Plasmodium falciparum Merozoite Surface Protein 8 Is a Ring-Stage Membrane Protein That Localizes to the Parasitophorous Vacuole of Infected Erythrocytes

Abstract

To date, the following seven glycosylphosphatidylinositol (GPI)-anchored merozoite antigens have been described in Plasmodium falciparum: merozoite-associated surface protein 1 (MSP-1), MSP-2, MSP-4, MSP-5, MSP-8, MSP-10, and the rhoptry-associated membrane antigen. Of these, MSP-1, MSP-8, and MSP-10 possess a double epidermal growth factor (EGF)-like domain at the C terminus, and these modules are considered potential targets of protective immunity. In this study, we found that surprisingly, P. falciparum MSP-8 is transcribed and translated in the ring stage and is absent from the surface of merozoites. MSP-8 is the only GPI-anchored protein known to be expressed at this time. It is synthesized as a mature 80-kDa protein which is rapidly processed to a C-terminal 17-kDa species that contains the double EGF module. As determined by a combination of immunofluorescence and membrane purification approaches, it appears likely that MSP-8 initially localizes to the parasite plasma membrane in the ring stage. Although the C-terminal 17-kDa fragment is present in more mature stages, at these times it is found in the food vacuole. We successfully disrupted the MSP-8 gene in P. falciparum, a process that validated the specificity of the antibodies used in this study and also demonstrated that MSP-8 does not play a role essential to maintenance of the erythrocyte cycle. This finding, together with the observation that MSP-8 is exclusively intracellular, casts doubt over the viability of this antigen as a vaccine. However, it is still possible that MSP-8 is involved in an early parasitophorous vacuole function that is significant for pathogenesis in the human host.

The erythrocytic life cycle of Plasmodium falciparum necessitates a series of highly specific, sequential interactions between merozoite and erythrocyte surface proteins. It is well established that antibodies targeting merozoite antigens confer a level of immunity to P. falciparum malaria. One target of these inhibitory antibodies is membrane-associated merozoite surface protein 1 (MSP-1) (13, 14, 16, 24, 25). MSP-1 is expressed as a 200-kDa protein on the surface of merozoites, and it is attached to the merozoite surface by a C-terminal glycosylphosphatidylinositol (GPI) anchor. Proteolytic processing of MSP-1 generates a 19-kDa, C-terminal, membrane-associated fragment termed MSP-1₁₉ that is present on the surface of merozoites during erythrocyte invasion and remains associated with the parasite plasma membrane upon formation of the parasitophorous vacuole in an infected erythrocyte (4). MSP-1₁₉ is composed almost entirely of two cysteine-rich epidermal growth factor (EGF)-like domains, which form a flat, disk-like structure with the two domains folded back on one another (23, 26). While antibodies that target MSP-1₁₉ can significantly inhibit invasion, they do not completely eliminate parasite entry into erythrocytes. This observation led us to investigate whether there are other genes whose structures are related to the structure of MSP-1 in the parasite genome. We hypothesized that MSP-1₁₉ gene paralogs may encode proteins which either complement or circumvent the function of MSP-1₁₉ under immune pressure. Using a bioinformatic approach, we searched the P. falciparum genome for genes encoding proteins which are likely to be expressed as GPI-anchored proteins and which contain double EGF-like domains at the C terminus. We identified two genes, MSP-8 and MSP-10, both of which have been described recently (2, 3).

The PfMSP-8 gene contains a single open reading frame predicted to encode a 597-amino-acid protein with characteristics similar to those of PyMSP-8 (3). In previous studies, an antiserum against a recombinant form of the PfMSP-8 EGF-like domains was shown to react with 98-, 50-, 25-, and 19-kDa species in extracts from trophozoite- and schizont-stage P. falciparum (3). The same antiserum also reacted with protein bands in detergent-enriched phases following Triton X-114 fractionation, and protein localized to the parasite surface of trophozoites and schizonts, as well as the surface of free merozoites. Based on these observations, it was proposed that PfMSP-8 is a GPI-anchored MSP that undergoes proteolytic processing similar to the processing of MSP-1 (3).

To determine whether the double EGF-like domains of MSP-8 perform a function similar to the functions of MSP-1, we previously used allelic replacement to construct a chimeric P. falciparum parasite line in which the double EGF module of PfMSP-1 was replaced by the corresponding region of PbMSP-8 (11). This mutant parasite line invaded erythrocytes and grew as efficiently as the parental parasite line, despite the fact that the respective double EGF-like domains exhibited only low (∼20%) sequence identity. This functional complementarity was consistent with the possibility that other similarly functioning EGF-like domains may promote escape from MSP-1₁₉-targeted immunity.

To characterize the role of MSP-8 in the erythrocytic cycle of P. falciparum, in this study we successfully disrupted the MSP-8 gene in the P. falciparum D10 line. Using this line to validate MSP-8-specific reagents, unexpectedly we observed that MSP-8 is synthesized during the ring stage and is absent from the surface of P. falciparum merozoites. Moreover, MSP-8 does not play an essential role in erythrocyte invasion or blood-stage parasite growth in vitro. However, as MSP-8 is the only known GPI-anchored surface protein synthesized in this early stage of the cycle, a time at which the major virulence genes are being expressed and trafficked through the parasitophorous vacuole, it is still possible that MSP-8 plays a significant role in such a process.

MATERIALS AND METHODS

Expression and purification of recombinant protein.

The DNA sequence corresponding to the double EGF domain of P. falciparum MSP-8 (amino acids V492 to S584) was amplified from P. falciparum (D10 line) genomic DNA using oligonucleotides PfMSP8EGF.1 (CGCGACGCGTGGATCCATGGTATGTGAGAATACAAAGTGTCC) and PFMSP8EGF.2 (CTAGACTAGTCTCGAGCTAACTAGAGGAACAATATATTCCATCACCTT). The resulting PCR product was digested with restriction enzymes BamHI and XhoI (Promega), ligated into the appropriate pGEX vector, and expressed in Escherichia coli BL21 cells (Stratagene) as a glutathione S-transferase (GST) fusion protein (30). The double EGF domains of P. falciparum MSP-1 were expressed as GST fusion proteins as previously described (24).

Generation of antibodies.

To generate antisera, 6-week-old female BALB/c mice and 3-month-old New Zealand White rabbits were immunized with 40 μg and 150 μg GST fusion protein, respectively, in Freund's complete adjuvant. Animals were boosted three times with 35 μg and 120 μg protein in Freund's incomplete adjuvant at 5 weeks postinjection, after which the animals were bled for serum collection. Anti-GST antibodies were removed from rabbit serum using a GST-bound CNBr-activated Sepharose 4B column (Amersham Biosciences). Polyclonal rabbit antibodies were routinely used at a dilution of 1/2,000 for Western blotting and at a dilution of 1/1,000 for indirect immunofluorescence assays (IFA).

Transfection plasmid construction.

The PCR product MSP-8 3′, which encompasses the DNA sequence immediately 3′ of the predicted ATG translation start site and secretion signal for P. falciparum MSP-8 to the TAG stop codon for P. falciparum MSP-8, was amplified from D10 genomic DNA using the following oligonucleotides (a restriction site is underlined): PfMSP8.1 (CCGCGGATCCGGGGAGAATGGAACTACAAATATCGAAAATAATCC) and PfMSP8.2(CTAGACTAGTCTCGAGCTATAAAATAAAAATAAATAGACACAATATAAAAAAG). The PCR product ΔMSP-8, which encompasses the DNA sequence immediately 3′ of the predicted ATG translation start site and secretion signal for P. falciparum MSP-8 to the DNA sequence immediately 5′ of the sequence for the double EGF-like domains of P. falciparum MSP-8, was amplified from D10 genomic DNA using the following oligonucleotides (a restriction site is underlined): PfMSP8.1 and PfMSP8.3 (CCGCTCGAGCTATACTTTATTATTATTAAAAAATTCGACTACAAAAG). The resulting DNA sequences were digested with restriction enzymes BamHI and XhoI and then ligated into BglII/XhoI-digested vector pHH1 (27) to create plasmids pMSP8 3′ and pΔMSP-8, respectively.

P. falciparum culture and transfection.

P. falciparum strain D10 parasites were cultured and synchronized by using standard procedures (18, 32). Ring-stage parasites (∼5% parasitemia) were transfected with 100 μg of purified plasmid DNA (Plasmid Maxi kit; QIAGEN) as described previously (9), except that modified electroporation conditions were used (15).

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis and Western blotting.

To obtain pure ring-stage populations of P. falciparum, parasites were double synchronized 4 h apart with 5% sorbitol (Sigma). Parasite proteins were obtained by saponin (0.15%) lysis of infected red blood cells at several times following synchronization. At each time after synchronization, microscopic examination of Giemsa-stained blood smears was used to determine the approximate developmental stage of parasite cultures. To obtain P. falciparum merozoites, erythrocytes infected with schizont-stage parasites were magnet purified on Macs separation columns (Miltenyi Biotech Gmbh), eluted into culture media, and then incubated until schizont rupture occurred. Free merozoites were collected from culture media by centrifugation. Proteins were separated on 4 to 20% polyacrylamide gradient sodium dodecyl sulfate-polyacrylamide gel electrophoresis gels under nonreducing conditions and transferred to Immobilon-P transfer membranes (Millipore) for Western blotting as described previously (24).

Analysis of transcription.

P. falciparum (D10 line) total RNA was islolated from saponin-lysed (0.15% saponin) infected erythrocytes using an RNeasy minikit according to the manufacturer's instructions (QIAGEN). Preparation of first-strand cDNA was performed with the extracted total RNA using the Superscript First Strand synthesis system (Invitrogen). The concentrations of cDNA samples were determined based on the optical density at 260 nm/optical density at 280 nm and were normalized to a concentration of 100 ng/μl. To detect the presence of transcripts in cDNA samples, PCR was performed with 250-ng P. falciparum cDNA samples using the following gene-specific oligonucleotide primers: PfMSP-1 EGF.1 (CAATGCGTAAAAAAACAATGTCCAG), PfMSP-1EGF.2 (CGATGGTATTTTCTGCAGTTCCTCTTAGTC), PfMSP-8 EGF.1 (GTATGTGAGAATACAAAGTGTCC), and PfMSP-8 EGF.2 (GGAACAATATATTCCATCACCTT).

For Northern blotting, ring-stage populations of P. falciparum parasites were double synchronized 4 h apart with 5% sorbitol (Sigma). Parasites were cultured in several 10-ml dishes at identical parasitemia levels. Individual 10-ml dishes were harvested at several times following synchronization. Samples were obtained by saponin (0.15%) lysis of infected red blood cells, washed in phosphate-buffered saline (PBS), and split into two pellets, one of which was used for analysis of protein expression by Western blotting and one of which was used for analysis of transcription by Northern blotting. Parasite total RNA was isolated from parasites as described previously (17). Pellets used for Northern blotting were resuspended in 1 ml of TRIzol (Invitrogen), vortexed, and stored at −70°C. Parasite total RNA was isolated as follows. TRIzol samples were thawed on ice and extracted with 200 μl of chloroform, and the aqueous supernatant was collected. Total RNA was precipitated by addition of 1 volume of isopropanol and incubation at −20°C overnight. Precipitated RNA was pelleted by centrifugation, resuspended in 10 μl of formamide (molecular biology grade), and incubated at 50°C for 10 min and then at 65°C for 10 min before it was placed immediately on ice. For Northern analysis, 3 μl of RNA from each sample (representing 30% of the total RNA from each time) was electrophoretically separated on a 1.0% agarose (5 mM guanidine thiocyanate) gel and transferred to Hybond-XL membranes (Amersham). Membranes were hybridized with gene-specific probes as previously described (33).

Growth, invasion, and invasion inhibition assays.

The growth rates of parasite lines were compared in a 5-day growth rate assay as previously described (1). Parasite invasion assays with enzyme-treated erythrocytes and antibody inhibition assays were performed as previously described (11).

Indirect immunofluorescence assays.

Parasitized erythrocytes and purified merozoites were washed once in PBS and then fixed in PBS containing 4.0% formaldehyde (diluted from 16% EM grade paraformaldehyde; Electron Microscopy Services) and 0.0075% glutaraldehyde (Merck) for 30 min at room temperature. Fixed cells were washed once in PBS, extracted for 10 min with 0.1% Triton X-100 (Sigma) in PBS, washed again in PBS, treated for 5 min with 0.1 mg/ml NaBH₄ in PBS, washed three times in PBS, and then blocked overnight at 4°C in 0.5% bovine serum albumin (Sigma) in PBS. Fixed erythrocytes infected with P. falciparum and P. falciparum merozoites were resuspended in PBS containing 1 μg/ml 4′,6′-diamidino-2-phenylindole (DAPI), 1/1,000 rabbit anti-PfMSP-8 EGF, and 1/20,000 mouse anti-PfMSP-1₁₉ 4H9/19 (7). Fixed cells were incubated with primary antibodies for 1 h at room temperature. After washing in PBS, erythrocytes infected with P. falciparum were resuspended in PBS containing 10 mg/ml Alexa Fluor 568 goat anti-mouse immunoglobulin G(H+L) [IgG(H+L)] and 10 mg/ml Alexa Fluor 488 goat anti-rabbit IgG(H+L) (Molecular Probes). Fixed P. falciparum merozoites were then resuspended in PBS containing 10 mg/ml Alexa Fluor 488 goat anti-mouse IgG(H+L) and 10 mg/ml Alexa Fluor 568 goat anti-rabbit IgG(H+L) (Molecular Probes). Fixed cells were incubated with secondary antibodies for 1 h at room temperature and then washed four times in PBS prior to mounting on slides and microscopic visualization. Dual color fluorescence images were captured using a Carl Zeiss Axioskop microscope with a PCO SensiCam and Axiovision 2 software.

Preparation of detergent-resistant membranes (DRMs).

Synchronized ring-stage 3D7 parasites (8 to 10% parasitemia) were washed twice in culture medium prior to lysis of erythrocyte membranes in 0.15% saponin on ice for 10 min. Samples were pelleted by centrifugation at 2,800 rpm for 10 min (Beckman GS-6 centrifuge) and then washed three times in morpholineethanesulfonic acid (MES)-buffered saline (25 mM MES [Sigma] [pH 6.5], 150 mM NaCl). Parasites were resuspended in 1.5 ml of MES with a Roche Complete protease inhibitor cocktail tablet and chilled to 0°C on an ice-water slurry. Once parasite samples reached 0°C, an equal volume of 1% Triton X-100 (SigmaUltra) in MES at 0°C was added (resulting in a final Triton X-100 concentration of 0.5%). The sample was maintained at 0°C for 30 min and resuspended every 10 min. Parasite samples (total volume, 3 ml) were transferred to 3-ml TLA 100.3 Beckman open-top thick-walled polycarbonate tubes and pelleted by centrifugation at 2°C for 30 min at 30,000 rpm in a prechilled TLA 100.3 rotor in a Beckman Optima MAX-E ultracentrifuge. Following ultracentrifugation, pellet samples were placed at 0°C, and the supernatant was discarded. Pelleted material was resuspended thoroughly in 184 μl 0.5% Triton X-100 in MES buffer containing protease inhibitors, after which an equal volume of 80% sucrose (ultrapure; Bethesda Research Laboratories, Inc.) in MES buffer was added and the material was thoroughly resuspended. The pellet material (total volume, 368 μl; 40% sucrose) was transferred to a Beckman TLS-55 2.2 ml open-top thin-walled polyallomer tube at 0°C. This material was overlaid with 1.1 ml of 35% sucrose in MES buffer, followed by 733 μl of 5% sucrose in MES buffer to form a sucrose step gradient. The gradient was transferred to a chilled Beckman TLS-55 swing rotor, and the preparation was ultracentrifuged for 18 h at 55,000 rpm at 2°C with low acceleration (9) and no braking (no deceleration). Following ultracentrifugation the gradient was removed and maintained at 0°C. Fifteen 146-μl fractions were removed from the top of the gradient, snap frozen on dry ice, and maintained at −80°C. Prior to Western blotting, fractions were thawed on ice, and 10 μl of each fraction was resuspended in an equal volume of 2× nonreducing sample buffer and placed at 70°C for 10 min.

RESULTS

Conserved features of the MSP-8 gene family.

Prior to this study, MSP-8-like genes had been identified in P. falciparum (3), Plasmodium yoelii (6), and Plasmodium berghei (11). Here, we identified three additional MSP-8 homologues in the Plasmodium chabaudi, Plasmodium vivax, and Plasmodium knowlesi genomes by searching the Plasmodb database (www.plasmodb.org). An alignment of the predicted protein sequences is shown in Fig. 1. One common feature is the presence of an N-terminal signal sequence immediately followed by an asparagine-rich domain, which varies in length from 222 amino acids in PfMSP-8 to 35 amino acids in PcMSP-8. This variability accounts for most of the difference in the predicted molecular weights of the various MSP-8-like proteins. All of the proteins have double EGF-like domains at the C terminus, followed by a series of hydrophobic amino acids, which agrees with the consensus for attachment of a GPI-anchored moiety, as determined by the Detection of GPI tool at www.expasy.org. The double EGFs of MSP-8 are relatively conserved compared to the remainder of the molecule.

FIG. 1.

Alignment of MSP-8 protein sequences from several human, primate, and rodent Plasmodium species. All MSP-8 proteins encode an N-terminal secretion signal that is between 22 and 25 amino acids long (underlined with dashes), a C-terminal double EGF-like domain (underlined), and a C-terminal GPI anchor attachment sequence. All protein sequences were aligned following the removal of an N-terminal asparagine-rich repeat region of variable length (indicated by the solid triangle).

To examine expression of PfMSP-8, polyclonal antibodies were raised against a recombinant form of the EGFs. The specificity of the antiserum was initially confirmed by Western blotting with a panel of recombinant EGFs derived from different MSPs (data not shown). To investigate expression of the MSP-8 protein in P. falciparum parasites, extracts of mixed-stage parasite lines D10, 3D7, HB3, and W2Mef were analyzed by probing Western blot membranes with the anti-PfMSP-8 EGF antibodies (Fig. 2). All four lines expressed two reactive species of MSP-8 (approximately 80 and 17 kDa).

FIG. 2.

PfMSP-8 protein expression in four P. falciparum laboratory cultured lines: Western blot of protein extracts from mixed-stage P. falciparum parasites of the D10, 3D7, HB3, and W2Mef lines. The blot was probed with rabbit antisera specific for the EGF-like domains of PfMSP-8 (αPfMSP-8). The positions of the full-length 80-kDa and processed 17-kDa species are indicated on the right.

Targeted disruption of the PfMSP-8 gene.

To determine whether MSP-8 performs a function essential to the erythrocytic cycle, we attempted to disrupt the PfMSP-8 gene. The following two transfection plasmids were generated: pMSP-8 3′, designed to reconstitute the complete PfMSP-8 gene, and pΔMSP-8, designed to truncate PfMSP-8. Upon transfection into P. falciparum line D10, both pMSP-8 3′ and pΔMSP-8 were shown to have integrated into the PfMSP-8 gene in bulk populations (data not shown). Both transfected populations were cloned, and randomly selected clones were analyzed. Southern blot analysis of EcoRV/NcoI-restricted genomic DNA from representative clones showed that the plasmids had integrated into the PfMSP-8 target site via the expected recombination event (Fig. 3A and B). EcoRV/NcoI digestion indicated the loss of the 6.6-kb endogenous fragment and the appearance of the expected 5.1- and 3.9-kb fragments in the D10-MSP-8 3′ population and 4.7- and 3.9-kb fragments in the D10-ΔMSP-8 population. To examine the expression of PfMSP-8 in the two parasite populations, extracts of mixed-stage parasites were collected and analyzed by Western blotting. Anti-PfMSP-1₁₉ antibodies detected 200-, 42-, and 19-kDa species of MSP-1 in both the parental and MSP-8 disrupted parasite lines (Fig. 3C, left panel). In contrast, anti-PfMSP-8 antibodies reacted with the expected 80- and 17-kDa species in the parental parasite line but did not detect these species in the MSP-8 disrupted line (Fig. 3C, right panel). To confirm the loss of PfMSP-8 expression throughout the erythrocytic cycle, both the D10-MSP-8 3′ and D10-ΔMSP-8 parasite lines were synchronized, and protein extracts were collected at eight times throughout the erythrocytic life cycle. These extracts were analyzed by Western blotting (Fig. 3D). As expected, the 80- and 17-kDa species of PfMSP-8 were detected exclusively in extracts collected from the control transfectant D10-MSP-8 3′. Notably, PfMSP-8 protein expression appeared to peak in ring-stage parasites (8 to 16 h after invasion). A corresponding Western blot membrane probed with heat shock protein 70 antibodies was used to track the total amount of protein loaded from each parasite population throughout the erythrocytic cycle, and the results confirmed that for each time approximately equal amounts of total protein from each parasite population were loaded (Fig. 3D). The reduced HSP-70 signal observed in ring-stage parasites compared to the signal observed in trophozoite-stage and schizont-stage parasites was due to the reduced expression of the HSP-70 protein early in the erythrocytic cycle

FIG. 3.

Targeted disruption of the PfMSP-8 gene in P. falciparum line D10. (A) Schematic diagram outlining the strategy used to target the endogenous PfMSP-8 gene and generate the parasite lines D10-MSP-8 3′ and D10-ΔMSP-8. The sequences encoding the signal peptide and the EGF-like domains are indicated by the black and grey boxes, respectively. hDHFR, human dihydrofolate reductase. (B) Southern blot of restricted genomic DNA isolated from the parental P. falciparum D10, D10-MSP-8 3′, and D10-ΔMSP-8 parasites. The blot was probed with a PfMSP-8 targeting sequence (solid line in panel A). (C) Western blots of protein extracts from mixed schizont- and ring-stage D10-MSP-8 3′ (WT) and D10-ΔMSP-8 (Δ8) parasites probed with mouse monoclonal antibody 4H9/19 specific for PfMSP-1₁₉ (αPfMSP-1) (left panel) and rabbit antisera specific for the EGF-like domains of PfMSP-8 (αPfMSP-8) (right panel). The locations of full-length 200-kDa MSP-1 and the processed 42-kDa and 19-kDa forms of PfMSP-1 are indicated on the left, while the positions of the 80-kDa and 17-kDa PfMSP-8 species are indicated on the right. (D) Western blots of protein extracts from several times throughout the erythrocytic life cycle of D10-MSP-8 3′ and D10-ΔMSP-8 parasites. The numbers at the top indicated the times (in hours) after erythrocyte invasion, and the letters indicate the erythrocytic life cycle stages, as follows: R, ring stage; T, trophozoite stage; S, schizont stage. Identical blots were probed with a polyclonal rabbit antiserum specific for the EGF-like domains of PfMSP-8 (αPfMSP-8) and polyclonal rabbit antisera specific for the PfHSP-70 protein (αPfHSP-70).

Using an antiserum directed against the N terminus of MSP-8 (3), we detected a weak 65-kDa species estimated to be present at about one-tenth the level of endogenous MSP-8 (data not shown). This species was not observed in control parasites and appeared to represent the truncated form of MSP-8 predicted to be synthesized following the single crossover event shown in Fig. 3A. It was almost certainly a nonfunctional form of the protein as, in addition to its reduced expression and/or increased rate of degradation, it lacked both a double EGF module and a GPI anchor attachment signal.

As the predicted molecular mass of PfMSP-8 is approximately 69.4 kDa, it is likely that the 80-kDa PfMSP-8 (PfMSP-8₈₀) protein species represents full-length PfMSP-8. Also, we propose that the 17-kDa, C-terminally processed species of PfMSP-8 (PfMSP-8₁₇) is likely to be analogous to MSP-1₁₉ (that is, composed almost entirely of the EGFs of PfMSP-8 and the GPI anchor moiety). This is especially likely given that antibodies raised against the N terminus of PfMSP-8 failed to detect the 17-kDa species and that the anti-PfMSP-8-EGF antibodies that we used in this analysis recognized only reduction-sensitive epitopes that presumably require appropriate folding of the complete EGF domains (data not shown).

Analysis of P. falciparum MSP-8 transcription and translation confirmed that MSP-8 is a ring-stage protein.

Given the unexpected observation that PfMSP-8 protein expression appeared to peak in ring stages, we investigated the timing of PfMSP-8 transcription and translation in more detail. cDNA was obtained from ring-synchronized D10 parasite cultures at six times during the erythrocytic cycle. Gene-specific PCR performed with cDNA samples detected the PfMSP-8 transcript maximally at 16 h (ring stage) after invasion (Fig. 4A, top panel). The PfMSP-8 transcript was also detected to a limited extent in a sample collected 48 h after invasion, which contained a mixture of schizont-stage parasites, free merozoites, and ring-stage parasites. In contrast, MSP-1 transcripts were detected only in cDNA samples collected 24 to 48 h after invasion, correlating with the trophozoite to schizont stages (Fig. 4A, top panel). In a separate experiment, extracts obtained from synchronized D10 parasite cultures at five times throughout the erythrocytic cycle were analyzed by probing Western blots with anti-PfMSP-8 antibodies. While both PfMSP-8₈₀ and PfMSP-8₁₇ were present in ring-stage parasites (8 to 16 h after invasion), only PfMSP-8₁₇ was present in trophozoite- and schizont-stage parasites (Fig. 4A, bottom panel).

FIG. 4.

Transcription of the PfMSP-8 gene is limited to ring-stage parasites and is tightly linked to translation. (A) (Top panel) Gene-specific reverse transcription (RT)-PCR performed with cDNA isolated at several times throughout the P. falciparum erythrocytic cycle to detect PfMSP-8 and PfMSP-1 transcripts. (Bottom panel) Western blot of parasite samples collected at several times throughout the P. falciparum erythrocytic life cycle probed with antisera specific for the EGF-like domains of PfMSP-8. The positions of PfMSP-8₈₀ (80) and PfMSP-8₁₇ (17) are indicated on the right. (B) (Top panel) Northern blots of total parasite RNA isolated at several times throughout the erythrocytic cycle of P. falciparum were probed with PfMSP-8, PfMSP-1, or 28S RNA sequences. (Bottom panel) Western blots of protein extracts taken from samples obtained at the same times as the samples analyzed by Northern blotting, as described above. Identical blots were probed with polyclonal rabbit antisera specific for the EGFs of PfMSP-8 (αPfMSP-8), monoclonal antibody 4H9/19 specific for PfMSP-1₁₉ (αPfMSP-1), and polyclonal rabbit antisera specific for PfHSP-70 (αPfHSP70). The numbers above the blots indicated the times (in hours) after erythrocyte invasion, and the letters indicate the erythrocytic life cycle stages, as follows: R, ring stage; T, trophozoite stage; S, schizont stage.

The tight linkage of PfMSP-8 transcription and translation was confirmed by Northern and Western blotting of identical parasite extracts collected from synchronized D10 parasites at eight times throughout the erythrocytic life cycle (Fig. 4B). Northern blots hybridized with a PfMSP-8-specific probe confirmed that PfMSP-8 transcription was confined to ring-stage parasites. PfMSP-8 transcription began as early as 4 h after invasion, peaked at 12 to 20 h after invasion, and was complete by 24 h after invasion (Fig. 4B, top panel). Western blot analysis of samples taken at the same times revealed an almost identical pattern of expression, although the PfMSP-8 protein appeared to persist longer than the PfMSP-8 mRNA (Fig. 4B, bottom panel). As expected, PfMSP-1 transcription and translation were completely out of phase with PfMSP-8, occurring only in schizont-stage parasites. Together, these data indicate that the PfMSP-8 gene is transcribed exclusively in ring-stage P. falciparum parasites and that the full-length form of the PfMSP-8 protein is synthesized at this time. Only the smaller C-terminal PfMSP-8₁₇ species was maintained until schizogony.

MSP-8 is localized to the parasitophorous vacuole of ring-stage parasites.

Localization of PfMSP-8 throughout the erythrocytic cycle of P. falciparum was determined by performing IFA with synchronized wild-type D10-MSP-8 3′ and D10-ΔMSP-8 parasite lines which were harvested at the early ring stage (12 h after invasion), late ring stage (20 h after invasion), and late schizont stage (44 h after invasion). Parasites were fixed with formaldehyde and probed with a combination of the PfMSP-1 (monoclonal antibody 4H9/19) and PfMSP-8 EGF antibodies (Fig. 5A). It was apparent that both reagents discontinuously stained the circumference of early-ring-stage parasites and were visualized as discrete dots which showed frequent, but not exclusive, colocalization. This pattern of staining suggests that in early-ring-stage parasites, PfMSP-8 is associated with the parasite plasma membrane and is often in close proximity to PfMSP-1₁₉. In late-ring-stage parasites, anti-PfMSP-1₁₉ antibodies did not stain the circumference of parasites, and staining was limited to a discrete point in the parasite cytoplasm adjacent to the nascent hemazoan crystal, suggesting that PfMSP-1₁₉ is located in the food vacuole at this time. In contrast, anti-PfMSP-8 antibody staining of late-ring-stage parasites was observed as a smooth rim-like pattern around the circumference of parasites and as a discrete dot adjacent to the nascent hemazoan crystal, which colocalized with PfMSP-1₁₉ antibody staining. This pattern of staining suggests that there is redistribution of PfMSP-8 on the parasite plasma membrane between the early and late ring stages and that eventually PfMSP-8 is trafficked to the food vacuole for digestion. In late-schizont-stage parasites, PfMSP-1₁₉ antibody stained the circumference of individual merozoites within the maturing parasite, which is consistent with the known localization of newly synthesized PfMSP-1 on the merozoite surface. In contrast, PfMSP-8 staining at this time was limited to the circumference of the large hemazoan crystal, which is consistent with remnant PfMSP-8 protein localizing in the food vacuole. IFA of D10-ΔMSP-8 parasites performed in parallel revealed no staining for PfMSP-8 and a similar reactivity with PfMSP-1₁₉antibodies, confirming the specificity of the PfMSP-8 reactivity described above and indicating that the loss of PfMSP-8 expression does not alter the localization and trafficking of PfMSP-1₁₉ (Fig. 5A).

FIG. 5.

PfMSP-8 localizes to the parasitophorous vacuole of ring-stage parasites and is absent from the surface of merozoites. (A) Double-labeling IFA of fixed early-ring-stage, late-ring-stage, and schizont-stage D10-MSP-8 3′ and D10-ΔMSP-8 parasites, using a mixture of rabbit antisera specific for the EGF-like domains of PfMSP-8 (αPfMSP-8) and monoclonal antibody 4H9/19 specific for PfMSP-1₁₉ (αPfMSP-1). DAPI was used for nuclear staining (blue). PfMSP-8 (green) and PfMSP-1 (red) colocalized in the parasitophorous vacuole of early-ring-stage P. falciparum parasites but not in the parasitophorous vacuole of late-stage P. falciparum parasites. (B) Double-labeling IFA of fixed merozoite-stage D10-MSP-8 3′ and D10-ΔMSP-8 parasites, using a mixture of rabbit antisera specific for the EGF-like domains of PfMSP-8 and monoclonal antibody 4H9/19 specific for PfMSP-1₁₉ as described above. schiz, schizont.

To investigate whether any PfMSP-8 protein expression occurs in merozoite-stage parasites, infected erythrocytes containing synchronized D10-MSP-8 3′ and D10-ΔMSP-8 schizont-stage parasites were purified and placed back into a culture until schizont rupture occurred. Free merozoites were then collected and subjected to formaldehyde fixation. IFA of D10-MSP-8 3′ and D10-ΔMSP-8 merozoites showed that PfMSP-1₁₉ antibodies stained the entire circumference of merozoites in a continuous rim-like pattern, as expected (Fig. 5B). In contrast, PfMSP-8 antibodies did not stain either D10-MSP-8 3′ or D10-ΔMSP-8 merozoites.

MSP-1₁₉ and MSP-8 are both present in the detergent-resistant membrane fraction of the ring-stage parasite membrane.

GPI-anchored proteins associate with cellular membranes and can be isolated as part of the DRM fraction from both erythrocytes (20, 28, 29) and P. falciparum extracts (19, 34). While the IFA analysis of PfMSP-8 localization was consistent with the predicted localization of the molecule to the parasite plasma membrane as a GPI-anchored protein, we tried to obtain further evidence for this by preparing DRMs from purified ring-stage parasites. For this, synchronized ring-stage P. falciparum 3D7 line parasites were lysed with saponin in order to remove much of the erythrocyte and parasitophorous vacuole membranes and were subjected to ice-cold Triton X-100 extraction. Following sucrose gradient floatation, DRM fractions of increasing density were collected from the top to the bottom of the gradient (Fig. 6). PfMSP-1₁₉, PfMSP-8₈₀, and PfMSP-8₁₇ each predominated in the floating fractions, which is consistent with the hypothesis that PfMSP-8 is present as a GPI-anchored protein in the parasite plasma membrane. The absence of the full-length MSP-1 species in these fractions confirmed the lack of significant schizont contamination in this preparation. Despite their cofractionation here and their partial colocalization by IFA, we found no evidence for association of PfMSP-1₁₉ and PfMSP-8 in coimmunoprecipitation experiments (data not shown).

FIG. 6.

Both MSP-1 and MSP-8 are found in DRMs of ring-stage P. falciparum: Western blots of sucrose gradient-fractionated DRM-associated proteins prepared from ring-stage P. falciparum 3D7 line parasites. Identical blots were probed with monoclonal antibody 4H9/19 specific for PfMSP-1₁₉ (αPfMSP-1) or with rabbit antisera specific for the EGF domains of PfMSP-8 (αPfMSP-8). The locations of PfMSP-1₁₉ (19), PfMSP-8₈₀ (80), and PfMSP-8₁₇ (17) are indicated on the left.

Loss of PfMSP-8 expression does not affect red blood cell invasion pathways or blood-stage growth rates.

The absence of the PfMSP-8 protein on the surface of invading merozoites suggests that PfMSP-8 does not play a role in erythrocyte invasion by P. falciparum merozoites. In accordance with this, loss of PfMSP-8 expression did not alter erythrocyte invasion pathways, as the D10-MSP-8 3′ and D10-ΔMSP-8 parasite lines were indistinguishable in the ability to invade erythrocytes treated with various enzyme combinations (Fig. 7A). As described previously for the P. falciparum line D10 parasites (12), both the D10-MSP-8 3′ and D10-ΔMSP-8 parasite lines invade red blood cells using primarily a trypsin-dependent, neuraminidase- and chymotrypsin-independent invasion pathway. PfMSP-1₁₉ and PfMSP-8₁₇ are both predicted to consist almost entirely of GPI-anchored double EGF-like domains that are functionally complementary (11). To confirm that loss of PfMSP-8 expression did not result in an increased dependence on PfMSP-1₁₉ for erythrocyte invasion, we measured the inhibitory effect of anti-PfMSP-1₁₉ antibodies on both the D10-MSP-8 3′ and D10-ΔMSP-8 parasite lines and found no significant difference (Fig. 7B). We also found no significant differences in the in vitro growth rates of D10-MSP-8 3′ and D10-ΔMSP-8 parasites (Fig. 7C).

FIG. 7.

Disruption of PfMSP-8 does not affect parasite growth, red blood cell invasion pathways, or the invasion-inhibiting effects of PfMSP-1₁₉-specific antibodies. (A) Invasion of enzyme-treated erythrocytes by D10-MSP-8 3′ and D10-ΔMSP-8 parasites. The erythrocyte treatments included NaHCO₃ (0.2%)-buffered medium (Control) and buffered medium supplemented with neuraminidase (N), trypsin (T), or chymotrypsin (C). (B) Invasion-inhibiting effects on the D10-MSP-8 3′ and D10-ΔMSP-8 parasite lines of either heat-inactivated human sera (Control) or rabbit monoclonal antibody 4H9/19 specific for PfMSP-1₁₉ (αPf-MSP-1). (C) Five-day growth assay of the D10-MSP-8 3′ and D10-ΔMSP-8 parasites.

Antibodies to MSP-8 EGFs are elicited in response to infection with P. berghei but not in response to infection with P. falciparum.

In order to examine whether antibodies against PfMSP-8 were elicited by natural infection with P. falciparum, a panel of recombinant proteins were examined for reactivity with pooled sera from naturally exposed individuals (Fig. 8) (22, 24). Human immune sera reacted with the recombinant PfMSP-1 EGFs but not with recombinant PfMSP-8 EGFs or control fusion proteins. All of the recombinant EGFs are correctly folded at least to a degree, as each has been used to successfully raise antisera that react appropriately with the respective parasite protein under nonreducing conditions. While it is possible that natural infection with P. falciparum induces antibody responses to PfMSP-8 which fail to react with the recombinant PfMSP-8 EGF domains or which have such low titers that they are undetectable by Western blotting, the lack of reactivity observed here with pooled human sera is consistent with the intracellular location of this protein and hence its lack of exposure to the immune system. In contrast, probing identical Western blots with pooled sera collected from BALB/c mice semi-immune to P. berghei ANKA infection (10) resulted in detection of both PbMSP-1 and PbMSP-8 EGFs. Hence, unlike the reaction to P. falciparum infection, antibodies to MSP-8 appear to be elicited by infection with P. berghei ANKA.

FIG. 8.

Analysis of immune responses to MSP-8 elicited by infection with P. falciparum and P. berghei ANKA. Pooled human sera from individuals clinically immune to P. falciparum infection and pooled sera from BALB/c mice made semi-immune to P. berghei ANKA infection were used to probe Western blots loaded with a panel of recombinant EGFs from PfMSP-1, PfMSP-8, PbMSP-1, and PbMSP-8. Blots were also probed with pooled human sera from individuals not exposed to P. falciparum infection (Non-immune human pool) and pooled sera from naive mice (Non-immune mouse pool).

DISCUSSION

PfMSP-8 was initially described by Black and colleagues (3) as a 98-kDa GPI-anchored membrane protein expressed throughout much of the asexual life cycle of P. falciparum. This initial characterization suggested that PfMSP-8 is processed into a number of fragments in a manner reminiscent of PfMSP-1 proteolysis. The apparent similarities between PfMSP-1 and PfMSP-8 were strengthened by our observation that the C-terminal double EGF-like domains of PfMSP-1 and PbMSP-8 are functionally complementary (11). Surprisingly, however, the present study firmly established that in contrast to MSP-1, PfMSP-8 is expressed predominantly during the first half of the P. falciparum erythrocytic cycle and is not detected on the surface of merozoites.

We successfully disrupted the PfMSP-8 gene by transfection. Using this parasite line, we were able to validate the specificity of an antiserum raised in this study against a recombinant form of the PfMSP-8 double EGF-like module. It is possible that the discrepancies between the results of our study and those of Black and colleagues were due to nonspecific cross-reactivity of an antiserum generated in the earlier study with species of unknown origin. Our anti-PfMSP-8 antibody reacted with both the MSP-8₈₀ and MSP-8₁₇ species of PfMSP-8 in extracts from the parental parasite line (although a weakly cross-reactive 90-kDa species was observed in some Western blot experiments). In combination with transcriptional analysis of the entire erythrocytic cycle of P. falciparum we found that the PfMSP-8 protein is expressed after erythrocyte invasion in ring stages. This conclusion is supported by recent studies of the P. falciparum transcriptome by Bozdech and colleagues (5), which also indicated that PfMSP-8 transcription peaks in ring-stage parasites.

Processing of PfMSP-8 appears to occur rapidly after the translation of full-length PfMSP-8 protein, as the MSP-8₁₇ species is present together with the 80-kDa form at early times. As parasites mature through the erythrocytic cycle, the level of MSP-8₈₀ decreases. In contrast, MSP-8₁₇ was detected until the schizont stage; however, at this time it appears to localize to the food vacuole. The absence of MSP-8 in merozoites discounts the possibility that PfMSP-8 may play a role in erythrocyte invasion and is supported by the observation that sera from individuals immune to P. falciparum does not contain antibodies against the MSP-8 EGFs. Having established that PfMSP-8 is synthesized as a ring-stage protein, we propose that its name should be changed to P. falciparum ring-stage membrane protein 1 (PfRMP-1).

IFA analysis of the localization of PfMSP-8 throughout the erythrocytic cycle highlights the dynamic nature of the parasitophorous vacuole of ring-stage parasites. In early ring stages, PfMSP-8 is located in discrete clusters around the circumference of parasites, in an area broadly defined as the parasitophorous vacuole. As PfMSP-8 remains associated with DRMs prepared from saponin-lysed ring-stage parasites, it appears to be most likely that PfMSP-8 is attached to the parasite plasma membrane of early ring stages via a GPI anchor. However, the exact localization of PfMSP-8 in either the parasite membrane or the parasitophorous vacuole membrane of ring-stage parasites remains uncertain pending immunoelectron microscopy to better probe the site of its localization. The clustering of PfMSP-8 observed in earlier rings is also relatively transient, as within 8 h PfMSP-8 becomes evenly distributed across the surface of the parasite membrane of maturing rings, before it is removed and trafficked to the food vacuole. The dynamic protein compartmentalization of the parasitophorous vacuole of ring-stage parasites is also apparent in the parasitophorous vacuole membrane. For example, the early ETRAMP proteins also appear to define subdomains of the parasitophorous vacuole membrane of early rings, before they are removed and replaced by the late ETRAMP proteins during the transition from rings to trophozoites (31). Hence, the parasitophorous vacuole and associated membranes appear to be sites of significant change during ring-stage development.

Given the absence from the surface of merozoites, the hypothesis that MSP-8 EGF domains allow invasion to proceed in the face MSP-1₁₉ inhibitory antibodies now seems unlikely. This is despite our previous observation that the EGFs of PbMSP-8 can functionally complement the function of EGFs from PfMSP-1 (11). Hence, the roles played by these EGFs remain unclear. Prior to invasion, it is probable that the MSP-1 EGFs have a structural role, and indeed this may be this domain's only role at this stage and may explain why it can be replaced with structurally similar domains, such as the MSP-8 EGFs. Importantly, as both the MSP-1 and MSP-8 EGFs are present in ring stages, it is still possible that at this time they perform the same (or a related) role in the establishment and functioning of the parasitophorous vacuole. Given that it can be deleted, PfMSP-8 is unlikely to be involved in the establishment of essential transport systems, such as those involved in the import of nutrients or the export of waste products from infected erythrocytes. However, there are several proteins whose functions are probably essential to in vivo survival and/or virulence of P. falciparum parasites but which are not required for in vitro growth (for example, the knob-associated histidine-rich protein that is involved in cytoadherence) (8). Many virulence proteins must be transported across both the parasite and parasitophorous vacuole membranes to localize to the erythrocyte cytosol and in some instances (e.g., PfEMP-1 proteins) to the infected erythrocyte surface (21). It is possible that PfMSP-8 is involved in the establishment of machinery (perhaps by mediating interactions with parasitophorous vacuole membrane proteins) which facilitates transport of at least some of these proteins.