INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

The merozoite surface protein 1 (MSP1) of Plasmodium spp. is synthesized during schizogony as a large 190- to 250-kDa polypeptide which is later processed into four of the major merozoite surface proteins. Upon release of the merozoite, the N-terminal parts of the proteins are subsequently cleaved off so that only a short 19-kDa C-terminal fragment (MSP1₁₉) remains anchored on the mature merozoite's surface at the time of red blood cell invasion (3). While extensive sequence heterogeneity has been described for most parts of the protein, the C-terminal portion shows a highly analogous structure, including two epidermal growth factor-like domains in all Plasmodium species (13).

Therefore, MSP1₁₉ is considered an important candidate in developing a subunit vaccine against the asexual blood stages of malaria. Thus, in rodent models, immunization with affinity-purified and recombinant MSP1 from Plasmodium yoelii yielded protection, as did the passive transfer of monoclonal antibodies and immune serum against PyMSP-1 (9). In monkeys, immunization trials with proteic MSP1₁₉ also elicited protection and determined that the conformation of the recombinant protein was crucial for protection. Indeed, while the baculovirus-produced MSP1₁₉ protected monkeys against challenge with Plasmodium falciparum (8), the Escherichia coli-derived MSP1₁₉ fused to glutathione S-transferase (GST) did not (6). More recent studies in Macaca sinica with Plasmodium cynomolgi, also revealed that immunization of monkeys with baculovirus-derived recombinant P. cynomolgi-MSP1₁₉ or MSP1₄₂ led to a high degree of protection against infection, which was mostly antibody dependent (20).

An alternative to the use of recombinant proteins and toxic adjuvants is immunization with genetic vaccines (27). For MSP1, it was recently shown that mice immunized with MSP1-coding plasmids showed decreased peak parasitemias after challenge with a nonlethal P. yoelii strain (2). Moreover, genetic vaccines containing CpG motifs (15, 22) produce in mice a Th1-like response considered essential for the control of a primary blood stage infection (23). Recently, genetic immunizations, including multiple genes expressed during different stages of the human malarias caused by P. falciparum and P. vivax, were tested (reviewed in reference 12).

In this study, we tested the protective potential of a plasmid bearing the gene coding for a fusion protein of MSP1₁₉ of P. chabaudi and the small hepatitis B virus surface protein (HBs). A similar construct for the P. vivax MSP1₁₉ had been previously shown to be highly antigenic (10) and to form hybrid viral particles (28). Moreover, since an effective immune response against blood stage infection of mice with P. chabaudi has a Th1 profile (24), we also coinjected plasmid vectors coding for Th1-associated cytokines, namely, murine interleukin-2 (IL-2), gamma interferon (IFN-), and granulocyte-macrophage colony-stimulating factor (GM-CSF). Additionally, we tested a plasmid coding for a selectin-MSP1₁₉-HBs hybrid which was supposed to target the MSP1₁₉-HBs hybrid to lymph nodes (4) and therefore enhance the immune reaction.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Construction of plasmid vectors used in genetic immunization trials. A fragment coding for the small HBs was excised from the plasmid pSV33M* by restriction with BamHI and EcoRV and inserted into pVXORF1, resulting in pVXORF-S (10). The MSP1₁₉ coding fragment was amplified from genomic P. chabaudi chabaudi PC1 DNA by standard PCR procedures. Complete sequencing of the fragment revealed that the sequence was identical to the sequence of P. chabaudi chabaudi strain CB (GenBank accession no. L22984). The EcoRI/BglII-restricted fragment was cloned into pVXORF-S via BamHI and EcoRI sites. A plasmid solely coding for MSP1₁₉ was cloned by insertion of an EcoRI-restricted and 5'-phosphorylated PCR fragment using the oligonucleotides described in Table 1 in pVXORF1 via EcoRI and EcoRV sites. In parallel, the MSP1₁₉ coding fragment was also inserted in the vector pGEX3X (Pharmacia, Uppsala, Sweden). Genes coding for IFN-, IL-2, and the ectodomain of L-selectin were cloned by standard reverse transcription-PCR from total RNA of 10⁸ concanavalin A-stimulated BALB/c mouse splenocytes; the primers are listed in Table 1. IL-2 and IFN- fragments were subcloned via EcoRI/BamHI digestion into pVXVR, a pVXORF1 variant lacking its tissue plasminogen activator secretion signal. The L-selectin fragment was subcloned via BamHI/BglII into pVXORF1- and pPcMSP1₁₉-HBs, resulting in pSel and pSel-PcMSP1₁₉-HBs, respectively. A functional plasmid coding for GM-CSF (29), termed pGMCSF, was kindly provided by M. M. Rodrigues (Escola Paulista de Medicina, São Paulo, Brazil). All plasmid constructs were checked for the correctness of their inserts by manual dideoxy sequencing (21). Plasmids were purified using the Qiagen Mega-Prep columns (Qiagen, Hilden, Germany), according to the manufacturer's instructions.

Expression of recombinant constructs in COS7 cells. Plasmids were transfected in COS7 cells by the standard DEAE-dextran method, and metabolically labeled proteins were immunoprecipitated as described elsewhere (5). Briefly, at 48 h posttransfection, 50% confluent COS7 cells in 3-cm-diameter dishes were pulsed with 80 µCi of Tran³⁵S-label (ICN, Costa Mesa, Calif.) for 15 min. After a 24-h chase period, culture supernatants were removed and immunoprecipitated for 2 h at room temperature with protein A-Sepharose-bound polyclonal goat anti-HBs (Dako, Hamburg, Germany) or immunoglobulins derived from the acute sera of P. chabaudi-infected trial mice. Proteins were resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis on a standard 12% polyacrylamide gel, fixed, stained with fluorographic reagent (Amplify; Amersham), dried, and autoradiographed.

Production of recombinant P. chabaudi MSP1₁₉-GST and total schizont proteins. Expression and purification of rPcMSP1₁₉-GST or rGST from pGEX3X-PcMSP1₁₉ or pGEX3X-transformed E. coli DH5 was done as described previously (17). Before use, the recombinant proteins were dialyzed extensively against phosphate-buffered saline (PBS). P. chabaudi proteins were obtained by lysis of IRBC containing schizont stage parasites with PBS-0.1% saponin for 5 min at room temperature. After three washing steps (ice-cold PBS-0.05% saponin), parasites were lysed in PBS-0.5% Triton X-100 for 10 min on ice in the presence of 1 mM phenylmethylsulfonyl fluoride. After pelleting of the insoluble debris, the supernatant was stored in liquid N₂ until use.

ELISA, Western blot, and isotyping. Proteins were analyzed by enzyme-linked immunosorbent assay (ELISA) and Western blot assays as described elsewhere (17). Isotyping of immunoglobulin G (IgG) was performed with sera at dilutions which resulted in an optical density at 450 nm (OD₄₅₀) of approximately 0.5 and using the Bio-Rad antigen-dependent isotyping kit (Bio-Rad, Hercules, Calif.) according to the manufacturer's instructions.

Immunization and challenge regimens. All animal experiments were conducted in accordance with local rules for animal housing. Female 6-week-old BALB/c mice received, at 3-week intervals, three intramuscular injections of 50 µl of plasmid solution (1 µg/µl) in PBS in each tibialis anterior muscle using a Becton Dickinson 0.5-ml insulin syringe armed with a 31.5-gauge needle. When animals were immunized with recombinant proteins, 50 µg per dose of purified PcMSP1₁₉-GST or GST emulsified in Freund complete (first dose) or Freund incomplete adjuvant (second and third doses) were injected intraperitoneally (i.p.). P. chabaudi chabaudi PC1 parasites were maintained by weekly reinfection using 10⁵ infected red blood cells (IRBC). The challenge of animals was performed by i.p. injection of 2.5 × 10⁴ P. chabaudi chabaudi PC1 IRBC, obtained from one infected animal with 20 to 30% parasitemia. Under these challenge conditions, all naive mice died by days 8 to 10. Serum samples were taken 3 days before challenge and from day 5 on after challenge by tail bleeding. Parasitemias were determined by counting 1,000 red blood cells per Giemsa-stained blood smear and day.

Statistical analyses. A two-tailed Student's t test was used to evaluate the significance of differences between antibody titers determined on the day of infection for all mice that were still alive. The significance of the proportions of survivors in different groups of mice was calculated by applying the ² method, considering Fisher's exact P values and treating all negative-group mice as the expected outcome group (n = 72).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

N-terminal fusion of PcMSP1₁₉ to HBs results in HBsAg-like particles. In the first experiment, we sought to determine whether a fusion protein containing PcMSP1₁₉ and HBs was released from transfected COS7 cells. To do so, immunoprecipitations from supernatants of transfected COS7 cells with plasmids coding for HBs or PcMSP1₁₉-HBs were performed using either anti-HBs or anti-PcMSP1₁₉ specific antibodies. As predicted, anti-HBs antibodies immunoprecipitated wild-type HBs (24-kDa and glycosylated 27-kDa HBs-protein) from the supernatant (Fig. 1A, lane 2). Using this same antibody, cells transfected with pPcMSP1₁₉-HBs secreted a hybrid protein with an expected monomer size of ~45 kDa (lane 4). In contrast, anti-PcMSP1₁₉ antibodies from a reconvalescent mouse immunoprecipitated only the hybrid protein (lanes 10), whereas they failed to precipitate wild-type HBs (lane 8). Significantly, when the supernatant fraction of pPcMSP1₁₉-HBs-transfected COS7 cells was fractionated on 10 to 40% CsCl gradients, fractions with 1.22 g/ml contained particle-like structures resembling those described earlier for the P. vivax MSP1₁₉-HBs particle (Fig. 1B and reference 28). Taken together, these data strongly indicate that PcMSP1₁₉-HBs is secreted from COS7 transfected cells in the form of a hybrid viral particle which exposes HBs and PcMSP1₁₉ epitopes on its surface.

View larger version (40K): [in this window] [in a new window]   FIG. 1.   PcMSP119-HBs is secreted from transfected COS7 cells. (A) COS7 cells were transfected with plasmids coding for HBs (lanes 1, 2, 7, and 8) or PcMSP119-HBs (lanes 3, 4, 9, and 10) or mock transfected with the expression vector pVXORF1 (lanes 5 and 6). After pulse-labeling and a 24-h chase, lysates (L) and supernatants (S) of cells were immunoprecipitated with antiHBs (diluted 1:500) or hyperimmune serum of a reconvalescent mouse (diluted 1:2,000). (B) Particle-like structures containing MSP119-HBs proteins after fractionation of transfected COS7 cell supernatants by CsCl gradient, resolved by electron microscopy. Material from the 1.21-g/ml fraction is shown at ×100,000 magnification.

Immunization of mice with pPcMSP1₁₉-HBs or pSel-PcMSP1₁₉-HBs but not pPcMSP1₁₉ results in the generation of antibodies against PcMSP1₁₉. We then tested which of our MSP1₁₉ constructs resulted in the highest antibody titers against PcMSP1₁₉ (Table 2). Injection of pPcMSP1₁₉-HBs results in the seroconversion of most mice after the first boost and antibody titers of 1:4,000 to 1:25,600 after the second boost, whereas injection of pPcMSP1₁₉ led to very low anti-MSP1₁₉ antibody titers, which remained near the detection limit (endpoint dilution titer of <1:200). The injection of pSel-PcMSP1₁₉-HBs in mice caused the production of anti- PcMSP1₁₉ specific antibodies after the first boost, but the titers were never higher than those obtained upon genetic immunization with pPcMSP1₁₉-HBs. When an IL-2 coding plasmid was coinjected with pPcMSP1₁₉-HBs, seroconversion was detected after the first boost, and the final titers were as high as 1:51,600. In contrast, coinjection of plasmids coding for IFN- or GM-CSF with PcMSP1₁₉-HBs delayed seroconversion, and the final titers were lower than in the immunizations with all other DNA constructs except pPcMSP1₁₉. Immunization of mice with a recombinant GST-PcMSP1₁₉ fusion protein resulted in the generation of high ELISA titers (1:512,000 to 1:1,000,000). In order to confirm the titers detected in the ELISA, we pooled the sera of each trial group and tested them in Western blots against native PcMSP1 from crude P. chabaudi schizont extracts. While the anti-PcMSP1 titers from sera of genetically immunized mice were similar in both systems, the protein-immunized mice showed dramatically lower titers in Western blots against native MSP1 (230-kDa band [Table 2]). Therefore, the sera of mice immunized with rPcMSP1₁₉-GST were only tested from here on as pools in Western blot assays.

BALB/c mice immunized with pPcMSP1₁₉-HBs or pSel- PcMSP1₁₉-HBs, but not pPcMSP1₁₉ or pPcMSP1₁₉-HBs, coinjected with cytokine genes are partially protected from death. We then tested the protective effect of each construct. Groups of mice were immunized as described above, and after three doses mice were challenged with 2.5 × 10⁴ P. chabaudi IRBC. As shown in Fig. 2, only in the pPcMSP1₁₉-HBs (Fig. 2A) and in the pSel-PcMSP1₁₉-HBs (Fig. 2B) were there significant numbers of survivors (P = 0.00002 [² = 31.2] and P = 0.002 [² = 18.2], respectively). Groups of mice immunized with pPcMSP1₁₉-HBs and pIL2 (Fig. 2C) showed few survivors (2 of 12 mice recovered from infection; P = 0.019, ² 12.0). Coinjection of pPcMSP1₁₉-HBs plus pGM-CSF (Fig. 2D) or pPcMSP1₁₉-HBs plus pIFN (Fig. 2E), and each of their respective control plasmids (pVXORF1+ pIFN or pGM-CSF, respectively) showed no survivors (not shown). Surprisingly, the injection of rPcMSP1₁₉-GST (Fig. 2G) did not provide significant protection; indeed, only 1 mouse out of 12 in two independent trials survived a 25,000 IRBC challenge (P > 0.05). The protection results are summarized in Table 3.

View larger version (33K): [in this window] [in a new window]   FIG. 2.   Course of infection in mice immunized with different plasmid constructs expressing PcMSP119. Groups of mice (n = 12) were immunized with different plasmids as described and infected i.p. with 25,000 IRBC. The parasitemia is indicated in a dotted line with squares, the percentage of survivors is marked as a line with triangles (left y axis), and the actual anti-PcMSP119 titer is indicated as a thick gray line with diamonds (right y axis). All values are geometric mean values, and were plotted against the day of infection. Mice were immunized with pVXORF-PcMSP119-HBs (A), pSel-PcMSP119-HBs (B), pVXORF-PcMSP119-HBs plus pIL2 (C), pVXORF-PcMSP119-HBs plus pGMCSF (D), pVXORF-PcMSP119-HBs plus pIFN (E), rPcMSP119-GST (G), or pVXORF1 (F) as described in Materials and Methods. All negative control mice injected with cytokine vectors and pVXORF1 or rGST, reacted essentially as shown in panel F.

Only discontinuous epitopes are mainly recognized in protected animals. In order to define whether discontinuous or linear PcMSP1 epitopes are preferentially recognized by the sera of protected mice, they were analyzed in immunoblots by using reducing and nonreducing SDS-PAGE conditions. As shown in Fig. 3, the sera of mice immunized with DNA vaccines recognized only unreduced MSP1, indicating that pPcMSP1₁₉-specific antibodies are elicited against discontinuous epitopes (signal for second boost). When the recombinant protein was injected, both forms of MSP1 were recognized (Fig. 3C). Importantly, before infection all immunized mice recognized several forms of MSP1 (nonreduced forms), representing its processed products, but not the 33-kDa product, which results from cleavage of the 42-kDa protein into the 19- and 33-kDa forms. During infection, however, antibodies were developed against other parts of the molecule, leading to a positive signal against reduced MSP1 (Fig. 3A and B, day 9, +DTT lanes). These linear epitopes seem to be localized exclusively in the N-terminal region of MSP1, since only the full-length form was recognized by the sera of DNA-immunized mice (Fig. 3A and B, signal for day 9, +DTT) and not, for example, the biggest cleavage product in the range of 130 to 150 kDa (compare with Fig. 3A and B, second boost, DTT lanes).

View larger version (51K): [in this window] [in a new window]   FIG. 3.   Genetic immunization elicits mostly anticonformational PcMSP119-specific epitopes. Whole extracts of P. chabaudi schizonts were electrophoresed with (+) or without () dithiothreitol (DTT), blotted onto nitrocellulose membranes and detected with the sera of individual mice of three selected groups of six mice each. Sera were taken before infection (2.boost) or on day 8 (d8) or day 9 (d9) after infection. (A) Sera of six mice immunized with pVXORF-PcMSP119-HBs were incubated in a dilution of 1:1,000 (2.boost) or 1:5,000 (d9). (B) Sera of six mice immunized with pSel-PcMSP119-HBs were incubated in a dilution of 1:1,000 (2.boost) or 1:5,000 (d9). (C) Sera of six mice immunized with rPcMSP119-GST were incubated in a dilution of 1:500 (2.boost) or 1:5,000 (d8). Lanes , reaction of anti-mouse IgG peroxidase conjugate with extracts and recombinant proteins.

Genetic immunization with pVXORF-PcMSP1₁₉-HBs induces mostly subclass IgG2a anti-PcMSP1₁₉. To determine if the previous induction of a specific PcMSP1₁₉ antibody subclass was responsible for the protection of mice, we determined the immunoglobulin subclasses in all vaccine trials (Fig. 4). No striking correlation could be found between survival and the appearance of a specific IgG subclass pattern, since all mice immunized with PcMSP1₁₉-HBs or coinjected with IL-2, IFN-, GM-CSF, and pSel-PcMSP1₁₉-HBs produced mostly IgG2a and IgG2b. In contrast, immunization with rPcMSP1₁₉-GST induced mainly a Th2-like response, as characterized by high titers of IgG1, IgG2a, and IgG2b and the almost complete absence of all other subclasses and subtypes.

View larger version (22K): [in this window] [in a new window]   FIG. 4.   Anti-PcMSP119-specific immunoglobulin isotypes after immunization with different plasmids. Mice were immunized with plasmids coding for the indicated products, and immunoglobulin subclasses were determined before infection. The serum dilution of each individual serum was adjusted so that an OD450 of 0.5 was expected in the ELISA. Geometric mean OD450 values with standard deviations are shown.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

In this study, we demonstrate that the genetic immunization of mice with recombinant plasmids encoding the C terminus of the P. chabaudi merozoite surface protein 1, PcMSP1₁₉, is protective against death caused by experimental blood stage infections with the highly virulent strain P. chabaudi PC1. The most efficient construct used in our study encodes a fusion between PcMSP1₁₉ and the small hepatitis B virus surface antigen, PcMSP1₁₉-HBs, which forms particle-like structures upon synthesis in transfected cells. In an attempt to increase the protective capacity of PcMSP1₁₉-HBs, we tried the coinjection of different cytokines of the Th1 subset, namely, IFN--, IL-2, and GM-CSF, and the inclusion of a lymph node-targeting signal in PcMSP1₁₉-HBs. None of them, however, increased the numbers of survivors after challenge compared to the use of PcMSP1₁₉-HBs alone. To the best of our knowledge, this report represents the first protective MSP1₁₉ DNA vaccine in the P. chabaudi-mouse model.

In earlier studies, others have shown that genetic immunizations of plasmids encoding MSP1 or portions of it, including MSP1₁₉, provided protection in BALB/c and C57BL/6 mice against infection with the nonlethal P. yoelii NL17 strain (2). Moreover, Kang and collaborators compared the protective capacity of a fusion between the MSP1₁₉ of P. yoelii and GST delivered as a DNA vaccine or as protein (14). They found no protection against lethal challenge with P. yoelii after immunization with the DNA construct. In our studies, DNA vaccination with the plasmid pPcMSP1₁₉-HBs provided a better protective effect than any other anti-blood stage malaria MSP1-based DNA vaccine so far reported (14). Survival occurred in 5 of 12 animals. Most likely, the HBs moiety plays a crucial role in this protection since we had previously demonstrated that HBs augmented the antigenicity of a plasmid encoding the MSP1₁₉ from P. vivax (10) and that it was able to form a viral hybrid particle (28). Moreover, a widely recognized T-helper-cell epitope is contained in the HBs domain (11).

In order to increase the degree of protection initially found for the PcMSP1₁₉-HBs construct, we opted for the coinjection of Th1-type response-associated cytokines, previously described as being essential in combating primary malaria infections in mice (16, 25). Specifically, IFN- was reported to help suppress malarial blood stage infection in the early phases (19), injection of recombinant IL-2 reduced parasitemias in murine malaria infections (18), and coinjection of GM-CSF-coding plasmids increased protection against sporozoite challenge, mediated by genetic immunization with a CS construct (26). None of the coinjected cytokine-coding plasmids (pIFN, pIL2, or pGMCSF) increased the level of protection during blood stage infection as opposed to the use of pPcMSP1₁₉-HBs alone. In fact, while the IL-2 coinjection resulted only in slightly increased prechallenge anti-MSP1₁₉ titers, all tested cytokines exhibit a more or less pronounced inhibitory effect on the boosting of anti-MSP1₁₉ titers. This suggests that these cytokines, in concert with the cytokine repertoire induced during malarial blood stage infection, inhibited (IL-2) or abrogated (GM-CSF or IFN-) antibody boosting. Together with recent reports on systemic side effects upon overexpression of these cytokines (see, for example, references 7, 30 and 31), the application of GM-CSF or IFN- in the development of malarial DNA vaccines may have to be reconsidered.

We also attempted to increase the protective efficacy of pPcMSP1₁₉-HBs by including a lymph node targeting signal. Previous reports have demonstrated that the fusion of L-selectin to an antigen led to an increase of two orders of magnitude in the antibody response against the antigen (4). Unfortunately, no significant increase in protection or in anti-MSP1₁₉ titers in comparison with those obtained with pPcMSP1₁₉-HBs could be observed; yet, the construct was still capable of inducing antibodies against discontinuous MSP1₁₉ epitopes and significant numbers of survivor mice were found. Taken together, the protection results demonstrated that pPcMSP1₁₉-HBs conferred the highest degree of protection, followed by pSel-MSP1₁₉-HBs, pPcMSP1₁₉-HBs+pIL2, and rPcMSP1₁₉-GST. In contrast, coinjection of pIFN and pGMCSF offered no protection, most likely due to the low antibody titers on the day of death (Fig. 2 and Table 3).

It has been well established that protection against asexual blood stages is antibody dependent, that discontinuous epitopes are the main target of protective antibodies, and that such antibodies are cytophilic (1). Significantly, the PcMSP1₁₉-HBs DNA construct induced IgG2a antibodies mainly against nonlinear epitopes, as demonstrated in immunoblots using reduced or nonreduced whole schizont antigen as a substrate. These results reference MSP1₁₉-HBs constructs for inclusion in future trials of anti-blood stage malaria DNA vaccines.