1. Am. J. Trop. Med. Hyg., 77(5), 2007, pp. 977–983
Copyright © 2007 by The American Society of Tropical Medicine and Hygiene
Differential Antibody Responses to Plasmodium falciparum Invasion Ligand Proteins
in Individuals Living in Malaria-Endemic Areas in Brazil and Cameroon
Louise Ford, Cheryl A. Lobo, Marilis Rodriguez, Mariano G. Zalis, Ricardo L. D. Machado, Andréa R. B. Rossit,
Carlos E. Cavasini, Alvaro A. R. Couto, Peter A. Enyong, and Sara Lustigman*
Laboratory of Molecular Parasitology, Lindsley F. Kimball Research Institute, New York Blood Center, New York, New York;
University Hospital, Federal University of Rio De Janeiro, Rio de Janeiro, Brazil; Faculty of Medicine of São José Do Rio Preto, São
José Do Rio Preto, São Paulo, Brazil; SEAMA College, Macapa, Amapa, Brazil; Tropical Medicine Research Station,
´ ´
Kumba, Cameroon
Abstract. Antibody responses to malaria invasion ligands and proteins on the merozoite surface have been shown to
interfere with red cell invasion and correlate with immunity to malaria. The current study is the first to characterize the
antibody responses to EBA-140 and EBA-181, Plasmodium falciparum invasion ligands implicated in the alternative
pathways of invasion, in age-matched populations of individuals living in endemic areas in both Brazil and Cameroon.
Antibody responses to the proteins screened were different between populations. The African individuals reacted
strongly with most fragments of these two EBAs, while the majority of the individuals from Mato Grosso, Brazil, reacted
weakly and those from the Amazon had elevated responses to these EBA proteins. When compared with the responses
against MSP-119 and EBA-175, it appeared that the Brazilian population has a variable ability to recognize P. falciparum
invasion ligand proteins and that these responses are distinct from the African population.
INTRODUCTION family characterized and shown to bind to the major glyco-
protein found on human erythrocytes, glycophorin A (GPA),
Falciparum malaria is one of the most significant causes of during invasion. Recombinant fragments of EBA-175 are rec-
morbidity and mortality in the developing world, with up to ognized by human sera from malaria-endemic areas.17 Anti-
500 million new cases annually.1 With the increase in antima- bodies raised in rabbits against EBA-peptide 4 blocked bind-
larial drug resistance, there is a growing requirement for an ing of native EBA-175 to human erythrocytes and inhibited
effective malaria vaccine. Epidemiologic surveys performed merozoite invasion in vitro.18,19 Additionally, IgG1 antibodies
in areas of high malaria transmission have shown that persons to EBA-175 peptide 4 are associated with protection against
who are continuously exposed to repeated malaria infection clinical malaria.20 We and others have shown that the major-
gradually develop clinical immunity.2,3 Experiments with an- ity of P. falciparum parasites in the endemic areas of India
tibodies purified from the sera of African adults who were and Brazil are not dependent on neuraminidase-sensitive in-
clinically immune to malaria and given by passive transfer to vasion pathways, of which the interaction of EBA-175 with
susceptible children have established that immunoglobulin G GPA is the major pathway.21–23 Importantly, the targeted
(IgG) is at least a main component of defense against the disruption of the eba-175 gene was not lethal but instead
asexual blood stage of Plasmodium falciparum.4–6 Thus, an- caused switching to a GPA-independent alternative invasion
tibodies that target the asexual blood-stage parasites seem to pathway.24 The parasite ligands that mediate the alternative
be of central importance,5,7 and several blood-stage antigens invasion pathways are postulated to belong to the EBL and
have been implicated as targets for protection.8,9 Vaccine de- RH families of invasion ligands. Although most of the immu-
velopment will be facilitated by studying the naturally ac- nogenicity studies have focused on EBA-175, little is known
quired immune responses that mediate protection. regarding the acquisition of natural antibodies to the other
Antibodies may act in different ways, by preventing mero- members of the EBL family. The rationale behind this study
zoite invasion of red blood cells (RBCs), by attacking infected was to evaluate for the first time the role of EBA-140 and
RBCs and facilitating phagocytosis, or by preventing cytoad- EBA-181, parasite ligands used in the alternative invasion
hesion of infected RBCs.10 Molecules important to all 3 of pathways as targets of naturally acquired immune responses.
these processes have been identified as candidate vaccines, Such data will be relevant to the design of vaccines aimed to
but those involved in merozoite invasion have been the most inhibit erythrocyte invasion by merozoites. Additionally,
studied. Repeated cycles of merozoite invasion of RBCs study of the naturally acquired antibodies present in endemic
quickly amplify the blood-stage parasitemia, and this contrib- populations may reflect the choice of invasion ligands that
utes to symptomatic disease. Successful invasion requires suc- locally circulating parasites use, which in turn may be dictated
cessful receptor–ligand interactions on the RBC surface. by the repertoire of host RBC receptors available in that
These invasion ligands can be present on the merozoite sur- specific study population.
face (MSP family) or be harbored in apical organelles like the In this paper, the naturally acquired antibody responses to
micronemes (EBL family) and the rhoptries (RH family). different regions of these two relatively unstudied invasion
P. falciparum uses a wide variety of RBC receptors for ligands, EBA-140 and EBA-181, were analyzed in popula-
invasion.11–16 Erythrocyte-binding antigen 175 (EBA-175) tions of individuals living in endemic areas of Brazil and Cam-
was the first member of the erythrocyte-binding ligand (EBL) eroon, known to be of differing endemicity.
MATERIALS AND METHODS
* Address correspondence to Sara Lustigman, Laboratory of Molecu-
lar Parasitology, Lindsley F. Kimball Research Institute, New York
Study areas, subjects, and blood sample collection. We ana-
Blood Center, 310 E. 67th St., New York, NY 10065. E-mail: lyzed groups of subjects who had been exposed to malaria
slustigman@nybloodcenter.org transmission in endemic areas in both Brazil and Cameroon.
977

2. 978 FORD AND OTHERS
Two study groups from hypoendemic regions for malaria in INKK) and EBA-175 region II (aa 145–760; 3D7 allele) were
Brazil were studied: The serum samples of the first group expressed in yeast and were a kind gift from D. Narum and L.
were collected in 1995 from 17 adult male individuals (median Miller (LMV, NIH). The 19-kDa processed fragment of the
age ‫ 52 ס‬years) living in Peixoto de Azevedo, a municipality merozoite surface protein 1, MSP-119, was obtained as a re-
of Mato Grosso state, located in the southern part of the combinant yeast protein from the MR4 (ATCC) and also
Amazon region in Brazil. The individuals were migrant gold corresponds to the 3D7 allele. These 3 recombinant proteins
mining workers without documented ethnicity who had come were properly folded.
to the local health service center, run by the Ministry of Measurement of antigen-specific antibody responses. Hu-
Health, and were found to be P. falciparum positive and P. man serum samples were analyzed for IgG1 and IgG3 reac-
vivax negative. These individuals resided in the endemic re- tivity to the recombinant malaria proteins by a defined en-
gion for only a few years and had no more than 1–3 malaria zyme-linked immunosorbent assay (ELISA), as described
episodes. The serum samples of the second group were col- previously.26 Sera were analyzed for IgG antibody isotypes.
lected in 2004 and originated from individuals who have lived The subclass of immunoglobulin determines antibody func-
in the Amazon regions (Macapá, Belém, Porto Velho, and tion (e.g., complement fixation or the activation of phago-
Rio Branco) of Brazil for most of their life and have reported cytes), and in humans, immunoglobulin G1 (IgG1) and IgG3
multiple episodes (range 1 to > 4 episodes) of P. falciparum are important mediators of malaria parasite clearance.27 To
malaria infection and were P. vivax-negative at the time of determine which E. coli-expressed EBA protein fragments
blood collection. This group consisted of 37 individuals (4 were antigenic, we performed an initial ELISA screen using
females and 33 males; median age ‫ 52 ס‬years, range 18–76 the various recombinant EBA fragments with a small group
years). of sera from Cameroon. Based on their antigenicity, for all
The serum samples from West Africa were collected in further experiments we focused on the antibody responses
1996 and 1997 and originated from individuals who live in generated to 3 EBA-140 protein fragments: portion of region
villages in the Kumba region, an area of hyperendemic ma- II, region VI and the C-terminal region, and to the C-terminal
laria in the southwest province of Cameroon. The individuals region fragment of EBA-181.
who participated in the study were born or had resided for Microtiter ELISA plates (Costar, Corning Life Sciences,
> 10 years in villages: Marumba I, Marumba II, Boa Bakundu, Corning, NY) were coated with 1 ␮g/mL of recombinant pro-
Bombanda, and Bombele. The group consisted of individuals tein diluted in 0.05 M carbonate buffer, pH 9.6. After incu-
who were both P. falciparum blood smear positive and nega- bation overnight at 4°C, the plates were washed 5 times with
tive at the time of bleeding. The Cameroonian study group phosphate-buffered saline (PBS) with 0.05% Tween 20
comprised 180 individuals (118 males and 62 females; median (PBST) and blocked with blocking buffer (3% BSA in PBST)
age ‫ 71 ס‬years; range, 3–75 years). An age-matched subset of for 1.5 hr at 37°C. Serum samples were pre-incubated with E.
these individuals was selected (N ‫ ;82 ס‬median age ‫23 ס‬ coli extract prior to dilution and incubation with the bound
years; range 8–38 years) to compare their responses with antigen to remove antibodies to potential E. coli contami-
those from the individuals from Brazil. nants in the recombinant protein preparations. Serum
All blood samples were obtained from volunteers with in- samples, diluted 1:500 in blocking buffer, were reacted with
formed consent using protocols approved by both local and the bound antigens by incubating for 2 hr at 37°C, in duplicate
NYBC IRB committees. wells. The bound IgG1 and IgG3 antibodies were detected
Recombinant malaria protein expression and purifica- after incubation for 1 hr at 37°C with mouse monoclonal
tion. Recombinant malaria proteins were expressed in Es- antibodies against different human IgG subclasses (Hybri-
cherichia coli BL21 (DE3) (Sigma, St. Louis, MO) as fusion doma Reagent Laboratory, Baltimore, MD) diluted 1:1000 in
proteins with glutathione S-transferase (GST) using the blocking buffer, followed by incubation for 1 hr at 37°C with
pGEX vector (Amersham Biotech, Piscataway, NJ). horseradish peroxidase-conjugated goat anti-mouse IgG (H +
The 3D7 strain of P. falciparum was used as template to L) (Kirkegaard & Perry Laboratories, Inc., Gaithersburg,
derive the various fragments that were used for cloning and MD) diluted 1:1250 in blocking buffer. Tetramethylbenzi-
expression. Fragments of the erythrocyte-binding proteins dine (Sigma, St Louis, MO) was used as the substrate for all
EBA-140 and EBA-181 were expressed as soluble GST- ELISAs, and the optical density (OD) was read at 450 nm on
fusion proteins and corresponded to fragments within the fol- a SpectraMax 190 ELISA Reader (Molecular Devices,
lowing regions: portion (aa 350–440) of the Duffy binding-like Sunnyvale, CA).
domain (DBL) also known as region II, portions of region VI Each serum sample was tested against GST as a negative
(aa 799–963) and the C-terminal region (aa 967–1140) of control for the GST-fusion proteins. Positive responders were
EBA-140, and portions of the N-terminal (aa 33–171) and the defined as those that gave an OD value greater than the
C-terminal (aa 1235–1345) of EBA-181. These regions do not cut-off OD, which was defined as the mean plus 3 standard
exhibit sequence polymorphism among the various P. falci- deviations of OD values from control sera. If the cut-off was
parum strains. Recombinant proteins were purified by affinity < 0.1, positive responders were defined as those with an OD
chromatography on glutathione agarose (Sigma) as previ- value > 0.1. Control sera used were collected from 5 healthy
ously described.25 Control GST was purified from E. coli adult volunteers living in New York who had never been
BL21 transformed with the pGEX vector alone. Fusion pro- exposed to malaria. Samples from the different study sites
teins were assessed by Coomassie blue staining of SDS- were run in parallel to ensure reliable comparison of ODs.
PAGE gels, and protein concentrations were measured using Statistical analysis. The ␹2 test was used to compare pro-
the Bio-Rad protein assay (Bio-Rad, Hercules, CA) accord- portions of antibody responders in different groups, whereas
ing to the manufacturer’s instructions. the differences in the immunoglobulin levels between groups
Full length EBA-140 region II (aa 141–756; the 3D7 allele, were compared using the two-tailed nonparametric Mann–

3. ANTIBODY RESPONSES TO MALARIA INVASION LIGANDS 979
Whitney U test. Spearman’s rank correlation test was used to ing in Macapá, Belém, Porto Velho, and Rio Branco, the
test the significance of the correlation between the age of the antigen-specific antibody responses differed between the 2
individuals and their antibody responses, with the correlation endemic regions within Brazil (Figure 1B, C). The antibody
coefficient being expressed as r. A P value of < 0.05 was responses in these Amazonian individuals, to the same panel
considered statistically significant. Statistical analysis was per- of EBA proteins, were elevated (percent positive responders:
formed using GraphPad PRISM software (GraphPad Soft- IgG1, 7.7–30.8%; IgG3, 28.5–69.2%) when compared with
ware Inc., San Diego, CA). those from Mato Grosso, although lower in magnitude than
the responses in the Cameroonian sera (Table 1). Amazonian
individuals had significantly higher antibody responses than
RESULTS
individuals from Mato Grosso to EBA-181 C-terminal frag-
Differential antibody responses to EBA-140 and EBA-181 ment (P < 0.001 for both IgG1 and IgG3), EBA-140 region II
fragment proteins in individuals living in Brazil and in Cam- fragment (P < 0.001 IgG1 and P ‫ 7200.0 ס‬IgG3), EBA-140
eroon. Analysis of the IgG1 and IgG3 responses to different region VI fragment (P ‫ 4230.0 ס‬IgG1), and EBA-140 C-
fragments of EBA-140 and EBA-181 in individuals from terminal fragment (P ‫ 1500.0 ס‬IgG3). Regardless, the anti-
Mato Grosso and the Amazon region, Brazil, and Cameroon body responses in the Amazon individuals was significantly
revealed a dramatic difference in both the prevalence and lower than the responses in the Cameroon individuals to
magnitude of responses between the populations (Figure 1; EBA-140 full-length region II (P < 0.001 for both IgG1 and
Table 1). Whereas the Cameroon individuals mounted an ef- IgG3), EBA-140 region II fragment (P ‫ 2610.0 ס‬IgG1 and
ficient antibody response to most of the recombinant EBA- P ‫ 3000.0 ס‬IgG3), EBA-140 region VI fragment (P ‫8800.0 ס‬
140 and EBA-181 proteins (percent positive responders: IgG3), and EBA-140 C-terminal fragment (P ‫ 6000.0 ס‬IgG1
IgG1, 38.1–96.4%; IgG3, 65–100%), the antibody response to and P < 0.001 IgG3). Antibody responses to the EBA-181
the same panel of proteins in residents of Mato Grosso was C-terminal fragment in the Amazonian individuals were also
much lower (percent positive responders: IgG1, 5.9–29.4%; lower than in the Cameroon individuals; however, these dif-
IgG3, 5.9–29.4%) (Figure 1A,B; Table 1). Only 1 serum ferences were not statistically significant (P ‫ 541.0 ס‬IgG1 and
sample from Mato Grosso in each assay showed a strong P ‫ 671.0 ס‬IgG3).
response (OD > 0.3) to the EBAs. All of the IgG1 and IgG3 Antibody responses to MSP-119 and EBA-175 region II. To
responses in Mato Grosso, with the exception of the IgG1 determine if the low humoral response seen in the Mato
response mounted to a fragment within the EBA-140 region Grosso population against the EBA proteins was antigen-
II, were significantly lower than the responses in Cameroon specific, we examined the antibodies present in these endemic
(EBA-140 proteins—full-length region II P < 0.001 for both, populations to 2 other known highly antigenic malaria pro-
region II fragment P < 0.001 IgG3, region VI fragment P ‫ס‬ teins: the GPA-binding domain of EBA-175, EBA-175 region
0.0097 IgG1 and P ‫ 4330.0 ס‬IgG3, C-terminal fragment P ‫ס‬ II, another member of the EBL family and the 19-kDa pro-
0.0087 IgG1 and P < 0.001 IgG3; EBA-181 C-terminal frag- cessed fragment of the merozoite surface protein 1 (MSP-119).
ment P < 0.001 for both). The EBA-140 proteins correspond- These results are presented in Table 1 and Figure 2. In both
ing to full-length region II, fragment of region VI, and frag- the Mato Grosso inhabitants and the individuals from Cam-
ment of the C-terminal were found to be the most antigenic eroon, the IgG3 responses to MSP-119 were prevalent (76.5%
ligands for the Mato Grosso population (11.8–29.4% positive IgG3-positive responders in Mato Grosso versus 95% IgG3-
responders), although the reactivity was much lower than positive responders in Cameroon; P ‫ 80.0 ס‬using ␹2 analysis).
found in the African population (80–100% positive respond- The prevalence of the IgG1 responses against MSP-119 was,
ers). In the African population, the full-length region II frag- however, significantly lower in Mato Grosso inhabitants than
ment, the RBC-binding domain of EBA-140, is the most im- in the individuals from Cameroon (70.6% IgG1-positive re-
munoreactive in all individuals, regardless of age. sponders in Mato Grosso versus 100% IgG1-positive re-
Interestingly, when we analyzed the profile of antibodies sponders in Cameroon, P ‫ 2␹ 700.0 ס‬test) (Table 1). Both the
present in other populations of the Brazilian Amazon, resid- IgG1 and IgG3 antibody responses to MSP-119 by individuals
FIGURE 1. Comparison of isotype-specific antibody responses to EBA-140 and EBA-181 fragment proteins in age-matched individuals living
in (A) Kumba, Cameroon (N ‫( ,)82 ס‬B) Mato Grosso, Brazil (N ‫ ,)71 ס‬and (C) the Amazon, Brazil (N ‫ .)93 ס‬Median values are indicated by
horizontal bars.

4. 980 FORD AND OTHERS
TABLE 1
Prevalence and magnitude of antibody responses to EBA fragment proteins and MSP-119 from age-matched individuals living in Kumba,
Cameroon,* Mato Grosso, Brazil,† and the Amazon, Brazil‡ (measured by ELISA)
Kumba, Cameroon Mato Grosso, Brazil Amazon, Brazil
IgG1 IgG3 IgG1 IgG3 IgG1 IgG3
Positive Positive Positive Positive Positive Positive
sera Median OD sera Median OD sera Median OD sera Median OD sera Median OD sera Median OD
Antigen (%)§ (95% CI)¶ (%) (95% CI) (%) (95% CI) (%) (95% CI) (%) (95% CI) (%) (95% CI)
EBA-181 38.1 0.07 65.0 0.28 5.9 0.001 5.9 0.020 10.3 0.04 69.2 0.14
aa1235–1345 (0.02–0.51) (0.29–0.93) (0–0.07) (0–0.04) (0.03–0.11) (0.13–0.25)
EBA-140 96.4 1.23 100 1.17 11.8 0.014 29.4 0.042 25.6 0.02 38.5 0.08
aa141–756 (0.90–1.66) (1.07–1.69) (0–0.08) (0–0.24) (0.01–0.34) (0.10–0.34)
EBA-140 20.0 0 70.8 0.14 5.9 0.001 5.9 0.001 7.7 0.04 28.5 0.03
aa350–440 (0–0.05) (0.13–0.71) (0–0.01) (0–0.03) (0.04–0.06) (0.04–0.17)
EBA-140 80.0 0.14 80.0 1.33 11.8 0.001 23.5 0.001 30.8 0.06 30.8 0
aa799–963 (0–0.81) (0–2.50) (0–0.20) (0–0.43) (0.08–0.29) (0.01–0.13)
EBA-140 80.0 0.27 81.8 0.79 29.4 0.050 11.8 0.001 12.8 0.04 38.5 0.04
aa967–1140 (0.11–0.42) (0.63–1.16) (0.03–0.11) (0–0.07) (0.04–0.06) (0.06–0.13)
EBA-175 92.9 1.19 100 1.34 11.8 0.001 5.9 0.001 5.1 0 15.4 0.01
aa145–760 (0.90–1.46) (1.07–1.72) (0–0.04) (0–0.004) (0–0.08) (0.01–0.06)
MSP-119 100 1.42 95.2 1.59 70.6 0.84 76.5 0.30 38.5 0.03 33.3 0
(1.03–1.54) (1.19–1.89) (0.45–1.17) (0.29–0.96) (0.26–0.72) (0.10–0.34)
* 28 serum samples collected from Kumba, Cameroon.
† 17 serum samples collected from Mato Grosso, Brazil.
‡ 39 serum samples collected from Amazon, Brazil.
§ Percentage of positive responders (% positive sera) are defined as those with an OD value greater than the cut-off OD which was defined as the mean + 3 standard deviations for the ODs
of control sera or OD values > 0.1.
¶ Median ODs are shown with 95% confidence interval (CI).
in Mato Grosso were significantly lower than the responses Mato Grosso to EBA-175 region II (P < 0.001 for both IgG1
mounted by Cameroon individuals (P ‫ 2940.0 ס‬for IgG1 and and IgG3), the antibody response in the Amazon individuals
P ‫ 8000.0 ס‬for IgG3) (Figure 2). The antibody responses to to EBA-175 region II was also significantly lower than those
MSP-119 in the Amazon individuals were also significantly in the Cameroon individuals (P < 0.001 for both IgG1 and
lower than the responses in the Cameroon individuals (P < IgG3) (Figure 2).
0.001 for both IgG1 and IgG3) (Figure 2). Although the responses to the EBA proteins in the Ama-
Notably, we found a significantly lower prevalence of both zon individuals were elevated compared with those from
IgG1 and IgG3 responses to EBA-175 region II (11.8% IgG1- Mato Grosso (Figures 1 and 2), both IgG1 and IgG3 re-
positive responders and 5.9% IgG3-positive responders) in sponses to MSP-119 in individuals from the Amazon region
Mato Grosso, which was in sharp contrast to the 93% IgG1- were significantly lower than those in Mato Grosso (P ‫ס‬
positive responders and 100% IgG3-positive response in 0.0438 IgG1 and P ‫ 6000.0 ס‬IgG3) (Figure 2).
Cameroon (P < 0.0001 for both ␹2 test) (Table 1). Both the Antibody responses to the C-terminal regions of EBA-181
IgG1 and IgG3 antibody responses to EBA-175 region II by and EBA-140 in Cameroonians correlate with age. Screening
individuals in Mato Grosso were significantly lower than the a small sample of adult individuals from Cameroon (N = 28,
responses mounted by Cameroon individuals (P < 0.0001 for median age ‫ 23 ס‬years) revealed that the most antigenic
both) (Figure 2). Although the Amazonian individuals had fragments of the EBA proteins, in addition to EBA-140 re-
significantly higher antibody responses than individuals from gion II, were the C-terminal regions of both EBA-181 and
FIGURE 2. Comparison of isotype-specific antibody responses to EBA-175 region II and MSP19 in age-matched individuals living in living in
(A) Kumba, Cameroon (N ‫( ,)82 ס‬B) Mato Grosso, Brazil (N ‫ ,)71 ס‬and (C) the Amazon, Brazil (N ‫ .)93 ס‬Median values are indicated by
horizontal bars.

5. ANTIBODY RESPONSES TO MALARIA INVASION LIGANDS 981
EBA-140 (Figure 1). We extended our study to a larger popu- Grosso individuals had detectable antibody responses. Thus,
lation (N ‫ ,081 ס‬median age ‫ )71 ס‬to perform an age cor- the Mato Grosso population had an apparently dichotomous
relation study in the Cameroonian population. Both the IgG3 response to the 2 sets of ligands, the MSPs versus the EBLs.
antibody responses to the C-terminal regions of both EBA- Further analysis of a second set of Brazilian individuals living
181 and EBA-140 were slightly increased with age (r ‫,391.0 ס‬ in the Amazon revealed that the individuals from Mato
P ‫ 10.0 ס‬and r ‫ ,681.0 ס‬P ‫ ,210.0 ס‬respectively) (data not Grosso were not typical of the Brazilian immune response, as
shown). EBA-140 region II-specific IgG3 responses were elevated EBA-140 and EBA-181 antibody responses were
highly elevated in all Cameroon individuals, at a 1:500 dilu- found in other regions of the Amazon, although with much
tion, regardless of age (data not shown). lower responses than those found in the Cameroonian popu-
In addition, none of the antibody responses in Cameroon lation. Regardless, most of the Amazonian population was
were correlated to active infection with P. falciparum as in- still nonresponders to the EBA-175 region II protein similarly
dicated by a positive blood smear (data not shown). as those from Mato Grosso.
We put forward 2 hypotheses to explain these differences:
(1) Exposure to the parasite in terms of both overall years
DISCUSSION living in endemic areas and number of malaria infections has
a direct effect on antibodies to various parasite antigens. A
Malaria is a unique example of an infection where people careful look at the history of malaria in Brazil reveals many
exposed multiple times to the blood-stage parasite acquire epidemics associated with migration of a nonimmune popu-
immunity with age and/or number of episodes. This ability to lation to malaria-endemic areas.34–36 Malaria in Brazil, which
combat malaria is an important adaptive trait of populations accounts for 90% of the cases in South America, has in-
living in endemic areas. The detection of significant differ- creased 3-fold during the past decade, and nearly 631,000
ences in the expression of this trait and the identification of cases were reported in 2002. Of these infections, 80% are P.
the factors involved should improve the understanding of the vivax and 20% are P. falciparum infections. Almost all re-
host–parasite relationship and thus lead to advances in con- ported cases of malaria in Brazil originate in the Amazon
trol strategies. region, where migrant populations and poor access to diag-
This is the first study to examine the humoral response to nosis and treatment are major barriers to effective control of
the invasion ligands EBA-140 and EBA-181 induced by natu- malaria. Migrant populations help to spread parasites
ral infection. It is believed that effective immunity to P. fal- throughout the endemic regions and also contribute to the
ciparum blood-stage parasites involves the acquisition of in- spread of the disease in malaria-free areas, thus increasing the
hibitory antibodies targeted to various antigens of the inva- risk for malaria outbreaks. It is believed that malaria in Mato
sive merozoite. The obligatory interaction between invasion Grosso could have originated from such migrant populations
ligands and their cognate receptors on the RBC represents a and is characterized by a hypoendemic pattern of infection,
potential target for inhibition by vaccine-induced antibodies. due mainly to its low demographic index. Thus, in Mato
Some of the current most promising vaccine candidates are Grosso, antibodies against only the immunodominant anti-
indeed such invasion ligands as exemplified by MSP-1, AMA- gens like MSP-119 and other MSPs were found, as exposure to
1, and EBA-175.28 In this study, recombinant proteins repre- P. falciparum in these inhabitants would not be as high or
senting different fragments of P. falciparum EBA-140 and repeated as in the rest of the Amazon (relatively stable popu-
EBA-181 were used to analyze the profile of naturally ac- lation), whose inhabitants are subjected to multiple malaria
quired human antibodies to these proteins. The amino acid episodes in short periods of time. Malaria in the Brazilian
sequence homology between region II, which is the RBC- Amazon, on the other hand, was characterized by elevated
binding domain, or region VI of the 2 EBL proteins is very antibody responses to all parasite ligands, including both
low, < 25% and < 20%, respectively. The similarity between MSP-119 and the EBLs, although overall magnitudes were
the EBA members of the EBL family is attributed to their less than what was found in Africa. This substantial difference
conserved structures.29 Although we recognize that the tested in exposure levels to the parasite thus impacted antibody re-
recombinant proteins may not represent all the conforma- sponse to these 2 sets of ligands in these 2 Brazilian popula-
tional epitopes on the native EBLs, this study serves as a tions. Future studies in a larger population in the Brazilian
springboard for a future large-scale analysis of the antibody Amazon could confirm this hypothesis. (2) The repertoire of
responses in these two populations. host RBC receptors available in specific populations will,
We found that the Mato Grosso individuals did not mount over time, select for expression of specific parasite ligands
a strong humoral response against these two EBA ligands. that can mediate successful RBC invasion, and thus the im-
This was in contrast to the African individuals in whom, con- mune responses to such ligands may be differential, correlat-
siderable IgG1 and IgG3 antibodies were detectable to all the ing to their expression. In a recent study in Tanzania,37 it was
recombinant EBA-140 and 180 fragments tested. In addition shown that expression levels of some of EBA and PfRH
to the full-length region II of EBA-140, a fragment of the ligands correlated with the specific invasion pathway that the
C-terminus of both EBA-140 and EBA-181 was also the most corresponding parasites used for invasion. The authors also
antigenic in this Cameroonian population. However, both hypothesized that the ligand expression patterns may repre-
populations mounted equally robust responses to the various sent frequency-dependent selection either by polymorphic
MSPs (data not shown), which have been shown to be anti- host receptors or by immune responses. Therefore, our re-
genic in a variety of different endemic populations.30–33 When sults, showing differences in the immunoreactivity, could be
the same sera were analyzed for immunoreactivity to the also linked to RBC heterogeneity between these 2 popula-
GPA-binding domain of EBA-175 (EBA-175 region II), a tions which could have dictated parasite choice of invasion
dominant member of the EBL family, only a few Mato pathways and thus expression of specific EBLs. We have pre-

6. 982 FORD AND OTHERS
viously characterized the invasion pathways used by field iso- 2. Marsh K, 1992. Malaria—a neglected disease? Parasitology 104
lates from Mato Grosso and found that EBA-175–mediated (Suppl): S53–S69.
3. Trape JF, Rogier C, Konate L, Diagne N, Bouganali H, Canque
invasion was not the pathway of choice for the majority of
B, Legros F, Badji A, Ndiaye G, Ndiaye P, Brahimi K, Ous-
these strains. Although we do not yet know the preferred mane F, Druilhe P, Pereira Da Silva L, 1994. The Dielmo
invasion pathways used by isolates from the regions of the project: a longitudinal study of natural malaria infection and
Amazon, the low anti-EBA-175 response we see in both Bra- the mechanisms of protective immunity in a community living
zilian populations could allude to a common low usage of the in a holoendemic area of Senegal. Am J Trop Med Hyg 51:
123–137.
EBA-175–GPA invasion pathway. On the other hand, a study 4. Bouharoun-Tayoun H, Attanath P, Sabchareon A, Chong-
in Gambia found that invasion via GPA and EBA-175 was suphajaisiddhi T, Druilhe P, 1990. Antibodies that protect hu-
frequently used by field strains,23 and if this is reflective of mans against Plasmodium falciparum blood stages do not on
malaria on the whole in Africa, then the high anti-EBA-175 their own inhibit parasite growth and invasion in vitro, but act
in cooperation with monocytes. J Exp Med 172: 1633–1641.
reactivities that were observed in the Cameroonian popula-
5. Cohen S, McGregor IA, Carrington S, 1961. Gamma-globulin
tion could be predicted. A study in Western Kenya found that and acquired immunity to human malaria. Nature 192: 733–
98.7% of samples tested had detectable total IgG against 737.
EBA-175 region II,38 while in 2 areas of West Africa—The 6. Druilhe P, Perignon JL, 1994. Mechanisms of defense against P.
Gambia and Nigeria—the proportion of individuals with an- falciparum asexual blood stages in humans. Immunol Lett 41:
115–120.
tibodies against the same molecule was 60–70%.39 Notably, a
7. Sabchareon A, Burnouf T, Ouattara D, Attanath P, Bouharoun-
retrospective analysis of naturally acquired antibodies to a Tayoun H, Chantavanich P, Foucault C, Chongsuphajaisiddhi
synthetic peptide from EBA-175 peptide 4 carried out in Ga- T, Druilhe P, 1991. Parasitologic and clinical human response
bon has shown that the overall prevalence rates of antibodies to immunoglobulin administration in falciparum malaria. Am J
to a linear epitope of EBA-175, EBA-175 peptide 4, were Trop Med Hyg 45: 297–308.
8. Deitsch KW, Hviid L, 2004. Variant surface antigens, virulence
85.2% and that the antibody response was associated with genes and the pathogenesis of malaria. Trends Parasitol 20:
immunity against clinical malaria.20 Moreover, antibodies to 562–566.
EBA-175 peptide 4 were inhibitory in growth-inhibition as- 9. Mahanty S, Saul A, Miller LH, 2003. Progress in the development
says in vitro.19 Future studies using full-length EBA-181 re- of recombinant and synthetic blood-stage malaria vaccines. J
gion II recombinant protein, that is functional in binding to Exp Biol 206: 3781–3788.
10. Miller LH, Hoffman SL, 1998. Research toward vaccines against
RBCs, in growth-inhibition assays with serum samples con- malaria. Nat Med 4: 520–524.
taining differential patterns of anti-EBA antibodies on labo- 11. Binks RH, Conway DJ, 1999. The major allelic dimorphisms in
ratory and field isolates from distinct geographic areas that have four Plasmodium falciparum merozoite proteins are not asso-
defined invasion pathways, in addition to determining antibody ciated with alternative pathways of erythrocyte invasion. Mol
Biochem Parasitol 103: 123–127.
responses by ELISA, would discern the role of receptor hetero-
12. Dolan SA, Proctor JL, Alling DW, Okubo Y, Wellems TE, Miller
geneity on the EBA ligand-specific antibody responses. LH, 1994. Glycophorin B as an EBA-175 independent Plas-
modium falciparum receptor of human erythrocytes. Mol Bio-
Received May 4, 2007. Accepted for publication August 13, 2007. chem Parasitol 64: 55–63.
Acknowledgments: We thank Meagan B. Gallagher, who optimized 13. Hadley TJ, Klotz FW, Pasvol G, Haynes JD, McGinniss MH,
the ELISA assays used in these studies during her internship. Okubo Y, Miller LH, 1987. Falciparum malaria parasites in-
vade erythrocytes that lack glycophorin A and B (MkMk).
Financial support: Work reported in this paper was funded in part by Strain differences indicate receptor heterogeneity and two
a grant from the NIH-P50 HL 54459 and in part by the New York pathways for invasion. J Clin Invest 80: 1190–1193.
Blood Center. The research of Ricardo Machado was supported by 14. Mitchell GH, Hadley TJ, McGinniss MH, Klotz FW, Miller LH,
Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) 1986. Invasion of erythrocytes by Plasmodium falciparum ma-
and Conselho Nacional de Pesquisa (CNPq). laria parasites: evidence for receptor heterogeneity and two
Authors’ addresses: Louise Ford, Cheryl A. Lobo, Marilis Rodriguez, receptors. Blood 67: 1519–1521.
and Sara Lustigman, Laboratory of Molecular Parasitology, Lindsley 15. Perkins ME, Holt EH, 1988. Erythrocyte receptor recognition
F. Kimball Research Institute, New York Blood Center, 310 East varies in Plasmodium falciparum isolates. Mol Biochem Para-
67th St., New York, NY 10065, Telephone: +1 (212) 570-3000, Fax: +1 sitol 27: 23–34.
(212) 570-3195, E-mail: slustigman@nybloodcenter.org. Mariano G. 16. Narum DL, Haynes JD, Fuhrmann S, Moch K, Liang H, Hoffman
Zalis, University Hospital, Federal University of Rio de Janeiro, 255 SL, Sim BK, 2000. Antibodies against the Plasmodium falci-
Rua Professor Rodolpho Paulo Rocco, Rio de Janeiro, 21949-900, parum receptor binding domain of EBA-175 block invasion
Brazil, Telephone: +55-21-2562-2617, Fax: + 55-21-2280-8193. pathways that do not involve sialic acids. Infect Immun 68:
Ricardo L.D. Machado, Andréa R.B. Rossit, and Carlos E. Cavasini, 1964–1966.
Faculty of Medicine of São José do Rio Preto, Avenida Brigadeiro 17. Daugherty JR, Murphy CI, Doros-Richert LA, Barbosa A,
Faria Lima, São José do Rio Preto, São Paulo, 15090-000, Brazil, Kashala LO, Ballou WR, Snellings NJ, Ockenhouse CF, Lanar
Telephone: +55-17-3201-5736, Fax: +55-17-3201-5911. Alvaro A.R. DE, 1997. Baculovirus-mediated expression of Plasmodium
Couto, SEAMA College, Macapa, 1201 Avenida Nacoes Unidas,
´ falciparum erythrocyte binding antigen 175 polypeptides and
Amapa 68908-170, Brazil, Telephone and Fax: +55-91-3223-7393. Pe-
´ their recognition by human antibodies. Infect Immun 65: 3631–
ter A. Enyong, Tropical Medicine Research Station, PO Box 55, 3637.
Kumba, Cameroon, Telephone/Fax: +237-33-54231. 18. Orlandi PA, Sim BK, Chulay JD, Haynes JD, 1990. Character-
Reprint requests: Sara Lustigman, Laboratory of Molecular Para- ization of the 175-kilodalton erythrocyte binding antigen of
sitology, Lindsley F. Kimball Research Institute, New York Blood Cen- Plasmodium falciparum. Mol Biochem Parasitol 40: 285–294.
ter, 310 East 67th St., New York, NY 10065, Telephone: +1 (212) 570- 19. Sim BK, Orlandi PA, Haynes JD, Klotz FW, Carter JM, Camus
3119, Fax: +1 (212) 570-3121, E-mail: slustigman@nybloodcenter.org. D, Zegans ME, Chulay JD, 1990. Primary structure of the
175K Plasmodium falciparum erythrocyte binding antigen and
REFERENCES identification of a peptide which elicits antibodies that inhibit
malaria merozoite invasion. J Cell Biol 111: 1877–1884.
1. Snow RW, Guerra CA, Noor AM, Myint HY, Hay SI, 2005. The 20. Toure FS, Deloron P, Migot-Nabias F, 2006. Analysis of human
global distribution of clinical episodes of Plasmodium falci- antibodies to erythrocyte binding antigen 175 peptide 4 of
parum malaria. Nature 434: 214–217. Plasmodium falciparum. Clin Med Res 4: 1–6.

7. ANTIBODY RESPONSES TO MALARIA INVASION LIGANDS 983
21. Okoyeh JN, Pillai CR, Chitnis CE, 1999. Plasmodium falciparum 31. al-Yaman F, Genton B, Kramer KJ, Chang SP, Hui GS, Baisor
field isolates commonly use erythrocyte invasion pathways that M, Alpers MP, 1996. Assessment of the role of naturally ac-
are independent of sialic acid residues of glycophorin A. Infect quired antibody levels to Plasmodium falciparum merozoite
Immun 67: 5784–5791. surface protein-1 in protecting Papua New Guinean children
22. Lobo CA, de Frazao K, Rodriguez M, Reid M, Zalis M, Lustig- from malaria morbidity. Am J Trop Med Hyg 54: 443–448.
man S, 2004. Invasion profiles of Brazilian field isolates of 32. Wang L, Crouch L, Richie TL, Nhan DH, Coppel RL, 2003.
Plasmodium falciparum: phenotypic and genotypic analyses. Naturally acquired antibody responses to the components of
Infect Immun 72: 5886–5891. the Plasmodium falciparum merozoite surface protein 1 com-
23. Baum J, Pinder M, Conway DJ, 2003. Erythrocyte invasion phe- plex. Parasite Immunol 25: 403–412.
notypes of Plasmodium falciparum in The Gambia. Infect Im- 33. Wang L, Richie TL, Stowers A, Nhan DH, Coppel RL, 2001.
mun 71: 1856–1863. Naturally acquired antibody responses to Plasmodium falci-
24. Reed MB, Caruana SR, Batchelor AH, Thompson JK, Crabb BS, parum merozoite surface protein 4 in a population living in an
Cowman AF, 2000. Targeted disruption of an erythrocyte area of endemicity in Vietnam. Infect Immun 69: 4390–4397.
binding antigen in Plasmodium falciparum is associated with a 34. McGreevy PB, Dietze R, Prata A, Hembree SC, 1989. Effects of
switch toward a sialic acid-independent pathway of invasion. immigration on the prevalence of malaria in rural areas of the
Proc Natl Acad Sci USA 97: 7509–7514. Amazon basin of Brazil. Mem Inst Oswaldo Cruz 84: 485–491.
25. Lobo CA, Rodriguez M, Reid M, Lustigman S, 2003. Glyco- 35. Sawyer D, 1993. Economic and social consequences of malaria in
phorin C is the receptor for the Plasmodium falciparum eryth- new colonization projects in Brazil. Soc Sci Med 37: 1131–1136.
rocyte binding ligand PfEBP-2 (baebl). Blood 101: 4628–4631. 36. Tauil PL, 1992. Intervention possibilities on the biologic cycle of
26. Joseph GT, Huima T, Lustigman S, 1998. Characterization of an malaria towards endemism control. Rev Inst Med Trop Sao
Onchocerca volvulus L3-specific larval antigen, Ov-ALT-1. Paulo 34 (Suppl 9): S21–S23.
Mol Biochem Parasitol 96: 177–183. 37. Bei AK, Membi CD, Rayner JC, Mubi M, Ngasala B, Sultan AA,
27. Garraud O, Mahanty S, Perraut R, 2003. Malaria-specific anti- Premji Z, Duraisingh MT, 2007. Variant merozoite protein
body subclasses in immune individuals: a key source of infor- expression is associated with erythrocyte invasion phenotypes
mation for vaccine design. Trends Immunol 24: 30–35. in Plasmodium falciparum isolates from Tanzania. Mol Bio-
28. Reed ZH, Friede M, Kieny MP, 2006. Malaria vaccine develop- chem Parasitol 153: 66–71.
ment: progress and challenges. Curr Mol Med 6: 231–245. 38. Ohas EA, Adams JH, Waitumbi JN, Orago AS, Barbosa A, La-
29. Liang H, Sim BK, 1997. Conservation of structure and function of nar DE, Stoute JA, 2004. Measurement of antibody levels
the erythrocyte-binding domain of Plasmodium falciparum against region II of the erythrocyte-binding antigen 175 of
EBA-175. Mol Biochem Parasitol 84: 241–245. Plasmodium falciparum in an area of malaria holoendemicity
30. al-Yaman F, Genton B, Anders R, Taraika J, Ginny M, Mellor S, in western Kenya. Infect Immun 72: 735–741.
Alpers MP, 1995. Assessment of the role of the humoral re- 39. Okenu DM, Riley EM, Bickle QD, Agomo PU, Barbosa A,
sponse to Plasmodium falciparum MSP2 compared to RESA Daugherty JR, Lanar DE, Conway DJ, 2000. Analysis of hu-
and SPf66 in protecting Papua New Guinean children from man antibodies to erythrocyte binding antigen 175 of Plasmo-
clinical malaria. Parasite Immunol 17: 493–501. dium falciparum. Infect Immun 68: 5559–5566.