The binding of erythrocytes infected with mature blood stage parasites to the vascular bed is key to the pathogenesis of malignant malaria. The binding is mediated by members of Plasmodium falciparum erythrocyte membrane protein 1 (PfEMP1) family. PfEMP1s can be divided into groups, and it has previously been suggested that parasites expressing group A or B/A PfEMP1s are most pathogenic. To test the hypothesis that the first malaria infections in infants and young children are dominated by parasites expressing A and B/A PfEMP1s, we measured the plasma Ab level against 48 recombinant PfEMP1 domains of different groupings in 1342 individuals living in five African villages characterized by markedly different malaria transmission. We show that children progressively acquire a broader repertoire of anti-PfEMP1 Abs, but that the rate of expansion is governed by transmission intensity. However, independently of transmission intensity, Abs are first acquired to particular duffy binding ligand-like domains belonging to group A or B/A PfEMP1s. The results support the view that anti-PfEMP1 Ab responses effectively structure the expenditure of the repertoire of PfEMP1 maintained by the parasite. Parasites expressing certain group A and B/A PfEMP1s are responded to first by individuals with limited previous exposure, and the resulting Abs reduce the fitness and pathogenicity of these parasites during subsequent infections. This allows parasites expressing less pathogenic PFEMP1s to dominate during later infections. The identification of PfEMP1 domains expressed by parasites causing disease in infants and young children is important for development of vaccines protecting against severe malaria.
In areas of stable Plasmodium falciparum transmission, individuals develop immunity to malaria (1). The pace at which protection is acquired depends on the intensity of malaria transmission, but in general terms the acquisition of immunity is characterized by several phases. Relatively rapidly, toddlers or young children first develop protection against severe noncerebral malaria (2), and subsequently over a period of years children acquire the ability to control the multiplication of asexual blood stage parasites and suffer fewer and fewer attacks of uncomplicated febrile malaria episodes (3).
P. falciparum erythrocytes membrane protein 1 (PfEMP1)3 is a parasite encoded protein expressed on the surface of infected erythrocytes (4). The protein mediates binding between the infected erythrocytes and receptors on the vascular lining and allows parasites to sequester in the peripheral circulation (5). Sequestration allows parasites to avoid passage through the spleen, where they are destroyed. PfEMP1s are large polymorphic multidomain proteins encoded by var genes (6). Each parasite genome contains ∼60 var genes (7), but by a not fully understood process of allelic exclusion, a parasite will only transcribe one var gene and express one variant of PfEMP1 at a time (8). Furthermore, the vast majority of the progeny of a dividing parasite will express the same PfEMP1 as the mother cell (9).
PfEMP1 molecules can be divided into different groups on the basis of the position of the encoding gene on the chromosomes, their 5′ upstream regions, and domain structure (10, 11, 12). Similarly, the two main PfEMP1 domain types, cysteine-rich interdomain region (CIDR) and duffy binding-like domain (DBL), can be divided into groups and subgroups on the basis of sequence analyses (13). All parasite genomes studied contain representatives of all main PfEMP1 groups, and the distributions of genes between the different groupings seem to be similar (14). Parasites expressing different PfEMP1s differ in phenotype. They bind different receptors on the vascular lining (differ in adhesion phenotype) and are recognized by human IgG in different ways (serological phenotype). An example of this are parasites expressing the relatively conserved PfEMP1, VAR2CSA, which binds to chondroitin sulfate A in the placenta (15) and to which high-titered Abs only are present in women who have suffered a placental infection (16). Similarly, in malaria endemic areas, parasites isolated from young children suffering from severe disease tend to express PfEMP1 molecules to which Abs are acquired early in life, and parasites isolated from older children tend to express PfEMP1s to which Abs are developed later in life (17, 18). These findings strongly suggest that parasites expressing certain PfEMP1 types have a selective advantage over parasites expressing other types in the naive host, but that this selective advantage is lost as the host develops adhesion-blocking Abs against the most pathogenic types. The Abs are thought to prevent sequestration, and erythrocytes infected with parasites expressing PfEMP1 targeted by these Abs will be destroyed in the spleen. Several studies using transcriptional analyses or in vitro manipulation have indicated that PfEMP1 belonging to groups A and B/A are expressed by parasites causing severe disease early in life, whereas the less severe disease episodes in older children are linked to groups B and C (19, 20, 21, 22). In this study, we have performed an extensive analysis on the acquisition of IgG Abs to 48 PfEMP1 domains by children living under markedly different intensity of malaria transmission. We show that children gradually acquire a broader repertoire of anti-PfEMP1 Abs, that the pace at which the Abs are acquired is governed by the transmission intensity, and that independent of malaria endemicity Abs are first acquired to DBL domains expressed by PfEMP1 molecules belonging to group A or B/A.
Materials and Methods
The study was conducted in the Tanga region in north-eastern Tanzania. This area is characterized by marked variations in intensity of P. falciparum transmission related to variations in altitude (23). Five study villages located within short geographical distances but at different altitudes and thus with different malaria transmission intensity were included in the study. Mgome and Mkokola villages are at low altitude (200–300 m), Ubiri and Kwamasimba villages are at intermediate altitude (1000–1200 m), and Magamba village is at high altitude (1600–1700 m). In Mgome, Ubiri, and Magamba villages, venous blood samples were collected from cohorts of ∼250 healthy individuals between 0 and 19 years randomly recruited at cross-sectional surveys in April 2001(24). The blood samples from Kwamasamba and Mkokola were collected during cross-sectional surveys in March 2004 from 320 individuals aged 0–59 years (25). After receiving written consent from the study participants or from their parents/guardians, a venous blood sample was collected. Plasma was stored at −20°C until analysis. A thin film blood slide was also prepared, stained with Giemsa, and investigated for the presence of malaria parasites. At the time of blood sampling the point prevalence of parasitemia among study participants aged 2.0–3.9 years was 91, 84, 51, 17, and 4% in Mgome, Mkokola, Ubiri, Kwamasimba, and Magamba, respectively.
The study protocols were approved by the Ethical Committee of the National Institute for Medical Research and Ministry of Health, Dar es Salaam, Tanzania.
Protein expression was performed as described previously (19, 26). Primer pairs designed to contain restriction enzyme sites (supplemental data)4 were used to amplify DBL and CIDR domain-encoding fragments from 3D7 genomic DNA. The digested PCR products were cloned into the baculovirus vector pAcGP67-A (BD Biosciences), which was designed to contain the V5 epitope upstream of a histidine tag in the C-terminal end of the construct. The identity of the cloned fragments was verified by sequencing. Linearized Bakpak6 baculovirus DNA (BD Biosciences/Clontech) was cotransfected with pAcGP67-A into Sf9 insect cells for generation of recombinant virus particles, and histidine-tagged proteins secreted into the supernatant of infected High-Five insect cells were purified on Co2+ metal-chelate agarose columns. Eluted products were dialyzed overnight in PBS and verified by SDS-PAGE and Western blotting using anti-V5 Abs. Rabbit Abs for each protein domain were raised and tested in ELISA for reactivity against the immunizing Ag. All rabbit Abs showed high reactivity to their respective domains.
Covalent coupling of recombinant PfEMP1 proteins to microspheres
Carboxylated Luminex microspheres were covalently coated with the different PfEMP1 protein domains through an interaction of their carboxyl groups and the amino groups on the proteins following the procedure suggested by the manufacturer. Microspheres (1.25 × 107 microspheres/ml) were brought to room temperature, vortexed for 1 min, and transferred to Eppendorf tubes. The supernatant was removed after centrifugation for 1 min at 16,000 × g. One milliliter of distilled water was added to the microspheres, vortexed for resuspension, followed by centrifugation for 1 min at 16,000 × g. The microspheres were sonicated in a water bath sonicator into suspension and centrifuged for 1 min at 16,000 × g. The supernatant was removed and 1 ml of activation buffer (0.1 M NaH2PO4 (pH 6.2)) was added to the pellet and vortexed for resuspension. In separate tubes 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide hydrochloride (EDC) and N-hydroxysulfosuccinimide (Sulfo-NHS; Pierce Biotechnology) were reconstituted to 50 mg/ml, and 125 μl of each was added to the microspheres, vortexed, and incubated at room temperature for 20 min with inversions in the dark. The microspheres were centrifuged for 1 min at 16,000 × g, resuspended in 1 ml of 50 mM MES (pH 5.0), centrifuged for 1 min at 16,000 × g, and the supernatant was removed. The MES wash was repeated. The microspheres were resuspended in 500 μl of MES. In separate tubes, the different protein samples (100 μg per ml of microspheres) were mixed with MES to a final volume of 500 μl and each was added to a separate microsphere population and incubated at room temperature for 2 h in the dark with inversions. The microspheres were centrifuged for 1 min at 16,000 × g and the supernatant was removed. The microspheres were washed twice in 1 ml of PBS/TBN (0.02% Tween 20, 0.1% BSA, and 0.05% sodium azide in PBS (pH 7.4)). The microspheres were resuspended in 1 ml of PBS/TBN and stored at 4°C in the dark. To determine whether coupling was effective, aliquots of the different microsphere sets were prepared for analysis as described below and analyzed on the Luminex instrument.
Multiplexing and lyophilization of microspheres
Equal volumes of the coated microspheres were pooled together and mixed by vortexing. Sucrose and Tween 20 were added to 3% and 0.05%, respectively, mixed by vortexing, and single-use aliquots were lyophilized (AdVantage, Wizard 2.0; VirTis) in polypropylene vials, sealed under nitrogen gas, and stored at −80°C. Immediately before use, lyophilized microspheres were reconstituted with distilled water and used for analysis (27) as described below.
Analysis of coupled microspheres on the Luminex
The coated microspheres were diluted 1/333 in assay buffer E (ABE buffer, 0.1% BSA, 0.05% Tween 20, 0.05% sodium azide in PBS (pH 7.4)) and 50-μl aliquots were dispensed into the wells of a 1.2-μm filter bottom 96-well microtiter plate (MSBVS 1210; Millipore) pre-wetted with ABE buffer. The microspheres in 96-well plates were washed three times with ABE using a vacuum manifold (Millipore). Frozen plasma samples were thawed at room temperature, mixed by vortexing, and spun at 16,000 × g for 5 min to remove particulates. Plasma samples were diluted 1/80 in ABE buffer and 50-μl aliquots of diluted sample were added to the microspheres and incubated in the dark on a shaking platform at 1100 rpm for 30 s followed by 300 rpm for 30 min. Excess Ab was removed using a vacuum manifold followed by three washes in ABE. Twenty-five microliters of biotinylated human IgG (Sigma-Aldrich) detection Ab diluted 1/500 in ABE was added to the microspheres, incubated in the dark with shaking at 1100 rpm for 30 s, followed by 300 rpm for 30 min and washed three times in ABE. Fifty microliters of streptavidin-PE (Sigma-Aldrich) diluted 1/500 in ABE was added to the microspheres and incubated in the dark with shaking at 1100 rpm for 30 s, followed by 300 rpm for 10 min. Excess streptavidin-PE was removed followed by three washes in ABE. The microspheres were then resuspended in 125 μl of ABE and analyzed on the Luminex instrument. The reader was set to read a minimum of 100 microspheres per microsphere region and results were expressed as median fluorescent intensity.
Demographic data was double entered into Epi Info and exported to Stata. The Ab levels measured as fluorescence emission from the PfEMP1 domain-coupled microspheres were collected in Excel and exported to Stata. The study participants were categorized as responders if the measured Ab level was higher than the mean ± 2 SD of 20 Danish controls. To evaluate the effect of age and village of residence on Ab recognition, linear regression was used to model the association between the number of domains to which an individual had Abs (number of domains recognized by IgG Abs) and the age (included as the square root of age measured in years) and village of residence of the study participants (included as a variable numbered 1–5, where the village of lowest endemicity was assigned the value 1, and the village of the highest endemicity was assigned the value 5). ANOVA was used to compare the number of domains recognized by individuals in specific age groups residing in different villages. Paired t test was used to compare the percentage of CIDR and the percentage of DBL domains recognized in each of the study participants. Spearman rank sum test was used to evaluate the association between the recognition of domains between the different villages; p values were Bonferroni adjusted. Mann-Whitney rank sum test was used to compare the percentage of responders to group A and B/A domains vs the percentage of responders to group B, C, and B/C domains. Stata software was used for the statistical analyses.
The anti-PfEMP1 Ab repertoire is acquired by age at a pace governed by the intensity of transmission
Ab levels to 48 (16 CIDR and 32 DBL) recombinant PfEMP1 domains were measured in plasma collected from 1342 individuals living in five African villages characterized by markedly different malaria transmission intensity. The likelihood of responding to a given DBL or CIDR domain increased with age in the four villages characterized by medium or high malaria transmission (p < 10−4, linear regression), but not in the village situated at 1700 m altitude and characterized by very low malaria transmission (Fig. 1⇓). The number of PfEMP1 domains recognized by an individual plasma was statistically significantly associated with the malaria endemicity in the study participants home village (p < 10−4 for both CIDR and DBL recognition, linear regression in a model correcting for age). In the village with the highest point prevalence rate (Mgome, malaria parasite point prevalence (PP) among children of 2.0–3.9 years of 91%), children aged 2–3 years responded to >50% and 25% of the CIDR and DBL domains, respectively. A corresponding level of reactivity was not acquired until the age of 10–15 years in the medium transmission village Kwamasimba (PP of 17%). Although Abs were acquired most rapidly in the village with highest transmission (Mgome), the breadth of the Ab repertoire seemed to be sealed. In the three villages with highest transmission (Mgome, Mkokola, Ubiri), ∼20–25% of the DBL domains and 50–60% of the CIDR domains were recognized by the children aged 10–15 years, and these percentages did not increase further in the older age groups. Furthermore, in the age group >15 years there was no significant difference between the likelihood of responding to CIDR or DBL domains between individuals living in the three villages with highest transmission (p > 0.05, ANOVA).
The fluorescence values obtained with each of the 48 PfEMP1 domains was added for each individual to estimate the total level of domain-specific Abs. This consolidated measure of anti-PfEMP1 Ab level (Fig. 2⇓) was also found to reflect subject age and village of residence (p < 10−4 for both CIDR and DBL domains, linear regression).
PfEMP1 Abs are acquired in a specific sequence independent of transmission intensity
There was a marked difference between the prevalence of Abs to the different domains. In general, CIDR domains were more likely to be recognized than DBL domains (p < 10−5, paired t test), and this difference was most pronounced in the older age groups.
The mean number of responders in each village was calculated for each domain, and for each village each domain was ranked relative to the other domains based on the percentage of individuals, who had a measurable IgG response. The ranking of the domains was then compared and it was found that there was a strong correlation between the domain ranking obtained in each village (Ks between 0.84 and 0.98, p < 10−4 for all comparisons, Spearman’s rank correlation test).
IgG against group A or B/A DBL domains are more prevalent than IgG against domains from group B or C PfEMP1s
The DBL and CIDR domains were ranked according to how many of the donors had a measurable Ab response to the domain (Tables I⇓ and II⇓). Among the 32 DBL domains tested, of the 10 best recognized domains, 9 belonged to group A or group B/A, and among the 10 least well-recognized domains 6 belonged to group B, B/C, or C, and when considering all of these domains there was a significantly better recognition of group A and B/A domains than B, B/C, and C domains (p < 0.05, Mann-Whitney’s rank sum test). The relationship was not absolute since several group A and group B/A domains were poorly recognized. Neither the position of the DBL domain within the native PfEMP1 nor the type predicted how well a domain was recognized (Table I⇓). The domains representing VAR1 and VAR3, coded by var genes conserved between genomes, were recognized by between 6 and 21% of the study participants (Table I⇓).
Abs against DBL domains are acquired in sequence and first to domains belonging to group A or group B/A
The acquisition of Ab to the individual domains by age was evaluated by comparing the percentage of responders in different age groups. For these analyses villages were divided into those with high transmission (Mgome, PP of 91%; Mkokola, PP of 84%) and those with medium transmission (Ubiri, PP of 51%; Kwamasimba, PP 17%). In the high transmission villages (Fig. 3⇓A), an Ab response to the best recognized DBL domain (domain 20, DBL2γ of PF11_0008) was acquired early, and at the age of 2–3 years 94% of the children had a measurable Ab response to this domain. Children then acquired Abs to a group of nine domains, which were recognized by between 13 and 25% of the 1- to 2-year-old infants. These nine domains included seven domains from group A or B/A PfEMP1 proteins and two from group C. Perhaps as a result of less intense exposure, the pattern of Ab acquisition was less complex in the medium transmission villages (Fig. 3⇓B). Here, three domains were recognized by >10% of the newborns (0–6 mo of age), and these Abs, which probably were of maternal origin, could not be detected in the infants aged 6 mo to 1 year, but thereafter the children steadily acquired a broader repertoire of IgG Abs to DBL domains. Interestingly, Abs were first acquired to A or B/A domains. At the age of 3–4 years, nine domains were recognized by >10% of the children, and these all belonged to group A or B/A PfEMPs. As for the children in the high transmission villages, responses to domain 20 were most prevalent, but the responses to domains 49 (DBL5ε of MAL6P1.4), 71 (DBL6ε of MAL6P1.4), and 69 (DBL2β of PFL0020w) were also acquired early in life. The best recognized DBL domains representing group B or group C PfEMP1s were 115 (DBL2δ of PFD1000c) and 105 (DBL2δ of PFD0630c), which were recognized by 10% of the children aged 4 to 5 years.
As with anti-DBL Abs, IgG against CIDR domains was detected in the newborns, and the percentages of responders were lower in the infants aged 6 mo to 1 year (Fig. 4⇓). After this age children acquired a broader and broader Ab repertoire against CIDR domains. Although the best recognized CIDR domain (CIDR1α of PF11_0521) represented a group A PfEMP1, there was no indication that domains belonging to group A or B/A molecules were better recognized by younger children than the CIDR domains belonging to group B or C. In fact, six group B or C CIDR domains were recognized by >10% of the children in the medium transmission villages at the age of 1–2 years, and it was only in the children aged 2–3 years of age that the first group B/A domain was recognized by >10% of the children (Fig. 4⇓B).
Until recently, most studies characterizing the Ab response to variable surface Ags (VSA) on erythrocytes infected with P. falciparum were performed by incubating infected erythrocytes in plasma from malaria-exposed donors and measuring binding by flow cytometry (28) or agglutination assays (29). PfEMP1 is the dominant Ab target in these assays (16, 19), but other VSA are also present on infected erythrocytes (30). Using parasite-based assays it has previously been shown that the repertoire of anti-VSA Abs increases with age (18, 31, 32) and that these Abs are important mediators of immunity acquired by individuals living in malaria endemic areas (33, 34, 35, 36). Furthermore, parasites from individuals with a narrow repertoire of anti-VSA Abs tend to express VSA types to which children in endemic areas develop Abs early in life, whereas parasites infecting individuals with a broad anti-VSA IgG repertoire express VSA types to which Abs develop later in life (18, 37). Taken together, these data imply that VSA expression is not random. In the absence of an acquired immune response, some VSA types are preferably expressed, but the presence of an anti-VSA Ab response forces the parasites into expression of other types. The simplest explanation for this phenomenon is based on an assumption that parasites expressing different PfEMP1 molecules have different effective growth rates. The growth rate of a parasite with a particular phenotype is governed by the intracellular growth rate, the merozoite reinvasion success rate, and how effectively the parasite sequesters and avoids splenic killing. Since there is little to suggest that the division inside the erythrocytes or the fitness for reinvasion rate differs markedly between parasites expressing different PfEMP1s, it can be assumed that the effective growth rate of parasites of different phenotype is governed by how effectively the expressed PfEMP1 is able to sequester the parasite in the peripheral circulation. Parasites released from the liver probably express a range of different PfEMP1s, but the infection rapidly becomes dominated by parasites expressing the PfEMP1 types that mediate the most effective sequestration and confer the highest effective growth rate (38). In endemic areas individuals would develop Abs against these PfEMP1 types first. To test this hypothesis and to identify PfEMP1 molecules linked to early infections, we produced a large number of recombinant PfEMP1 domains and used a high-throughput assay to monitor the acquisition of the anti-PfEMP1 Ab repertoire among children living under different malaria transmission. Our results demonstrate that children living in areas of high malaria transmission develop a broad repertoire of Abs at an earlier age than do children living under lower transmission intensity. Abs to CIDR domains were developed faster than the Abs to DBL domains, and this probably reflects that there is a broader cross-reactivity between CIDR domains than between DBL domains. This notion is supported by previous findings using competition ELISA indicating that different recombinant CIDR domains based on 3D7 sequence could out-compete each other, but that there was little cross-reactivity between 3D7 DBL domains (39). None of the children in this study was likely to have been infected with 3D7, and yet many developed Abs reacting with recombinant DBL domains based on this genome. Thus, there must be a widespread serological cross-reactivity between DBL domains coded from different genomes. The prevalence of Abs to the different domains varied dramatically. Interestingly, there was a strong association between how well domains were recognized in the different villages. This structuring has been noted before (40), and it indicates that the forces driving the acquisition of Abs at the different sites are similar, although the acquisition rates vary. DBL domains belonging to group A and B/A PfEMP1 molecules were in general better recognized than domains belonging to groups B and C PfEMP1. More importantly, we could identify a set of DBL domains to which Abs consistently were acquired early in life before Abs had been developed to other DBL domains. These DBL domains belonged to group A or B/A PfEMP1 molecules, supporting earlier findings linking expression of groups A and B/A PfEMP1 to severe malaria infections in young children (19, 20, 21, 22).
To understand the biology of PfEMP1 it is necessary to understand how the PfEMP1 repertoire is structured and how it evolves. The major PfEMP1 groups can be identified in Plasmodium reichenovii 14, a species that diverged from P. falciparum ∼5 million years ago, and this indicates that the PfEMP1 repertoire evolves no randomly and slowly, at a pace very different from that in viruses like HIV or influenza. The most conserved PfEMP1s are VAR1, VAR2CSA, and VAR3. The genes coding these proteins are found in all or most P. falciparum parasites, and stretches of both 5′ and 3′ base sequence allow identification of these genes. VAR2CSA is involved in the pathogenesis of placental malaria (15), whereas the function of VAR1 and VAR3 remain unknown. Indeed, it is uncertain whether the transcription of these genes results in protein expression. In this study, we found that a large proportion of individuals had Abs against recombinant proteins representing VAR1 and VAR3. These could have been cross-reactive Abs induced by other PfEMP1, but the fact that three of the four VAR1 domains were recognized by almost the same percentage of individuals and that the recognition of these domains in an individual was tightly associated suggest that these genes are transcribed and translated into protein in some parasites. We have previously proposed that the gene PFD1235w belong to a conserved var gene family and named it var4 (19). Homologs of var4 cannot be identified in all parasites (14), but large stretches of sequence highly similar to the 5′ region of PFD1235w and coding the first four domains can be identified in some parasites (19). The presence of anti-VAR4 Abs has also been associated with protection against malaria anemia (25). In this study the DBL4γ domain of the 3D7 VAR4, which is not located in the part of VAR4 that is conserved between parasites, was recognized by 30% of the donors. Other domains were recognized to a higher degree. These included the DBL2γ domain of PF11_0008 and the DBL5ε of MAL6P1.4, which were recognized by many individuals and perhaps more interestingly also very often were the first domains that individuals developed anti-DBL Abs against. This indicates that parasites expressing PfEMP1 molecules, which induce Abs against these domains, have a selective advantage in infants and young children. The DBL2γ domain of PF11_0008 was the best recognized domain, and after the first year of life >60% of the children in the high transmission villages had acquired Abs against this domain. We have previously shown that the presence of Abs to PF11_0008 was associated with protection against malaria fevers (41). In the age group 2–3 years, DBL5ε of MAL6P1.4 was recognized by about half of the children in the high transmission village and by more than a third of the children in the medium transmission villages. MAL6P1.4 is coded by a B/A gene, and the most striking feature about the protein is that the C-terminal domains consist of a run of three DBLε domains. Within 3D7 PfEMP1s, it only shares this feature with VAR2CSA, but this cassette of domains can be found in PfEMP1s from HB3. Six DBL domains from MAL6P1.4 were included in the Ab array, and these were recognized to very different degrees. Many individuals had Abs against the three DBLε domains (25–55%) and <1% against the DBL3γ domain. This indicates that domains cross-reacting with the DBLε are often expressed by parasites, whereas parasites less often express domains inducing Abs against the DBL3γ. From this it can be inferred that the parasites expressing PfEMP1 inducing the Abs against the DBLε domains did not express a domain inducing Abs reacting with the DBL3γ, indicating that these domain types do not usually appear together in one PfEMP1 molecule. The same type of disassociation between the serological recognition of domains appearing in the same 3D7 PfEMP1 was noted for other proteins in which several domains were represented in the Ab array (PF08_0141, PFD1000c, PF11_0521, PFD0005w). Collectively, the present and previous data support the notion that the PfEMP1 repertoire slowly develops as a result of point mutations and recombination between groups (A, B, C) or more likely subgroups (e.g., subgroups of A genes) of var genes (11, 12). The data also suggest that caution is required when typing PfEMP1 genes based on small stretches of N-terminal sequence, as the functional regions may be located distal from the sequenced region and these regions may not be inherited together.
In this study Ab and parasitemia levels were measured at a single time point and age-group changes were used as a proxy for changes in Ab levels that an individual would experience over time. We are currently conducting a longitudinal study of the acquisition of anti-PfEMP1 Abs in a cohort of young children, which will allow us to assess both the acquisition and stability of these responses. A second limitation of the study is that the assay used to detect anti-PfEMP1-specific Abs could have differential sensitivity and be more sensitive at detecting Abs to group A and B/A domains compared with groups B and C. However, animals (rats, mice, and rabbits) immunized with different PfEMP1 constructs develop comparable levels of domain-specific IgG (our unpublished data). Thus, there is little to suggest that the Ab assay employed systematically is better in detecting some group-specific responses than others, or that group A and A/B domains inherently are more potent immunogens than domains belonging to groups B and C.
In conclusion, this study has demonstrated that the acquisition of Abs to PfEMP1 is structured so that Abs to some group A and B/A DBL domains consistently are developed first. This indicates that parasites expressing PfEMP1 domains cross-reacting serologically with the identified domains dominate infections in young children. Immunity to severe noncerebral malaria is thought to be developed after a few infections (2), and the identification of the antigenic targets of this protection is important for the development of malaria vaccines to protect children from malaria deaths.
We are grateful to the study participants, including their parents/guardians as well as village helpers and health management teams in Tanga region, Tanzania. We acknowledge with thanks Juma Akida, Zacharia Savael, Susanne Pedersen, Dominic Shauri, Fabio-Avit Massawe, John Hiza, William Chambo, and Seth Nguhu for excellent technical assistance throughout the study.
The authors have no financial conflicts of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
↵1 This study received financial support from a grant from the Foundation for the National Institute of Health through the Grand Challenges in Global Health initiative and the Danish International Development Agency (DANIDA) and the Novo Nordisk Foundation (Grant 10335). G. Kennedy and J. Lusingu were supported by the Gates Malaria Partnership. The study was conducted under the auspices of the Joint Malaria Programme, a collaborative research initiative between National Institute for Medical Research, Tanzania, Kilimanjaro Christian Medical College, London School of Hygiene and Tropical Medicine, and the University of Copenhagen.
↵2 Address correspondence and reprint requests to Gerald K. K. Cham, Centre for Medical Parasitology, University of Copenhagen, CSS Building 22-23, ⊘ster Farimagsgade 5, PO Box 2099, 1014 Copenhagen K, Denmark. E-mail address:
↵3 Abbreviations used in this paper: PfEMP1, P. falciparum erythrocyte membrane protein 1; CIDR, cysteine-rich interdomain region; DBL, duffy binding ligand; PP, point prevalence; VSA, variant surface Ag.
↵4 The online version of this article contains supplemental material.
- Received April 27, 2009.
- Accepted June 30, 2009.
- Copyright © 2009 by The American Association of Immunologists, Inc.