears to be pypAg-2.

pypAg-2 is distinct from both pypAg-1 and pypAg-3 relative to their DNA and deduced amino acid sequences and patterns of expression. PypAg-2 is a GPI-anchored protein, expressed in association with the trophozoite and merozoite membranes. In contrast to many malaria vaccine candidate antigens, pypAg-2 lacks a characteristic immunodominant repeat domain. The most striking feature of pypAg-2 is the presence of two EGF-like domains at its C terminus. These EGF-like domains have now been identified in at least five P. falciparum vaccine candidate antigens including MSP-1 and generally appear to be targets of protective antibodies (17, 25, 29, 35, 36).

The C-terminal EGF-like domains of P. falciparum MSP-1 are well characterized in terms of structure, function, sequence polymorphism, and vaccine potential (25, 39). The combined data make P. falciparum MSP-1 the leading blood-stage vaccine candidate antigen at present. Antibody to the EGF-like domains of MSP-1 blocks merozoite invasion of erythrocytes in vitro (6, 13) and confers immunity to challenge infection in animal models (15). The similarity between the EGF-like domains of pypAg-2 and MSP-1, combined with the expression of pypAg-2 on merozoites, raise the possibility that pypAg-2 is also involved in parasite invasion of erythrocytes. MSP-1 and pypAg-2 have little sequence similarity beyond the C-terminal EGF-like domains. Likewise, the sequence of pypAg-2 differs significantly from those of P. falciparum MSP-4 (35) and MSP-5 (36) as well as of the single fused protein MSP-4/5 of P. chabaudi (5), P. yoelii, and P. berghei (30); additional merozoite surface antigens that each contain a single EGF-like domain. Like MSP-4, however, pypAg-2 appears to be expressed on the trophozoite membrane early in the erythrocyte cycle, prior to its association with the membrane of merozoites (5, 35).

The immunization and challenge experiments provided the key data to establish the protective potential of pypAg-2. Mice immunized with rpAg-2 by using RAS or a combination of CFA and IL-12 as adjuvant resisted an otherwise lethal P. yoelii 17XL challenge. RAS was chosen as a good adjuvant for the overall induction of antibody responses. The CFA-plus-IL-12 combination was selected to promote the production of Th1-associated IgG isotypes (IgG2a and IgG3). Both adjuvants were equally efficient in promoting protective pypAg-2 responses. While a significant increase in the prepatent period was not observed, rpAg-2-immunized mice largely controlled their P. yoelii 17XL infection for the first 8 to 10 days. Parasitemia increased shortly thereafter but was ultimately suppressed. This breakthrough in parasitemia may have resulted from the consumption of immunization-induced antibodies to pypAg-2. It is also worth noting, however, that the late rise in parasitemia occurred as the infection shifted from mature erythrocytes to reticulocytes. This suggests that the level of protection induced by pypAg-2 immunization may in part depend on the type of host erythrocyte infected. This may be particularly relevant to human malaria, given that P. falciparum also infects mature erythrocytes as well as reticulocytes. In the end, we believe that immunization with a combination of pypAg-1, pypAg-2, and pypAg-3 will significantly increase overall protective efficacy. Since these three antigens were identified based on the same protective immunization strategy, it should be possible to optimally formulate such a combination.

Finally, one hallmark of the EGF-like domains of malarial proteins characterized thus far is their conservation across species of Plasmodium, including the rodent malarial parasites. For example, the EGF-like domains of P. falciparum MSP-1 and MSP-4 are conserved in the homologous antigens of P. yoelii, P. chabaudi, and P. berghei (5, 9, 30, 38, 46). By cross-hybridization with pypAg-2 probes, we have isolated genomic clones encoding related P. falciparum antigens. Based on the conservation of the two C-terminal EGF-like domains, we believe that we have cloned the P. falciparum homologue of pypAg-2 (C. C. Belk, P. D. Dunn, and J. M. Burns, Jr., Abstr. 48th Annu. Meet. Am. Soc. Trop. Med. Hyg., abstr. 128, 1999). Furthermore, two partial sequences of this antigen gene have been deposited in the P. falciparum genome sequencing database at the Sanger Center (chromosomes 5 and 6). Therefore, we anticipate that our studies of pypAg-2 in the P. yoelii model will contribute to ongoing efforts to develop a P. falciparum blood-stage vaccine.