INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Malaria infection of humans remains one of the most important public health problems of tropical regions. Plasmodium falciparum is the most important causative agent of malaria and is responsible for millions of human deaths each year. Improvements in malaria control measures are urgently needed, and one such improvement is the development of a vaccine against asexual stages of the parasite. The selection of protein antigens for such a vaccine has been hampered by the lack of a reliable and readily accessible challenge system for P. falciparum. Accordingly, much attention has focused on the study of laboratory rodents infected by murine malaria parasite species, such as Plasmodium chabaudi, Plasmodium yoelii, Plasmodium berghei, and Plasmodium vinckei. Although not perfect models for the human infection, these systems have proved useful, and important advances in our understanding of the principles of vaccine design have followed their use. In particular, the discovery of merozoite surface protein 1 (MSP1) as a useful component of a subunit vaccine was made in rodent systems (21).

Accordingly, efficacy in such systems is now considered an important criterion for selection of vaccine candidate antigens. Additionally, it is possible to recognize important functional regions of proteins by examining regions of homology between murine and human malaria parasite proteins. Unfortunately, many potential candidate antigens have not been assessed in murine systems because it has not proved possible to identify the homologous gene in these parasites. For example, homologues for the vaccine candidates RAP1 and RAP2 (3, 9, 22, 43, 46), ABRA (51), MSP2 (19, 35, 48, 50), and MSP3 (37, 38, 40) have not yet been identified. This may be because the homologous gene can be quite distantly related and hybridization studies are not sufficiently sensitive to detect these sequences. It has thus been impossible to assess the efficacy of antigen combinations against challenge, particularly by heterologous strains.

We have recently identified a region of P. falciparum chromosome 2 that encodes three distinct merozoite surface proteins in tandem: MSP2, MSP5, and MSP4 (36). MSP4 and MSP5 encode glycosylphosphatidylinositol (GPI)-anchored proteins with observed molecular masses of 40 kDa, and both proteins are predicted to contain a single epidermal growth factor (EGF)-like domain. The locations and similar structural features of these genes indicate that they have probably arisen from a gene duplication event. However, they are actually quite dissimilar at the protein level, the only significant homology being in the EGF-like domain. MSP2 is a well described GPI-anchored protein that, although highly variable in sequence, has been suggested as a possible vaccine component (41). We report here the identification of the syntenic region of the genome in the murine malaria parasite Plasmodium chabaudi adami DS. We detect only one gene in this region, which encodes a novel integral membrane protein with a single EGF-like domain. The gene has an intron-exon structure similar to those of MSP4 and MSP5, and the encoded protein has major structural and immunochemical properties in common with MSP4 and MSP5. We report the detailed characteristics of the gene and its protein product.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Parasites. P. chabaudi adami DS was obtained from Terry Spithill (Monash University, Victoria, Australia). The parasites were maintained by passage in female BALB/c mice, aged 6 to 8 weeks. Stabilates were stored at 70°C. To observe schizonts in the peripheral blood, the mice were maintained in a controlled reversed light-dark cycle with illumination between 1530 and 0630 h, and blood samples were taken at 1300 h. Parasitemia was monitored by Giemsa-stained thin blood smears obtained from the tail.

PCR amplification and chromosome walking. An internal fragment of the gene encoding adenylosuccinate lyase (ASL) was amplified from P. chabaudi adami DS genomic DNA with primers (p423 and p426 [Table 1]) which were designed from the P. falciparum ASL gene sequence (33). PCRs were performed with Taq DNA polymerase (Boehringer Mannheim, Mannheim, Germany) under the following temperature conditions: 35 cycles (each) of 94°C for 1 min, 45°C for 1 min, and 72°C for 2 min and one cycle of 94°C for 1 min, 45°C for 1 min, and 72°C for 5 min. Chromosome walking was performed by either inverse PCR (39) or vectorette PCR (1). For inverse PCR, genomic DNA was digested with various restriction enzymes and religated under dilute conditions to form circular molecules. PCRs were performed with inverted primers derived from the known sequence. Vectorette libraries were constructed from P. chabaudi adami DS genomic DNA digested with various restriction enzymes and ligated to compatible vectorette ends provided by D. Peterson (University of Georgia, Athens). Subsequent PCRs were performed on these libraries with a primer derived from the known sequence and a vectorette-specific primer. PCRs were performed with AmpliTaq Gold (Perkin-Elmer, Branchburg, N.J.) under the following temperature conditions: one cycle of 94°C for 9 min, 40 cycles (each) of 94°C for 1 min, 50°C for 1 min, and 72°C for 1 min, and one cycle of 72°C for 7 min. More extensive chromosome walking was achieved by screening a P. chabaudi 96V partial Sau3A genomic library constructed in plasmid vector pBluescript KS(+) (53) with probes derived from the P. chabaudi adami DS sequence.

DNA sequencing and analysis. DNA sequencing of PCR products and recombinant plasmids was performed with the PRISM Ready Reaction DyeDeoxy Terminator cycle sequencing kit (Applied Biosystems, Inc., Foster City, Calif.) as described by the manufacturer. Sequencing reaction products were separated on a model 373A automated DNA sequencer (Applied Biosystems), and the sequences were analyzed with Sequencher software.

Reverse transcription (RT)-PCR experiments. Total RNA was isolated from P. chabaudi adami DS parasites with TRIZOL reagent (Gibco BRL, Bethesda, Md.) and treated with RQ1 RNase-free DNase (Promega, Madison, Wis.). Reverse transcriptase reactions were performed with Superscript II (Gibco BRL) according to the manufacturer's instructions. Primers p498 and p501 (Table 1) correspond to the predicted 5' and 3' ends of the PcMSP4/5 gene, respectively. Primers p771 and p772 (Table 1) correspond to sequences flanking the intron of the P. chabaudi adami DS erythrocyte membrane antigen 1 (PcEMA-1) gene (18) (GenBank accession no. L27592). PCRs were conducted with Taq polymerase (Boehringer Mannheim) under the following temperature conditions: one cycle of 94°C for 3 min, 39 cycles (each) of 94°C for 1 min, 50°C for 1 min, and 72°C for 1 min, and one cycle of 72°C for 7 min.

Expression and purification of recombinant proteins. A region of genomic DNA containing the PcMSP4/5 gene, lacking the predicted hydrophobic signal and GPI anchor sequences, was amplified from plasmid pMC346 by using primers p525 and p526 (Table 1). The forward primer (p525) contains a BspHI restriction site and an ATG translational start codon. The reverse primer (p526) contains a BglII site, a termination codon, and sequences encoding a C-terminal hexa-His tag. The resulting PCR product was subcloned into the plasmid vector pTrcHis-A (Invitrogen, Carlsbad, Calif.), from which the existing N-terminal hexa-His tag was removed by restriction endonuclease digestion with NcoI and BglII. The recombinant plasmid was introduced into Escherichia coli BL21 (DE3) (Novagen, Milwaukee, Wis.) for protein expression, and large-scale purification of the fusion protein was performed with TALON metal affinity resin (Clontech, Palo Alto, Calif.) according to the manufacturer's instructions. The purity and integrity of the fusion proteins were assessed on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gels stained with Coomassie blue. Mark12 standards (Novex, San Diego, Calif.) were used as protein molecular mass markers. Protein concentrations were measured with the Bio-Rad (Hercules, Calif.) protein assay, as described in the manufacturer's instructions.

Production of antibodies. Female BALB/c mice aged 6 to 8 weeks were immunized with 25 µg of PcMSP4/5 hexahistidine fusion protein emulsified in complete Freund's adjuvant (Difco Laboratories, Detroit, Mich.) and injected intraperitoneally. Two subsequent boosters of 25 µg of antigen with incomplete Freund's adjuvant (Difco Laboratories) were administered at monthly intervals. Sera were collected 2 weeks after each booster.

SDS-PAGE and immunoblotting. P. chabaudi lysates were fractionated by SDS-PAGE on 12% (vol/vol) polyacrylamide gels and immunoblotted onto PolyScreen polyvinylidene difluoride transfer membrane (NEN Life Science Products, Boston, Mass.) according to the manufacturer's instructions. Reduction of samples was achieved by the inclusion of 100 mM dithiothreitol in the SDS sample buffer (125 mM Tris-HCl [pH 6.8], 4% [wt/vol] SDS). SeeBlue prestained standards (Novex) were used as protein molecular mass markers. Reactive antibodies were detected with anti-mouse immunoglobulin (Ig) conjugated to horseradish peroxidase (Silenus Laboratories, Victoria, Australia) with Renaissance Western blot chemiluminescence reagent (NEN Life Science Products).

Triton X-114 partitioning. P. chabaudi parasites were isolated from infected erythrocytes with 0.15% saponin (44) and subjected to Triton X-114 phase separation as previously described (49). Briefly, P. chabaudi parasites were lysed in the presence of 0.5% Triton X-114 (Sigma Chemical Company, St. Louis, Mo.), the parasite lysate was centrifuged to remove insoluble materials, and the supernatant was loaded onto a cushion of 6% sucrose in 0.06% Triton X-114. Phase separation was conducted by incubation at 37°C for 5 min followed by centrifugation at 5,000 × g for 5 min. The aqueous phase was washed with Triton X-114 three times to remove any hydrophobic material, and the detergent phase was washed three times with phosphate-buffered saline to deplete any hydrophilic material. The resulting samples were fractionated by SDS-PAGE, and the proteins were electrophoretically transferred to PolyScreen polyvinylidene difluoride transfer membrane. Reactive antibodies were detected with horseradish peroxidase-conjugated anti-mouse Ig followed by autoradiography with Renaissance Western blot chemiluminescence reagent.

Indirect-immunofluorescence assays. Indirect-immunofluorescence assays were performed as previously described (29). The primary antibody was mouse anti-PcMSP4/5 antiserum raised against the hexa-His-tagged fusion protein, and the secondary antibody was fluorescein isothiocyanate (FITC)-conjugated anti-mouse Ig (Sigma Chemical Company) or Alexa 488-conjugated anti-mouse Ig (Molecular Probes, Inc., Eugene, Oreg.). Confocal microscopy was performed with a Bio-Rad MRC1024ES laser scanning confocal microscope equipped with a krypton-argon laser with main emissions at 488 and 568 nm. The confocal scan head was mounted on a Nikon eclipse TE300 inverted microscope equipped with a 60× (1.4 NA) objective. The FITC- and Alexa 488-labelled secondary antibodies were imaged in the green channel with the 522/35 filter set.

Nucleotide sequence accession numbers. The nucleotide sequence reported in this manuscript has the following GenBank accession numbers: AF080446 and AF080447.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Isolation of a P. chabaudi gene homologous to MSP4 and MSP5. In P. falciparum, the MSP4, MSP5, and MSP2 genes are located in tandem in a head-to-tail configuration on a region close to the left-hand end of chromosome 2, downstream from the gene encoding the enzyme ASL (36) and upstream from a gene encoding a large asparagine-rich protein of unknown function. The reported conserved relative location of many genes among different Plasmodium species (6, 23) and the highly conserved nature of the ASL protein were utilized in a strategy to isolate the merozoite surface protein homologues in the murine malaria parasite P. chabaudi. Homology PCR was performed with primers designed from the P. falciparum ASL gene sequence (33). A 1-kb PCR product was obtained, and sequence analysis revealed an open reading frame (ORF) showing a high degree of similarity to the P. falciparum ASL protein sequence (data not shown).

View larger version (29K): [in this window] [in a new window]   FIG. 1.   (A) Nucleotide and predicted amino acid sequences of PcMSP4/5 from P. chabaudi adami DS. Nucleotide and amino acid numbers are shown. Noncoding DNA sequences are shown in lowercase. The cysteine residues and the single glycine residue of the EGF-like domain are shown in boldface. The putative GPI attachment sequence is underlined. (B) Sequence alignments comparing the EGF-like domains in PcMSP4/5 with those of MSP4 and MSP5. The six conserved cysteine residues are shown in boldface, and the single conserved glycine residues are underlined. (C) Comparison of the predicted protein sequences of PcMSP4/5 with the N-terminal region of MSP2 from an FC27 variant of P. falciparum (GenBank accession no. Y09246). Identical amino acid residues are indicated by asterisks, and conserved amino acid changes are indicated by dots.

View larger version (16K): [in this window] [in a new window]   FIG. 2.   (A) Schematic diagram (to scale) showing a comparison of the arrangement of the MSP4, MSP5, and MSP2 genes in P. falciparum with the syntenic region in P. chabaudi. The flanking genes encoding ASL and an asparagine-rich protein are also shown. (B) Schematic representation (to scale) of the intron-exon arrangement and predicted polypeptide structure of PcMSP4/5. The position of the EGF-like domain is indicated by a hatched box. Signals for secretion and GPI attachment are represented by solid boxes.

Features of the PcMSP4/5 gene. The PcMSP4/5 gene is predicted to encode a 210-amino-acid protein with a calculated molecular mass of 22 kDa. The P. chabaudi gene encodes a protein with a N-terminal signal sequence and a C-terminal hydrophobic region typical of a GPI attachment signal. There is also a single EGF-like domain located in a position similar to those of MSP4 and MSP5, near the C terminus of the protein (Fig. 2B). The typical spacing of cysteines is extremely well conserved in PcMSP4/5, and 51 and 46% of residues in this region are identical to MSP4 and MSP5, respectively (Fig. 1B). Outside of the EGF-like domain there is only very limited similarity between the P. falciparum and P. chabaudi sequences. The short peptide sequences ENGR (residues 34 to 37) and RILG (residues 42 to 45) in MSP4 match identical tetrapeptide sequences in PcMSP4/5 at residues 163 to 166 and 43 to 46, respectively.

View larger version (70K): [in this window] [in a new window]   FIG. 3.   RT-PCR analysis of total RNA isolated from P. chabaudi ring and trophozoite stage parasites. (A) Primers corresponding to the predicted 5' and 3' ends of the PcMSP4/5 ORF were used in PCRs on DNase-treated reverse-transcribed total RNA from rings (lane 1) and trophozoites (lane 3). Negative controls (no reverse transcriptase) for each RNA sample are shown in lanes 2 and 4. As a positive control, an identical PCR was performed with P. chabaudi genomic DNA as a template (lane 5). (B) Primers designed from the sequences flanking the intron of the PcEMA-1 gene (GenBank accession no. L27592) were used in PCRs on DNase-treated reverse-transcribed total RNA from rings (lane 1) and trophozoites (lane 3). Negative controls (no reverse transcriptase) for each RNA sample are shown in lanes 2 and 4. As a positive control, an identical PCR was performed with P. chabaudi genomic DNA as a template (lane 5). An additional negative control (no template) is shown in lane 6. The size standard (in kilobases) is indicated on the left of each panel.

PcMSP4/5 is only expressed in the mature form of the parasite. In order to study the properties of the protein encoded by PcMSP4/5, we expressed the gene as a near-full-length protein, lacking signal and GPI attachment sequences but with a C-terminal hexahistidine tag. This recombinant protein was used to raise specific antisera in mice. Ring and trophozoite stage parasites were harvested from mice infected with P. chabaudi adami DS. Parasite lysates were subjected to SDS-PAGE and electroblotted. Mouse antibodies raised against the PcMSP4/5 hexahistidine fusion protein reacted with a single band with a molecular mass of approximately 36 kDa (Fig. 4B). This is considerably greater than the predicted molecular mass but typical of what has been observed for many P. falciparum asexual-stage proteins, including both MSP4 and MSP5. Further evidence for the aberrantly slow mobility of PcMSP4/5 in SDS-polyacrylamide gels is given by the observation that the full-length recombinant protein has an apparent molecular mass of 36 kDa (Fig. 4A). Expression of the PcMSP4/5 protein was observed only in the mature form of the parasite, a finding that is consistent with the results obtained for P. falciparum MSP4 and MSP5, which are most abundant in trophozoites and schizonts (34). This result is consistent with the RT-PCR results (Fig. 3), which demonstrate the presence of more spliced message in trophozoite stage parasites. The reactivity of the antibodies was lessened if the parasite lysate was exposed to a reducing agent prior to electrophoresis.

View larger version (20K): [in this window] [in a new window]   FIG. 4.   (A) SDS-PAGE analysis of hexahistidine-tagged PcMSP4/5 fusion protein. Molecular mass standards (in kilodaltons) are shown on the left of the panel. (B) Immunoblot analysis of P. chabaudi adami DS parasite lysates. Uninfected mouse erythrocytes (lane 1), ring stage parasite lysates (lanes 2 and 3), and trophozoite stage parasite lysates (lanes 4 and 5) were separated by SDS-PAGE, electroblotted, and probed with mouse anti-PcMSP4/5 antibodies. Parasite lysates were either untreated (lanes 2 and 4) or treated with a reducing agent (lanes 3 and 5). Molecular mass standards (in kilodaltons) are shown on the left of the panel. (C) Phase separation studies of P. chabaudi adami DS. Trophozoite stage parasite proteins were lysed in the presence of Triton X-114, and after phase separation at 37°C, aliquots of various samples were subjected to SDS-PAGE and electroblotted. The immunoblots were probed with mouse anti-PcMSP4/5 antibodies. The fractions shown are uninfected mouse erythrocytes (lane 1), a Triton X-114 lysate of infected erythrocytes (lane 2), the Triton X-114-insoluble pellet (lane 3), the Triton X-114 depleted aqueous phase (lane 4), and the Triton X-114 detergent phase (lane 5). Molecular mass standards (in kilodaltons) are indicated on the left of the panel.

Evidence for the location of PcMSP4/5 on the parasite surface. Parasitized erythrocytes were harvested from P. chabaudi-infected mice and subjected to Triton X-114 partitioning to separate proteins into detergent-soluble and aqueous-phase proteins. Phase-separated proteins were electroblotted and probed with mouse anti-PcMSP4/5 antibodies (Fig. 4C). The abundance of PcMSP4/5 in the detergent phase indicates the presence of a hydrophobic region that could interact with the cell membrane. Indirect-immunofluorescence confocal microscopy was used to analyze P. chabaudi parasites in acetone-fixed thin blood films with mouse antibodies raised to the PcMSP4/5 hexahistidine-tagged fusion protein. Sera from nonimmune mice were used as controls for indirect-immunofluorescence experiments, and no fluorescence was detected with these sera. With the anti-PcMSP4/5 sera, no fluorescence was detected in ring stage parasites, but trophozoite stage parasites, schizonts, and free merozoites were observed to fluoresce. Of interest, there was prominent fluorescence around the rim of the parasites, particularly in free merozoites, an appearance indicative of the membrane location. Areas of the parasite cytoplasm in trophozoites were also stained (Fig. 5).

View larger version (117K): [in this window] [in a new window]   FIG. 5.   Indirect-immunofluorescence assay with mouse anti-PcMSP4/5 antibodies showing surface labelling of a single P. chabaudi trophozoite (top left panel) with FITC-conjugated anti-mouse Ig secondary antibody and surface labelling of a P. chabaudi schizont (top right panel) and free merozoites (bottom panels) with Alexa 488-conjugated anti-mouse Ig secondary antibody.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Only two surface proteins of merozoites have previously been identified in rodent malaria parasite species. The homologues of MSP1 and AMA1 have been found in P. chabaudi (10, 15), P. yoelii (25, 30), and P. berghei (24, 25). Our results indicate that we have discovered a single protein in P. chabaudi which is essentially equally similar to both MSP4 and MSP5 of P. falciparum. In P. falciparum these genes are flanked by the genes encoding the enzyme ASL and an asparagine-rich protein. We have demonstrated that this gene arrangement is preserved in P. chabaudi, and it is consistent with evidence that gene arrangement is often syntenic in the two species (6, 23). Furthermore, the PcMSP4/5 gene is a two-exon structure that encodes a polypeptide similar in size to MSP4 and MSP5. PcMSP4/5 shares all the major structural features of MSP4 and MSP5, including a hydrophobic N-terminal signal sequence, a predicted signal for GPI attachment at the C-terminus, and a single EGF-like domain. Triton X-114 phase separation demonstrated that PcMSP4/5 partitioned into the detergent phase, a property typical of membrane-associated proteins, including MSP4, MSP2, and AMA1 (11, 34, 47). Indirect-immunofluorescence assays showed staining that is consistent with a surface location for PcMSP4/5 in trophozoites, schizonts, and free merozoites. Thus, PcMSP4/5 is the third merozoite surface protein to be identified in rodent malaria parasite species.

The results of RT-PCR analysis indicated that there is differential splicing of PcMSP4/5 mRNA transcripts, depending on the stage of parasite examined. Several lines of evidence suggest that the unspliced mRNA detected cannot be due to genomic DNA contamination. Firstly, DNase was used to treat all RNA samples and no genomic product was detectable in samples not incubated with reverse transcriptase. Secondly, RT-PCR analysis of an unrelated multiexon gene, PcEMA-1, on the same mRNA samples did not show any product equal in size to that of genomic DNA; rather, all mRNA was found to be spliced. In ring stage parasites, where there is little or no detectable PcMSP4/5, the majority of mRNA is found in the unspliced form. Translation of this mRNA, assuming it reaches the cytoplasm, would result in premature termination and probable destruction of the incomplete polypeptide. Conversely, in trophozoites, the mRNA is found predominantly in the spliced form, at a time when PcMSP4/5 protein is readily detected. The presence of a small proportion of spliced mRNA in ring stage parasite samples but the absence of protein on immunoblots may be due to the differences in the detection limits of each analysis, with RT-PCR being the more sensitive. Alternatively, there may have been slight differences in the abundance of parasites at each stage of the growth cycle between the parasite samples used for each method. The quantitation of PCR samples can be problematic; however, in this case the products arise from a single pair of primers so that competition is occurring between the two mRNA splice variants. We have repeated these experiments using low numbers of PCR cycles, and the differences in the ratios of mRNA are similar to those shown here (data not shown).

Differential splicing has been reported for genes encoding asexual-stage proteins of P. falciparum. However, to our knowledge this is the first description of apparent control of stage-specific expression of protein by differential mRNA splicing. The mRNA precursor transcribed from the 41-3 gene yields at least three distinct mRNA species encoding different isoforms of the antigen in the asexual blood stage (28). The encoded protein was reported to be located in the erythrocyte cytoplasm, but its stage-specific expression is not known. There was also no data on whether the different splice variants were preferentially present at different stages. The presence of an intron in the 5' upstream region has been reported in the genes encoding hypoxanthine-guanine phosphoribosyl transferase (5) and the histidine-rich proteins II and III; however, their presence does not alter the ability to translate the encoded polypeptide (52). Using RT-PCR, we have not been able to detect an intron within a 400-bp region upstream of the ATG initiation codon of PcMSP4/5 (our unpublished results).

In P. falciparum, the MSP2 gene is located between ASL and MSP5 on chromosome 2 (36). The homologue of MSP2 is not present at the corresponding position in P. chabaudi. It is possible that MSP2 and MSP5 have been translocated to another chromosomal location. We have been unable to detect either the MSP2 or the MSP5 gene by PCR or by hybridization with the P. falciparum sequences as probes, nor do any of a panel of anti-MSP2 or anti-MSP5 reagents react with P. chabaudi lysates by immunoblotting. Southern analysis with a PcMSP4/5 probe on P. chabaudi genomic DNA reveals only a single gene when washes are performed at low stringency (our unpublished results). This may be because the sequences have substantially diverged, but we favor the suggestion that MSP2 and MSP5 are not present in the P. chabaudi genome. Consistent with this is the presence of sequences somewhat similar to MSP2 in PcMSP4/5, which may represent functional domains that in P. falciparum have been elaborated into a separate protein. The only definitive determination of an MSP2 homologue in a malaria parasite species other than P. falciparum is the closely related primate malaria parasite Plasmodium reichnowi (16). A study has been reported of P. chabaudi (later stated to be P. berghei) in which immunization of mice by synthetic peptides of PfMSP2 conferred protection against subsequent challenge (45). This result suggested the presence in P. berghei of a protein containing sequences homologous to that of MSP2. We have recently identified the MSP4/5 gene in P. berghei, and sequence analysis of the region between ASL and PbMSP4/5 showed that no ORFs are present (27a). Antisera raised to PfMSP2 peptides reacted with a diffuse band of 40 to 60 kDa in immunoblots of parasite protein lysates, and it is possible that protection was due to the raising of an immune response cross-reactive to an as-yet-unidentified protein. Thus, it appears that PcMSP4/5 functionally replaces all three merozoite surface proteins in P. chabaudi. However, the possibility that the homologue of MSP2 is situated at a different chromosomal locus in P. chabaudi cannot be eliminated until the entire genome sequence of P. chabaudi is available. It is unknown why there are different numbers of merozoite surface proteins in the two species. There are clear differences between the two species in host range and ability to interact with erythrocytes. P. falciparum has been shown to have a number of invasion pathways and to be capable of entering erythrocytes by sialic acid-dependent and -independent pathways (20). No such parallel invasion pathways have been described for murine malaria parasites. It is possible that the multiplicity of merozoite surface proteins in P. falciparum may reflect involvement in alternative pathways of invasion.

PcMSP4/5 contains a single EGF-like domain located in a position similar to those of MSP4 and MSP5, near the carboxyl terminus of the protein. Such a structure has been identified in other malaria parasite surface proteins, including MSP1, Pfs25 (27), and Pgs28 (17). The EGF-like domains in several P. falciparum proteins are capable of eliciting antibodies which interfere with the parasite life cycle (2, 7, 8, 17, 26). In the case of MSP1, studies with the P. yoelii homologue have shown that immunization with the EGF-like domains is capable of inducing host protective immunity (4, 13, 14, 31, 32, 42). Recent work in our laboratory has provided evidence that the antigenicity of MSP4 is dependent on correct folding of the EGF-like domain (52a). The observed decrease in reactivity to PcMSP4/5 following reduction of parasite lysates indicates that the immunizing protein is in a conformation capable of eliciting antibodies to conformational epitopes.

One of the commonly held tenets of malaria vaccine design is that an asexual-stage vaccine will contain multiple antigens. This, it is suggested, will overcome difficulties in protecting against challenge by heterologous parasite isolates. In a murine challenge system, lack of protection against heterologous challenge is already described for AMA1 (12) and MSP1 (42). We believe the availability of the P. chabaudi MSP4/5 homologue as a recombinant full-length protein will greatly facilitate the pretrial testing of various aspects of host protectiveness of MSP4 and MSP5 and of antigen combinations in a convenient animal model system.