Recombinant Plasmodium falciparum reticulocyte homology protein 4 binds to erythrocytes and blocks invasion

Abstract

Plasmodium falciparum invasion of human erythrocytes involves several parasite and erythrocyte receptors that enable parasite invasion by multiple redundant pathways. A key challenge to the development of effective vaccines that block parasite infection of erythrocytes is identifying the players in these pathways and determining their function. Invasion by the parasite clone, Dd2, requires sialic acid on the erythrocyte surface; Dd2/NM is a variant selected for its ability to invade neuraminidase-treated erythrocytes that lack sialic acid. The P. falciparum protein, reticulocyte homology 4 (PfRH4), is uniquely up-regulated in Dd2/NM compared with Dd2, suggesting that it may be a parasite receptor involved in invasion. The aim of the present study was to determine the role of PfRH4 in invasion of erythrocytes and to determine whether it is a target of antibody-mediated blockade and thus a vaccine candidate. We show that both native PfRH4 and a recombinant 30-kDa protein to a conserved region of PfRH4 (rRH4₃₀) bind strongly to neuraminidase-treated erythrocytes. rRH4₃₀ blocks both the erythrocyte binding of the native PfRH4 and invasion of neuraminidase-treated erythrocytes by Dd2/NM. Taken together, these results indicate that PfRH4 is a parasite receptor involved in sialic acid-independent invasion of erythrocytes. Although antibodies to rRH4₃₀ block binding of the native protein to erythrocytes, these antibodies failed to block invasion. These findings suggest that, although PfRH4 is required for invasion of neuraminidase-treated erythrocytes by Dd2/NM, it is inaccessible for antibody-mediated inhibition of the invasion process.

Unlike Plasmodium vivax, which depends completely on the single interaction of the parasite's Duffy binding protein with the Duffy blood group antigen on erythrocytes (1, 2), Plasmodium falciparum exploits multiple parasite receptors to invade erythrocytes. The redundancy in molecular interactions allows P. falciparum to use alternate pathways for invasion of human erythrocytes. The full repertoire of parasite receptors is not yet identified, and the role in alternate invasion pathways of those identified still remains to be fully defined (3–5).

Most of the parasite receptors that are known to play a role in erythrocyte binding and invasion of Plasmodium can be classified into two families. First, the Duffy binding-like (DBL) family that includes the P. vivax/Plasmodium knowlesi Duffy binding proteins and the P. falciparum erythrocyte binding-like proteins (EBA-175, BAEBL, JESEBL, EBL-1, and PEBL). Second, the reticulocyte binding-like (RBL) family that includes the Plasmodium yoelii 235-kDa rhoptry proteins, the P. vivax reticulocyte binding proteins (PvRBP-1 and -2), and the P. falciparum reticulocyte homology (PfRH) proteins (PfRH1, PfRH2a, PfRH2b, PfRH3, PfRH4, and PfRH5) (3–5).

Here, we focus on one member of the PfRH family of parasite receptors, P. falciparum reticulocyte homology 4, PfRH4 (6), to determine its role in invasion and the potential of its antibodies to block invasion. It was shown 17 years ago that the P. falciparum clone Dd2 that is unable to invade neuraminidase-treated erythrocytes switched to sialic acid-independent invasion when cultured with neuraminidase-treated erythrocytes (7). This switched parasite was named Dd2/NM. Expression profiling of the Dd2/NM and Dd2 parasites revealed that two genes, PfRH4 and the pseudogene PEBL, were up-regulated in Dd2/NM (8, 9). Disruption of the PfRH4 gene in the Dd2 parasite clone blocked the ability to switch from sialic acid-dependent to sialic acid-independent invasion of erythrocytes (9), thus providing evidence that PfRH4 is required for the sialic acid-independent erythrocyte invasion.

To further understand how PfRH4 functions in erythrocyte invasion by P. falciparum, we studied the erythrocyte binding activity of PfRH4. Our studies showed that PfRH4 binds to the erythrocyte surface in a sialic acid-independent manner, consistent with a role for PfRH4 in sialic acid-independent invasion of neuraminidase-treated erythrocytes by the Dd2/NM clone. We also identified a 261-aa region in the PfRH4 protein that binds erythrocytes with the same specificity as the native protein and blocks invasion by Dd2/NM. Furthermore, antibodies to PfRH4 block binding of the native protein to erythrocytes, but these PfRH4-specific antibodies do not block invasion. Thus, although PfRH4 plays a role in the sialic acid-independent invasion pathway, it does not appear to be accessible to antibodies during the invasion process.

Results

Expression and Refolding of a Recombinant 30-kDa Protein from PfRH4 (rRH4₃₀).

The interest in the PfRH family of proteins derives from the finding that two P. vivax RBL proteins bind to reticulocytes that are preferentially invaded by P. vivax (10). Thus, this family is implicated as parasite receptors during erythrocyte invasion. The RBL family of proteins does not have obvious domain structures, such as the cysteine-rich regions in the DBL family (e.g., P. vivax Duffy binding protein). The 261-aa sequence of PfRH4 from Asn-328 to Asp-588 (GenBank accession no. AAM 47174) was chosen on the basis of a clustal alignment between PfRH4 and the phylogenetically close PvRBP1 (10) (Fig. 1), which shows homology between PfRH4 and PvRBP1 (Fig. 1). Similar pairwise alignment of PfRH4 with PfRH1 (11) also exhibits regions of homology (data not shown). In designing the region of PfRH4 for recombinant expression, we confined ourselves to the 261-aa region, because the N-terminal and C-terminal to this sequence are stretches of low complexity that we did not want to include in our recombinant protein (Fig. 1). The sequence of this 261-aa region is identical in the P. falciparum clones, Dd2 (GenBank accession no. AAM 47174), Dd2/NM (GenBank accession no. DQ100425), HB3 (GenBank accession no. 47173), and 3D7 (GenBank accession no. 47192). The recombinant 30-kDa fragment of PfRH4 (rRH4₃₀) was overexpressed in E. coli as inclusion bodies, refolded from the inclusion body in an l-Arg rich buffer and purified to homogeneity, using hydrophobic interaction chromatography. rRH4₃₀ was expressed with a C-terminal 6 His-tag that allowed rRH4₃₀ to be identified in an immunoblot, using a His-tag specific antibody. rRH4₃₀ was confirmed to consist of the full 261 aa by N-terminal sequencing and reactivity with the C-terminal 6 His-tag specific antibody in immunoblots. Both rats and rabbits were immunized with rRH4₃₀ to produce PfRH4-specific antibodies.

Fig. 1.

Expression of a recombinant 30-kDa protein of PfRH4 (rRH4₃₀) chosen on the basis of homology with P. vivax reticulocyte binding protein 1 (PvRBP1). Clustal alignment of the PfRH4 (GenBank accession no. AAA47174) protein sequence (amino acids 283–642) with PvRBP-1 (GenBank accession no. AAA29743). The 261-aa sequence of PfRH4 selected for recombinant expression is highlighted in the boxes. *, identical residues; :, conserved substitutions; ., semiconserved substitutions.

PfRH4 Binds Neuraminidase-Treated Erythrocytes.

The native full-length PfRH4 was studied from culture supernatants that contained merozoites released from infected erythrocytes in the absence of any target erythrocytes for their invasion. Studies have shown that extracellular merozoites release parasite proteins into the culture, and such culture supernatants can be a source of parasite receptors that bind erythrocytes (12–15). Culture supernatants from the PfRH4 expressing P. falciparum clones, HB3 and Dd2/NM, were assayed by immunoblotting and immunoprecipitation, respectively, using the anti-PfRH4 antibody. A protein of ≈250 kDa was found in the P. falciparum HB3 culture supernatant (Fig. 2A), the size corresponding to the predicted molecular mass of PfRH4 (6). PfRH4 in Dd2/NM culture supernatant was detected as a doublet of ≈150 kDa (Fig. 2C). Presumably, the smaller size reflects proteolytic cleavage of PfRH4 produced during the incubation of the Dd2/NM culture supernatant.

Fig. 2.

Erythrocyte binding activity of native PfRH4 and recombinant rRH4₃₀ protein. (A) Binding of the native PfRH4 protein in the HB3 culture supernatant incubated with untreated (U) erythrocytes, different enzyme-treated erythrocytes (Nm, neuraminidase; T, trypsin; C_T, chymotrypsin), and rabbit (Rab) erythrocytes. The PfRH4 parasite protein was detected in the eluate fractions by immunoblotting. (B) Binding of the rRH4₃₀ protein with a similar set of erythrocytes. (C) Binding of native PfRH4 in the ³⁵S-metabolically labeled culture supernatant (Total Sup) of Dd2/NM. Proteolytically cleaved fragments of PfRH4 were detected in the eluate by immunoprecipitation. The cleaved fragments of PfRH4 bind to both untreated and neuraminidase-treated erythrocytes. (D) The same eluate samples were used for the detection of EBA-175 by immunoprecipitation. EBA-175 binds to untreated erythrocytes but not to neuraminidase-treated erythrocytes.

To study the binding characteristics of PfRH4 to erythrocytes, culture supernatants were incubated with erythrocytes. For HB3, the erythrocytes were washed and spun through oil, and bound proteins were eluted with 1.5 M NaCl. The binding persisted after three washes with RPMI medium 1640. After the washes, the erythrocytes were spun through oil to separate the liquid from the erythrocytes for elution of parasite proteins from the erythrocytes. PfRH4 was detected by immunoblot of eluate from the unlabeled HB3 culture supernatant. The amount of PfRH4 bound with neuraminidase-treated erythrocytes was greater than the amount bound with untreated erythrocytes (Fig. 2A). The binding of PfRH4 was greatly reduced by chymotrypsin or trypsin treatment of erythrocytes, indicating that the protein was removed by these enzymes (Fig. 2A). No protein bound to rabbit erythrocytes was detected (Fig. 2A). PfRH4 in the ³⁵S-metabolically labeled culture supernatant of Dd2/NM was detected by immunoprecipitation. The 150-kDa fragments of PfRH4 from P. falciparum clone Dd2/NM did not remain bound to erythrocytes if the cells were washed with RPMI medium 1640 and was only seen bound to erythrocytes if they were only run through oil without washing (Fig. 2C). Similar to the HB3 culture supernatant, the native PfRH4 proteins found in the culture supernatant of Dd2/NM bound to neuraminidase-treated erythrocytes more strongly than to untreated erythrocytes (Fig. 2C). The same eluate samples showed that EBA-175 bound with untreated erythrocytes and not neuraminidase-treated erythrocytes (Fig. 2D). EBA-175 is known not to bind neuraminidase-treated erythrocytes indicating that unbound supernatant is not carried through the oil along with the erythrocytes. Thus, the binding of PfRH4 from the Dd2/NM culture supernatant with neuraminidase-treated erythrocytes is specific. Because the full-length PfRH4 native protein was detected only in the HB3 culture supernatant, all subsequent studies were done with the HB3 culture supernatant. Because the region of rRH4₃₀ was identical in HB3 and Dd2/NM, the antibodies to rRH4₃₀ should bind PfRH4 from both parasites equally well.

rRH4₃₀ Binds Erythrocytes in a Receptor-Specific Manner.

The specificity of rRH4₃₀ for binding erythrocytes was studied in two ways: (i) the binding of rRH4₃₀ to various modified erythrocytes was compared with the binding of the native PfRH4, and (ii) the ability of rRH4₃₀ to competitively inhibit the binding of the native PfRH4 to erythrocytes was tested. rRH4₃₀ was incubated with erythrocytes in the erythrocyte binding assay described above. The quantity of rRH4₃₀ that bound with neuraminidase-treated erythrocytes was greater than with normal erythrocytes (Fig. 2B), similar to the findings with the native protein (Fig. 2A), and binding was greatly reduced to chymotrypsin- and trypsin-treated erythrocytes (Fig. 2B). Furthermore, both rRH4₃₀ and the native PfRH4 protein failed to bind rabbit erythrocytes (Fig. 2 A and B). rRH4₃₀ bound with the same specificity as the native PfRH4, indicating that rRH4₃₀ has an intact receptor binding domain similar to the native protein.

To prove that rRH4₃₀ and the native PfRH4 bound to the same molecule on the erythrocyte, we investigated whether the binding of rRH4₃₀ could block the erythrocyte binding of native PfRH4. rRH4₃₀ competed for the binding of PfRH4 to erythrocytes in a dose-dependent manner (Fig. 3A), with an IC₅₀ of 0.37 μM (Fig. 3B). We conclude from these studies that rRH4₃₀ binds to the same erythrocyte molecule as native PfRH4.

Fig. 3.

Recombinant PfRH4 (rRH4₃₀) blocks the erythrocyte binding of native PfRH4. (A) Competition binding assay in which the binding of native PfRH4 (HB3) is inhibited by rRH4₃₀ in a dose-dependent manner. (B) Inhibition curve derived from scans of two independent assays shows that rRH4₃₀ blocks the binding of native PfRH4 at an IC₅₀ of 0.37 μM.

rRH4₃₀ Blocked Invasion of Neuraminidase-Treated Erythrocytes by Dd2/NM.

Erythrocytes were preincubated with rRH4₃₀ at concentrations that blocked the erythrocyte binding of the native PfRH4 and then tested for invasion by Dd2/NM. rRH4₃₀ blocked Dd2/NM invasion of neuraminidase-treated erythrocyte (Fig. 4B). However, at the concentration of 3.4 μM, which completely blocked the binding of the native PfRH4 protein, invasion was reduced by only 70%, thus implying that more recombinant protein was needed to completely block parasite invasion than was required for blocking binding of the native PfRH4 protein. The blockade was specific for Dd2/NM in that rRH4₃₀ did not block invasion of the P. falciparum clone 3D7 into neuraminidase-treated erythrocytes (Fig. 4C). Furthermore, rRH4₃₀ did not block the invasion of untreated erythrocytes by Dd2/NM (Fig. 4A), indicating that PfRH4 is redundant for parasites that can use sialic acid for invasion.

Fig. 4.

Recombinant PfRH4 (rRH4₃₀) blocks erythrocyte invasion of Dd2/NM through the sialic acid-independent pathway. (A and B) The ability of the erythrocyte binding rRH4₃₀ protein to block invasion of normal untreated (A) and neuraminidase-treated erythrocytes (B) by the Dd2/NM parasite clone was studied in invasion inhibition assays in which the target erythrocytes were preincubated with rRH4₃₀ at different final concentrations (0.34, 0.87, 1.70, and 3.4 μM). The bound rRH4₃₀ reduced invasion of only the neuraminidase-treated erythrocytes. (C) Invasion of another sialic acid-independent parasite clone, 3D7, with neuraminidase-treated erythrocytes preincubated with similar amounts of rRH4₃₀ did not block the invasion of the parasite.

Antibodies to rRH4₃₀ Block Native PfRH4 Binding but Not Invasion.

After demonstrating that rRH4₃₀ blocks invasion of neuraminidase-treated erythrocytes by Dd2/NM, we determined whether antibodies to rRH4₃₀ could block invasion. We demonstrated that the antibodies raised against rRH4₃₀ blocked binding of the native PfRH4 to the erythrocyte surface (Fig. 5 and Table 1). The anti-PfRH4 antibodies blocked binding of the native protein to erythrocytes in a dose-dependent manner (Fig. 5). At concentrations used in the growth inhibition assays, the anti-PfRH4 antibodies blocked the erythrocyte binding of the native PfRH4 protein by 22–90%. However, we found that antibody concentrations that would block binding of the native protein to erythrocytes did not block invasion (Table 1). Both anti-PfRH4 IgG purified from rats and rabbits did not block invasion of either untreated or neuraminidase-treated erythrocytes (Table 1). Thus, although PfRH4 is essential for invasion, it does not appear to be accessible to antibody-mediated blockade of invasion.

Fig. 5.

Anti-PfRH4 antibodies block erythrocyte binding of the native PfRH4 protein. Total IgG from rat 38, immunized with rRH4₃₀, blocks erythrocyte binding of the PfRH4 protein. HB3 culture supernatant was incubated with normal erythrocytes in the presence of purified IgG from rat sera at different final concentrations of 0.5–5.0 μg/μl. The percentage of blocking of the erythrocyte binding of the native PfRH4 is shown in the lower numbers.

Table 1.

Growth inhibition assay (GIA) of the anti-PfRH4 IgG against the Dd2/NM clone of P. falciparum and antibody mediated blocking of RBC binding of native PfRH4

Anti-PfRH4 IgG	Animal no.	Bleed	IgG in GIA well, μg/μl	ELISA units* in GIA well	Blocking, %†	Inhibition, %
						Untreated RBC	Neuraminidase-treated RBC
Rat	35	day 70	5.0	33,323	56	−2	5
Rat	36	day 70	5.0	21,552	41	0	5
Rat	37	day 70	5.0	22,330	41	2	−7
Rat	38	day 70	5.0	62,832	90	0	−3
Rat	Pool‡	day 0	5.0	13	ND	0	0
Rabbit	1	day 98	6.2	22,010	22	−7	16
Rabbit	2	day 98	6.2	18,685	30	−8	10
Rabbit	3	day 98	6.2	24,098	66	−13	19
Rabbit	Pool‡	day 0	6.2	44	ND	−5	12

ND, not determined; RBC, red blood cell.

*ELISA units represent the reciprocal of the dilution required to obtain an OD of 1 in a standardized ELISA assay.

^†Percentage of blocking in binding of native PfRH4 to RBCs by the anti-PfRH4 IgG.

^‡Total IgG purified from preimmune sera pooled together from the four rats or three rabbits.

Localization of PfRH4 in the Merozoite.

The localization of the PfRH4 native protein in the merozoite was studied by confocal immunofluorescence microscopy, using the anti-PfRH4 antibody. The native PfRH4 was found to be localized at the apical pole of the merozoite (Fig. 6 A and B). Because there was not perfect colocalization with known rhoptry and microneme marker proteins, RON4 and AMA-1, respectively (Fig. 6 A and B), further analysis by immunoelectron microscopy is essential for determining the precise location of PfRH4 in the merozoite. Two rhoptry marker proteins, RhopH1 and RAP-1 that reside in the body of the rhoptry, were also used in the study to check whether PfRH4 localizes in the body of the rhoptry. However, no colocalization was observed with these two rhoptry proteins (data not shown). In free merozoites, PfRH4 remained predominantly localized in the apical end; some was transferred to the merozoite plasma membrane (Fig. 6C), whereas AMA1, a micronemal protein, was observed to be translocated to the merozoite surface to a much greater extent (Fig. 6C).

Fig. 6.

Apical localization of PfRH4 in the merozoite by immunofluorescence confocal microscopy. (A) 3D7 schizonts were double-labeled with anti-PfRH4 IgG and anti-AMA1 monoclonal antibody (1G6). Mature schizonts immunolabeled with anti-PfRH4 were stained with Alexa 488 secondary antibody (green). Schizonts labeled with anti-AMA1 (1G6) were stained with Alexa 594 secondary antibody (red). The differential interference contrast (DIC) image also shows two early schizonts that do not express PfRH4 or AMA-1, because these proteins are expressed late in the asexual life cycle. (B) 3D7 mature schizonts were double-labeled with anti-PfRH4 IgG and anti-RON4 monoclonal antibody (24C6). Schizonts labeled with anti-RON4 were stained with Alexa 594 secondary antibody (red). PfRH4 did not colocalize with the known rhoptry and microneme marker proteins, RON4 and AMA-1, respectively. (C) A free merozoite was double-labeled with anti-PfRH4 IgG and anti-AMA1 monoclonal antibody (1G6).

Discussion

The identification of the parasite and erythrocyte receptors that mediate invasion of P. falciparum is complicated by the fact that the parasite has evolved multiple, redundant invasion pathways (3). Thus, to study a single pathway, it is necessary to manipulate the erythrocyte to restrict the invasion of the parasite through a single, nonredundant pathway. For example, the Duffy blood group antigen was discovered as an invasion receptor for P. vivax, because the parasite has only one nonredundant Duffy-dependent mode of invasion (1, 2). Likewise, the P. falciparum receptor, EBA-175 was discovered by the failure of certain parasite clones to invade glycophorin A null erythrocytes, En(a−) (16), although another P. falciparum clone could invade En(a−) erythrocytes (17). In the present case, we chose a system where the parasite clone Dd2 switched from requiring sialic acid for invasion to invading neuraminidase-treated erythrocytes devoid of sialic acid on their surface (7), a change that was associated with the up-regulation of the parasite receptor, PfRH4 (8, 9). To prove that no other parasite protein was redundant for PfRH4-mediated invasion of neuraminidase-treated erythrocytes by Dd2/NM, we showed that recombinant PfRH4 (rRH4₃₀), when bound to erythrocytes, blocked invasion of the Dd2/NM clone. Invasion by Dd2/NM of erythrocytes not treated by neuraminidase was unaffected by rRH4₃₀, indicating that PfRH4 was not required for invasion of untreated erythrocytes. The selected up-regulation of PfRH4 allowing Dd2/NM to invade neuraminidase-treated erythrocytes was apparently not accompanied by a down-regulation of parasite receptors used by the parental Dd2 clone to invade untreated erythrocytes (8, 9). Thus, in the special circumstances of studying Dd2/NM invasion of neuraminidase-treated erythrocytes, PfRH4 was essential. The P. falciparum clone 3D7 could invade neuraminidase-treated erythrocytes with rRH4₃₀ bound to the erythrocyte, indicating that multiple sialic acid-independent pathways of invasion do exist.

The requirement of PfRH4 for invasion of neuraminidase-treated erythrocytes by Dd2/NM establishes the conditions for testing the effect of anti-PfRH4 antibodies on invasion. Under these special conditions, we were unable to block invasion by anti-PfRH4 antibodies. Monoclonal antibodies against the first identified member of the RBL family, Py235, could confer protection against P. yoelii by restricting invasion to reticulocytes, the preferred target erythrocyte (18). Why, then, do anti-PfRH4 antibodies not block invasion? The possible reasons are as follows. First, antibodies to only a 30-kDa recombinant protein of PfRH4, rRH4₃₀, may not block erythrocyte invasion because of the quality or quantity of the antibody. This seems unlikely, because anti-PfRH4 antibodies block the binding of the native PfRH4 to the erythrocyte surface.

Second, the parasite receptor may be hidden from antibodies or only exposed for a brief period during invasion, thus requiring high titers of antibody to block invasion. Our immunofluorescence localization studies have shown that PfRH4 is located at the apical end but does not colocalize with the known rhoptry or microneme markers. Although this may imply that PfRH4 may be localized in a new organelle or differentially released from micronemes, further studies by immunoelectron microscopy are required to confirm the localization. In free merozoites, PfRH4 was observed to be primarily localized in an apical organelle and not released on the merozoite surface to the extent as in case of AMA-1, which is found in micronemes and is found on the surface of extracellular merozoites. Thus, it appears that PfRH4 may be released in the later stages of invasion. A similar situation exists for HIV, where the virus surface is heavily glycosylated, and the virus molecule that binds with the cells' chemokine receptor is only uncovered for a short period (19). Understanding why PfRH4 remains hidden will require two unknown facts to be elucidated: (i) in which organelle PfRH4 is located in the merozoite and (ii) when it is released during invasion. Py235, another RBL protein, is found in the rhoptries (18, 20) and may be more accessible to antibody-mediated parasite inhibition than PfRH4. We know that antibody cannot pass the apical junction. If the receptor is released after formation of the junction, then the antibody would have no access. Alternatively, the receptor may form the junction too rapidly after attachment to interact with antibody. Whatever the mechanism, we have learned that PfRH4 will probably not be an effective candidate in vaccine development against P. falciparum.

Our present data with the 3D7 parasite clone show that P. falciparum has redundancy even in the sialic acid-independent pathway of invasion. The redundancy in erythrocyte invasion of Plasmodium has been largely explained by the large repertoire of parasite receptors belonging to the DBL and RBL families. However, alternative invasion pathways may not always include both DBL and RBL members; instead they may include only one or possibly other parasite molecules. There are a number of examples of invasion pathways where DBL proteins are not involved. First, P. knowlesi cannot invade Duffy-negative human erythrocytes but invades Duffy-negative human erythrocytes treated with neuraminidase or trypsin (21). Second, P. vivax invades Saimiri (squirrel monkey) erythrocytes, but the P. vivax Duffy binding protein does not bind to Saimiri erythrocytes (22). Third, in Duffy blood group knockout mice it was reported that invasion of Duffy-negative mature erythrocytes by the P. yoelii 17X YM virulent strain was greatly reduced whereas invasion of the Duffy-negative reticulocytes was not affected (23). These three examples suggest that P. vivax/P. knowlesi are using a Duffy-independent pathway to invade that involves RBL receptors alone or with another yet-to-be-identified parasite molecule. The challenge now is to dissect the molecules involved in invasion by different Plasmodium species and different clones of P. falciparum.

Materials and Methods

P. falciparum Parasites.

P. falciparum clones used in this study were Dd2 (7), Dd2/NM1 (7), 3D7 (24), and HB3 (25). The parasites were grown as described in ref. 27. The identity of each clone was confirmed by microsatellite fingerprinting (26) and by studying the invasion phenotypes of the parasites with neuraminidase-treated erythrocytes (27). Enzymatic treatments of the erythrocytes were done as described in ref. 27.

Expression of the Recombinant PfRH4 Protein (rRH4₃₀).

The amino acid sequence of the PfRH4 protein in the Dd2/NM parasite clone was used to express a 30-kDa fragment of PfRH4 (rRH4₃₀), constituting 261 aa from Asn-328 to Asp-588 (GenBank accession no. AAM 47174). A 783-bp fragment of the PfRh4 gene encoding the 261 aa was amplified by using the following primers: 5′-GAGAGCTAGCAATATTCTTAATGCAGATCCTGATTTAAG-3′ and 5′-GAGAGGATCCTTAGTGATGGTGATGGTGATGCTCTAGATCGTTATAATACATATTAAA A G T A T T A ATTTTTGTATCG-3′. The PCR product was digested with NheI and BamHI (New England Biolabs, Beverly, MA) and inserted downstream of the T7 promoter in the E. coli expression vector, pET-11a (Novagen, San Diego, CA), to obtain the plasmid pRh4-1PET11a. The transcribed sequence of the pRh4-1PET11a insert contains an additional His-tag (LEHHHHHH) at the C terminus. Sequencing of the ligated plasmid confirmed the correct sequence of the PfRh4 gene fragment and that its insertion was also in the correct reading frame. E. coli BL21(DE3) cells (Novagen, San Diego, CA) were transformed with pRh4-1PET11a and used for the expression of rRH4₃₀. Expression was performed in a 1-liter culture, using Luria Bertani medium containing 100 mM ampicillin at 37°C. Once the optical density at 600 nm reached 1.0, the culture was induced by adding isopropyl 1-thio-β-galactopyranoside to a final concentration of 1 mM. Induction was continued for 3 h before harvesting by centrifugation, and the cell pellet was stored at −80°C.

Refolding and Purification of the Recombinant PfRH4 Protein (rRH4₃₀).

The frozen cell pellet was resuspended in 100 ml of 100 mM Tris·HCl, pH 8.0, mixed at 4°C for 1 h, and then lysed at 1.3 × 10⁸ Pa by using a microfluidizer (Microfluidics, Newton, MA). The resulting lysate was mixed with an equal volume of urea wash buffer (10 mM Tris·HCl, pH 8.0/5 mM EDTA/8 M urea/1% Triton X-100) and stirred for 2 h at 4°C. The lysate was centrifuged for 30 min at 25,000 × g, and the rRH4₃₀ protein was detected in the inclusion bodies. The inclusion bodies were resuspended in a solubilization buffer (10 mM Tris·HCl, pH 8.0/8 M guanidine·HCl) and stirred with a magnetic stirrer for 2 h at room temperature. The guanidine-solubilized material was clarified by centrifugation at 25,000 × g for 30 min at 4°C. The denatured supernatant was then refolded by a 30-fold rapid dilution in a refolding buffer (55 mM MES, pH 6.5/1.1 mM EDTA/264 mM NaCl/11 mM KCl/550 mM l-arginine/550 mM guanidine·HCl). The refolding solution was incubated for 24 h at 4°C with continuous stirring, then mixed with an equal volume of 100 mM phosphate buffer, pH 7.4, containing 2.0 M ammonium sulfate and 20% glycerol, for 1 h at 4°C. The mixed solution was then applied to a low substitution phenyl Sepharose hydrophobic interaction column (GE Healthcare, Piscataway, NJ) equilibrated with the binding buffer (50 mM phosphate buffer, pH 7.4, containing 1.0 M ammonium sulfate and 10% glycerol). After sample application, the column was washed with 10 column volumes of 50 mM phosphate buffer, pH 7.4, with 1.0 M ammonium sulfate, and then rRH4₃₀ was eluted with a linear gradient to 100% elution buffer (50 mM phosphate buffer/10% glycerol). The fractions containing rRH4₃₀ after hydrophobic interaction chromatography were pooled and further purified to homogeneity, using a size exclusion gel filtration column (GE Healthcare).

Preparation of Rat and Rabbit Antiserum Against Recombinant PfRH4 (rRH4₃₀).

Rats were immunized intramuscularly with 50 μg of rRH4₃₀ emulsified with complete Freund's adjuvant (Sigma, St. Louis, MO) in the primary immunization on day 0 and followed by subsequent immunizations with incomplete Freund's adjuvant on days 28 and 56. The rat sera were collected on day 70. Rabbits were immunized intramuscularly with 50 μg of rRH4₃₀ emulsified with Montanide ISA 720 (SEPPIC, Fairfield, NJ) on days 0, 28, 56, and 84. The rabbit sera were collected on day 98. Antibody levels against rRH4₃₀ were measured in the sera by a standardized ELISA (28).

Erythrocyte Binding Assays (EBA).

Soluble parasite proteins were obtained from parasite culture supernatants of schizont-infected erythrocytes as described in refs. 13–15. Hot ³⁵S-metabolically labeled and cold unlabeled parasite culture supernatants were obtained from the Dd2/NM and HB3 parasite clones, respectively. Culture supernatant (100 μl) was incubated with 100 μl of packed erythrocytes and incubated at 37°C for 1 h. After incubation, the mixture was centrifuged at 14,000 × g, and the supernatant was removed. The erythrocytes were washed three times with RPMI medium 1640 incomplete medium at pH 6.7, layered on dibutyl phthalate (Sigma), and centrifuged at 14,000 × g for 1 min. The supernatant and oil were removed by aspiration. Bound parasite proteins were eluted from the erythrocytes with 1.5 M NaCl. The cold eluate fractions (HB3 supernatant) were analyzed for the presence of the PfRH4 protein by immunoblotting, using the anti-PfRH4 rat sera. Eluate fractions were loaded on a Nupage 4–12% bis-Tris gel (Invitrogen, Carlsbad, CA) and then transferred onto a nitrocellulose membrane (Invitrogen). The membrane was probed with primary rat anti-PfRH4 sera (1:1,000) followed with the donkey anti-rat secondary antibody (1:20,000) conjugated to HRP (Jackson Immunoresearch, West Grove, PA). Bands were visualized by using an ECL kit (Pierce, Rockford, IL). PfRH4 was detected in the hot eluate samples of Dd2/NM by immunoprecipitation as described in refs. 13–15.

In EBAs with rRH4₃₀, 0.3 μg of protein in 100 μl of PBS at pH 7.4 was incubated with 100 μl of packed erythrocytes at 37°C for 1 h. Thereafter, the EBA was followed as described above. rRH4₃₀ was detected in the eluate by immunoblotting.

Competition Binding Assays.

HB3 culture supernatant (100 μl) was incubated with 100 μl of PBS at pH 7.4, containing increasing amounts of rRH4₃₀, at final concentrations of 0.17, 0.42, 0.85, 1.7, 2.1, 2.5, and 3.4 μM. The mixture was further incubated with 100 μl of packed erythrocytes at 37°C for 1 h. Thereafter, the EBA was followed as described above. The bound PfRH4 native protein was quantified from immunoblots, using ImageJ software (National Institutes of Health).

Invasion Assays Using Target Erythrocytes Bound with rRH4₃₀.

The target erythrocytes (100 μl of packed volume) were incubated with rRH4₃₀ at final concentrations of 0.34, 0.85, 1.7, and 3.4 μM as described above. Invasion assays were done as described in ref. 27. After a 20-h incubation period, thin smears were made and stained with Giemsa. Forty hours after invasion, the parasite-infected erythrocytes were collected, washed, and stained with ethidium bromide. The ethidium bromide stained parasites were measured with a FACScan (Becton Dickinson, Franklin Lakes, NJ) as described in ref. 29. The rate of invasion obtained from the FACScan results were validated by counting ring stage parasites in Giemsa stained smears.

Antibody-Mediated Blocking of Binding and Growth Inhibition Assays.

The IgG fraction of each rat sera was obtained as described in ref. 30. Total IgG from four rats and three rabbits were used to block the erythrocyte binding of the native PfRH4 protein. HB3 culture supernatant (100 μl) was incubated with 100 μl of packed normal erythrocytes in the presence of total IgG at a final concentration ranging from 0.5 to 5.0 μg/μl (rat IgG) and 0.5–15 μg/μl (rabbit IgG) at 37°C for 1 h. Thereafter, the EBA was followed as described above.

In the growth inhibition assays, 6 × 10⁴ schizont-infected erythrocytes were incubated with 2 × 10⁷ target erythrocytes in 100 μl of complete medium containing total IgG from either rat sera at a concentration of 5.0 μg/μl or from rabbits at a concentration of 6.2 μg/μl. Forty hours after invasion, the growth of the parasites was evaluated by biochemical assay of parasite lactate dehydrogenase as described in ref. 30.

Immunofluorescence Localization of PfRH4.

Schizont-enriched parasites were smeared on slides and stored at −80°C. Slides were thawed, acetone-fixed, reacted with the primary antibodies at room temperature for 1 h, and then reacted with the Alexa fluor dye conjugated secondary antibodies (Invitrogen) at room temperature for 1 h. PfRH4 localization was detected with the anti-PfRH4 rat IgG. AMA-1 was detected with the monoclonal antibody, 1G6 (Malaria Vaccine Development Branch, National Institutes of Health), and RON4 was detected with the monoclonal antibody 24C6 (a kind gift from J.-F. Dubremetz, University of Montpellier, Montpellier, France). The differential interference contrast and fluorescent-stained images were collected by using a confocal microscope (Leica, Wetzlar, Germany; catalog no. SP2-AOBS).

Abbreviations

DBL

Duffy binding-like

EBA

erythrocyte binding assay

PfRH

P. falciparum reticulocyte homology

RBL

reticulocyte binding-like.