Comparison of the impact of allelic polymorphisms in PfAMA1 on the induction of T Cell responses in high and low malaria endemic communities in Ghana

Malaria is a deadly vector-borne tropical disease of great public health concern. The disease is caused by the apicomplexan parasite Plasmodium, which is transmitted through the bite of an infected female Anopheles mosquito during a blood meal. With the increase in the number of existing tools recommended by the World Health Organization (WHO) for malaria control, malaria-related deaths have decreased over the last two decades in sub-Saharan Africa [1]. However, with these interventions in place, people, including children and adults, are still getting sick and dying from malaria. Hence, WHO has recommended the development of new interventions including vaccines to reinforce our hope of eradicating the disease [2]. RTS,S/AS01E (Mosquirix™) is the first and only malaria vaccine to date to demonstrate protection against malaria in children in phase 3 clinical trials [3]. However, its efficacy has been reported to wane with time [4] and this needs to be addressed in order to stay on track to meet the malaria eradication goal. Alternative vaccine design strategies are being explored for the development of more efficacious vaccines.

An effective malaria vaccine should be capable of inducing protective immune responses against variant forms of the parasite and in a genetically diverse population. Plasmodium falciparum apical membrane antigen 1 (PfAMA1) has been investigated as a vaccine candidate in several clinical trials [5‐8]. PfAMA1 is found in the sporozoite, liver, and blood stages of the parasite [9] and therefore represents a multi-target antigen for vaccine development. It has been shown to be a target of both antibody [6, 10, 11] and T cell [12‐16] responses. Antibodies to PfAMA1 have been demonstrated to inhibit red blood cell invasion by merozoites [17].

In populations with high malaria endemicity, there is a measured high meiotic recombination rate and nucleotide substitutions in P. falciparum and these have been associated with increased drug and immune pressure on the parasites [18, 19]. For PfAMA1, multiple nucleotide substitutions within the protein sequence result in the formation of allelic or polymorphic forms of the antigen in different parasite variants. As a direct result of these polymorphisms, antibodies to one allelic form of PfAMA1 have been reported to bind less with other PfAMA1 alleles because a substantial fraction of antibodies are towards strain-specific epitopes [17, 20]. While the effect of antigen polymorphisms has been clearly demonstrated for antibody responses, there is much less data on whether there is a similar effect of polymorphism on T cell recognition of epitopes within polymorphic antigens such as PfAMA1. Sedegah et al. [21] have investigated the effect of single amino acid substitutions in immunodominant PfAMA1 allelic epitopes on T cell response induction and report a significant effect of these single amino acid substitutions on TCR recognition of peptide-MHC complexes, and hence IFN-γ induction.

T cells are classically stimulated in the context of major histocompatibility complex (MHC) molecules and are known to bind to specific amino acid residues on the peptide sequence [22]. Furthermore, T cell receptors (TCRs) are typically known to interact with both the MHC molecules and the peptides being presented by the MHC molecules through specific residue identification [23‐25]. MHC genes are amongst the most diverse human genes with thousands of variants within any population [26]. Hence, polymorphism within the peptides and/or within the MHC molecule will most likely affect MHC binding and TCR recognition [27]. This study therefore investigated the impact of allelic polymorphisms in selected PfAMA1 peptide sequences from three strains of P. falciparum (3D7, 7G8 and FVO) on their function as targets of immunodominant T cell responses in high and low malaria transmission communities in southern Ghana.

T lymphocyte recognition of sequences within malaria antigens is an essential feature of the adaptive immune response for limiting Plasmodium infection in humans. Interferon-gamma (IFN-γ) is a key immune molecule that is secreted by parasite-specific activated T cells and has a liver stage parasite killing effect. Recognition of parasite antigen peptide-MHC complexes by T cells involves molecular interactions between amino acid residues on both the TCR and the parasite peptides, and polymorphisms within either could potentially affect peptide recognition. In this study, we assessed the IFN-γ induction potential of predicted polymorphic PfAMA1 peptides using a standard ELISpot assay and compared responses between individuals living in high and low malaria transmission areas.

Based on the peptides tested in this study, positive response frequency was higher in Obom relative to Legon. This may be a reflection of the fact that Obom is a high malaria transmission community, and subjects may have been exposed to multiple strains circulating in the community. The observation correlated with the expectation that high transmission most likely drives increased parasite diversity [37, 38]. The multiplicity of infection with parasites has generally been reported to be greater in high transmission areas compared with low transmission areas [39, 40]. It is therefore expected that, persons living in high endemic communities would most likely have concomitant infection with diverse parasite strains, and this will result in the activation of several clones of T cells committed to responding to the different parasite strains. These T cell clones can survive for years [41] and upon re-infection, they can be recalled to mount immune attack on the new parasites. However, populations with low endemicity have a lower proportion of mixed-genotype infections, due to a low rate of superinfection, hence limited diverse clones of T cells. It must be noted that our earlier studies have picked up relatively higher numbers of anti-PfAMA1 responses with subjects recruited from the Legon area, but these studies have mostly tested PfAMA1 peptide pools that span the entire PfAMA1 sequence [32, 42, 43], while the current study tested a limited number of 15 single 9-10mer HLA-restricted minimal epitopes. Five of the eight (63%) peptides that made positive responses in this study were 3D7 PfAMA1 variants (Table 2), and while this could suggest 3D7-like parasites to be the commonest circulating P. falciparum strains in the study communities, this observation could also simply be due to the fact that our selection of peptides for testing was based on epitope predictions for the 3D7 variant peptides.

For both study sites, the number of positive peptide responses increased upon CD8 + enrichment of PBMCs. We have previously reported similar findings with CSP peptides [36] and this could be due a depletion of CD4 + regulatory T cell subsets (T regs) from PBMCs. Classically, T regs are a subset of CD4 + T cells [44] and have been shown to increase in frequency during natural malaria infections [45‐47]. The hallmark effect of Tregs are the impairment of T cell proliferation and cytokine production from other T cell subsets following engagement of their antigen-specific TCRs [41, 48, 49]. Hence depletion of these subsets may have led to removal of the stimulation restriction effect on CD8 + T cells, resulting in higher IFN-γ activities in some subjects. It is worth noting that the subjects who made positive responses in assays with CD8 + enriched PBMCs (s3, s5, s7 and s17) did not make significant responses with the unfractionated PBMCs. Conversely, in subjects (s2 and s6) whose unfractionated PBMC were positive to some peptides, these effects were lost upon depletion of the CD4 + fraction of PBMCs. On this basis, it is possible that these positive responses seen were CD4 + T cell-specific although we could not confirm this due to insufficient cell numbers to test the CD4 + enriched PBMC.

Amino acid changes that affected peptide positivity in assays mostly occurred at residue positions 4, 5 and 6 (Fig. 3). Jordan et al. [50] reported that substitutions that suitably change the spatial structure of peptides may enhance the immunogenicity of epitopes and improve the binding of TCRs to MHC ligands. Moreover, Calis et al. [51] have reported a significant impact of amino acid residue, weighted by the position at which it is found within the peptide, on the peptide’s immunogenicity. They found that T cells have preference for aromatic and large residues and position 4 to 6 were shown to be most important especially in MHC class I antigen presentation to CD8 + T cells. In this study, substitutions within the peptides that made positive responses mostly occurred within these positions.

In contrast, all peptides within two allelic sets elicited positive responses (Figs. 1 and 2), suggesting that the substitutions within these peptides, which occurred in positions 4 and 6 respectively (Fig. 3), were not enough to abrogate T cell response induction. Thus there was no clear relationship between amino acid residue substitutions within the peptide sequences and their effect on peptide positivity in assays.

This study was limited in two ways. First, very few peptides were tested. Using a greater number of predicted minimal epitopes could have increased our chances of identifying additional immunogenic peptides whose reactivities may be altered by the changes in variant parasite strains. Second, subjects from both sites were not HLA-typed hence the peptides tested were not selected on the basis of their predicted recognition of subjects’ HLA types. Selection of peptides based on study subject HLA types may have increased the positivity rate of the tested peptides [13]. These findings are however relevant for the concept of constituting several T cell epitopes that are identified to be presented by specific HLA alleles into a multi-epitope vaccine for broad population use or as biomarkers for protective T cell immunity. Promiscuity in HLA recognition and binding by immunodominant peptides [52, 53] would contribute to the understanding of how this could work and make it possible for such potential multi-epitope vaccines to work in persons with HLA types that are not those upon which the vaccine epitopes were identified.

This study has provided some evidence that polymorphisms existing in peptides from malaria vaccine candidate antigens, like PfAMA1, can affect peptide recognition by T cells and hence the immune response that will be elicited. Furthermore, high endemicity which is expected to drive greater parasite diversity, may result in a broadening of the repertoire of T cells that can recognize specific peptides beyond what can be achieved with a whole parasite vaccine containing a single parasite strain. We can also infer the possibility of broad HLA recognition of peptides originally predicted to be presented by specific HLA alleles, and this could further strengthen the multi-epitope vaccine design concept. These findings will further our understanding of cellular immune mechanisms that govern anti-Plasmodium T cell responses and help direct the selection of peptides for inclusion in multi-epitope vaccines that can offer cross-strain protection as well as biomarkers for assessing protective T cell immunity.