Conserved residues in the Plasmodium vivax Duffy-binding protein ligand domain are critical for erythrocyte receptor recognition Kelley M. VanBuskirk*, Elitza Sevova, and John H. Adams† Department of Biological Sciences, Center for Tropical Disease Research and Training, University of Notre Dame, Notre Dame, IN 46556 Edited by Louis H. Miller, National Institutes of Health, Rockville, MD, and approved September 22, 2004 (received for review July 27, 2004)
Malaria merozoite invasion of human erythrocytes depends on recognition of specific erythrocyte surface receptors by parasite ligands. Plasmodium vivax merozoite invasion is totally dependent on the recognition of the Duffy blood group antigen by the parasite ligand Duffy-binding protein (DBP). Receptor recognition by P. vivax relies on a cysteine-rich domain, the DBL domain or region II, at the N terminus of the extracellular portion of DBP. The minimal region of the DBP implicated for receptor recognition lies between cysteines 4 and 8 of the DBL domain, which is a region that also has the highest rate of allelic polymorphisms among parasite isolates. We previously found that allelic polymorphisms in this region altered the P. vivax DBL domain antigenic character, which contrasts with changes in receptor specificity attributed to polymorphisms in some homologous ligands of Plasmodium falciparum. To further investigate the relative importance of conserved and polymorphic residues within this DBL central region, we identified residues critical for receptor recognition by site-directed mutagenesis. Seventy-seven surface-predicted residues of the Sal-1 DBL domain were substituted with alanine and assayed for erythrocyte binding activity by expression of the mutant proteins on the surface of transiently transfected COS cells. The functional effect of alanine substitution varied from nil to complete loss of DBL erythrocyte-binding activity. Mutations that caused loss of ligand function mostly occurred in discontinuous clusters of conserved residues, whereas nearly all mutations in polymorphic residues did not affect erythrocyte binding. These data delineate DBL domain residues essential for receptor recognition. malaria 兩 erythrocyte-binding protein
M
alaria caused by Plasmodium vivax is responsible for substantial morbidity in developing countries, accounting for 70–90% of infections acquired in Asia and the Americas as well as 12.4% of infections acquired in Africa (1). Malaria merozoites must attach to and invade new blood cells to begin each cycle of parasite development, making this brief event a critical phase of the parasite’s life cycle. Invasion must occur quickly through a complex, multistep process with a distinct progression of events, involving numerous molecules expressed on the surface of the merozoite and in the apical organelles (2–7). Merozoites of P. vivax and Plasmodium knowlesi are not able to invade human erythrocytes lacking the Duffy blood group antigen because they are unable to proceed through the critical step of junction formation between the merozoite and erythrocyte. The ligands that mediate this are the Duffy-bindinglike (DBL) family of erythrocyte-binding proteins (DBL–EBP), expressed from the erythrocyte-binding-like (ebl) genes (8). The DBL–EBP are type I membrane proteins sequestered in the micronemes of invasive merozoites during late schizont development and then presumably released during the invasion process (9–12). Total reliance of P. vivax merozoite recognition on a single erythrocyte receptor for junction formation appears to be due to its expression of only a single ebl orthologue, the Duffy-binding protein (DBP), and a lack of other functionally equivalent ligands. Multiplicity of DBL–EBPs generates alter-
15754 –15759 兩 PNAS 兩 November 2, 2004 兩 vol. 101 兩 no. 44
native pathways of invasion for Plasmodium falciparum and P. knowlesi. Merozoites of these parasites express erythrocyte binding antigen-175 (EBA175) and BAEBL (EBA140, EBP2) or multiple DBP-like ligands, respectively, which have significant roles in determining host cell specificity and providing alternative invasion pathways for this malaria parasite species (13–23). Erythrocyte receptor recognition by the DBL–EBP depends on a cysteine-rich region known as region II (DBPII) or the DBL ligand domain. This extracellular domain of the DBP consists of 330 aa with 12 cysteines and numerous aromatic residues, suggestive of a domain with a higher-ordered structure (8). Indeed, biochemical analyses of refolded recombinant DBPII revealed substantial helical content (24). Other secondary structure, such as disulfide bonds, is likely important because no free thiols are present in the native domain, indicating that all 12 cysteines are involved in disulfide bonds (25). No data on tertiary structure of any DBL domain are available. Several lines of evidence have identified the central region of the DBL domain, between cysteines 4 and 8, as the segment containing the principal determinants for receptor recognition (25–28). The DBL domains of the P. vivax and P. knowlesi EBPs share ⬇70% identity. This includes the Pk-␣ gene, which encodes a DBP homologue that binds to the Duffy antigen on human and rhesus erythrocytes. The Pk- and Pk-␥ proteins do not bind the Duffy blood group antigen, but instead recognize different receptors on rhesus erythrocytes (28, 29). A series of experiments tested chimeric DBL domain constructs of P. vivax and P. knowlesi EBPs that bind different human and rhesus erythrocyte receptors. Binding of P. vivax DBPII to the human Duffy blood group receptor was localized to a 170-aa segment between cysteines 4 and 7 (28). The functional importance of this central segment of the DBL domain was corroborated by a recent study of deletion constructs of DBL domains from the P. knowlesi EBPs and P. vivax DBP (25). Interestingly, this region approximately corresponds to a minimal recombination unit of the P. knowlesi EBPs, identified as the central of three subregions created by multiple intergenic recombinations (30). These data on the importance of the central DBL domain in determining receptor specificity are supported by experiments that have demonstrated that polymorphisms within the homologous subregion of another member of the P. falciparum DBL–EBP family, BAEBL, result in different erythrocyte-binding phenotypes (17). These changes in receptor profile of variant forms of BAEBL can be mediated by single amino acid changes in the DBL domain. Despite this apparent importance of the central DBL domain in receptor recognition, this region in P. vivax DBP is hyperThis paper was submitted directly (Track II) to the PNAS office. Abbreviations: EBP, erythrocyte-binding protein; DBP, Duffy-binding protein; EBA175, erythrocyte-binding antigen 175; DBPII, DPB region II; DBL, Duffy binding-like. *Present address: Seattle Biomedical Research Institute, 307 Westlake Avenue North, Suite 500, Seattle, WA 98109-5219. †To
whom correspondence should be addressed. E-mail:
[email protected].
© 2004 by The National Academy of Sciences of the USA
www.pnas.org兾cgi兾doi兾10.1073兾pnas.0405421101
Fig. 1. Schematic drawing of the important structural domains of the DBP. There are six separate regions of the DBP extracellular domain as defined by differing degrees of identity to DBL–EBP of other Plasmodium species and a seventh region consisting of the transmembrane and cytoplasmic domains (3). These genes are type I membrane proteins characterized by an N-terminal cysteine-rich ligand domain, the DBL domain (region II), and a second conserved cysteine-rich domain (region VI). Region II (amino acids 206 –530) when expressed on the surface of transfected COS7 will bind Duffy-positive human erythrocytes. The central portion of DBPII between cysteines 4 –7 is the principal site for receptor recognition and also has more polymorphisms than any other part of the ORF.
Methods Selection of Residues for Mutation. Residues from the Cys 4–8 segment of DBPII predicted to be solvent-accessible were selected for mutation to alanine to identify amino acids that are critical to the binding function of region II. Prediction of surface exposure was initially performed by analysis of the P. vivax DBL domain with the Surface Probability algorithm (MACVECTOR 6.5.3, Accelrys). PREDICT PROTEIN (www.predictprotein.org) was also used, and similar predictions were observed across the programs. The algorithm sums values across hexapeptide windows and divides by 6 for an average of surface probability expressed as a fraction for each residue. Residues with a surface probability value of ⱖ0.500 were designated as surfacepredicted; this threshold value was chosen to include all solventaccessible residues likely to be critical to binding function. Selected surface-predicted residues were initially mutated to alanine as sets of adjacent pairs. Single substitution constructs were then created for residues within selected double mutations that affected erythrocyte binding. DBPII-pEGFP Construct. DBPII was cloned into the pEGFP-N1
plasmid with flanking signal sequence from the herpes simplex virus glycoprotein D1 to target expression to the surface of transfected COS cells as a GFP fusion protein (36). Recombinant plasmids were purified by using an endotoxin-free plasmid DNA purification system (Qiagen, Valencia, CA). Primer Design and Site-Directed Mutagenesis. Primers were designed to be 25–45 bp in length and have a GC content of ⱖ40%. When possible, the 3⬘ end of a primer terminated in one or more G or C bases. Target melting temperature (TM) was ⱖ75°C by the following formula: TM ⫽ 81.5 ⫹ 0.41 (%GC) ⫺ 675兾n ⫺ % mismatch, where n ⫽ primer length. Site-directed mutagenesis was conducted according to the QuikSite (Stratagene) protocol. Complementary primer pairs were designed to contain the desired nucleotide changes for chosen mutation sites in dbp. pEGFP-N1兾HSVGD1兾DBP SalI was used as a template. Reaction mixes contained 50–100 ng of template, 125 ng of each primer, 2.5 units of Pfu Turbo polymerase, 1⫻ reaction buffer, and 10 mM dNTP (2.5 mM each). Thermocycling was performed VanBuskirk et al.
on a GeneAmp 9600 for 3 min at 94°C, followed by 16–18 cycles of 94°C for 30 sec, 55°C for 1 min, and 72°C for 12 min 24 sec. Reaction products were digested for 2–4 h with 2.5 units of DpnI, which is specific for methylated DNA. This step selected for newly generated mutant DNA strands by specifically cleaving nonmutant parental template strands. Constructs were sequenced to confirm generation of intended mutations. COS7 Cell Transfection and Erythrocyte-Binding Assay. Recombinant plasmids were purified and COS7 cell cultures were transfected by using described methods (37–39). A Lipofectamine (GIBCO兾 BRL) solution was mixed with plasmid DNA to transfect adherent COS7 cells. The binding assay was performed 42 h after transfection. Human erythrocytes were added to each well in a 10% suspension, incubated at room temperature for 2 h, and washed three times with PBS. Binding was quantified by counting rosettes observed over 30 fields of view at ⫻200 magnification. Transfected COS7 cells with five or more attached erythrocytes were defined as positive rosettes, which is an accepted value for this type of assay (23). In each experiment, three wells of COS7 cells were transfected for each mutation, and the data shown are from at least two separate experiments. Additional experiments were done for a given mutation in cases where there was a discrepancy in the results such that the values of the two experiments were not in the same binding category. Transfection efficiency was assayed by detecting GFP fluorescence of transfected COS7.
Results Prediction of Solvent-Accessible Residues in the DBP Ligand Domain.
The DBP is a membrane protein with a large extracellular domain containing two conserved cysteine-rich regions, DBPII and VI (Fig. 1). Function of DBPII appears to depend on tertiary structure cysteine-rich domains similar to other putative malarial ligands. The principal site for receptor recognition lies between cysteines 4 and 7, which is a region under immune selective pressure, creating frequent polymorphisms (Fig. 1). To identify contact residues in this region of DBPII that are most likely to be responsible for binding the receptor, we designed a mutagenesis PNAS 兩 November 2, 2004 兩 vol. 101 兩 no. 44 兩 15755
MICROBIOLOGY
variable and has a high ratio of nonsynonymous to synonymous mutations (26, 27, 31, 32). These polymorphisms in DBP cluster within the critical erythrocyte-binding segment in a pattern consistent with a high selection pressure on DBP (26, 33, 34). Unlike its P. falciparum paralogues BAEBL and JESEBL, significant differences in erythrocyte-binding specificity were not evident for allelic polymorphisms of P. vivax DBP and P. falciparum EBA175 (27, 35). A separate mechanism appears to operate in P. vivax, where allelic variation functions as a mechanism of immune evasion. In a previous study, we determined that common allelic polymorphisms of the DBP ligand domain alter its antigenic character, conferring resistance to inhibitory antibody without altering erythrocyte-binding properties of the allelic variants (35). However, little is yet known about the residues within the DBL domain that are critical for interacting with the erythrocyte receptor. The aim of this study was to identify contact residues of the DBP ligand domain important for receptor recognition. An in vitro cytoadherence assay expressing P. vivax DBP ligand domains on the surface of transfected COS7 cells was used to identify residues that control erythrocyte-binding properties (29). Our strategy targeted surface-predicted residues of the DBL domain for mutation to alanine to identify amino acids that are directly involved in the interaction with its erythrocyte receptor. Understanding the molecular basis of the parasite ligand–erythrocyte receptor interaction is essential for designing a vaccine that targets essential functional residues of P. vivax DBP.
for each mutant DBP were scored in four categories of erythrocyte binding as insignificant effect or normal (⫹⫹⫹) at ⬎90% relative to wild-type DBPII SalI, partial loss of binding (⫹⫹) at ⬎50–90%, severe disruption of binding (⫹) at ⬎5–50%, or complete abrogation of binding (⫺) at 0–5% (Fig. 4). Identification of DBPII Residues Important for Erythrocyte Receptor Recognition. Seventy-seven surface-predicted residues between
Fig. 2. Surface residues of the ligand domain are expected to be important for recognition and binding of the erythrocyte receptor. We analyzed residues of P. vivax DBPII ligand domain for their potential solvent-accessibility within the native molecule. Approximately two-thirds of the 330-aa ligand were predicted to be solvent-accessible or surface-exposed (underlined). Conserved cysteines are highlighted and numbered. The critical binding domain in the central region, corresponding to region II.2 (8,13), is shaded.
strategy to target surface residues to minimize the probability to disrupt the DBPII tertiary structure. Surface residues were defined as those highly solvent-accessible by surface prediction algorithms. This profile method gave a fractional probability of surface exposure calculated by the ratio of percent exposed to the sum of percents exposed and buried (40), based on values determined for solvent accessibility of amino acids in a set of proteins for which atomic coordinates are known (41). Of 330 residues forming P. vivax DBPII, 202 (61.2%) were predicted to be exposed on the surface of the molecule (Fig. 2). Of 181 aa that form the segment from cysteines 4–8 of DBPII, 106 (58.6%) were predicted to have surface exposure. These surface-predicted residues occurred in patches of varying length that may reflect the purported ␣-helical nature of this ligand domain. The segment between cysteines 7 and 8 was included as an experimental control, because it is not supposed to contribute to receptor recognition (28). It is a region that consists almost exclusively of predicted solvent-accessible residues. Expression of DBPII Mutants on Transfected COS7 Cells. We used a modified version of the standard in vitro assay designed to express the DBL domain as the major part of the HSVgD1 membrane protein on the surface of transiently transfected COS7 cells (29). In our method, the HSVgD1兾DBPII ORF has GFP attached at the C terminus of the HSVgD cytoplasmic domain to provide a direct measure of transfection efficiency by means of fluorescence microscopy in every assay (data not shown and Fig. 4, which is published as supporting information on the PNAS web site). Also, this C-terminal arrangement of GFP allowed full-length expression of alanine-substituted DBPII mutants in transfected cells to be confirmed. Transfection efficiencies and expression levels of DBPII containing sitedirected mutations appeared to be similar in all constructs tested except for two, which were excluded from the results. Cells transfected with these constructs exhibited a diffuse pattern of fluorescence, suggesting that structure disruption or misfolding of the protein resulted in unstable proteins that were rapidly degraded. In an attempt to create a structurally unstable DBP product, conserved cysteine 6 (residue C422) was mutated to alanine, but, interestingly, this mutation did not affect erythrocyte binding. Analysis of the functional impact created by the alanine substitutions was determined by counting the erythrocyte rosettes on transfected COS cells, rosettes being defined as COS cells with 50% or greater cover by bound erythrocytes. Results 15756 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0405421101
DBPII cysteines 4–7 were altered to alanine by site-directed mutagenesis. Our initial series of mutagenesis constructs was done by changing pairs of adjacent amino acids, and those residues associated with altered erythrocyte binding were then individually changed to alanine. Nearly half of all alanine substitutions in this principal region for receptor recognition had no effect on erythrocyte binding (Fig. 3). Residues important for erythrocyte receptor binding occurred in eight discontinuous contact clusters. Each patch of surface-predicted residues between cysteines 4 and 7 contained at least one residue that completely abrogated binding when substituted with alanine. None of the substitutions between cysteines 7 and 8 had any deleterious effect on binding. Twenty residues were identified as essential for erythrocyte binding among the eight contact clusters (Fig. 3). These primary contact residues of the DBL domain were flanked by 22 other less important contact residues, which showed a partial loss of function when substituted with alanine. Thirty-five residues had no negative effect on binding when substituted, and some of these residues actually enhanced binding of the mutant DBP. As expected, changes made in DBL polymorphic residues generally did not affect erythrocyte binding when substituted with alanine. In particular, even substitution of charged, polar polymorphic residues within the hypervariable region (E385, K386, and H390) with the nonpolar alanine did not affect binding. This finding was expected because many of these polymorphisms display a wide range of biochemical classes in DBP from natural isolates. Alanine substitution of only one variant residue altered erythrocyte binding. This residue occurs only as D or N and lies adjacent to the hypervariable region. Biochemical grouping of residues that affected binding were distributed fairly evenly across the three categories of polar charged, polar uncharged, and nonpolar residues (Table 1). Further substitution analysis to create conserved and radical amino acid changes will better elucidate underlying patterns in functional classes at these critical resides. For example, the substitution F344A abrogated binding completely, but no observable change in binding occurred when the phenylalanine residue was substituted with another aromatic amino acid, tyrosine (F344Y, data not shown). This result indicates that the shape or other properties of the side group at this residue may be important to function of the molecule. Discussion The virulent bloodstage growth of malaria parasites depends on the infection of new blood cells during each growth cycle, and parasite ligands play a critical role in merozoite invasion of erythrocytes. DBL–EBPs are a family of erythrocyte-binding proteins that have key functions recognizing erythrocyte surface receptors necessary for invasion of host erythrocytes. Various Plasmodium species use different receptors for erythrocyte invasion, and P. vivax has an absolute requirement for the presence of the Duffy blood group antigen as a surface receptor to invade human erythrocytes. This receptor is recognized by DBP, the only DBL–EBP homologue present in P. vivax. Initially sequestered in the micronemes of mature merozoites, DBP molecules are most likely released onto the merozoite surface early during the invasion process, similar to Toxoplasma gondii microneme proteins, and its interaction with the Duffy receptor then makes possible the junction formation and subsequent VanBuskirk et al.
invasion. This strategy of a just-in-time release of the ligand seemingly minimizes the direct exposure of the functional ligand to antibody inhibition and enhances parasite survival. Because the DBPII is a prime target for vaccine development against P. vivax malaria, we need to identify those contact residues vital for its erythrocyte recognition to optimally develop an effective DBP vaccine. Alanine-scan site-directed mutagenesis identified multiple residues in the P. vivax DBPII ligand domain that are essential or important in receptor recognition. Solvent-accessible Table 1. Summary of alanine-scan mutation data classified by functional group of substituted residues Class
Amino acid
Complete loss
Partial loss
Neutral
Acidic
D E H K R N C S T Y Q F G I L P V W
0 2 0 4 2 2 0 0 0 2 1 2 0 1 2 0 0 2
3 1 0 1 2 5 0 1 2 1 0 0 1 1 1 0 3 0
0 4 0 5 2 2 1 1 3 2 3 1 0 2 4 1 4 0
Basic
Uncharged polar
Nonpolar
Residues that affected binding function of DBPII were distributed across each category of biochemical character. The sums of complete and partial loss of function and of neutral function were 20, 22, and 35, respectively.
VanBuskirk et al.
epitopes of the DBPII were targeted for mutagenesis because these residues are most likely to be surface exposed on the DBL ligand domain and involved in direct interaction with the polypeptide receptor of the Duffy blood group antigen. Each patch of surface-predicted residues within the central part of the DBL domain between cysteines 4 and 7 contained residues that were observed to be important for erythrocyte binding. Mutations that completely abrogated ligand function occurred in eight discontinuous clusters, whereas mutations in f lanking residues resulted in partial loss of function. There was no obvious pattern in the biochemical class of amino acid that was involved in receptor recognition (Table 1). Of the 77 targeted residues, changes in 20 generated complete loss of function, and another 22 partially abrogated erythrocyte binding of the DBL domain (Fig. 3). Somewhat surprisingly, more than half of the residues caused no significant loss of erythrocytebinding activity when substituted, even though population genetic studies have not found these to be variant and alanine was significantly different biochemically than the wild-type amino acid. It is possible that these conserved residues have another important function not directly involved in receptor binding. Residues between cysteines 7 and 8 were predicted to be almost exclusively surface-exposed and not important for receptor binding, confirming a previous prediction (28). Our results emphasize that the central region of the DBPII ligand between cysteines 4 and 7 is critical for receptor recognition (25, 28). There are eight discontinuous clusters within the key binding segment of the DBL domain that are important for interacting with the erythrocyte receptor. The discontinuous nature of these regions indicates that receptor recognition occurs over a broad area of the ligand domain, and that they probably come together in the native conformation to form the receptor-binding site. These mutagenesis data confirm that recognition of the polypeptide receptor on the Duffy blood group antigen takes place within a region of the P. vivax DBP, where the highest rate of polymorphisms in the molecule also occurs (26, 27). Substitution of polymorphic PNAS 兩 November 2, 2004 兩 vol. 101 兩 no. 44 兩 15757
MICROBIOLOGY
Fig. 3. Site-directed mutation analysis of the DBP ligand surface residues showed that conserved residues, but not variant residues, are essential for erythrocyte binding. Residues are shown coded by size proportional to effect on erythrocyte binding relative to wild-type SalI DBL. The largest residues represent alanine substitutions that caused complete loss of function and mutations with partial loss of function decreasing in size proportionally. Residues shown as strikethrough were substituted without loss of function. Shaded residues represent allelic polymorphisms of the DBL domain. Only residues identified as surface-predicted are shown, whereas other residues are indicated only as a nonproportional space.
residues within the hypervariable segment of DBPII generally did not affect erythrocyte-binding activity, with one exception in a residue with restricted variation. This finding is consistent with previous data indicating that naturally occurring variant allelic forms of DBP do not show significant differences in erythrocyte-binding activity. Analyses of DBPII and the homologous P. falciparum EBA175 DBL domains have found that the naturally occurring polymorphisms do not alter erythrocyte-binding properties of these ligands (27, 42). Instead, there is considerable evidence that acquired immunity is a driving force that selects for amino acid variants at certain residues within the DBL ligand of the P. vivax DBP (32, 34). Consistent with this concept, we have found that allelic polymorphisms can alter the antigenic character of the DBL domain, increasing refractoriness to antibody inhibition of ligand-erythrocyte binding (35). In contrast to this lack of effect on receptor recognition for polymorphisms in DBP, each allelic mutation in the DBL ligand domains of the P. falciparum paralogues BAEBL and JESEBL altered their receptor recognition properties (17, 23). BAEBL and JESEBL are members of the DBL–EBP family in P. falciparum with potentially important roles in determining specificity of invaded erythrocytes (15, 17–19, 21, 22). The functional effects of polymorphisms for BAEBL and JESEBL contrast sharply with those observed for DBP and EBA175, because each single amino acid change in the DBL domain of BAEBL and JESEBL leads to recognition of different erythrocyte surface receptors. Because there are numerous differences between the P. falciparum DBL–EBPs and P. vivax DBP, it is not possible to directly compare or correlate positional effects of these polymorphisms. Still, we can conclude that there are two mechanisms providing selective pressure to establish and maintain polymorphisms of the ligand domains of the DBL–EBP family. One type of selection exemplified by BAEBL and JESEBL occurs when random point mutations alter receptor specificity for another erythrocyte surface molecule, which permits use of alternative invasion pathways when the primary receptor is not available. Such a mechanism would seem particularly advantageous for highly virulent parasite species such as P. falciparum that have a strong selective pressure on humans for genetic traits that impart a survival advantage against this much more lethal form of malaria. From
this, we can speculate that some multigene families in more virulent malaria parasites evolve as a mechanism to overcome genetic resistance that emerges in humans. Another type of selective pressure creating parasite polymorphisms occurs through antibody inhibition of their erythrocyte-binding function. Such immune-selective pressure occurs when changes in key antigenic determinants allow escape from a neutralizing acquired immunity. This type of variation appears to be exemplified by the P. vivax DBP and P. falciparum EBA175. Our site-directed mutagenesis results indicate that the variant residues of DBP are predominantly not contact residues directly involved in binding the receptor, but f lank these functionally important residues. The observed pattern of DBP polymorphisms is consistent with antibody inhibition having primarily a steric hindrance on receptor recognition. The one exception we observed was in a residue that has a very restricted polymorphism, occurring either as a charged (D) or polar (N) amino acid. Because of its limited variation and because the residue lies adjacent to the hypervariable region (26), we believe that it is involved in receptor recognition and at least partially comprises an antibody-accessible neutralizing epitope. Such an occurrence is consistent with a receptorbinding site that is an open groove, which is expected for binding to a polypeptide receptor. Further work remains to be done to determine the effect of alanine substitution on variant residues outside of the hypervariable segment of DBPII. They are not expected to be contact residues directly involved in DBPII–erythrocyte binding, but the evidence suggests that the polymorphisms are driven by host immune selection. Analysis is also needed to more accurately measure quantitative changes in binding affinity of mutants relative to wt DBPII. These data will enable us to pinpoint residues that mediate the differences in receptor specificity seen between P. vivax and other DBL–EBPs. Such information is important for the design of an anti-DBP vaccine to elicit an immune response to residues that are essential for function of P. vivax DBPII.
1. Mendis, K., Sina, B. J., Marchesini, P. & Carter, R. (2001) Am. J. Trop. Med. Hyg. 64, 97–106. 2. Bannister, L. H. & Dluzewski, A. R. (1990) Blood Cells 16, 257–292. 3. Barnwell, J. W. & Galinski, M. R. (1995) Ann. Trop. Med. Parasitol. 89, 113–120. 4. Dvorak, J. A., Miller, L. H., Whitehouse, W. C. & Shiroishi, T. (1975) Science 187, 748–750. 5. Perkins, M. E. (1989) Exp. Parasitol. 69, 94–99. 6. Ward, G. E., Chitnis, C. E. & Miller, L. H. (1994) in Strategies for Intracellular Survival of Microbes, ed. Russell, D. G. (Bailliere Tindall, London), Vol. 1, pp. 155–187. 7. Guarda, J. A., Asayag, C. R. & Witzig, R. (1999) Emerg. Infect. Dis. 5, 209–215. 8. Adams, J. H., Sim, B. K. L., Dolan, S. A., Fang, X., Kaslow, D. C. & Miller, L. H. (1992) Proc. Natl. Acad. Sci. USA 89, 7085–7089. 9. Haynes, J. D., Dalton, J. P., Klotz, F. W., McGinniss, M. H., Hadley, T. J., Hudson, D. E. & Miller, L. H. (1988) J. Exp. Med. 167, 1873–1881. 10. Wertheimer, S. P. & Barnwell, J. W. (1989) Exp. Parasitol. 69, 340–350. 11. Adams, J. H., Hudson, D. E., Torii, M., Ward, G. E., Wellems, T. E., Aikawa, M. & Miller, L. H. (1990) Cell 63, 141–153. 12. Fang, X. D., Kaslow, D. C., Adams, J. H. & Miller, L. H. (1991) Mol. Biochem. Parasitol. 44, 125–132. 13. Miller, L. H., Hudson, D. & Haynes, J. D. (1988) Mol. Biochem. Parasitol. 31, 217–222. 14. Adams, J. H., Kaneko, O., Blair, P. L. & Peterson, D. S. (2001) Trends Parasitol. 17, 297–299. 15. Narum, D. L., Fuhrmann, S. R., Luu, T. & Sim, B. K. (2002) Mol. Biochem. Parasitol. 119, 159–168. 16. Michon, P., Stevens, J. R., Kaneko, O. & Adams, J. H. (2002) Mol. Biol. Evol. 19, 1128–1142.
17. Mayer, D. C., Mu, J. B., Feng, X., Su, X. Z. & Miller, L. H. (2002) J. Exp. Med. 196, 1523–1528. 18. Thompson, J. K., Triglia, T., Reed, M. B. & Cowman, A. F. (2001) Mol. Microbiol. 41, 47–58. 19. Mayer, D. C., Kaneko, O., Hudson-Taylor, D. E., Reid, M. E. & Miller, L. H. (2001) Proc. Natl. Acad. Sci. USA 98, 5222–5227. 20. Gilberger, T. W., Thompson, J. K., Triglia, T., Good, R. T., Duraisingh, M. T. & Cowman, A. F. (2003) J. Biol. Chem. 278, 14480–14486. 21. Maier, A. G., Duraisingh, M. T., Reeder, J. C., Patel, S. S., Kazura, J. W., Zimmerman, P. A. & Cowman, A. F. (2003) Nat. Med. 9, 87–92. 22. Lobo, C. A., Rodriguez, M., Reid, M. & Lustigman, S. (2003) Blood 101, 4628–4631. 23. Mayer, D. C., Mu, J. B., Kaneko, O., Duan, J., Su, X. Z. & Miller, L. H. (2004) Proc. Natl. Acad. Sci. USA 101, 2518–2523. 24. Singh, S., Pandey, K., Chattopadhayay, R., Yazdani, S. S., Lynn, A., Bharadwaj, A., Ranjan, A. & Chitnis, C. (2001) J. Biol. Chem. 276, 17111–17116. 25. Singh, S. K., Singh, A. P., Pandey, S., Yazdani, S. S., Chitnis, C. E. & Sharma, A. (2003) Biochem. J. 374, 193–198. 26. Tsuboi, T., Kappe, S. H., al-Yaman, F., Prickett, M. D., Alpers, M. & Adams, J. H. (1994) Infect. Immun. 62, 5581–5586. 27. Xainli, J., Adams, J. H. & King, C. L. (2000) Mol. Biochem. Parasitol. 111, 253–260. 28. Ranjan, A. & Chitnis, C. E. (1999) Proc. Natl. Acad. Sci. USA 96, 14067–14072. 29. Chitnis, C. E. & Miller, L. H. (1994) J. Exp. Med. 180, 497–506. 30. Prickett, M. D., Smarz, T. R. & Adams, J. H. (1994) Mol. Biochem. Parasitol. 63, 37–48. 31. Cole-Tobian, J. L., Cortes, A., Baisor, M., Kastens, W., Xainli, J., Bockarie, M., Adams, J. H. & King, C. L. (2002) J. Infect. Dis. 186, 531–539.
15758 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0405421101
We thank Neil Lobo and Breandon Kennedy for helpful suggestions on mutagenesis methods and Doug Shoue for technical assistance. This work was supported by National Institutes of Health Grant R01 AI33656. K.M.V. has been supported by a Clare Boothe Luce Fellowship.
VanBuskirk et al.
37. Fraser, T. S., Kappe, S. H., Narum, D. L., VanBuskirk, K. M. & Adams, J. H. (2001) Mol. Biochem. Parasitol. 117, 49–59. 38. Michon, P. A., Arevalo-Herrera, M., Fraser, T., Herrera, S. & Adams, J. H. (1998) Am. J. Trop. Med. Hyg. 59, 597–599. 39. Michon, P., Woolley, I., Wood, E. M., Kastens, W., Zimmerman, P. A. & Adams, J. H. (2001) FEBS Lett. 495, 111–114. 40. Emini, E. A., Hughes, J. V., Perlow, D. S. & Boger, J. (1985) J. Virol. 55, 836–839. 41. Janin, J. & Wodak, S. (1978) J. Mol. Biol. 125, 357–386. 42. Liang, H. & Sim, B. K. L. (1997) Mol. Biochem. Parasitol. 84, 241–245.
MICROBIOLOGY
32. Baum, J., Thomas, A. W. & Conway, D. J. (2003) Genetics 163, 1327– 1336. 33. Conway, D. J., Roper, C., Oduola, A. M., Arnot, D. E., Kremsner, P. G., Grobusch, M. P., Curtis, C. F. & Greenwood, B. M. (1999) Proc. Natl. Acad. Sci. USA 96, 4506–4511. 34. Cole-Tobian, J. & King, C. L. (2003) Mol. Biochem. Parasitol. 127, 121–132. 35. VanBuskirk, K. M., Cole-Tobian, J., Baisor, M., Sevova, E. S., Bockarie, M., King, C. L. & Adams, J. H. (2004) J. Infect. Dis., 190, 1556–1562. 36. Michon, P., Fraser, T. & Adams, J. H. (2000) Infect. Immun. 68, 3164 – 3171.
VanBuskirk et al.
PNAS 兩 November 2, 2004 兩 vol. 101 兩 no. 44 兩 15759
