Background
Despite several countries having reduced malaria incidence by more than 75%, and a reduction in mortality by 48% globally, more than 3 billion people are still at risk of contracting malaria and some 438,000 deaths still occur every year [
1]. Current malaria control and elimination efforts would be greatly enhanced by the development of novel and more sensitive surveillance tools. For instance, serological markers that can be used to estimate exposure to malaria parasites and/or indicate a person’s immune status would help to identify populations at risk, and to direct resources to areas in more need [
2‐
4]. Additionally, the development and deployment of highly efficacious vaccines against the two major malaria parasites,
Plasmodium falciparum and
Plasmodium vivax, would certainly accelerate malaria elimination [
2,
5].
Identifying optimal antigenic targets for evaluating exposure or for vaccine development, however, remains a huge challenge due to the complexity of malaria parasites biology and epidemiology [
6]. As the dynamics of antibody acquisition and maintenance vary based on exposure intensity, which serologic markers are informative of exposure or immunity is likely to differ by age group and transmission setting [
4,
7,
8]. A better understanding of the human immune responses to malaria parasites is thus essential for biomarker discovery, and very useful in guiding rational vaccine design [
4,
7,
8]. To date, relatively little is known about the early acquisition and role of anti-
Plasmodium spp. antibodies in young children, how such responses compare to responses in older children/adults, or those from different transmission intensity areas [
7‐
11]. The investigation of antigenic targets and their potential as vaccine candidates or biomarkers of exposure in naturally exposed populations has been mainly restricted to
P. falciparum and very few
P. vivax merozoite proteins [
7‐
11].
In malaria parasites, glycosylphosphatidylinositol (GPI) is a glycolipid highly conserved across different species [
12]. In
Plasmodium spp., GPI can be found both free and as an anchor sustaining many proteins on the parasite’s membrane, including merozoite surface and rhoptry proteins, as well as many other vaccine candidates and proteins of unknown function [
12]. In humans, GPI is known to induce strong humoral response, promote the expression of genes of pro-inflammatory compounds (including tumour-necrosis factor (TNF), interleukin-1 [IL-1] and IL-12), nitric oxide, and adhesion molecules on the surface of the vascular endothelium, which can be recognized by
P. falciparum erythrocyte membrane protein 1 (
PfEMP1), contributing to the development of anaemia and severe malaria [
13,
14].
It has been consistently demonstrated that GPIs purified from
P. falciparum are recognized by plasma/serum from people living in malaria-endemic areas however, the quality of GPIs purified from
P. falciparum might have led to controversial results [
15,
16]. Cross-reactivity between antibodies raised against
P. falciparum GPI and
P. vivax is expected, as despite having a high complexity that allows various chemical modifications and high functional diversity, the core of the GPI glycan structure is evolutionary highly conserved in different species [
17]. Only limited structural variability (in fatty-acid composition or glycosylation) or antigenic variation have been described [
18‐
21] in comparison to the many allelic polymorphisms identified in merozoite surface proteins [
22‐
24], and the consequent high antigenic variation [
25‐
27].
To date, the association between the levels of antibodies to GPI and the risk of malaria clinical disease remains poorly explored. To address this gap, this study aimed to measure total IgG levels to a synthetic glycan corresponding to
P. falciparum GPI (
PfGPI) in a cohort of children aged 1–3 years from Papua New Guinea (PNG), exploring the associations between antibody levels and prospective risk of malaria. Individual differences in exposure to
Plasmodium spp. blood-stage infections have been well characterized by molecular genotyping [
28,
29], and children have been shown to had acquired immunity to
P. vivax, but no yet to
P. falciparum [
28‐
30]. The potential use of IgG to
PfGPI as a serological biomarker of immune status to both
P. falciparum and
P. vivax parasites was investigated.
Discussion
A better understanding of the acquisition of immunity to malaria parasites in different age groups and transmission settings is essential for the identification of antigens useful as biomarkers of exposure/immunity, or with potential for vaccine development—especially for
P. vivax, since a continuous in vitro culture system is still inexistent [
4,
5]. In the present study, antibody levels to a synthetic glycan correspondent to
PfGPI [
31] was measured in a cohort of children 1–3 years old from PNG [
30], exploring the associations between antibody levels and risk of
P. falciparum and
P. vivax-malaria.
Despite the very high transmission intensity in East Sepik Province when the cohort study was conducted [
30], seroprevalence of antibodies to
PfGPI was low in this age group. Similar low seroprevalence have been described in children < 6 years from Madang Province in PNG [
32], as well as in Indonesia [
33], Kenya [
18] and Gambia [
34]. One explanation for this is the low ability of the immune system of very young children (< 2 years old) in producing antibodies against carbohydrate antigens [
35]. This also suggests that the majority of GPI that the immune system has access to and thus can produce antibodies to is the free form, rather than the form that anchors proteins to the parasite membrane. If physically attached to their GPI anchors, parasite surface proteins might be expected to provide T cell help for anti-GPI antibody production [
15]. Although not observed in the young children included in the study, seroprevalence and magnitude of antibody responses to
PfGPI have been described to increase with age and decline with parasite density in PNG [
32] and Kenyan adults [
18].
For the young children included in this study, recent
P. falciparum and
P. vivax infections were the main determinant of antibody levels to GPI. The rapid although transient peaks in antibody levels in the presence of a current infection might suggest that they are generated by the differentiation of naive B-cells into short-lived plasma cells driven by the concurrent infection rather than by long-lived plasma cells generated from previous infections, as previously described for malarial protein antigens [
36]. Given the absence of peptide epitopes for conventional T cells, antibodies to free GPI are likely to be T cell-independent during the first malaria infections [
15]. Although they can stimulate antigen-specific B cells, memory is not generated, and accessory cells (e.g. macrophages and dendritic cells) and co-stimulatory signals (e.g. IL-1) are thus required for an effective immune response [
37]. Later with increasing exposure, or if attached to an immunogenic carrier, however, GPI might be taken up by follicular B cells, be processed and presented on cell surface major histocompatibility complex class II (MHCII) molecules, where they may engage peptide-specific T cells [
15,
35]. Memory B cells can thus be generated during this T cell-dependent process, and be re-activated upon future stimulation [
35].
Children with homozygote Gerbich blood type (Gerbich negative) had higher antibody levels to
PfGPI than heterozygote or wild type children. The Gerbich antigen is expressed on glycophorins C (GPC) and D (GPD) [
38], and both GPC/D interact with the 4.1 R protein complex and contribute to the stability of the erythrocyte membrane [
38,
39]. A high incidence of Gerbich negative in PNG been hypothesized as an advantage against infection and severe malaria [
40,
41]. While it was found that deletion of the exon 3 result in Gerbich negativity and make
P. falciparum unable to invade erythrocytes using the erythrocyte binding protein 140 [EBA140] [
39,
42], to date, clinical studies have not been able to show a consistent association between risk of malaria and this phenotype [
43‐
45]. Further in-depth studies will be required to elucidate whether the interaction between Gerbich genotype, reduced parasite invasion and slower parasite growth result in increased host immune-responses (including to
PfGPI), and whether this may indeed combine to provide protection against
P. falciparum or
P. vivax malaria [
46].
In young PNG children, high antibody levels to
PfGPI were associated with higher risk of
P. falciparum malaria. In contrast, they were also associated with reduced risk of
P. vivax malaria. This accurately reflects the different levels of naturally acquired immunity to the two species in this cohort: while in these children incidence of
P. vivax episodes significantly decreases starting in the 2nd year of life, the burden of
P. falciparum infection continues to increase until the 4th year of life [
30]. This difference is related to a significantly higher exposure to
P. vivax than
P. falciparum blood-stage infections, i.e.
P. vivax molFOB was considerably higher than
P. falciparum molFOB (14 versus 5.5 parasite clones/child/year-at-risk, respectively). This high number of
P. vivax clones that infect children in early childhood thus contribute to a very rapid acquisition of immunity to clinical
P. vivax malaria, not yet reached for
P. falciparum [
29,
30]. Acquisition of immunity to
P. falciparum in high transmission settings such as PNG is achieved a number of years later (~ 10 years old) with increasing exposure to
P. falciparum infections [
7,
8]. Anti-
PfGPI antibodies in this age group seem to be an accurate reflection of the children’s current immune-status to both
P. falciparum and
P. vivax malaria, acting as both a biomarker of increased risk of
P. falciparum, able to identify individuals with the highest level of exposure to
P. falciparum recent infections, as well as a biomarker of acquired immunity to
P. vivax.
In 2002, a study in rodent models firstly showed that antibodies raised against
PfGPI were able to delay mortality by
Plasmodium berghei, demonstrating proof of concept for a GPI-based anti-toxic malaria vaccine [
31]. The antagonists of GPI-mediated signaling and murine monoclonal antibodies against
PfGPIs were shown to be able to block the induction of toxic responses, also suggesting that GPI-based therapy is possible [
47,
48]. In more recent studies, GPI was found to be present across all stages of the malaria parasites life cycles. Furthermore, in a pre-clinical evaluation of a GPI-based vaccine in
P. berghei models, the vaccine showed efficacy in sporozoite challenges, was able to reduce parasite replication and transmission to mosquitoes (unpublished data, Schofield.) Altogether, these findings suggest that a GPI vaccine may be able to prevent both blood-stage and liver infections, disease and block transmission of parasite from human to mosquito, thus acting as a unique carbohydrate multi-stage, multi-parasite vaccine. Consistent with this, high levels of the anti-GPI antibodies have been correlated with resistance to clinical symptoms, such as anaemia and fever [
18], and lower levels observed among Senegalese adults with cerebral malaria compared to individuals with uncomplicated malaria [
49]. Although anti-
PfGPI antibodies are short-lived or intermittent in very young children, older children and adults seem to be able to sustain high antibody levels for longer [
18,
32‐
34,
50]. Furthermore, GPI low immunogenicity in young children and can be overcome if the antigen is conjugated to a protein carrier, which can also help stimulation of B-cell memory formation [
35]. Future functional studies are now necessary to confirm whether anti-
PfGPI antibodies contribute to the protection observed against
P. vivax, or only act as a mirror of the protection conferred by antibodies to other antigenic targets.
This study highlights anti-PfGPI antibodies as a possible biomarker of anti-malaria immunity in very young children. Further studies including older age groups will confirm its utility as a biomarker of immunity for P. vivax, and whether they will indeed also reflect acquired immunity to P. falciparum.
Authors’ contributions
The following authors contributed with: samples and reagents used in this study—LS, EL, BK, PS; study design—IM, LS; data generation—CTF and AC; statistical analysis—CTF, CSNLWS, IM; manuscript writing—CTF, and IM. All authors read and approved the final manuscript.