MAJOR ARTICLE

Dynamics of Anti-VAR2CSA Immunoglobulin G

Response in a Cohort of Senegalese Pregnant Women
N. G. Tuikue Ndam,1 A. Salanti,2 J.-Y. Le-Hesran,1 G. Cottrell,1 N. Fievet,1 L. Turner,2 S. Sow,3 J.-M. Dangou,4
T. Theander,2 and P. Deloron1
1
Institut de Recherche pour le Développement, UR010, Mother and Child Health in the Tropics, Université Paris Descartes, Paris, France;
2
Centre for Medical Parasitology at Rigshospitalet and University of Copenhagen, Copenhagen, Denmark; 3Thiadiaye Hospital, Thiadiaye,
and 4Histopathology Laboratory, Cheick Anta Diop University, Dakar, Senegal

Downloaded from http://jid.oxfordjournals.org/ at Chulalongkorn University on June 4, 2014

Background. Pregnancy-associated malaria (PAM) is precipitated by the accumulation of parasites in the
placental intervillous spaces and causes maternal anemia and low birth weight. In PAM, placental parasites adhere
to chondroitin sulfate A (CSA) through a unique set of variant surface antigens (VSAPAM). Several studies have
shown that 1 var gene, var2csa, is transcribed at high levels and expressed in CSA-binding Plasmodium falciparum
parasites.
Methods. Plasma levels of anti-VAR2CSA immunoglobulin G (IgG) in Senegalese women were measured during
pregnancy by enzyme-linked immunosorbent assay, using 3 recombinant proteins representing 3 domains of the
var2csa gene product.
Results. The 3 recombinant proteins were specifically recognized by plasma from pregnant women but not
by control plasma. A parity-dependent recognition pattern was observed with 2 of the 3 VAR2CSA antigens. A
kinetic study demonstrated that a single P. falciparum infection was able to trigger a VAR2CSA-specific antibody
response. Among women with infected placentas, women with high anti-VAR2CSA IgG levels at enrollment were
more likely to present with a past infection than with an acute/chronic infection.
Conclusions. Anti-VAR2CSA IgGs are involved in clinical protection against pregnancy-associated malaria and
strengthens the hope for making a VAR2CSA-based vaccine.

Malaria parasites are major human pathogens occurring
throughout the tropical world. In areas endemic for
parasite transmission, the consequences for the preg-
nant woman (maternal anemia) and her fetus (reduc-
tion of birth weight) also highlight the importance of
this infection [1]. In sub-Saharan Africa, at least 25
million pregnant women are exposed to malaria every
year [2]. Pregnancy-associated malaria (PAM) is par-
ticularly frequent during first pregnancies in women
living in areas of intense falciparum malaria transmis-
Received 28 June 2005; accepted 29 August 2005; electronically published 30
January 2006.
Potential conflicts of interest: none reported.
Financial support: Commission of the European Union (grant QLK2-CT-2001-
01302, PAMVAC and QLK2-CT-2002-01197, EUROMALVAC2); Danish Medical Re-
search Council (Statens Sundhedsvidenskabelige Forskningsråd; grants 22-02-0220
and 22-03-0333); French Ministry of Research (ACI Pal+/2001). N.G.T.N. is sup-
ported by PhD studentships from the French Ministry of Education and Research.
A.S. was supported by PhD studentships from the Gates Malaria Partnership.
Reprints or correspondence: Dr. Philippe Deloron, Institut de Recherche pour le
Développement, UR010, Paris, France (Philippe.Deloron@ird.fr).
The Journal of Infectious Diseases 2006; 193:713–20
 2006 by the Infectious Diseases Society of America. All rights reserved.
0022-1899/2006/19305-0014$15.00
sion [3]. PAM is caused by Plasmodium falciparum, which
expresses unique variant surface antigens (VSAPAM) that
allow the parasite to sequester in the placenta [4] by
binding to chondroitin sulfate proteoglycan (CSPG)
receptors on syncytiotrophoblasts [5]. Women who
have had PAM develop VSAPAM-specific antiadhesive
antibodies, which are associated with protection from
malaria during subsequent pregnancies [6, 7]. These
protective antibodies are thought to recognize reason-
ably conserved parasite antigens, because serum and
parasites from pregnant women from different malarial
areas cross-react [8]. These observations have raised
hope for the development of a vaccine to prevent PAM.
Adhesion of infected erythrocytes to host endothelium
and uninfected erythrocytes has been shown to be me-
diated by var-encoded members of the P. falciparum
erythrocyte membrane protein–1 (PfEMP1) family [9,
10]. Later, it was shown that var2csa, but not var1csa,
both having been previously associated with the chon-
droitin sulfate A (CSA)–binding phenotype, was up-
regulated in parasite strains selected for CSA binding
[11, 12] as well as in field placental parasite isolates

Anti-VAR2CSA IgG Response • JID 2006:193 (1 March) • 713

[13]. The corresponding VAR2CSA protein was shown to be
expressed on the surface of infected erythrocytes, and anti-
VAR2CSA IgGs were specifically found in pregnant women [13,
14]. In the present study, we measured the anti-VAR2CSA IgG
response in a cohort of Senegalese pregnant women and an-
alyzed the acquisition of VAR2CSA-specific IgGs during the
course of pregnancy, in relation to clinical and parasitological
follow-up.
SUBJECTS, MATERIALS, AND METHODS
Study site, field surveillance, and plasma sample collection.
The present study was conducted in Thiadiaye Hospital (Thia-
diaye is a town situated 130 km east of Dakar, the political capital
of Senegal) and in 2 corresponding health centers in the neigh-
boring villages of Fissel and Sandiara. Malaria there is seasonal-
ly transmitted, during the rains from September to December,
with an estimate of 10 infected bites/individual/year [15].
Pregnant women were enrolled in a cohort study during an
antenatal consultation (ANC) during the period 30 July–15
October 2001. Women !6 months pregnant were enrolled if
(1) they were not infected with malaria parasites at the time
of inclusion and declared not to have had malaria since being
pregnant and (2) they were likely to be exposed to infective
mosquito bites during their pregnancy. A total of 306 pregnant
women (found to be negative for malaria by immunochro-
matographic test and thick blood smear) with a mean  SD
age of 24.1  6.1 years were followed for evidence of malaria
by active and passive detection through monthly ANC visits
and through weekly home visits until delivery. At each ANC
visit, a blood sample was obtained from all women, regardless
of whether they presented with fever. During home visits, a
capillary blood sample was collected from women presenting
with fever (axillary temperature 137.5C). All women present-
ing with fever and a positive thick blood smear were given
curative treatment with chloroquine, the first-line antimalar-
ial drug in Senegal at that time. However, most (72/119) of the
malarial infections were symptomless. At delivery, peripheral
and placental blood was investigated by microscopy for the
presence of malaria parasites. Placental biopsy specimens were
also collected and processed. Malarial infections were classified
as described elsewhere [16]. Plasma was isolated and stored at
⫺20C until use. Blood samples from 9 nulligravid women and
8 men living in the same area were also collected.
The ethical committee, Ministry of Health, Senegal, provided
ethics approval for this research. All participants gave informed
consent.
Flow cytometry. Six P. falciparum isolates were collected in
November 2003 from infected placentas from women in ma-
ternity wards in Guediawaye, a suburb of Dakar, who had type
O red blood cells [17]. Antibodies (IgG) to VSAs expressed by
these 6 placental isolates were measured in the plasma samples
714 • JID 2006:193 (1 March) • Tuikue Ndam et al.
collected at enrollment and delivery by flow cytometry. Samples
were coded and subjected to blinded testing and analysis, using
a FACSCalibur cytometer (Becton Dickinson) [17]. The level
of IgG recognizing VSAs was expressed as median fluorescence
intensity (MFI) in a channel of ethidium bromide–labeled in-
fected erythrocytes. For each isolate, the threshold for positivity
was defined as 2 SDs above the mean MFI from 30 nulligravid
women from Thiadiaye.
Cytoadhesion inhibition assays. CSPG purified from hu-
man placental tissue was coated as circular spots onto Falcon
Petri dishes (Becton Dickinson), as described elsewhere [17].
Serum samples from 30 pregnant women, selected for their
VSA recognition ability, were used in CSPG binding competi-
tion assays against the 6 placental parasite isolates. Antiadhe-
sion activity was expressed as the percentage of binding com-
Downloaded from http://jid.oxfordjournals.org/ at Chulalongkorn University on June 4, 2014
pared with the adhesion control, defined by the mean adhesion
occurring with plasma from 6 Senegalese nulligravid women.
This binding did not differ significantly from the binding ob-
tained in assays without plasma.
Recombinant antigens. Three domains (DBL1-x, DBL5-␧,
and DBL6-␧) of the synthetic var2csa gene previously generated
[14] were cloned into the pBAP-TOPO vector (Invitrogen) by
polymerase chain reaction. Recombinant proteins from these
constructs were produced in baculovirus-infected Sf9 cells and
purified as described elsewhere [18].
ELISA. After the optimal coating concentration for each
VAR2CSA antigen was determined, plasma levels of specific
IgG were measured by ELISA. A recombinant MSP1 (yPfMSP1-
19) protein was used for measuring antibodies against a P.
falciparum antigen not related to PAM. The optical density was
converted into arbitrary units (AUs), as described elsewhere
[19]. Antibody responders were defined as those having an
antibody level (in AUs) 12 SDs above the mean absorbance
yielded by 55 negative control plasma samples from unex-
posed French pregnant women. A pool of these plasma sam-
ples was used as a negative control, whereas a pool of plasma
samples from multigravid pregnant Senegalese women, pre-
viously demonstrated to have high levels of anti-VSA IgGs
against placental isolates, was used as a positive control. A
group of 20 negative control samples from unexposed French
men were also included in the study.
Statistical analysis. Differences between groups were tested
by the appropriate nonparametric Mann-Whitney test. The x2
test for unpaired samples (or Fisher’s exact test, when required)
was used to test for differences between categorical variables.
Correlations (r) were assessed by Spearman test. Associations
between the level of antibodies and placental infection when
controlling for parity were tested using multivariate logistic re-
gression. For paired analysis, absolute differences in antibody
levels between time points were calculated for each pair of
samples. The variations in antibody levels between time points
 Downloaded from http://jid.oxfordjournals.org/ at Chulalongkorn University on June 4, 2014
Figure 1. Sex and pregnancy specificity of plasma IgG reactivity to VAR2CSA recombinant proteins. ELISA was performed on DBL1-x–, DBL5-␧–,
and DBL6-␧–coated plates. Plasma levels of IgG are expressed as arbitrary units (AUs) and are shown for unexposed men (n p 20 ) and pregnant
women (n p 55) from France, malaria-exposed Senegalese men (n p 9 ), nonpregnant nulligravid Senegalese women (n p 8 ), and pregnant women
whose blood was collected at enrollment early during pregnancy (n p 265 ) and at delivery (n p 275 ). The top, bottom, and line through the middle
of the box correspond to the 75th percentile, 25th percentile, and 50th percentile (median), respectively. The whiskers extend from the 10th percentile
to the top 90th percentile.

were tested by the Wilcoxon signed rank test for matched pairs. single peripheral thick blood smear was positive during the
The significance limit was P ! .05 . SAS (version 8.2; SAS In- entire follow-up for 54 of the 119 women; 14 of the 119 also
stitute) software was used. presented at delivery with an active placental infection, and 5
presented with a past infection. Twenty-five other women pre-
RESULTS sented with an active placental infection but either never were
found to be infected (n p 10) or were found to be positive
Study population. Of the 306 women enrolled, 12 were lost to
more than once (n p 15) before delivery.
follow-up, and the delivery serum sample was lacking for 19.
Serological reactivity to VAR2CSA domains. The 3 recom-
Plasma samples were available from 275 pregnant women at
binant VAR2CSA domains were specifically recognized by ELISA
delivery, of whom 60 were primigravid, 52 were secundigravid,
in plasma samples from pregnant women (figure 1), and the
and 163 were multigravid. Plasma samples were available from
265 of these 275 women at enrollment, making paired analysis
possible. Placental histological data were available for 226 wom-
en; of these women, 172 were found to be negative for infec-
tion, whereas 54 presented with signs of placental infection.
Malarial infection during follow-up and at delivery.
During follow-up, 119 women experienced at least 1 positive
peripheral parasitemia episode between enrollment and deliv-
ery. Of these women, 47 also presented with fever at this time
and received a curative regimen of chloroquine (the drug ad-
vocated in Senegal at the time of study for both prophylaxis
and treatment). At delivery, placental infection was observed
in 15% of women by microscopy and in 24% by a combination
of microscopy and histologic analysis. The prevalence rate of Figure 2. Parity dependency of plasma IgG reactivity to VAR2CSA re-
placental infection was highest in primigravid women (28%), combinant proteins. ELISA was performed on DBL1-x–, DBL5-␧–, and DBL6-
gradually decreased to the 15%–22% range between the second ␧–coated plates. Plasma levels of IgG are expressed as arbitrary units (AUs)
and are shown at enrollment for primigravid (n p 60), secondigravid (n
and fourth pregnancies, then reached very low rates (2%) after
p 52), and multigravid (n p 163) women from Senegal. The top, bottom,
the fifth pregnancy. Of the 54 women with a placental infec- and line through the middle of the box correspond to the 75th percentile,
tion, 25 presented with an acute infection, 14 presented with 25th percentile, and 50th percentile (median), respectively. The whiskers
a chronic infection, and 15 presented with a past infection. A extend from the 10th percentile to the top 90th percentile.

Anti-VAR2CSA IgG Response • JID 2006:193 (1 March) • 715

Figure 3. Relationship between placental histologic results and plasma anti-VAR2CSA and anti-MSP1 IgG levels at enrollment. Data are segregated
into acute/chronic (A/C) infection (n p 39 ) and past infection (n p 15 ). The top, bottom, and line through the middle of the box correspond to the
75th percentile, 25th percentile, and 50th percentile (median), respectively. The whiskers on the bottom extend from the 10th percentile to the top
90th percentile. AU, arbitrary units.
plasma antibody levels were higher in pregnant women than in
control groups of European adults not exposed to malaria (P !
.0001 for all comparisons). In the individual pregnant women,
there was a tight correlation between the antibody levels to the
3 VAR2CSA domains (0.6 ! r ! 0.8; P ! .0001 for all compari-
sons). The relationship between pregnancy and VAR2CSA ex-
pression was demonstrated by the markedly higher levels of anti–
DBL5-␧ and anti–DBL6-␧ antibodies in pregnant women than
in the men and nulligravid women from Senegal (P ! .001 for
all comparisons). The plasma levels of anti-MSP1 antibodies were
higher in the malaria-exposed individuals than in the donors
from France, but there were no statistically significant differences
in the plasma levels of anti-MSP1 antibodies between the dif-
ferent groups of donors from Senegal (data not shown).
A parity-dependent recognition profile of IgG against 2 of the
3 recombinant domains was seen both at enrollment (DBL5-␧,
r p 0.23, P p .03; DBL6-␧, r p 0.20, P p .03) and at delivery
(DBL5-␧, r p 0.27, P p .03; DBL6-␧, r p 0.25, P p .03) (fig-
ure 2). No parity dependency was seen with anti-MSP1 IgG levels.
Plasma samples from these pregnant women were also assessed
for PAM-specific anti-VSA IgG, using 6 freshly collected placen-
Table 1. Levels of antibody to 3 VAR2CSA domains (DBL1-x, DBL5-␧, and DBL6-␧) in
a cohort of pregnant women from Senegal.
Group, period
Uninfected during pregnancy (n p 156)
Enrollment
Delivery
Infected during pregnancy (n p 119)
Enrollment
Delivery
NOTE. AU, arbitrary units.
a
P ! .02, enrollment vs. delivery.
716 • JID 2006:193 (1 March) • Tuikue Ndam et al.
Downloaded from http://jid.oxfordjournals.org/ at Chulalongkorn University on June 4, 2014
tal isolates. The levels of specific anti-VSA IgG, as measured
by flow cytometry, correlated with all 3 VAR2CSA-specific IgG
levels (0.3 ! r ! 0.7; P ! .0001 for all comparisons).
To examine the relationship between antibody level and pla-
cental infection, women were divided into those who presented
with an infected placenta at delivery and those in whom there
was no evidence of placental infection. At delivery, levels of
antibody to the recombinant VAR2CSA domains were higher
in women with infected placentas than in those with uninfected
placentas (P ! .001 for all comparisons). No such relationship
was observed in plasma samples collected early during preg-
nancy. However, of the women presenting with a placental
infection, women with anti-VAR2CSA IgG levels at enrollment
that were above the median were more likely to present with
a past infection than with an acute/chronic infection (DBL1-
x, P p .03; DBL5-␧, P p .13; DBL6-␧, P p .03 [x2 test]). Sim-
ilarly, the mean levels of antibodies at enrollment were lower
when parasites were present in the placenta (acute/chronic in-
fection) than when they were absent (past infection) (DBL1-
x, P p .09; DBL5-␧, P p .07; DBL6-␧, P p .06) (figure 3). This
pattern was not observed with anti-MSP1 antibody. No sig-
Antibody level, mean AU (% of subjects)
DBL1-x DBL5-␧ DBL6-␧
17.9 (44.0) 26.3 (51.0) 28.1 (60.0)
19.9 (41.0) 20.1 (28.0)a 27.3 (52.0)
23.1 (46.0) 27.2 (47.0) 28.8 (60.0)
27.3 (53.5)a 38.4 (52.0)
a
38.7 (77.0)
a

Figure 4. Acquisition and decay of anti-VAR2CSA IgGs after a Plas-
modium falciparum infection during the course of pregnancy. Day 0 cor-
responds to the time that peripheral parasitemia was detected. Squares
denote DBL1-x, triangles denote DBL5-␧, and Xs denote DBL6-␧. The first
point of each graph corresponds to enrollment, and the last point cor-
responds to delivery. a, b, and c, Representative examples of how antibody
levels changed in time in individual women. In panel a, antibody levels
increase sharply at the time of infection and remain high (representative
of 16 women). In panel b, antibody levels increase and decrease (rep-
resentative of 26 women). In panel c, antibody levels are stable (rep-
resentative of 7 women). AU, arbitrary units.
nificant relationship was found between plasma levels of anti-
VAR2CSA IgGs and maternal anemia or birth weight of the
offspring either at enrollment or at delivery.
Among the 119 women who experienced at least 1 P. falci-
parum infection during their pregnancy, the levels of anti-
VAR2CSA IgGs were higher at delivery than at enrollment (P
! .001 for all comparisons, Mann-Whitney paired samples rank
sum test), even after parity was controlled for (DBL1-x, P p
.02; DBL5-␧, P ! .0001; DBL6-␧, P p .004). The levels of an-
tibodies against all 3 VAR2CSA domains were similar at delivery
among the 47 women who had received chloroquine treatment
and among the 72 women who did not (P 1 .74 for all com-
parisons). In women who never presented with a positive thick
blood smear during follow-up, anti-VAR2CSA IgG levels did
not change in a systematic pattern (table 1).
VAR2CSA-like immunogenic epitopes carried by PAM
parasites. The longitudinal study design made it possible to
monitor anti-VAR2CSA IgG levels before and after the detec-
tion of parasites in the 49 women in whom 1 peripheral in-
fection was detected during pregnancy. Three general pat-
terns were detected. In some women, the antibody levels
increased after detection of the infection and subsequently re-
mained fairly stable (16 women; figure 4a). In other women,
there was an increase in levels at the time of infection, followed
by a marked decrease later during pregnancy (26 women; figure
4b). The last pattern was found in 7 women (figure 4c), who
Downloaded from http://jid.oxfordjournals.org/ at Chulalongkorn University on June 4, 2014
had rather high anti-VAR2CSA IgG levels before infection, and
there was no change in the anti-VAR2CSA IgG levels in con-
nection to the demonstration of parasitemia. It is possible that
these women were infected with non-PAM parasites, because
none presented with a placental infection at delivery. In general,
antibody levels decreased more often when the infection oc-
curred during the second trimester, but this was probably the
consequence of the disappearance of parasites, because de-
creasing VAR2CSA levels were not observed in women pre-
senting with an active placental infection (table 2). The peak
in antibody levels often coincided with the detection of the
parasites (figure 4b) but also occurred after this point in time
(figure 4a). This delay in the occurrence of the boost was more
frequent in primigravid women than in multigravid women,
probably indicating a slower development of the primary re-
sponse to VAR2CSA in these women.
Relationship between anti-VAR2CSA IgGs and antibodies
inhibiting the in vitro adhesion of some placental isolates.
All 6 placental isolates bound to CSA and CSPG and tran-
scribed high levels of the var2csa gene [13, 17]. In plasma
samples from 30 pregnant women, selected according to their
VSA recognition, the level of anti-VAR2CSA IgGs correlated
with those of antibodies inhibiting the adhesion of 2 of these
6 isolates (table 3). The lack of correlation found for the oth-
er 4 isolates suggests a possible heterogeneity between pla-
cental isolates. Similar associations were observed between cy-
tometry-based data and adhesion inhibition activity of the
plasma samples studied [17].
DISCUSSION
Plasma antibodies from pregnant women recognize VSA of CSA-
selected lines or placental isolates in a parity-dependent man-
ner [17, 20]. These responses were measured toward the entire
repertoire of VSA present on the surface of infected erythro-
cytes. PAM parasites express antigenically distinct VSA, and
recent studies have shown that parasites expressing VAR2CSA,
Anti-VAR2CSA IgG Response • JID 2006:193 (1 March) • 717
 Table 2. Changes in the levels of anti-VAR2CSA IgGs during pregnancy, measured in 49 Sen-
egalese women in whom a single thick blood smear was found to be positive during follow-up,
grouped according to placental histologic assessment at delivery.

Change in antibody level,

a
mean AU (P value )
Between enrollment Between antibody
Group, antibody and antibody peak peak and delivery
No active placental infection at delivery (n p 40)
DBL1-x 11.8 (.005) ⫺11.9 (.001)
DBL5-␧ 25.8 (.001) ⫺25.2 (!.0001)
DBL6-␧ 12.2 (.004) ⫺14.2 (!.001)
b
Active placental infection at delivery (n p 9)
DBL1-x 13.7 (.04) ⫺2.2 (.55)
DBL5-␧ 27.7 (.03) 3.8 (.68)
DBL6-␧ 15.1 (.02) ⫺0.6 (.64)

NOTE. AU, arbitrary units.

Downloaded from http://jid.oxfordjournals.org/ at Chulalongkorn University on June 4, 2014

a
Wilcoxon matched pairs signed rank test.
b
Five women were excluded because of an insufficient no. of plasma samples.

a member of the PfEMP1 family, on the surface of infected
erythrocytes have the VSAPAM phenotype, which is characteristic
of placental parasites [11, 13, 14]. Unlike many var genes, the
var2csa gene shows a high level of homology in genetically
distinct isolates [12, 21]. In the present study, 3 individual
domains of the VAR2CSA protein were used to measure specific
antibodies present in plasma samples from pregnant women.
The recognition profiles of all 3 domains correlated with those
of the VSA expressed by 6 placental isolates. The difference in
levels of antibody to DBL5-␧ and DBL6-␧ between Senegalese
pregnant women, on the one hand, and Senegalese nulligravid
women and men, on the other, suggests that antibodies against
VAR2CSA are acquired during pregnancy; this is in line with
previous data regarding parity dependency of PAM-specific and
protective antibodies [6, 7]. However, the present study is the
first, to our knowledge, to investigate the development of anti-
VAR2CSA IgGs over the course of pregnancy, and the results
further strengthen the hypothesis that VAR2CSA is specifically
expressed by placental parasites. Anti-VAR2CSA IgG responses
increased between enrollment and delivery in women who had
a P. falciparum infection during their pregnancy, suggesting that
these women had been infected with parasites expressing epi-
topes similar to those expressed on VAR2CSA domains. The
decay of anti-VAR2CSA IgG after infection in the subgroup of
women with a single peripheral infection during pregnancy and
with no active placental infection (table 2) suggests that anti-
VAR2CSA IgGs are involved in parasite clearance—a notion
supported by the finding that high levels of anti-VAR2CSA IgGs
were associated with reduced subsequent risk of active placental
infections.
VSAPAM-specific IgGs seem to mediate protection against
PAM [6, 7, 8, 22], and women with chronic placental infec-
tion and low VSAPAM IgG levels are much more likely to be
anemic and to have babies with low birth weight than are
women with high antibody levels [7]. In a subsequent study
of the same women, similar associations were found be-
tween the anti–DBL5-␧ VAR2CSA IgG levels at delivery and
birth weight [14]. The lack of association between the levels of
anti-VAR2CSA IgGs at enrollment and clinical consequences
of PAM in this study is probably due to the limited number
of women presenting with chronic placental infection. The lack
of protective association with antibody levels at delivery may
also be due to epidemiological differences in malaria between
study sites, because kinetics of anti-VSA antibodies is likely to
differ between areas of stable/intense or low/seasonal trans-
mission. Antibody levels at delivery in our study were mostly
markers of an active placental infection, because infections oc-
curring close to the time of delivery were more often associated
with an acute infection in the placenta. In these women, es-
pecially in women with lower parity who are experiencing pri-
mary immune responses, the kinetics of the antibody response
may be still in the ascending period. Delivery may, thus, not
be the best time point at which to assess the protective role of
antibodies, particularly in low-transmission areas. In contrast,
Table 3. Association between plasma levels of anti-VAR2CSA
IgGs and the ability of plasma to inhibit the binding of placen-
tal isolates (47, 48, 49, 51, 53 and 55) to chondroitin-sulfate
proteoglycans.
Antibody 47 48 49 51 53 55
DBL1-x 0.02 0.32 ⫺0.21 ⫺0.01 0.37 0.2
DBL5-␧ 0.03 0.03 0.02 ⫺0.02 0.51 0.39
DBL6-␧ ⫺0.07 ⫺0.04 ⫺0.09 ⫺0.01 0.50 0.43
NOTE. Data are Spearman correlation coefficients (r). Significant (P ! .05)
results are shown in bold.

718 • JID 2006:193 (1 March) • Tuikue Ndam et al.

the level of antibodies at the beginning of pregnancy may yield
relevant information concerning the protectiveness of the an-
tibodies. Women with high levels of anti-VAR2CSA IgGs at
enrollment were likely to better control placental infections
than were those with lower levels. Anti-VAR2CSA IgGs can
persist for several months, as is shown by the parity dependen-
cy of the prevalence rates of such antibodies in women at
enrollment (a period free of transmission). The presence of
detectable amounts of such antibodies in some primigravid
women at enrollment was surprising. It may be hypothesized
that, once the placenta appears, it is able to select parasites with
the PAM phenotype from the parasite populations normally
present at the submicroscopic level in most individuals in areas
in which malaria is endemic [23, 24], even during periods of
nontransmission. This selection may generate small waves of
parasite populations that probably have low fitness but are able
to stimulate the immune system, since the measures yielded
parity-dependent profiles. Evidence from in vitro adhesion in-
hibition of some, but not all, placental isolates suggests that
anti-VAR2CSA IgGs can interfere with CSA-binding sites of
these isolates. The lack of a relationship between anti-VAR2CSA
IgGs and antiadhesion antibodies in 4 of 6 of these isolates
suggests the existence of heterogeneity among placental iso-
lates. This heterogeneity may involve multiple antigenic VSAs
with affinity to CSA or simply a polymorphism within the
var2csa gene product. VAR2CSA protein from CSA-adherent
parasites possesses multiple CSA-binding domains, and a do-
main carrying this property in 1 strain may not have the same
ability in another strain, as was recently shown for the DBL3-
X domain of the A4 and 3D7 strains [25]. This functional het-
erogeneity among VAR2CSA domains may explain differences
and similarities among placental isolates. Since it is likely that
isolates can use different VAR2CSA domains for binding CSA,
epitopes involved in CSA binding may differ between isolates.
In the Gamain et al. [25] study, DBL2-x, DBL3-x, and DBL6-␧
were shown to be able to bind CSA. In the present study, ad-
hesion-inhibitory antibody levels correlated with those of anti–
DBL5-␧ and anti–DBL6-␧ antibodies for 2 isolates, suggesting
that both domains carry CSA-binding sites. Alternatively, only 1
domain may carry the ligand, but the reactivities of both do-
mains may be linked because of their architectural proximity
in the protein. The finding that one particular var gene
(var2csa) is highly transcribed by placental isolates [13] and by
several CSA-selected strains [11, 12, 26] indicates that the rep-
ertoire of PfEMP1 molecules involved in PAM is probably re-
stricted, although isolates may use similar or different do-
mains for interaction with CSA. This is in line with the in-
creasing protection that corresponds to increasing gravidity
[27]. Taken together, because clinical consequences of PAM
are likely to be attributable to active chronic infection in the
placenta [7], high levels of antibodies to VAR2CSA early during
pregnancy are likely to contribute to the mechanisms of pro-
tection, possibly by limiting parasite adhesion to the placenta
or by a cytophilic process clearing the parasites that sequestered
in the placenta through other putative mechanisms. An ex-
panded panel of VAR2CSA domain constructs representing dif-
ferent epitopes will allow a detailed study of human antibodies
to polymorphic or conserved regions of the VAR2CSA protein,
as well as an analysis of their relative importance in protective
immunity. Such analyses will be useful for the logical design
of a VAR2CSA vaccine that will be suitable for efficient use in
different malaria-endemic populations.
Acknowledgments
We are indebted to all of the individuals who participated in this study,
especially the mothers. The support of the maternity staff of Thiadiaye
Downloaded from http://jid.oxfordjournals.org/ at Chulalongkorn University on June 4, 2014
Hospital in sample collection is acknowledged, as is the work of the field
team of the Institut de Recherche pour le Développement research unit in
Senegal. We thank Sandrine Houzé for plasma sample collection from
unexposed individuals at Hôpital Bichat-Claude Bernard, Paris.
References
1. Brabin BJ, Romagosa C, Abdelgalil S, et al. The sick placenta—the role
of malaria. Placenta 2004; 25:359–78.
2. World Health Organization. A strategic framework for malaria pre-
vention and control during pregnancy in African region. Report ARF/
MAL/04/01. Brazzaville: World Health Organization, 2004.
3. Brabin BJ. An analysis of malaria in pregnancy in Africa. Bull World
Health Organ 1983; 61:1005–16.
4. Fried M, Duffy PE. Adherence of Plasmodium falciparum to chondro-
itin sulfate A in the human placenta. Science 1996; 272:1502–4.
5. Achur RN, Valiyaveettil M, Alkhalil A, Ockenhouse CF, Gowda DC.
Characterization of proteoglycans of human placenta and identification
of unique chondroitin sulfate proteoglycans of the intervillous spaces
that mediate the adherence of Plasmodium falciparum-infected eryth-
rocytes to the placenta. J Biol Chem 2000; 275:40344–56.
6. Duffy PE, Fried M. Antibodies that inhibit Plasmodium falciparum ad-
hesion to chondroitin sulfate A are associated with increased birth weight
and the gestational age of newborns. Infect Immun 2003; 71:6620–3.
7. Staalsoe T, Shulman CE, Bulmer JN, Kawuondo K, Marsh K, Hviid L.
Variant surface antigen-specific IgG and protection against clinical con-
sequences of pregnancy-associated Plasmodium falciparum malaria. Lan-
cet 2004; 363:283–9.
8. Fried M, Duffy PE. Maternal malaria and parasite adhesion. J Mol
Med 1998; 76:162–71.
9. Baruch DI, Ma XC, Singh HB, Bi X, Pasloske BL, Howard RJ. Iden-
tification of a region of PfEMP1 that mediates adherence of Plasmo-
dium falciparum infected erythrocytes to CD36: conserved function
with variant sequence. Blood 1997; 90:3766–75.
10. Rowe JA, Moulds JM, Newbold CI, Miller LH. P. falciparum rosetting
mediated by a parasite-variant erythrocyte membrane protein and com-
plement-receptor 1. Nature 1997; 388:292–5.
11. Duffy MF, Byrne TJ, Elliott SR, et al. Broad analysis reveals a consistent
pattern of var gene transcription in Plasmodium falciparum repeated-
ly selected for a defined adhesion phenotype. Mol Microbiol 2005; 56:
774–88.
12. Salanti A, Staalsoe T, Lavstsen T, et al. Selective upregulation of a single
distinctly structured var gene in chondroitin sulphate A-adhering Plas-
modium falciparum involved in pregnancy-associated malaria. Mol Mi-
crobiol 2003; 49:179–91.
13. Tuikue Ndam NG, Salanti A, Bertin G, et al. High level of var2csa
Anti-VAR2CSA IgG Response • JID 2006:193 (1 March) • 719
 transcription by Plasmodium falciparum isolated from the placenta. J
Infect Dis 2005; 192:331–5.
14. Salanti A, Dahlbäck M, Turner L, et al. Evidence for the involvement
of VAR2CSA in pregnancy-associated malaria. J Exp Med 2004; 200:
1197–203.
15. Robert V, Dieng H, Lochouran L, et al. [Malaria transmission in the
rural zone of Niakhar, Senegal]. Trop Med Int Health 1998; 3:667–77.
16. Scott S, Cumberland P, Shulman CE, et al. Neonatal measles immunity
in rural Kenya: the influence of HIV and placental malaria infections
on placental transfer of antibodies and levels of antibody in maternal
and cord serum samples. J Infect Dis 2005; 191:1854–60.
17. Tuikue Ndam NG, Fievet N, Bertin G, Cottrell G, Gaye A, Deloron P.
Variable adhesion abilities and overlapping antigenic properties in pla-
cental Plasmodium falciparum isolates. J Infect Dis 2004; 190:2001–9.
18. Jensen AT, Magistrado P, Sharp S, et al. Plasmodium falciparum as-
sociated with severe childhood malaria preferentially expresses PfEMP1
encoded by group A var genes. J Exp Med 2004; 199:1179–90.
19. Rasheed FN, Bulmer JN, De Francisco A, et al. Relationships between
maternal malaria and malarial immune responses in mothers and neo-
nates. Parasite Immunol 1995; 17:1–10.
20. Beeson JG, Mann EJ, Elliott SR, et al. Antibodies to variant surface
antigens of Plasmodium falciparum–infected erythrocytes and adhe-
sion inhibitory antibodies are associated with placental malaria and
have overlapping and distinct targets. J Infect Dis 2004; 189:540–51.
21. Kraemer SM, Smith JD. Evidence for the importance of genetic struc-
720 • JID 2006:193 (1 March) • Tuikue Ndam et al.
turing to the structural and functional specialization of the Plasmodi-
um falciparum var gene family. Mol Microbiol 2003; 50:1527–38.
22. O’Neil-Dunne I, Achur RN, Agbor-Enoh ST, et al. Gravidity-depend-
ent production of antibodies that inhibit binding of Plasmodium falcip-
arum-infected erythrocytes to placental chondroitin sulfate proteogly-
can during pregnancy. Infect Immun 2001; 69:7487–92.
23. Adam I, A-Elbasit IE, Salih I, Elbashir MI. Submicroscopic Plasmodi-
um falciparum infections during pregnancy, in an area of Sudan with a
low intensity of malaria transmission. Ann Trop Med Parasitol 2005; 99:
339–44.
24. Walker-Abbey A, Djokam RR, Eno A, et al. Malaria in pregnant Ca-
meroonian women: the effect of age and gravidity on submicroscopic
and mixed-species infections and multiple parasite genotypes. Am J
Trop Med Hyg 2005; 72:229–35.
25. Gamain B, Trimnell AR, Scheidig C, Scherf A, Miller LH, Smith JD.
Identification of multiple chondroitin sulfate A (CSA)–binding do-
mains in the var2CSA gene transcribed in CSA-binding parasites. J
Infect Dis 2005; 191:1010–3.
26. Elliott SR, Duffy MF, Byrne TJ, et al. Cross-reactive surface epitopes
Downloaded from http://jid.oxfordjournals.org/ at Chulalongkorn University on June 4, 2014
on chondroitin sulfate A-adherent Plasmodium falciparum-infected
erythrocytes are associated with transcription of var2csa. Infect Immun
2005; 73:2848–56.
27. Brabin BJ, Johnson PM. Placental malaria and pre-eclampsia through
the looking glass backwards? J Reprod Immunol 2005; 65:1–15.