ARTICLE
https://doi.org/10.1038/s41467-021-26118-w OPEN
Afucosylated Plasmodium falciparum-specific IgG is
induced by infection but not by subunit vaccination
Mads Delbo Larsen 1,2,12, Mary Lopez-Perez 3,12, Emmanuel Kakra Dickson4, Paulina Ampomah5,
Nicaise Tuikue Ndam6, Jan Nouta7, Carolien A. M. Koeleman7, Agnes L. Hipgrave Ederveen 7,
Benjamin Mordmüller8,9, Ali Salanti 3, Morten Agertoug Nielsen 3, Achille Massougbodji10,
C. Ellen van der Schoot1,2, Michael F. Ofori4, Manfred Wuhrer 7, Lars Hviid 3,11,13✉ &
Gestur Vidarsson 1,2,13✉
Plasmodium falciparum erythrocyte membrane protein 1 (PfEMP1) family members mediate
receptor- and tissue-specific sequestration of infected erythrocytes (IEs) in malaria. Antibody
responses are a central component of naturally acquired malaria immunity. PfEMP1-specific
IgG likely protects by inhibiting IE sequestration and through IgG-Fc Receptor (FcγR)
mediated phagocytosis and killing of antibody-opsonized IEs. The affinity of afucosylated IgG
to FcγRIIIa is up to 40-fold higher than fucosylated IgG, resulting in enhanced antibody-
dependent cellular cytotoxicity. Most IgG in plasma is fully fucosylated, but afucosylated IgG
is elicited in response to enveloped viruses and to paternal alloantigens during pregnancy.
Here we show that naturally acquired PfEMP1-specific IgG is strongly afucosylated in a stable
and exposure-dependent manner, and efficiently induces FcγRIIIa-dependent natural killer
(NK) cell degranulation. In contrast, immunization with a subunit PfEMP1 (VAR2CSA) vac-
cine results in fully fucosylated specific IgG. These results have implications for under-
standing protective natural- and vaccine-induced immunity to malaria.
1 Department of Experimental Immunohematology, Sanquin Research, Amsterdam, The Netherlands. 2 Landsteiner Laboratory, Amsterdam UMC, University
of Amsterdam, Amsterdam, The Netherlands. 3 Centre for Medical Parasitology, Department of Immunology and Microbiology, Faculty of Health and
Medical Sciences, University of Copenhagen, Copenhagen, Denmark. 4 Department of Immunology, Noguchi Memorial Institute for Medical Research,
University of Ghana, Accra, Ghana. 5 Department of Biomedical Sciences, School of Allied Health Sciences, University of Cape Coast, Cape Coast, Ghana.
6 Université de Paris, MERIT, IRD, 75006 Paris, France. 7 Center for Proteomics and Metabolomics, Leiden University Medical Center, Leiden, The
Netherlands. 8 Department of Medical Microbiology, Radboud University Medical Center, Nijmegen, The Netherlands. 9 Institut für Tropenmedizin,
Universitätsklinikum Tübingen, Tübingen, Germany. 10 Centre d’Etude et de Recherche sur le Paludisme Associé à la Grossesse et à l’Enfance (CERPAGE),
Faculté des Sciences de la Santé, Université d’Abomey-Calavi, Godomey, Benin. 11 Centre for Medical Parasitology, Department of Infectious Diseases,
Rigshospitalet, Copenhagen, Denmark. 12These authors contributed equally: Mads Delbo Larsen, Mary Lopez-Perez. 13These authors jointly supervised this
work: Gestur Vidarsson, Lars Hviid. ✉email: lhviid@sund.ku.dk; g.vidarsson@sanquin.nl
NATURE COMMUNICATIONS |         (2021) 12:5838 | https://doi.org/10.1038/s41467-021-26118-w |www.nature.com/naturecommunications 1
1234567890():,;
ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-26118-w
The most severe form of malaria is caused by the protozoan in the IgG-Fc imposed by the core-fucose
32. Even more strikingly,
parasite Plasmodium falciparum. The disease is currently IgG-afucosylation can convert a non-functional ADCC potential
estimated to cost around 400,000 lives a year, mostly of to strong and clinically significant responses33–36. Increased
young children and pregnant women in sub-Saharan Africa. In galactosylation at N297 can further enhance affinity to FcγRIII by
addition, nearly 900,000 babies are born with low birth weight as additional twofold, and also increases the complement activating
a consequence of placental malaria (PM)1. The particular viru- capacity of the antibody. In contrast, no influence of bisecting N-
lence of P. falciparum is related to the efficient adhesion of the acetylglucosamine (GlcNAc) on antibody effector functions has
infected erythrocytes (IEs) to host receptors in the vasculature, been demonstrated so far35.
such as endothelial protein C receptor, intercellular adhesion Fc fucosylation of plasma IgG is near 100% at birth, and
molecule 1, and oncofetal chondroitin sulfate A2–4, mediated by although it decreases slightly with age, it normally remains high
members of the protein family P. falciparum erythrocyte mem- (~94%) in adults37,38. Nevertheless, very marked and clinically
brane protein 1 (PfEMP1), embedded in the membrane of IEs5. significant reductions (down to ~10%) in antigen-specific IgG-Fc
The sequestration of IEs can cause tissue-specific circulatory fucosylation are frequently observed after alloimmunization
compromise and inflammation, which in turn can lead to severe against erythrocyte and platelet alloantigens39–41. Afucosylation
and life-threatening complications such as cerebral malaria (CM) has also been observed for antigen-specific IgG to various
and PM6,7. Severe malaria in children has repeatedly been shown enveloped viruses33,34,42. In human immunodeficiency virus
to be associated with parasites expressing particular subsets of (HIV) infections, low Fc fucosylation has been proposed as a trait
PfEMP1, such as Group A and B/A4,8, whereas PM is strongly of elite controllers42, but it is associated with FcγRIIIa-mediated
associated with parasites expressing VAR2CSA-type PfEMP1 immunopathology in SARS-CoV-2 and secondary dengue virus
(refs. 9,10). infections33,34,43. Vaccination with the attenuated para-
Acquired protective immunity to P. falciparum malaria is myxoviruses measles and mumps also results in specific IgG with
mainly mediated by IgG with specificity for antigens expressed by reduced fucosylation similar to that acquired after natural
the asexual blood-stage parasites11. PfEMP1 is a key target5, infection34. In contrast, infection with the non-enveloped par-
although antibodies to other blood-stage antigens such as the vovirus B19, protein subunit vaccination against hepatitis B virus,
merozoite-specific antigens glutamate-rich protein (GLURP), vaccination with inactivated influenza virus, or vaccination
merozoite surface protein 1, and reticulocyte-binding protein against tetanus, pneumococcal, and meningococcal disease do not
homolog 5 also contribute to naturally acquired protection12–14. induce selectively afucosylated IgG34,44,45.
Importantly, the selective protection from severe malaria that The above findings have led us to propose that afucosylated
develops early in childhood is related to acquisition of IgG spe- IgG has evolved as a beneficiary immune response to foreign
cific for Group A and B/A PfEMP1 variants8,15,16. As a result, antigens expressed on host membranes in the context of infec-
life-threatening complications are rare in teenagers and beyond in tions, which is mimicked in alloimmunizations with devastating
P. falciparum endemic regions. PM, which is caused by selective consequences34,40,41,46,47.
accumulation of VAR2CSA-positive IEs in the placenta from In this study, we show that antibody responses to P. falciparum
early in pregnancy17,18, constitutes an important exception to this antigens expressed on the IE surface are also a subject to afuco-
rule. Only VAR2CSA mediates adhesion to placenta-specific sylation. Specifically, we show this for naturally acquired IgG
chondroitin sulfate9,19. Because of this, and because antibodies responses to the PfEMP1 antigens VAR6 and VAR2CSA as well
specific for non-pregnancy-related types of PfEMP1 do not cross- as its absences in responses to the merozoite antigen GLURP and
react with VAR2CSA9,20, primigravid women are immunologi- VAR2CSA-specific IgG induced by subunit vaccination.
cally naïve to VAR2CSA and therefore highly susceptible to PM
despite general protective immunity acquired during childhood.
However, substantial IgG-mediated protection against PM is Results and discussion
acquired in a parity-dependent manner, and PM is therefore Naturally acquired PfEMP1-specific IgG is highly afucosylated.
mainly a problem in the first pregnancy9,21–23. We first used a set of plasma samples collected from 127 pregnant
Acquired immunity mediated by PfEMP1-specific IgG is gen- Ghanaian women at the time of their first visit to antenatal
erally thought to rely on its ability to interfere directly with IE clinics48 to assess N297 glycosylation of IgG with specificity for
sequestration (i.e., neutralizing, adhesion-inhibitory antibodies). three P. falciparum recombinant antigens. We used the full
However, antibody-mediated opsonization of IEs is a likely ectodomains of VAR2CSA and the non-pregnancy-restricted
additional effector function of these antibodies, since the antibody Group A-type VAR6, which are both naturally expressed on the
response to most P. falciparum asexual blood-stage antigens IE surface. We also included the merozoite antigen GLURP,
(including PfEMP1) is completely dominated by the cytophilic which is not expressed on IE surface49 (Fig. 1).
subclasses IgG1 and (to a lesser extent) IgG3 (refs. 24,25). In line with our hypothesis suggesting that afucosylated IgG
Nevertheless, the relative importance of neutralization and response is restricted to foreign antigens expressed on host cells
opsonization remains largely unexplored. Complement-mediated (such as alloantigens and outer-membrane proteins of enveloped
destruction of IgG-coated IEs does not seem important26, sug- viruses34,40,41,46,47), IgG1 responses to VAR6 and VAR2CSA
gesting that IgG opsonization of IEs by IgG functions mainly were markedly Fc afucosylated (Fig. 2a). All individuals showed
through IgG-Fc receptor (FcγR)-dependent phagocytosis and lowered anti-VAR6 Fc fucosylation compared to total IgG1,
antibody-dependent cellular cytotoxicity (ADCC)27–29. The latter which remained high. The magnitude of the decreased Fc
involves FcγRIIIa30. Binding of IgG to FcγRIIIa critically depends fucosylation of VAR6-specific IgG1 exceeded any previously
on the composition of a highly conserved N-linked glycan at reported pathogen-derived immune response. The most similar
position 297 in the Fc region31. Different monosaccharides, such responses are against rhesus D on red blood cells and human
as galactose, fucose, and sialic acids, are added to the bi-antennary platelet antigen-1a on platelets. However, IgG1 responses to those
core structure of the Fc glycan (Fig. 1c). The level of fucosylation antigens display big variation in Fc fucosylation ranging from
is of particular significance, since afucosylated IgG has up to 20- almost 100 to 10%40,46. In contrast, GLURP-specific IgG1-Fc
fold increased affinity for FcγRIIIa32. The reason is believed to be fucosylation was generally high, also in line with our hypothesis
a steric clash between a unique N-linked glycan at position N162, (Fig. 2a). A few women showed marked afucosylation of GLURP-
not found in FcγRIa nor FcγRIIa/b, and the bi-antennary glycan specific IgG1 (Fig. 2a), possibly in response to GLURP deposited
2 NATURE COMMUNICATIONS |         (2021)1 2:5838 | https://doi.org/10.1038/s41467-021-26118-w | www.nature.com/naturecommunications
NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-26118-w ARTICLE
a b
Endothelial cell Purificaon Syncytiotrophoblast Purificaon of total of angen-
plasma IgG specific IgG
using 
protein G
1 2 3 4 5 6 7 8 9 10 11 12
A
B
C
D
E
F
GLURP+ P. falciparum G
merozoite VAR6
+ P. falciparum IE VAR2CSA+ P. falciparum IE H
c = N-Acetylglucosamine
= Mannose
= Galactose
= Fucose
= Sialic acid
3+ 3+
878.6929 986.7292
3+ 3+
100 3+ 884.0244 1035.0753
932.7110 3+ 3+
986.7292 3+ 938.042475 3+ 3+1035.0753 3+1083.7619
3+ 3+ 932.7110 1083.7619
50 884.0244
3+ 878.69293+
938.0424 1000.4044
25 3+ 3+
981.3962 1000.4044
0
900 950 1000 1050 1100 900 950 1000 1050 1100
m/z m/z
Total IgG Angen-specific IgG
Fig. 1 Background and study workflow. a IgG1 specific for the merozoite antigen GLURP and two members of the PfEMP1 family expressed on the surface
of IEs were analyzed in this study. Most PfEMP1 variants facilitate sequestration of IEs to vascular endothelium (exemplified here by VAR6), while
VAR2CSA-type PfEMP1 mediate IE sequestration in the placental syncytiotrophoblast and intervillous space. b Plasma samples were split and used to
purify total plasma IgG1 and antigen-specific IgG1, using protein G-coupled sepharose and solid-phase absorption with recombinant antigens, respectively.
Eluted IgG1 was digested with trypsin and the glycopeptides analyzed by liquid chromatography mass spectrometry (LC-MS). Examples of MS spectra of
total IgG1 (left) and antigen-specific (anti-VAR6) IgG1 (right) from one sample is shown. c The fractions of the different glycosylation traits of the Fc glycan
depicted were calculated from LC-MS spectra.
on the erythrocyte surface during invasion, as has been described negatively correlated with age (Supplementary Fig. 2A), most
for other merozoite-specific antigens50. IgG1 specific for all three likely reflecting the strong correlation between parity and age. In
P. falciparum antigens showed higher Fc galactosylation and contrast, no correlation was found between age and GLURP-
sialylation levels than total IgG1, similar to what is known for specific IgG Fc fucosylation (Supplementary Fig. 2). The
recent immunizations34,44 (Supplementary Fig. 1A, B). Levels of variability in afucosylation in the first pregnancy to VAR2CSA
bisecting GlcNAc were lower for VAR2CSA- and VAR6-specific is in line with the stochastic afucosylation in primary alloimmune
IgG1, and higher for GLURP-specific IgG1 compared to total responses to red blood cell- and platelet antigens in
IgG1 (Supplementary Fig. 1C). These results indicate that pregnancies40,47 as well as the stochastic levels of afucosylation
antigen-specific IgG levels are modulated in complex ways in responses against different antigens within an individual34.
according to exposure and antigen context. Together, these results suggest that despite the initial variability,
Afucosylation of VAR2CSA-specific IgG1 was generally less afucosylated IgG to PfEMP1 accumulates with repeated exposure.
pronounced than that of VAR6-specific IgG1 (Fig. 2a). Exposure No correlations between IgG levels and Fc fucosylation levels
to VAR2CSA-type PfEMP1 occurs later in life, as it is restricted to were observed (Supplementary Fig. 3). VAR2CSA-specific Fc
pregnancy, whereas P. falciparum expressing Group A PfEMP1 fucosylation was not significantly different between women with
(such as VAR6) are associated with severe malaria in children4,8. or without PM (P= 0.12; Mann–Whitney test, Supplementary
IgG responses to Group A PfEMP1 variants are acquired from Fig. 4), and VAR2CSA-specific IgG-Fc fucosylation levels did not
early in life in endemic areas through repeated exposure to add significant predictive strength to generalized linear models
parasites expressing those variants15,16,51,52. VAR6-specific IgG1 predicting birth weight (P= 0.071) or maternal hemoglobin levels
was consistently afucosylated in all tested individuals, probably as at delivery (P= 0.19) (Supplementary Tables 1 and 2). However,
a result of continuous exposure to Group A PfEMP1 in childhood the striking pattern of decreasing Fc fucosylation with increasing
(Fig. 2a), suggesting that afucosylation is a persistent phenotype parity provides a highly plausible molecular explanation of the
once acquired. This is further supported by the fact that VAR6- well-known protective effect of multiple pregnancies towards PM.
specific IgG-Fc fucosylation was negatively correlated with age, as Studies aimed to establish the protective effect of afucosylated
was total IgG-Fc fucosylation similar to previous reports37 VAR2CSA-specific IgG in PM should be carried out in the future.
(Supplementary Fig. 2). In contrast, the level of fucosylation of
VAR2CSA-specific IgG1 was more varied (Fig. 2a) and decreased Fc afucosylation of PfEMP1-specific IgG is stable. The above
with increased antigen exposure, using parity as proxy (Fig. 2b). findings support the hypothesis that afucosylated IgG specific for
This was not the case for VAR6- (Fig. 2c) or GLURP-specific host membrane-associated immunogens is attained following
IgG1 (Fig. 2d), and only marginally so for total plasma IgG1 repeated exposure and that the phenotype is stable once acquired.
(Fig. 2e). VAR2CSA-specific IgG Fc fucosylation was also To examine this hypothesis further, and to consolidate the
NATURE COMMUNICATIONS |   (      2021) 12:5838 | https://doi.org/10.1038/s41467-021-26118-w |www.nature.com/naturecommunications 3
Relave intensies (%)
ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-26118-w
Fig. 2 Fc fucosylation of naturally acquired P. falciparum-specific IgG depends on antigen location and exposure. a Fc fucosylation levels of total plasma
IgG1 (gray, n= 127) and IgG1 specific for VAR2CSA (orange, n= 117), VAR6 (green, n= 121), and GLURP (blue, n= 88) in Ghanaian pregnant women (left
four panels). Fc fucosylation levels of total plasma IgG1 from unexposed Dutch women (n= 5) were included for comparison (right panel). Medians and
densities are shown. Statistically significant pairwise differences between antigen-specific IgG and total IgG (multiple two-sided Wilcoxon signed-rank
tests with Bonferroni correction) are indicated (****P < 0.0001). b–e Correlations of b VAR2CSA-, c VAR6-, d GLURP-specific, and e total IgG1-Fc
fucosylation levels with parity. P values and correlation coefficients are shown. Statistical significance of correlations (Spearman’s correlations. *P < 0.05;
**P < 0.01; ***P < 0.001; ****P < 0.0001.
findings described above, we proceeded to determine the Fc women were recruited, respectively48,53. In this cohort, we were
fucosylation of IgG with specificity for the same three antigens, also able to purify low level of VAR2CSA-specific IgG from
using an availability-based subset (N= 72) of plasma samples nulligravidae, plausibly caused by unrecorded/undiscovered
from a previously published cohort of Ghanaian women sampled pregnancies. Although VAR2CSA-specific IgG levels decay
while not pregnant53. The findings regarding total and antigen- markedly within 6 months of delivery54,55, the parity-dependency
specific IgG1 (Fig. 3 and Supplementary Fig. 1D–F) were fully of the degree of VAR2CSA-specific IgG1-Fc afucosylation
consistent with those obtained with the samples from pregnant remained in these non-pregnant women (Fig. 3b). Furthermore,
women. The marked Fc afucosylation of VAR2CSA- and VAR6- there was no significant correlation between the time since last
specific IgG1 was more pronounced among this second group of pregnancy and Fc fucosylation levels of VAR2CSA-specific IgG1
women (Fig. 3a and Supplementary Fig. 5), probably reflecting (Fig. 3c). Taken together, these findings reinforce the inference
the more intense parasite transmission in the rainforest compared that PfEMP1-specific IgG1-Fc afucosylation remains stable in the
to the coastal savannah where the non-pregnant and pregnant absence of exposure to antigen. This conclusion is in line with our
4 NATURE COMMUNICATIONS |         (2021)1 2:5838 | https://doi.org/10.1038/s41467-021-26118-w | www.nature.com/naturecommunications
NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-26118-w ARTICLE
Fig. 3 Fc fucosylation levels of VAR2CSA-specifc IgG is temporally
stable. a Fc fucosylation levels of total plasma IgG1 (gray, n= 72) and IgG1
with specificity for VAR2CSA (orange, n= 50), VAR6 (green, n= 65), and
GLURP (blue, n= 43) in non-pregnant Ghanaian women exposed to
VAR2CSA during one or more previous pregnancies. Fc fucosylation levels
of total plasma IgG1 from unexposed Dutch females (n= 5) are included as
controls. Medians and densities are shown. Statistically significant pairwise
differences between antigen-specific IgG and total IgG (multiple two-sided
Wilcoxon signed-rank tests with Bonferroni correction) are indicated
(****P < 0.0001). b Correlation between fucosylation levels of VAR2CSA-
specific IgG1 and parity. c Correlation between fucosylation levels of
VAR2CSA-specific IgG1 and time since last pregnancy. Statistical
significance of correlations are shown (Spearman’s correlations. *P < 0.05;
**P < 0.01; ***P < 0.001; ****P < 0.0001).
the most recent exposure to parasites expressing VAR2CSA53.
This stable response is similar to HIV- and cytomegalovirus-
specific responses but markedly different from initial SARS-CoV-
2 responses, in which IgG is only afucosylated for a few weeks
after seroconversion in most patients34. This may suggest that the
initial SARS-CoV-2-specific antibodies were either secreted by
short-lived plasma cells/plasmablasts or that afucosylation in
those cells is reprogrammed by particular inflammatory
conditions.
Subunit VAR2CSA vaccination does not induce afucosylated
IgG. When measured at the time of delivery, high levels of IgG
recognizing placenta-sequestering IEs are strongly associated with
protection from adverse pregnancy outcome9,23,56. Many of these
antibodies interfere with placental IE sequestration21,22, and it is
therefore generally assumed that neutralizing (adhesion-blocking)
antibodies are required for clinical protection against PM57–59.
On this basis, development of vaccines to prevent PM, based on
the so-called minimal-binding-domain (MBD) of VAR2CSA, is
currently in progress60,61. To examine the levels of Fc fucosyla-
tion of VAR2CSA-specific IgG following subunit vaccination, we
tested plasma samples from the PAMVAC Phase 1 clinical trial,
which involved adult volunteers without previous P. falciparum
exposure, vaccinated with a recombinant VAR2CSA-MBD
construct60. In contrast to the results obtained with naturally
induced VAR2CSA IgG1, the PAMVAC vaccination induced
almost completely fucosylated IgG1; even significantly more
fucosylated than total plasma IgG from the same donors (Fig. 4a).
The remaining glycosylation traits showed changes of antigen-
specific IgG compared to total IgG, similar to naturally infected
subjects (Supplementary Fig. 1G–I). This is in line with our recent
comparison of naturally acquired and subunit vaccine-induced
IgG1 specific for hepatitis B virus34. To assess the possibility that
the full fucosylation of the vaccine-induced VAR2CSA-specific
IgG was due to the vaccinees’ lack of previous exposure to
P. falciparum, genetics, or other environmental parameters, we
also tested samples obtained from the parallel trial of the PAM-
VAC vaccine in Beninese nulligravidae, who were therefore
unexposed to VAR2CSA despite lifelong P. falciparum exposure.
The results (Fig. 4b and Supplementary Fig. 1J–L) were essentially
identical to those obtained with unexposed volunteers. Similar to
previous findings regarding fucosylation of IgG1 alloantibodies the Ghanaian cohorts described above, the Beninese cohort had
being stable for >10 years41,46. However, unlike the Fc afucosy- lower Fc fucosylation levels of total plasma IgG compared to
lation of PfEMP1-specific IgG1, which appeared to be exposure- previous reports of European cohorts and the unexposed vaccine
dependent, boosting with alloantigens was found to have no cohort consisting of Europeans, reaffirming previous reports from
apparent effect on the Fc fucosylation41,46. It also suggests that in rural areas62. This is likely due to accumulating afucosylated IgG
these cases, afucosylated IgG1 is secreted by long-lived plasma to both P. falciparum membrane antigens and enveloped
cells, which for VAR2CSA are sustained for up to a decade after viruses34,37.
NATURE COMMUNICATIONS |         (2021)1 2:5838 | https://doi.org/10.1038/s41467-021-26118-w |www.nature.com/naturecommunications 5
ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-26118-w
recombinant fucosylation variants of the VAR2CSA-specific
human monoclonal antibody PAM1.4. Whereas both bound
VAR2CSA similarly in ELISA (Fig. 5d), only the afucosylated
PAM1.4 induced marked NK-cell degranulation (Fig. 5e). Toge-
ther, these findings underscore the functional significance of Fc
afucosylation of PfEMP1-specific IgG, indicating that IgG
induced by PfEMP1 subunit vaccination lack potentially impor-
tant characteristics of the naturally acquired antibody response.
Our study supports the hypothesis that the immune system has
evolved a capacity to selectively modulate the glycosylation
pattern of the IgG Fc region, thereby fine-tuning the effector
response triggered by antibody-opsonized targets34. Specifically, it
appears that immunogens expressed on host membranes induce
afucosylated IgG, thereby increasing its ability to elicit FcγRIII-
dependent effector responses such as ADCC. In contrast,
immunogens in solution or present on the surface of pathogens
seem to mainly induce fucosylated IgG, thus steering the effector
response against IgG-opsonized targets towards other FcγR-
dependent effector functions. The plasticity in human immune
responses to modulate IgG effector functions by altered
fucosylation endows the immune system with a so far largely
unappreciated level of adaptability. While it is congruent with the
current understanding of how the immune system works, the
functional importance of afucosylated IgG in malaria remains to
be demonstrated. In the meantime, it should be emphasized that
the decrease in Fc fucosylation reported here exceeds any that has
previously been reported for pathogen-derived antigens. Indeed,
it also surpasses the clinically significant afucosylation of the IgG
response to alloantigens39,40,46 found in pregnancy where
mothers are naturally exposed to paternally inherited antigens
Fig. 4 VAR2CSA-specific IgG induced by subunit vaccination is not Fc on fetal cells such as blood cells. These antigens are, in essence,
afucosylated. Fc fucosylation levels of total (gray) and VAR2CSA-specific foreign to the maternal immune system and exposed on the cell
(orange) plasma IgG1 in German vaccinees (n= 32) without (a) and in surface in a way akin to truly foreign pathogen-derived antigens
Beninese vaccinees (n= 18) with (b) natural exposure to P. falciparum. by intracellular pathogens expressing proteins on the host surface
Medians and densities are shown. Statistically significant pairwise such as enveloped viruses34 and P. falciparum. This implies that
differences between antigen-specific IgG and total IgG (two-sided the immunopathogenic IgG raised in alloimmune-mediated
Wilcoxon signed-rank test) are indicated (****P < 0.0001). diseases in pregnancies is an unfortunate corollary of an
evolutionary conserved and advantageous immune mechanism
Only afucosylated VAR2CSA-specific IgG induces natural against intracellular pathogens, raising potent afucosylated IgG
killer cell degranulation. Afucosylation of IgG Fc improves the needed to eliminate these. Several immunomodulatory molecules
affinity of IgG for FcγRIII32,35, increasing NK-cell-mediated are likely to regulate this mechanism, but due to the apparent
ADCC against IgG-opsonized targets36. Recently it was reported necessity of self-membrane association of the antigen, we have
that IgG from individuals naturally exposed to P. falciparum previously proposed a model with a yet unknown membrane-
makes IEs susceptible to NK-cell-mediated ADCC, and that bound receptor on B cells recognizing self34. Signaling from this
PfEMP1-specific IgG is a major contributor to this response29. To receptor is likely to result in transcriptional changes, either
investigate the functional importance of afucosylation of decreasing the expression of fucosyltransferases or increasing the
PfEMP1-specific IgG for ADCC, we assessed the ten Ghanaian expression of molecular chaperones limiting the access of
plasma samples with the highest and lowest Fc fucosylation of fucosyltransferases to the N297 glycan. Nevertheless, the
VAR2CSA-specific IgG (Fig. 5a), respectively, for NK-cell mechanisms of this immunomodulation are beyond of the scope
degranulation efficiency. The samples had similar VAR2CSA- of this study. Finally, the data suggest that to induce afucosylated
specific IgG levels (P= 0.80, Mann–Whitney test; Fig. 5b). IgG responses with increased ADCC—and potentially protective
However, only VAR2CSA-specific IgG from individuals with low capacity—alternative vaccination strategies are required, mimick-
VAR2CSA-specific Fc fucosylation caused marked NK-cell ing the expression of antigens on host cells.
degranulation-induced expression of CD107a (P= 0.0003,
Mann–Whitney test; Fig. 5c). The only afucosylated VAR2CSA
IgG1 not stimulating strong NK-cell degranulation was the Methods
sample with the lowest IgG levels. In line with earlier work35,36, Human subjects. We used biological samples collected as part of the following
the fucosylation status of these antibodies proved to be a more studies: (i) A longitudinal study of malaria in pregnancy, conducted in Dodowa,
important predictor of NK-cell-mediated activity than their located in a coastal savannah area with stable, seasonal P. falciparum trans-
mission, approximately 40 km north of Accra, Ghana48. (ii) A cross-sectional
quantity (Fig. 5b). Apart from fucosylation, only anti-VAR2CSA study of immune responses to VAR2CSA in healthy non-pregnant women53,
IgG1-Fc galactosylation correlated with NK-cell degranulation conducted in Assin Foso, in a rainforest area with high and perennial
(r= 0.44, P= 0.05). However, this was only borderline statisti- P. falciparum transmission, located approximately 80 km north of Cape Coast,
63
cally significant, reaffirming previous observations of slightly Ghana . (iii) A phase 1 clinical trial of the VAR2CSA-based PAMVAC vaccine,
conducted in non-immune German volunteers and in adult, nulligravid
increased FcγRIII affinity of afucosylated IgG with increased Fc P. falciparum-exposed Beninese women volunteers60. Healthy blood donor
galactosylation35. To consolidate these results and to directly samples from Sanquin, Amsterdam, The Netherlands, were used as negative
compare the impact of Fc fucosylation, we next assayed control donors.
6 NATURE COMMUNICATIONS |         (2021)1 2:5838 | https://doi.org/10.1038/s41467-021-26118-w | www.nature.com/naturecommunications
NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-26118-w ARTICLE
Fig. 5 Afucosylated PfEMP1-specific IgG induces NK-cell-mediated ADCC. a VAR2CSA-specific IgG fucosylation levels of samples with low (open
symbols) and high fucose (filled symbols) levels. b Comparison of VAR2CSA-specific IgG levels and c CD107a expression between highly fucosylated
(filled symbols) and afucosylated anti-VAR2CSA IgG1 (open symbols) samples, respectively. Mean from three independent experiments and P values from
two-sided Mann–Whitney tests are shown. d VAR2CSA-specific, human monoclonal antibody PAM1.4 as either fucosylated or afucosylated IgG1 was
titrated in the same assay and VAR2CSA-binding or e degranulation activity (CD107a expression) on NK92-CD16a cells was measured. IgG: non-immune
human IgG. Data represent mean values ± SD from three independent experiments.
The Ghanaian donors all had serologic evidence of exposure to P. falciparum, was terminated after overnight incubation (37 °C) by acidifying to a final con-
with seropositivity rates above 90% in the non-pregnant cohort53 and above 70% in centration of 2% formic acid. Prior to mass spectrometry injection, sodium
the pregnant cohort48. deoxycholate precipitates, in samples where this was added, were removed by
A more detailed demographic description of the analyzed cohorts can be found centrifugation (3000 × g; 30 min), and filtering through 0.65 µm low protein
in the Supplementary Information (Supplementary Table 3). binding filter plates (Millipore, Burlington, USA).
Analysis of IgG Fc glycosylation was performed with nanoLC reverse phase-
electrospray mass spectrometry on an Impact HD quadrupole-time-of-flight mass
P. falciparum recombinant antigens. The full-length ectodomains of the
spectrometer (Bruker Daltonics, Bremen, Germany), and data were processed with
VAR2CSA-type PfEMP1 antigen IT4VAR04 (VAR2CSA) and of the Group A
Skyline software (version 4.2.0.19107) as described elsewhere34. Distinct samples
PfEMP1 antigen HB3VAR6 (VAR6) were expressed in baculovirus-infected insect
were measured once. Samples were considered seropositive if the intensity of
cells and purified as described previously58,64. The amino-terminal, non-repetitive
antigen-specific IgG1 glycopeptides was at least higher than the mean plus 10× the
R0 region of glutamate-rich protein (GLURP) was expressed in Escherichia coli and
65 standard deviation of Dutch seronegative control samples. The level of fucosylationpurified as described elsewhere .
and bisection were calculated as the sum of the relative intensities of glycoforms
containing the respective glycotraits. Galactosylation and sialylation levels were
Purification of IgG from plasma samples. Total IgG from individual donors was calculated as antenna occupancy. The relative intensities of the glycoforms were
purified from ~1 µL plasma using the AssayMAP Bravo platform (Agilent Technol- summed with mono-galactosylated/sialylated species only contributing with 50% of
ogies, Santa Clara, USA) with Protein G-coupled cartridges as described elsewhere34. their relative intensity. Details on analyzed glycopeptides can be found in the
P. falciparum antigen-specific IgG was purified from individual donors by Supplementary Information (Supplementary Table 4)
incubation (1 h, room temperature) of individual plasma samples (diluted 1:10 in
phosphate-buffered saline (PBS) supplemented with TWEEN 20 (0.05 %; PBS-T))
in 96-well Maxisorp plates (Nunc, Roskilde, Denmark) coated overnight (4 °C; Human monoclonal VAR2CSA-specific IgG. The human monoclonal IgG1
PBS) with VAR2CSA (2 µg/mL), VAR6 (2 µg/mL), or GLURP (1 µg/mL). antibody, PAM1.4, derived from an EBV-immortalized memory B-cell clone from
66
Following the incubation, the plates were washed 3× with PBS-T, 2× with PBS, and a Ghanaian woman with natural exposure to PM , recognizes a conformational
2× with ammonium bicarbonate (50 mM). Antigen-specific IgG were finally eluted epitope in several VAR2CSA-type PfEMP1 proteins, including IT4VAR04. In the
by formic acid (100 mM; 5 min shaking). present study, we used a non-modified recombinant version of PAM1.4 produced
in HEK293F cells with high Fc fucosylation and a glyco-engineered variant with
low Fc fucosylation26,67.
Mass spectrometric IgG Fc glycosylation analysis. Eluates of purified IgG were
collected in low-binding PCR plates (Eppendorf, Hamburg, Germany) and dried by
vacuum centrifugation (50 °C). The dried samples were dissolved in a reduction Quantification of VAR2CSA-specific IgG. Levels of VAR2CSA-specific IgG were
and alkylation buffer containing sodium deoxycholate (0.4%), tris(2-carboxyethyl) assessed by ELISA as previously described68. In brief, 96-well flat-bottom microtiter
phosphine (10 mM), 2-chloroacetamide (40 mM), and TRIS (pH 8.5; 100 mM), or plates (Nunc MaxiSorp, Thermo Fisher Scientific) were coated overnight at 4 °C with
ammonium bicarbonate (50 mM). After boiling the samples (10 min; 95 °C), full-length VAR2CSA (100 ng/well in PBS). Monoclonal antibody (0.08–10 μg/mL)
trypsin (5 µg/mL) in ammonium bicarbonate (50 mM) was added. The digestion or plasma samples (1:400) were added in duplicate, followed by washing and
NATURE COMMUNICATIONS |         (2021) 12:5838 | https://doi.org/10.1038/s41467-021-26118-w |www.nature.com/naturecommunications 7
ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-26118-w
horseradish peroxidase-conjugated rabbit anti-human IgG (1:3,000; Dako). Bound 8. Jensen, A. T. R. et al. Plasmodium falciparum associated with severe childhood
antibodies were detected by adding TMB PLUS2 (Eco-Tek), and the reaction stopped malaria preferentially expresses PfEMP1 encoded by Group A var genes. J.
by the addition of 0.2M H2SO4. The optical density (OD) was read at 450 nm and Exp. Med. 199, 1179–1190 (2004).
the specific antibody levels were calculated in arbitrary units (arb. units), using the 9. Salanti, A. et al. Evidence for the involvement of VAR2CSA in pregnancy-
equation 100 × [(ODSAMPLE−ODBLANK)/(ODPOS.CTRL−ODBLANK)]. associated malaria. J. Exp. Med. 200, 1197–1203 (2004).
10. Tuikue Ndam, N. G. et al. High level of var2csa transcription by Plasmodium
Antibody‑dependent cellular cytotoxicity (ADCC) assay. Degranulation- falciparum isolated from the placenta. J. Infect. Dis. 192, 331–335 (2005).
induced CD107a expression in response to IgG bound to plastic-immobilized 11. Cohen, S., McGregor, I. A. & Carrington, S. Gammaglobulin and acquired
antigen is a convenient marker of NK-cell ADCC69. Here, we coated 96-well flat- immunity to human malaria. Nature 192, 733–737 (1961).
bottom microtiter plates (Nunc MaxiSorp; Thermo Fisher Scientific) overnight at 12. Conway, D. J. et al. A principal target of human immunity to malaria
4 °C with full-length VAR2CSA (100 ng/well in PBS68). Following 1 h blocking identified by molecular population genetic and immunological analyses. Nat.
with PBS containing 1% Ig-free bovine serum albumin (BSA) (1% PBS-BSA), Med. 6, 689–692 (2000).
plasma samples (1:20) or PAM1.4 variants (0.08–10 µg/mL) were added for 1 h at 13. Douglas, A. D. et al. The blood-stage malaria antigen PfRH5 is susceptible to
37 °C. After washing, 1.6 × 105 NK92 cells stably expressing CD16a and GFP70 vaccine-inducible cross-strain neutralizing antibody. Nat. Commun. 2, 601
were added to each well. In addition, anti-human CD107a-PE (1:40; H4A3 clone; (2011).
BD Biosciences), 10 μg/mL brefeldin A (Sigma-Aldrich), and 2 µM monensin 14. Kana, I. H. et al. Naturally acquired antibodies target the glutamate-rich
(Sigma-Aldrich) were added, and the cells incubated for 4 h at 37 °C. Cells were protein on intact merozoites and predict protection against febrile malaria. J.
then centrifuged and stained with near-IR fixable Live/Dead dye (Invitrogen), Infect. Dis. 215, 623–630 (2017).
followed by data acquisition on a FACS LSRII flow cytometer (BD Biosciences), 15. Bull, P. C. et al. Plasmodium falciparum-infected erythrocytes: agglutination
and analysis with FlowLogic software (version8.3; Inivai Technologies, Australia). by diverse Kenyan plasma is associated with severe disease and young host
Wells with antigen and NK cells, but without antibody were included in all age. J. Infect. Dis. 182, 252–259 (2000).
experiments to control for unspecific activation. Plasma samples from four Danish 16. Cham, G. K. et al. Hierarchical, domain type-specific acquisition of antibodies
non-pregnant women without malaria exposure and purified human IgG (Sigma- to Plasmodium falciparum erythrocyte membrane protein 1 in Tanzanian
Aldrich) were included as negative controls. Example of the gating strategy is children. Infect. Immun. 78, 4653–4659 (2010).
provided as Supplementary Information (Supplementary Fig. 6) 17. Schmiegelow, C. et al. Plasmodium falciparum infection early in pregnancy
has profound consequences for fetal growth. J. Infect. Dis. 216, 1601–1610
Statistical tests. Statistical analyses were performed using R: A Language and (2017).
Environment for Statistical Computing (Version 3.5.2). Performed tests are men- 18. Ofori, M. F., Lamptey, H., Dickson, E. K., Kyei-Baafour, E. & Hviid, L.
tioned in the text. All performed tests were two-sided and tests for normal dis- Etiology of placental Plasmodium falciparum malaria in African women. J.
tribution of data were performed where applicable. Infect. Dis. 218, 277–281 (2018).
19. Viebig, N. K. et al. A single member of the Plasmodium falciparum var
Ethics statement. Collection of biological samples for this study was approved by multigene family determines cytoadhesion to the placental receptor
the Institutional Review Board of Noguchi Memorial Institute for Medical chondroitin sulphate A. EMBO Rep. 6, 775–781 (2005).
Research, University of Ghana (study 038/10-11) by the Regional Research Ethics 20. Tuikue Ndam, N. G. et al. Dynamics of anti-VAR2CSA immunoglobulin G
Committees, Capital Region of Denmark (protocol H-4-2013-083), by the Aca- response in a cohort of Senegalese pregnant women. J. Infect. Dis. 193,
demic Medical Center Institutional Medical Ethics Committee of the University of 713–720 (2006).
Amsterdam, by the Ethics Committee of the Medical Faculty and the University 21. Fried, M., Nosten, F., Brockman, A., Brabin, B. T. & Duffy, P. E. Maternal
Clinics of the University of Tubingen, and by the German Regulatory authorities. antibodies block malaria. Nature 395, 851–852 (1998).
The study was conducted in adherence to the International Council for Technical 22. Ricke, C. H. et al. Plasma antibodies from malaria-exposed pregnant women
Requirements for Human Use guidelines and the principles of the Declaration of recognize variant surface antigens on Plasmodium falciparum-infected
Helsinki. Written informed consent was obtained from all participants before erythrocytes in a parity-dependent manner and block parasite adhesion to
enrollment. chondroitin sulphate A. J. Immunol. 165, 3309–3316 (2000).
23. Staalsoe, T. et al. Variant surface antigen-specific IgG and protection against
the clinical consequences of pregnancy-associated Plasmodium falciparum
Reporting summary. Further information on research design is available in the Nature
malaria. Lancet 363, 283–289 (2004).
Research Reporting Summary linked to this article.
24. Piper, K. P., Roberts, D. J. & Day, K. P. Plasmodium falciparum: analysis of the
antibody specificity to the surface of the trophozoite-infected erythrocyte. Exp.
Data availability Parasitol. 91, 161–169 (1999).
The liquid chromatography mass spectrometry data generated in this study have been 25. Megnekou, R., Staalsoe, T., Taylor, D. W., Leke, R. & Hviid, L. Effects of
deposited in the MassIVE database under accession code MSV000088060. Further data pregnancy and intensity of Plasmodium falciparum transmission on
that support the findings of this study are available from the corresponding author upon immunoglobulin G subclass responses to variant surface antigens. Infect.
reasonable request. Data regarding the original vaccination study are available from Immun. 73, 4112–4118 (2005).
ClinicalTrial.gov ID: NCT02647489. Source data are provided with this paper. 26. Larsen, M. D. et al. Evasion of classical complement pathway activation on
Plasmodium falciparum-infected erythrocytes opsonized by PfEMP1-specific
IgG. Front. Immunol. 9, 3088 (2019).
Received: 24 May 2021; Accepted: 11 September 2021; 27. Marsh, K., Otoo, L., Hayes, R. J., Carson, D. C. & Greenwood, B. M.
Antibodies to blood stage antigens of Plasmodium falciparum in rural
Gambians and their relation to protection against infection. Trans. R. Soc.
Trop. Med. Hyg. 83, 293–303 (1989).
28. Ataide, R., Mwapasa, V., Molyneux, M. E., Meshnick, S. R. & Rogerson, S. J.
Antibodies that induce phagocytosis of malaria infected erythrocytes: effect of
References HIV infection and correlation with clinical outcomes. PLoS ONE 6, e22491
1. World Health Organization. World Malaria Report 2020 (2020). (2011).
2. Bengtsson, A. et al. A novel domain cassette identifies Plasmodium falciparum 29. Arora, G. et al. NK cells inhibit Plasmodium falciparum growth in red
PfEMP1 proteins binding ICAM-1 and is a target of cross-reactive, adhesion- blood cells via antibody-dependent cellular cytotoxicity. Elife 7, e36806 (2018).
inhibitory antibodies. J. Immunol. 190, 240–249 (2013). 30. Ravetch, J. V. & Perussia, B. Alternative membrane forms of FcgRIII (CD16)
3. Fried, M. & Duffy, P. E. Adherence of Plasmodium falciparum to chondroitin on human natural killer cells and neutrophils. Cell type-specific expression of
sulphate A in the human placenta. Science 272, 1502–1504 (1996). two genes that differ in single nucleotide substitutions. J. Exp. Med 170,
4. Turner, L. et al. Severe malaria is associated with parasite binding to 481–497 (1989).
endothelial protein C receptor. Nature 498, 502–505 (2013). 31. Vidarsson, G., Dekkers, G. & Rispens, T. IgG subclasses and allotypes: from
5. Hviid, L. & Jensen, A. T. PfEMP1—a parasite protein family of key importance structure to effector functions. Front Immunol. 5, 520 (2014).
in Plasmodium falciparum malaria immunity and pathogenesis. Adv. 32. Ferrara, C. et al. Unique carbohydrate-carbohydrate interactions are required
Parasitol. 88, 51–84 (2015). for high affinity binding between FcgRIII and antibodies lacking core fucose.
6. Rogerson, S. J., Hviid, L., Duffy, P. E., Leke, R. F. G. & Taylor, D. W. Malaria in Proc. Natl Acad. Sci. USA 108, 12669–12674 (2011).
pregnancy: pathogenesis and immunity. Lancet Infect. Dis. 7, 105–117 (2007). 33. Wang, T. T. et al. IgG antibodies to dengue enhanced for FcgRIIIA binding
7. Jensen, A. R., Adams, Y. & Hviid, L. Cerebral Plasmodium falciparum malaria: determine disease severity. Science 355, 395–398 (2017).
the role of PfEMP1 in its pathogenesis and immunity, and PfEMP1-based 34. Larsen, M. D. et al. Afucosylated IgG characterizes enveloped viral responses
vaccines to prevent it. Immunol. Rev. 293, 230–252 (2020). and correlates with COVID-19 severity. Science 371, eabc8378 (2021).
8 NATURE COMMUNICATIONS |         (2021)1 2:5838 | https://doi.org/10.1038/s41467-021-26118-w | www.nature.com/naturecommunications
NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-26118-w ARTICLE
35. Dekkers, G. et al. Decoding the human immunoglobulin G-glycan repertoire 60. Mordmuller, B. et al. First-in-human, randomized, double-blind clinical trial
reveals a spectrum of Fc-receptor- and complement-mediated-effector of differentially adjuvanted PAMVAC, a vaccine candidate to prevent
activities. Front. Immunol. 8, 877 (2017). pregnancy-associated malaria. Clin. Infect. Dis. 69, 1509–1516 (2019).
36. Temming, A. R. et al. Functional attributes of antibodies, effector cells, and 61. Sirima, S. B. et al. PRIMVAC vaccine adjuvanted with Alhydrogel or GLA-SE
target cells affecting NK cell-mediated antibody-dependent cellular to prevent placental malaria: a first-in-human, randomised, double-blind,
cytotoxicity. J. Immunol. 203, 3126–3135 (2019). placebo-controlled study. Lancet Infect. Dis. 20, 585–597 (2020).
37. de Haan, N., Reiding, K. R., Driessen, G., van der Burg, M. & Wuhrer, M. 62. de Jong, S. E. et al. IgG1 Fc N-glycan galactosylation as a biomarker for
Changes in healthy human IgG Fc-glycosylation after birth and during early immune activation. Sci. Rep. 6, 28207 (2016).
childhood. J. Proteome Res. 15, 1853–1861 (2016). 63. Afari, E. A., Appawu, M., Dunyo, S., Baffoe-Wilmot, A. & Nkrumah, F. K.
38. Bakovic, M. P. et al. High-throughput IgG Fc N-glycosylation profiling by Malaria infection, morbidity and transmission in two ecological zones in
mass spectrometry of glycopeptides. J. Proteome Res. 12, 821–831 (2013). southern Ghana. Afr. J. Health Sci. 2, 312–316 (1995).
39. Wuhrer, M. et al. Regulated glycosylation patterns of IgG during alloimmune 64. Stevenson, L. et al. Investigating the function of Fc-specific binding of IgM to
responses against human platelet antigens. J. Proteome Res. 8, 450–456 Plasmodium falciparum erythrocyte membrane protein 1 mediating
(2009). erythrocyte rosetting. Cell Microbiol. 17, 819–831 (2015).
40. Kapur, R. et al. Low anti-RhD IgG-Fc-fucosylation in pregnancy: a new 65. Theisen, M., Vuust, J., Gottschau, A., Jepsen, S. & Hogh, B. Antigenicity and
variable predicting severity in haemolytic disease of the fetus and newborn. Br. immunogenicity of recombinant glutamate-rich protein of Plasmodium falciparum
J. Haematol. 166, 936–945 (2014). expressed in Escherichia coli. Clin. Diagn. Lab Immunol. 2, 30–34 (1995).
41. Sonneveld, M. E. et al. Glycosylation pattern of anti-platelet IgG is stable 66. Barfod, L. et al. Human pregnancy-associated malaria-specific B cells target
during pregnancy and predicts clinical outcome in alloimmune polymorphic, conformational epitopes in VAR2CSA. Mol. Micobiol. 63,
thrombocytopenia. Br. J. Haematol. 174, 310–320 (2016). 335–347 (2007).
42. Ackerman, M. E. et al. Natural variation in Fc glycosylation of HIV-specific 67. Dekkers, G. et al. Multi-level glyco-engineering techniques to generate IgG
antibodies impacts antiviral activity. J. Clin. Invest. 123, 2183–2192 (2013). with defined Fc-glycans. Sci. Rep. 6, 36964 (2016).
43. Chakraborty, S. et al. Proinflammatory IgG Fc structures in patients with 68. Lopez-Perez, M. et al. IgG responses to the Plasmodium falciparum antigen
severe COVID-19. Nat. Immunol. 22, 67–73 (2021). VAR2CSA in Colombia are restricted to pregnancy and are not induced by
44. Selman, M. H. et al. Changes in antigen-specific IgG1 Fc N-glycosylation upon exposure to Plasmodium vivax. Infect. Immun. 86, e00136–00118 (2018).
influenza and tetanus vaccination. Mol. Cell Proteomics 11, 014563 (2012). 69. Jegaskanda, S. et al. Cross-reactive influenza-specific antibody-dependent
M111. cellular cytotoxicity antibodies in the absence of neutralizing antibodies. J.
45. Vestrheim, A. C. et al. A pilot study showing differences in glycosylation Immunol. 190, 1837–1848 (2013).
patterns of IgG subclasses induced by pneumococcal, meningococcal, and two 70. Snyder, K. M. et al. Expression of a recombinant high affinity IgG Fc receptor
types of influenza vaccines. Immun. Inflamm. Dis. 2, 76–91 (2014). by engineered NK cells as a docking platform for therapeutic mAbs to target
46. Kapur, R. et al. A prominent lack of IgG1-Fc fucosylation of cancer cells. Front. Immunol. 9, 2873 (2018).
platelet alloantibodies in pregnancy. Blood 123, 471–480 (2014).
47. Sonneveld, M. E. et al. Antigen specificity determines anti-red blood cell IgG-
Fc alloantibody glycosylation and thereby severity of haemolytic disease of the Acknowledgements
fetus and newborn. Br. J. Haematol. 176, 651–660 (2017). We are grateful to all the individuals donating blood samples for this study, and to the
48. Ofori, M. et al. Pregnancy-associated malaria in a rural community of Ghana. scientists and health workers participating in the studies for which they were originally
Ghana Med. J. 43, 13–18 (2009). collected. We thank Michael Theisen (University of Copenhagen and Statens Serumin-
49. Borre, M. B. et al. Primary structure and localization of a conserved stitut) for GLURP antigen and GLURP-reactive IgG preparation, and Bruce Walcheck
immunogenic Plasmodium falciparum glutamate rich protein (GLURP) and Geoff Hart (University of Minnesota) for the NK92-CD16a cell line. We also
expressed in both the preerythrocytic and erythrocytic stages of the vertebrate acknowledge Erik de Graaf (Sanquin Research) for optimization of the analysis pipelines
lifecycle. Mol. Biochem. Parasitol. 49, 119–132 (1991). used in this study, done in relation to previous projects. The study was funded by the
50. Awah, N. W., Troye-Blomberg, M., Berzins, K. & Gysin, J. Mechanisms of Landsteiner Foundation for Blood Transfusion Research grant number 1721 and the
malarial anaemia: potential involvement of the Plasmodium falciparum low Danish International Development Agency (Danida), 12-081RH and 17-02-KU. The
molecular weight rhoptry-associated proteins. Acta Trop. 112, 295–302 PAMVAC study (ClinicalTrials.gov ID NCT02647489) was sponsored by the Uni-
(2009). versitätsklinikum Tübingen and funded by the European Union Seventh Framework
51. Nielsen, M. A. et al. Plasmodium falciparum variant surface antigen expression Programme (FP7-HEALTH-2012-INNOVATION; under grant agreement 304815), the
varies between isolates causing severe and non-severe malaria and is modified Danish Advanced Technology Foundation (under grant number 005-2011-1), and a
by acquired immunity. J. Immunol. 168, 3444–3450 (2002). Medium Scale Collaborative Project supported by the German Federal Ministry of
52. Olsen, R. W. et al. Natural and vaccine-induced acquisition of cross-reactive Education and Research (Bundesministerium für Bildung und Forschung) through EVI,
IgG-inhibiting ICAM-1-specific binding of a Plasmodium falciparum KfW, and Irish Aid. The funders had no role in study design, data collection and analysis,
PfEMP1 subtype associated specifically with cerebral malaria. Infect. Immun. decision to publish, or preparation of the manuscript.
86, e00622–00617 (2018).
53. Ampomah, P., Stevenson, L., Ofori, M. F., Barfod, L. & Hviid, L. B-cell Author contributions
responses to pregnancy-restricted and -unrestricted Plasmodium falciparum M.D.L., C.E.v.d.S., L.H., and G.V. conceptualized the study. M.F.O., L.H., and G.V.
erythrocyte membrane protein 1 antigens in Ghanaian women naturally acquired funding. M.D.L., M.L.-P., J.N., M.W., L.H., and G.V. performed the investi-
exposed to malaria parasites. Infect. Immun. 82, 1860–1871 gations presented in this study. E.K.D., P.A., B.M., A.S., M.A.N., M.F.O., N.T.N., and
(2014). A.M. supplied study material. M.D.L., L.H., and G.V. wrote the original draft which was
54. Staalsoe, T. et al. Acquisition and decay of antibodies to pregnancy-associated reviewed and edited by all authors.
variant antigens on the surface of Plasmodium falciparum infected
erythrocytes that are associated with protection against placental parasitemia.
J. Infect. Dis. 184, 618–626 (2001). Competing interests
55. Ampomah, P., Stevenson, L., Ofori, M. F., Barfod, L. & Hviid, L. Kinetics of B The authors declare no competing interests.
cell responses to Plasmodium falciparum erythrocyte membrane protein 1 in
Ghanaian women naturally exposed to malaria parasites. J. Immunol. 192,
5236–5244 (2014). Additional information
56. Duffy, P. E. & Fried, M. Antibodies that inhibit Plasmodium falciparum Supplementary information The online version contains supplementary material
adhesion to chondroitin sulfate A are associated with increased birth weight available at https://doi.org/10.1038/s41467-021-26118-w.
and the gestational age of newborns. Infect. Immun. 71, 6620–6623 (2003).
57. Beeson, J. G. et al. Antibodies to variant surface antigens of Plasmodium Correspondence and requests for materials should be addressed to Lars Hviid or Gestur
falciparum-infected erythrocytes and adhesion inhibitory antibodies are Vidarsson.
associated with placental malaria and have overlapping and distinct targets. J. Peer review information Nature Communications thanks the anonymous, reviewer(s)
Infect. Dis. 189, 540–551 (2004). for their contribution to the peer review of this work. Peer reviewer reports are available.
58. Khunrae, P. et al. Full-length recombinant Plasmodium falciparum VAR2CSA
binds specifically to CSPG and induces potent parasite adhesion-blocking Reprints and permission information is available at http://www.nature.com/reprints
antibodies. J. Mol. Biol. 397, 826–834 (2010).
59. Srivastava, A. et al. Full-length extracellular region of the var2CSA variant of
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
PfEMP1 is required for specific, high-affinity binding to CSA. Proc. Natl Acad.
published maps and institutional affiliations.
Sci. USA 107, 4884–4889 (2010).
NATURE COMMUNICATIONS |         (2021) 12:5838 | https://doi.org/10.1038/s41467-021-26118-w |www.nature.com/naturecommunications 9
ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-26118-w
Open Access This article is licensed under a Creative Commons
Attribution 4.0 International License, which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to the Creative
Commons license, and indicate if changes were made. The images or other third party
material in this article are included in the article’s Creative Commons license, unless
indicated otherwise in a credit line to the material. If material is not included in the
article’s Creative Commons license and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder. To view a copy of this license, visit http://creativecommons.org/
licenses/by/4.0/.
© The Author(s) 2021
10 NATURE COMMUNICATIONS |         (2021) 12:5838 | https://doi.org/10.1038/s41467-021-26118-w | www.nature.com/naturecommunications