Identification and Disruption of the Gene Encoding the Third Member of the Low-Molecular-Mass Rhoptry Complex in Plasmodium falciparum

Abstract

The low-molecular-mass rhoptry complex of Plasmodium falciparum consists of three proteins, rhoptry-associated protein 1 (RAP1), RAP2, and RAP3. The genes encoding RAP1 and RAP2 are known; however, the RAP3 gene has not been identified. In this study we identify the RAP3 gene from the P. falciparum genome database and show that this protein is part of the low-molecular-mass rhoptry complex. Disruption of RAP3 demonstrated that it is not essential for merozoite invasion, probably because RAP2 can complement the loss of RAP3. RAP3 has homology with RAP2, and the genes are encoded on chromosome 5 in a head-to-tail fashion. Analysis of the genome databases has identified homologous genes in all Plasmodium spp., suggesting that this protein plays a role in merozoite invasion. The region surrounding the RAP3 homologue in the Plasmodium yoelii genome is syntenic with the same region in P. falciparum; however, there is a single gene. Phylogenetic comparison of the RAP2/3 protein family from Plasmodium spp. suggests that the RAP2/3 duplication occurred after divergence of these parasite species.

The merozoite form of the asexual life cycle in the blood stage of Plasmodium spp. attaches to the surface of the erythrocyte, thus initiating the invasion process (23). Ultrastructural and biochemical studies of merozoites have suggested that the rhoptries play an essential role in the invasion process. These electron-dense organelles, part of the apical complex, are connected to the surface of the apical end of the merozoite by a duct-like structure, and their contents are expelled during erythrocyte invasion (1). Proteins located on the surface of merozoites and those released from the apical organelles, the rhoptries and micronemes, are exposed to the immune system. Consequently, these proteins are potential vaccine candidates, and there is considerable interest in the identification and characterization of proteins in these organelles (2).

A number of proteins located in the rhoptries have been identified and characterized, such as the low-molecular-mass rhoptry protein complex of Plasmodium falciparum (28). This complex consists of various polypeptides, ranging in size from 86 to 37 kDa, that are immunoprecipitated by antibodies to rhoptry-associated protein 1 (RAP1) (7, 11, 29, 43). Several studies established that the largest product, reported to be 86 kDa, is a short-lived precursor of the 82-kDa RAP1 protein and that this 82-kDa protein is further processed to a 67-kDa protein (7, 26, 27, 29). The 42-kDa protein (also reported as 39 kDa) was found to be the product of a second gene, RAP2 (7, 11, 28, 39, 42). Debate remained regarding the origin of the 40-kDa (or 37-kDa) protein also immunoprecipitated as a part of the RAP1 complex. It was suggested that this protein was a breakdown product of either RAP1 or RAP2 (41). It was subsequently shown, however, that RAP1, RAP2, and RAP3 yield different V8 protease cleavage patterns and are thus products of three distinct genes (28). The RAP2 and RAP3 proteins, in particular, were hypothesized to be related molecules, possibly precursor and product. The V8 protease digestion patterns for these two proteins were readily distinguishable, and additionally it was shown via immunodepletion experiments that either RAP2 or RAP3, but not both molecules, were bound to a given RAP complex (28).

The precise function of the RAP complex is unknown, but it is thought to play a role in invasion. Disruption of the RAP1 gene showed that it was not required for merozoite invasion, although obtaining the gene knockout was difficult, suggesting some role for the complex in blood stage growth, albeit a nonessential one (4). Truncation of RAP1 disrupted the interaction with RAP2 and RAP3 in the complex. The truncated form of RAP1 was able to traffic to the rhoptries; however, RAP2 was apparently retained in the endoplasmic reticulum (ER). This suggests that a function of RAP1 is to act as an escorter protein of RAP2 to the rhoptries in P. falciparum (4).

The RAP1 and RAP2 genes have been characterized and sequenced (41, 42), but the RAP3 gene has yet to be identified. In this report, we describe the gene encoding RAP3 and show that this protein is closely related to RAP2. Disruption of the RAP3 gene showed that the RAP3 protein associated with RAP1 in the low-molecular-mass complex.

MATERIALS AND METHODS

Sequence analysis.

Preliminary sequence data at the Sanger Centre website (http://www.sanger.ac.uk) were searched for sequence homology to the RAP2 protein by using tblastn. Predicted protein sequences were aligned using ClustalX 1.81. The RAP2/RAP3 locus of chromosome 5 was analyzed using the resources at the PlasmoDB web site (http://plasmodb.org) (3).

Parasites and transfection.

P. falciparum asexual erythrocytic-stage parasites were cultivated (47) and synchronized by standard procedures (31). The parasite line W2mef from Thailand was used in this study (35). W2mef parasites were transfected, for stable transfection, according to previously published protocols (22, 49) with 80 μg of plasmid that had been purified using the Qiagen (Hilden, Germany) plasmid maxi kit. WR99210-resistant transfected parasites were detectable after 2 to 3 weeks of continuous culture. Parasites containing integrated forms of the plasmid were selected, and cloned lines were obtained by limiting dilution (15, 50).

The transfection plasmid pHH1ΔRAP3 was constructed by cloning an 810-bp RAP3 insert with a mutated start codon into the pHH1 transfection vector (38). This fragment was amplified by using a D10 genomic DNA template and the forward primer HHR35 (CCCCAGATCTTGTACAGGCGCGCCTAAAAAACGATTAGAAAATTTTTGATTTCG [nucleotides −6 to +24]) and the reverse primer HHR33 (CCCCCTCGAGCCCCCCTAGGGGGATCATCTAAAGAAAATGTTTTACTACG [nucleotides 779 to 804 of the coding region]). A BglII restriction enzyme site (shown in boldface for the forward primer) was introduced at the 5′ end of the fragment, and an XhoI restriction enzyme site (shown in boldface for the reverse primer) was introduced at the 3′ end of the fragment for cloning. The start codon was mutated from ATG to ACG, as indicated by the underlined base pair in the forward primer. (A schematic representation of the plasmid is shown in Fig. 3A.)

FIG. 3.

Disruption of the RAP3 gene. (A) Integration of the pHH1ΔRAP3 plasmid by single-site homologous recombination produces a “pseudodiploid,” in which the upstream copy of RAP3 is truncated while the downstream copy lacks a promoter element and has a mutated start codon. Two copies of the plasmid have been integrated as shown. E, EcoRI; B, BsrGI. (B) Hybridization of a pulsed-field gel electrophoresis blot, containing separated P. falciparum chromosomes, with pGem plasmid sequences alone (left panel) reveals multiple transgenic sequences in W2mefΔRAP3/0, corresponding to episomal plasmid DNA. After cycling of cultures with or without WR99210, stable parasite line W2mefΔRAP3/1 exhibits a band that comigrates with chromosome 5 along with a more slowly migrating band that does not correspond to any chromosome. Two independent clonal isolates derived from the W2mefΔRAP3/1 transfectants, W2mefΔRAP3c1 and W2mefΔRAP3c2, exhibit a single band that comigrates with chromosome 5. The upper band on both blots corresponds to DNA remaining in the wells. Hybridization of an identical blot with labeled RAP3 (right) confirms that the pGem sequences are present on the same chromosome as RAP3. (C) Digestion of genomic DNA with EcoRI or BsrGI confirms that the RAP3 gene has been disrupted in W2mefΔRAP3c1 and W2mefΔRAP3c2, yielding the expected restriction pattern.

Nucleic acids and pulsed-field gel electrophoresis.

Genomic DNA was extracted from trophozoite-stage parasites as described previously (12). The analysis of nucleic acids by Southern blot hybridization and recombinant DNA techniques were carried out by using standard procedures. Chromosomes were prepared from trophozoite- and schizont-stage parasites and separated by pulsed-field gel electrophoresis (9) as described previously (13).

RT-PCR.

Total RNA was extracted from synchronized parasite cultures (harvested at 10 to 15% parasitemia) by using the RNeasy mini kit (Qiagen) according to the manufacturer's protocol. DNase treatment was performed using the RNase-free DNase set from Qiagen on the RNeasy mini spin columns as prescribed by the manufacturer. For reverse transcriptase PCR (RT-PCR), 10 μl of total RNA from W2mef parasites was reverse transcribed to cDNA by using an oligo(dT) primer and the Superscript II First Strand cDNA Synthesis Kit (Gibco BRL, Gaithersburg, Md.). One microliter of cDNA was then used for each PCR. The oligonucleotides used for RT-PCR were as follows: forward primer R3 RTPCR, GTATATATTAACACATTTGAG (RAP3 5′ untranslated region); reverse primer R3int, CACAAAAAGCAAATATATTGG (RAP3 nucleotides 814 to 836 of the coding region); and reverse primer PBDT, CAGTTATAAATACAATCAATTGG (pHH1 Plasmodium berghei DT 3′ terminator).

Biosynthetic radiolabeling and immunoprecipitation.

Trophozoites from synchronized parasite lines were incubated with 300 μCi of [³⁵S]methionine (NEN, Boston, Mass.) per ml until multinucleated schizonts were apparent, and proteins were extracted as described previously (27) in 1-ml volumes of 1% T-NET (1% Triton X-100, 50 mM Tris-HCl [pH 7.4], 150 mM NaCl, 5 mM EDTA) with Complete (Roche, Mannheim, Germany) protease inhibitor. Immunoprecipitations were performed with the RAP1 monoclonal antibody 7H8/50 (43) and protein G-Sepharose (Pharmacia Biotech, Uppsala, Sweden). Proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and visualized by enhancement with Amplify (Amersham Pharmacia Biotech, Buckinghamshire, United Kingdom) and autoradiography.

Immunofluorescence.

Immunofluorescence of synchronized parasitized erythrocytes was performed on air-dried thin smears after fixation at room temperature in acetone-methanol (9:1) for 5 min. RAP1 was detected using rabbit antibodies raised to a glutathione S-transferase-RAP1 fusion protein and rhodamine-labeled goat anti-rabbit antibodies (Chemicon, Temecula, Calif.). RAP2 was detected using mouse monoclonal antibody 7B2/1H1 and fluorescein isothiocyanate-labeled sheep anti-mouse antibodies (Silenus, Melbourne, Australia). All parasites were also labeled with the 4′,6-diamidino-2-phenylindole (DAPI) DNA stain (Molecular Probes Inc., Eugene, Oreg.).

Enzymatic treatment of erythrocytes and invasion assays.

W2mef-infected erythrocytes are unable to be purified by Percoll gradient centrifugation, and as a result a novel method was devised to measure erythrocyte invasion (37). This method is dependent on the inability of W2mef merozoites to invade erythrocytes treated with both neuraminidase and trypsin. This method was also adapted to include the use of [³H]hypoxanthine incorporation to determine the invasion rate (10). Erythrocytes were treated as reported previously (17) with various concentrations of either a single enzyme or a combination of enzymes, i.e., 1 mg of TPCK (l-1-tosylamide-2-phenylethyl chloromethyl ketone)-treated trypsin (Sigma, St. Louis, Mo.) per ml, 0.5 mg of soybean trypsin inhibitor (Sigma) per ml, and 0.66 U of Vibrio cholerae neuraminidase (Calbiochem, La Jolla, Calif.) per ml. Enzyme-treated erythrocytes were then washed twice and adjusted to 4% hematocrit in hypoxanthine-free culture medium supplemented with 5% serum and 5% albumax. Parasites were sorbitol synchronized 48 h prior to the invasion assay. On the day of the assay, ring stage parasites were sorbitol synchronized once again and then placed back into culture for 2 h. Parasites were adjusted to 2.5% parasitemia in 4% hematocrit. Pellets of parasitized erythrocytes were then treated with 1 mg of TPCK-treated trypsin per ml and 0.66 U of neuraminidase per ml for 1 h at 37°C. After washing twice, the parasites were treated with 0.5 mg of soybean trypsin inhibitor per ml. Finally the parasites were washed and resuspended in hypoxanthine-free culture medium supplemented with 5% human serum and 5% albumax. The assay was then set up in triplicate, in a final volume of 100 μl, at 0.5% parasitemia in 4% hematocrit in a flat-bottomed 96-well microtiter plate (Becton Dickinson). To allow for reinvasion, parasites were incubated for 48 h before [³H]hypoxanthine (Amersham Pharmacia Biotech, Buckinghamshire, United Kingdom) was added at a final concentration of 1 μCi per well. After a further 24 h, the cells were frozen and thawed to lyse infected erythrocytes. Samples were then transferred to glass fiber filters via a cell harvester (Packard), and [³H]hypoxanthine incorporation was quantitated using a scintillation counter. The mean number of counts from the triplicate wells was calculated and expressed as a percentage of the mean counts observed in parallel cultures of each parasite line in untreated, control erythrocytes.

Nucleotide sequence accession number.

The GenBank accession number for RAP3 is AF516606.

RESULTS

The RAP3 gene is located on chromosome 5 adjacent to RAP2 in P. falciparum. As RAP2 and RAP3 have similar molecular masses and they both interact with RAP1 (7, 11, 29, 43), it was considered likely that they may share some homology. To test this hypothesis, the RAP2 gene sequence (42) was used to search the P. falciparum genome sequence database (http://www.sanger.ac.uk). RAP2 showed homology to a gene predicted to encode a protein of 44.4 kDa. An alignment of the putative RAP3 protein with RAP2 shows that these proteins are homologous, exhibiting 68% similarity and 44% identity (Fig. 1). The five cysteine residues present in RAP2 are conserved in the putative RAP3 protein, as is the signal sequence cleavage site (39, 42). The RAP3 gene from W2mef was cloned and sequenced and found to be identical to the 3D7 gene present in the genome database, suggesting that polymorphisms in this protein may be limited in the P. falciparum population.

FIG. 1.

Sequence alignment of PfRAP2, PfRAP3, and the RAP2/3 homologues from P. yoelii, P. berghei, P. knowlesi, and P. vivax. Identical residues are shown in black, whereas conserved residues are shown in gray. Conserved cysteine residues are indicated by asterisks. The putative signal cleavage site is indicated by an arrow. PLAFA, P. falciparum; PLAYO, P. yoelii; PLABE, P. berghei; PLAKN, P. knowlesi; PLAVI, P. vivax.

The RAP2-RAP3 locus of chromosome 5 from P. falciparum was further analyzed, and a schematic of the genes surrounding this locus is shown in Fig. 2. RAP2 and RAP3 are in a subtelomeric location, closely linked to the P. falciparum interspersed repeat antigen (FIRA) (45, 46) and skeleton binding protein (Pfsbp1) (6) genes. A gene predicted to encode a nuclear scaffold-like protein similar to nuc2+ of fission yeast (25) was also closely linked to RAP2 on the centromeric side of the locus. The close linkage of the RAP2 and RAP3 genes in a head-to-tail fashion as well as the high level of similarity suggests that they arose in a gene duplication event.

FIG. 2.

Organization of the chromosomal region around the RAP2, RAP3, and RAP2/3 genes of P. falciparum and P. yoelii. The nuc2+ gene encodes a nuclear scaffold-like protein of fission yeast. The FIRA gene encodes the P. falciparum interspersed repeat antigen. The sbp1 gene encodes skeleton binding protein 1.

Search analysis of the murine malaria databases (http://plasmodb.org) (The Plasmodium Genome Consortium, 2001) identified a full-length gene in Plasmodium yoelii (PyRAP2/3) and a partial gene in P. berghei (PbRAP2/3) that have similarity to RAP2 (PfRAP2) and RAP3 (PfRAP3) from P. falciparum. Additionally, searching of the Plasmodium vivax (PvRAP2/3) and Plasmodium knowlesi (PkRAP2/3) sequence databases has shown that both of these species also have a gene that is homologous to RAP2 and RAP3. Alignments of the protein sequences encoded by these genes with the protein sequences of PfRAP2 and PfRAP3 show that the P. yoelii, P. berghei, P. knowlesi, and P. vivax proteins have similarity (Fig. 1). The PyRAP2/3 protein shows 30% identity and 49% similarity to PfRAP2 and 32% identity and 55% similarity to PfRAP3. Similarly, PkRAP2/3 has 44% identity and 63% similarity to both PfRAP2 and PfRAP3. The signal sequence cleavage site is conserved in all proteins, as are the positions of three of the five cysteine residues present in RAP2 and RAP3.

The PyRAP2/3 from P. yoelii was found on a 14.7-kb contig from the sequence database. The genes surrounding PyRAP2/3 on this contig were predicted to encode homologues to FIRA (45, 46) and nuc2+ (25), the same genes that flank PfRAP2 and PfRAP3 in P. falciparum (Fig. 2). Therefore P. yoelii has a single RAP2/3 gene, compared to the duplicated PfRAP2 and PfRAP3 genes of P. falciparum.

Targeted disruption of the RAP3 gene in P. falciparum.

To disrupt the RAP3 gene, a transfection plasmid was constructed by cloning an 810-bp fragment from the RAP3 gene into the vector pHH1 (38) to produce pHH1ΔRAP3 (Fig. 3). The parasite line W2mef was transfected with pHH1ΔRAP3 and selected with WR99210. Drug-resistant parasites were selected for those containing integrated forms of the plasmid by one cycle of cultivation without drug (to permit loss of episomal plasmid) (Fig. 3). Chromosomes were analyzed from the W2mef parent and transfected populations W2mefRAP3/0 and W2mefΔRAP3/1, corresponding to parasites isolated shortly after transfection or after one round of cycling, respectively (Fig. 3B). As expected, the pGem probe did not recognize the parental W2mef line but hybridized with several bands in W2mefΔRAP3/0, corresponding to episomal plasmids that typically migrate near chromosomes 6 to 8 under these conditions (15, 34). In contrast, W2mefΔRAP3/1 contained a hybridizing band migrating with chromosome 5, after one cycle for integration, along with a more slowly migrating band that did not correspond to a chromosome. This suggested that there was some integration into the RAP3 gene after one cycle of selection. Hybridization of an identical filter with RAP3 confirmed that the pGem sequences are present on the same chromosome as RAP3 along with the more slowly migrating band (Fig. 3B). The latter band is likely to be a concatameric form of the plasmid, previously observed with other constructs, that has been shown to consist of several copies of the plasmid arranged in a head-to-tail orientation in a complex structure (34). The parasite line W2mefΔRAP3/1 was cloned by limiting dilution, and chromosomes from two independent clonal lines derived from W2mefΔRAP3/1 were analyzed. These lines, designated W2mefΔRAP3c1 and W2mefΔRAP3c2, showed a single hybridizing band migrating with chromosome 5 when hybridized with both plasmid pGem and RAP3 DNA, suggesting plasmid integration into chromosome 5 (Fig. 3B). These clonal lines were used for all further experiments.

In order to confirm that the transfected plasmid pHH1ΔRAP3 had integrated by homologous recombination into the RAP3 gene on chromosome 5, purified genomic DNAs from W2mef, W2mefΔRAP3c1, and W2mefΔRAP3c2 were digested with BsrGI and EcoRI prior to Southern blotting (Fig. 3C). The RAP3 probe in W2mef detected a single 14.6-kb BsrGI band, whereas W2mefΔRAP3c1 and W2mefΔRAP3c2 showed a 12.9-kb band and two smaller hybridizing bands of 8.3 and 6.6 kb. Similarly, the RAP3 probe detected a 14.8-kb EcoRI fragment in W2mef, whereas in W2mefΔRAP3c1 and W2mefΔRAP3c2, fragments of 11.5, 9.9, and 6.6 kb were obtained. The extra bands present for EcoRI-digested W2mef are probably the result of enzyme star activity. Two copies of the plasmid had integrated, as evidenced by the 6.6-kb BsrGI and EcoRI fragments, which correspond in size to an extra copy of the plasmid. These results are consistent with a single-site homologous recombination between the transfection plasmid and the endogenous locus, producing a “pseudodiploid” in which the upstream copy of RAP3 is truncated, while the downstream copy lacks a promoter element and has a mutated start codon (Fig. 3A) (14-16, 50).

Several attempts were made to produce antibodies specific to RAP3. Fusion proteins between RAP3 and glutathione S-transferase were expressed in Escherichia coli and injected into rabbits and mice (44), and synthetic RAP3 peptides were also injected into mice and rabbits. Unfortunately, none of these numerous attempts was successful. Therefore, in the absence of specific antisera, RT-PCR was used to analyze the mRNA transcripts produced by both wild-type and ΔRAP3 parasites to verify that the RAP3 gene was transcribed and had indeed been disrupted (Fig. 4). Two sets of primers were used to amplify cDNA synthesized from total RNA purified from schizont stages of W2mef, W2mefΔRAP3c1, and W2mefΔRAP3c2 parasites. One set of primers was specific for a wild-type RAP3 transcript (R3 RTPCR and R3int), while the other primers were specific for a transcript that would be produced only if the RAP3 gene was disrupted (R3 RTPCR and PBDT) (Fig. 4A). A product of approximately 916 bp was obtained using W2mef cDNA with the R3 RTPCR and R3int oligonucleotides (wild-type primers) (Fig. 4B, lane 1). A product of the same size was also obtained by PCR amplification of W2mef genomic DNA with the wild-type primers (Fig. 4B, lane 5). As a control to demonstrate that there was no genomic DNA contamination of the RNA preparation we used primers to the dihydropteroate synthase (DHPS) gene, which has a number of introns (48). A DNA band of 480 bp was obtained with these primers, which span an intron, showing no contamination of the RNA with genomic DNA (Fig. 4B, lane 3). Additionally, controls that did not include reverse transcriptase failed to show any bands when they were PCR amplified with wild-type primers (Fig. 4B, lane 4). As expected, there were no bands amplified from either wild-type RNA or genomic DNA when mutant-specific primers (R3 RTPCR and PBDT) were used (Fig. 4B, lanes 2 and 6). These results demonstrate that the RAP3 gene is transcribed in wild-type W2mef parasites.

FIG. 4.

Transcription of the RAP3 gene is disrupted in W2mefΔRAP3c1 and W2mefΔRAP3c2. (A) Schematic of the wild-type RAP3 gene, showing the locations of the wild-type-specific primers and the expected 916-bp transcript as well as the locations of the mutant-specific primers and the expected 1,044-bp transcript. Restriction sites shown: E, EcoRI; B, BamHI. (B) PCR products obtained from W2mef wild-type cDNA and genomic DNA show that the RAP3 gene is transcribed in W2mef parasites. Lanes 1, 4, and 5, wild-type primers; lanes 2 and 6, mutant primers; lane 3, DHPS control primers. (C) W2mefΔRAP3c1 cDNA and genomic DNA PCR products indicate that wild-type RAP3 transcripts are not made in ΔRAP3 mutant parasites. Instead, a mutant transcript is produced, which suggests that a truncated RAP3 protein may be expressed. Lanes 1 and 5, wild-type primers; lanes 2, 4, and 6, mutant primers; lane 3, DHPS control primers. (D) PCR products from W2mefΔRAP3c2 cDNA and genomic DNA also show that a mutant transcript but not a wild-type transcript is produced. Lanes 1 and 5, wild-type primers; lanes 2, 4, and 6, mutant primers; lane 3, DHPS control primers. gDNA, genomic DNA; RT, reverse transcriptase.

In contrast, there were no products amplified by using wild-type-specific primers with W2mefΔRAP3c1 and W2mefΔRAP3c2 RNA or genomic DNA, suggesting that the RAP3 gene had been disrupted (Fig. 4C and D, lanes 1 and 5). There was, however, a product of 1,044 bp amplified by using mutant-specific primers with both the cDNA and genomic DNA from ΔRAP3 mutants (Fig. 4C and D, lanes 2 and 6). Once again, control reactions with primers across an intron of DHPS and mutant primers with cDNA but no reverse transcriptase verified that there was no genomic DNA contamination of the RNA purified from the ΔRAP3 mutant parasites (Fig. 4C and D, lanes 3 and 4). These results confirm that the RAP3 gene has been disrupted in the W2mefΔRAP3c1 and W2m3fΔRAP3c2 parasite lines and show that the wild-type mRNA is not transcribed in these parasites.

Disruption of RAP3 results in loss of the RAP3 protein from the RAP complex.

Due to a lack of specific antisera to RAP3, it was not possible to directly analyze protein expression in the ΔRAP3 mutant parasites. Therefore, we used immunoprecipitation of the RAP complex with monoclonal antibodies specific to RAP1 (Fig. 5). As expected, in W2mef parasites, both RAP2 (42 kDa) and RAP3 (40 kDa) were coprecipitated with RAP1 (82 and 67 kDa) (27, 28, 43). However, in ΔRAP3 mutant parasites the 42-kDa RAP2 protein coprecipitated with RAP1, but the 40-kDa RAP3 band was no longer present (Fig. 5). Therefore, the lack of a 40-kDa band in the RAP complex of ΔRAP3 mutant parasites demonstrated that the RAP3 gene encodes the RAP3 component of the low-molecular-mass rhoptry protein complex.

FIG. 5.

Immunoprecipitation of the RAP complex from W2mef, W2mefΔRAP3c1, and W2mefΔRAP3c2. Purified trophozoites were metabolically labeled for 6 h with [³⁵S]methionine and immunoprecipitated with monoclonal antibody 7H8/50 (left panel); background binding to protein G beads alone is shown (center panel), and total labeled protein is also shown (right panel). Both RAP2 (42 kDa) and RAP3 (40 kDa) were coprecipitated with RAP1 (82 and 67 kDa) in wild-type W2mef parasites. In the ΔRAP3 mutant parasites the 40-kDa RAP3 protein no longer coprecipitates with RAP1 and RAP2.

The ΔRAP3 mutant parasites were analyzed by using immunofluorescence microscopy to determine if loss of RAP3 expression had an effect on RAP1 and RAP2 subcellular localization. Immunofluorescence microscopy of W2mef schizonts with either polyclonal rabbit anti-RAP1 or monoclonal anti-RAP2 antibodies revealed a punctate fluorescence pattern typical of rhoptry staining (Fig. 6, top row) (11, 27, 29, 41, 43). Immunofluorescence microscopy of W2mefΔRAP3c1 parasites with the same anti-RAP1 or anti-RAP2 antibodies showed a fluorescence pattern similar to that for wild-type parasites (Fig. 6, bottom row). Therefore, the subcellular localization of RAP1 and RAP2 was unaltered in ΔRAP3 mutant parasites, indicating that trafficking of these proteins to the rhoptries is not affected by disruption of the RAP3 gene.

FIG. 6.

Subcellular localization of RAP1 and RAP2 in W2mef and W2mefΔRAP3c1. Antibodies to RAP1 (green) or RAP2 (red) were incubated with either W2mef or W2mefΔRAP3c1. The yellow in the merged image indicates colocalization.

Invasion of erythrocytes by ΔRAP3 mutant parasites.

Erythrocyte invasion by P. falciparum is a complex process that is believed to incorporate redundancies (24, 32, 33). A number of invasion pathways have been described to date (8, 17, 24). These pathways are defined by entry via the erythrocyte surface proteins glycophorin A, glycophorin B, and an unknown receptor X. It has been shown previously that W2mef parasites are capable of invading erythrocytes via sialic acid residues present on both glycophorins A and B (17). To determine if invasion via these pathways was affected by disruption of RAP3, the ability of ΔRAP3 mutant parasites to invade human erythrocytes that had been pretreated with neuraminidase and/or trypsin was examined. As expected, W2mef parasites invade these erythrocytes inefficiently (approximately 7.6% of invasion into normal cells). No significant difference from W2mef controls in the ability of either the W2mefΔRAP3c1 or W2mefΔRAP3c2 parasites to invade untreated erythrocytes versus treated erythrocytes was observed (data not shown).

To determine if disruption of RAP3 had an effect on the growth rate of these parasites, we compared the growth of wild-type and mutant parasites over a 72-h period. The invasion rates of the W2mef parent and ΔRAP3 mutant parasites in normal erythrocytes showed no statistically significant difference. Therefore, disruption of RAP3, and loss of the RAP3 protein from the RAP complex, has not affected the efficiency of in vitro blood stage growth.

Phylogenetic analysis of the RAP2/3 proteins from plasmodia.

A phylogenetic analysis of the six RAP2/3 proteins from Plasmodium spp. was generated by using the neighbor-joining algorithm implemented in ClustalX 1.81 of the multiple alignment (Fig. 7). The tree shows the divergence between the primate and murine strains of Plasmodium spp. as determined by the bootstrap score. Sequence comparison of the C termini of the proteins in the alignment suggests that the P. knowlesi and P. vivax proteins have diverged little since speciation of these two parasites (85% identity and 98% similarity over 69 amino acids). This is in agreement with phylogenetic trees based on cytochrome b and rRNA sequences (19-21). The phylogenetic tree shows deep branching of PfRAP2 and PfRAP3, equal in depth to that of the P. knowlesi protein, indicating that there has been considerable mutation since the duplication of these genes and implying that the RAP2-RAP3 gene duplication was an ancient event.

FIG. 7.

Phylogenetic analysis of the RAP2/3 proteins from Plasmodium spp. A phylogram was generated by using the neighbor-joining algorithm implemented in ClustalX 1.81 of the multiple alignment depicted in Fig. 1. The bootstrap values are shown on the branches and indicate the number of times out of 1,000 replications that the branching was supported. This tree is unrooted, as no outgroup homologous protein exist.

DISCUSSION

Ultrastructural and biochemical studies have suggested that the rhoptry organelles play an essential role in the invasion process. The RAP complex, consisting of RAP1, -2, and -3, is localized to this organelle and may function in merozoite invasion (7, 11, 29, 43). This is supported by the observations that antibodies to RAP1 inhibit parasite invasion and immunization with purified RAP complex can protect against P. falciparum challenge in Saimiri monkeys (36, 40). It is likely that the RAP3 protein together with the other components of this complex contributed to the protective response observed in those studies, but without a known gene or protein sequence, further analysis has not been possible. We have identified the gene encoding RAP3 and have shown that this protein forms a complex with RAP1.

The identification of the RAP3 gene in the P. falciparum genome was based on the strong homology shown with RAP2 and the demonstration that the protein forms a complex with RAP1. We reasoned that RAP3 would share homology with RAP2, as they had approximately the same molecular mass and both interacted with RAP1 separately in the RAP complex. Importantly, disruption of the RAP3 gene resulted in the loss of the corresponding protein in the RAP complex as shown by immunoprecipitation with anti-RAP1 antibodies. This unequivocally demonstrates that the protein encoded by the RAP3 gene is the 40-kDa RAP3 protein that forms a complex with RAP1.

Previously, it has been shown that association with RAP1 is essential for trafficking of RAP2 to the rhoptries during development of these organelles (4). Disruption of RAP1-RAP2/3 interactions results in RAP2 being retained in the ER and in the absence of this complex within the rhoptries. It is likely that RAP3 was also mislocalized to the ER in parasites expressing truncated RAP1 (4); however, in the absence of specific antisera to RAP3, it is not possible to prove this unequivocally. These results have shown that the RAP complex is not essential for merozoite invasion in the D10 parasite line. Subsequent attempts to disrupt RAP1 and RAP2 in other P. falciparum lines have been unsuccessful, suggesting that the proteins encoded by these genes provide an important advantage for the parasite. This is consistent with the conservation of RAP2/3 within Plasmodium spp. and suggests that this complex plays a role in merozoite invasion of erythrocytes from the broad range of host organisms that these parasites infect.

It was possible to disrupt RAP3, and it is likely that the RAP2 protein is able to complement the loss of function of RAP3. This is consistent with RAP2 and RAP3 forming separate heterodimers with RAP1 rather than a trimeric complex (28). It is interesting that P. falciparum has two homologous RAP2/3 proteins compared to P. yoelii. This is analogous to MSP4/5, which is also conserved across Plasmodium spp. Only one gene has been identified in Plasmodium chabaudi, P. yoelii, and P. berghei whereas P. falciparum has the two related genes MSP4 and MSP5 (5, 30). The functional role of two independent RAP2/3 proteins in P. falciparum is unknown, but it is possible that they may play a role in alternate invasion pathways for human erythrocytes by P. falciparum. In order to pursue this possibility, it will be important to determine if the RAP1, -2, and -3 genes can be disrupted in different parasite lines by using the new negative-selection transfection vector that has recently become available (18).

The proteins of the RAP complex are vaccine candidates, and it is important to develop an understanding of their function and also their potential role in induction of host protective immune responses. The identification of the RAP3 gene in this study will provide the tools to address these questions and to help understand the functional relationship between RAP1, RAP2, and RAP3 in P. falciparum.