merozoite surface protein 9 is a promising vaccine candidate antigen. Previous studies have demonstrated that (i) PvMSP9 is conserved among mice, primate and human Plasmodium
]; (ii) PvMSP9 recombinant proteins induce high titers of antibodies [13
]; (iii) antibodies raised against PvMSP9 are capable of inhibiting merozoite invasion [12
]; and (iv) malaria-exposed individuals present high frequency of natural antibody and cellular immune response against different regions of PvMSP9 [14
]. Clinical trials based on a few selected malaria antigens have shown limited immunogenicity and a failure to induce long-lasting immunity, possibly due to the lack of effective T-cell epitopes in the constructs used as immunogens [16
]. Nevertheless, there have been only a few T-cell epitopes reported from malaria antigens [18
]. A major obstacle for identifying T-cell epitopes is the high level of polymorphism of HLA class II molecules. Thus, one of the most relevant steps for malaria vaccine development is to define T-cell epitopes that can interact promiscuously with a broad range of HLA-DR and/or HLA-DQ molecules.
Here we present the identification of five T cell epitopes in the vaccine candidate PvMSP9 that are capable of stimulating T cells from donors expressing various HLA genotypes and with confirmed exposure to P. vivax infections.
Experimental screening methods to evaluate the presence of HLA restriction in immune response to vaccine candidates are expensive and time consuming. Computational prediction methods complement experimental studies, minimize the number of validation experiments, and significantly expedite the epitope mapping process [11
]. Such methods have helped identify promiscuous epitopes within Leishmania
], Mycobacterium tuberculosis
] and HIV [27
] antigens. Several promiscuous epitopes from pre-erythrocytic [22
], asexual blood-stage [21
], and gametocyte [20
] antigens have been predicted and/or experimentally confirmed for P. falciparum
. In contrast, only limited studies have focused on promiscuous epitopes for P. vivax
In our study, eleven peptides were predicted by the ProPred algorithm to be promiscuous, but only five of them were recognized at high frequency by PBMCs from individuals living in malaria endemic areas. The recall response elicited by at least one of these five peptides was high for both IFN-γ (64.1%) and for IL-4 (50.7%) in comparison with the frequencies observed for other Plasmodium
antigens such as PvTRAg40 [33
], PfTRAP [34
], PvDBP [35
]. The frequency of T cells reactive to PvMSP9 is comparable to a study by Farouk et al. [36
] that measured the cellular response to crude P. falciparum
antigens by ELISPOT in a Malian population. Evaluating the numbers of IFN-γ and IL-4 SFCs, we observed that IFN-γ had higher numbers of secreting cells than IL-4 for all the synthetic peptides studied. No correlation between IFN-γ response and malaria exposure was observed. However, IL-4 SFC produced upon peptide pL stimulation correlated positively with time of residence in the endemic area and the number of IL-4 spots generated after stimulation with all overlapping peptides (pH, pK, pL) were higher in individuals who have lived in malaria endemic areas for more than 20 years when compared with those who have lived in such areas for less than 20 years. It is possible that variations in exposure may also explain variations in the type of naturally induced TH1 and TH2 immune responses to PvMSP9 [14
]. Indeed, data reported by Troye-Blomberg et al.
], showed a strong association between elevated IgG and IgE antibodies to blood-stage antigens with increased numbers of IL-4 secreting cells in individuals less susceptible to malaria infection. Similarly, correlations between the production of IL-4 in response to the P. falciparum
malaria antigen Pf155RESA and protection against malaria were also reported [38
The frequency and numbers of responders to overlapping peptides shows that the core sequence shared with peptides pH, pK and pL (ASIDSMI) is highly immunogenic. However the presence of 23 individuals who present cellular response only to peptide pL suggest that this peptide may have two immunodominant epitopes, one in the overlapping core region and the second one in the carboxy-terminal region that is not shared with pH or pK (DEIDFYEK).
The evaluation of IFN-γ and IL-4 production was used here to measure the recognition and activation of T cells by PvMSP9 putative promiscuous T-cell epitopes. To correlate the cellular response with the prevalence of MHC class II alleles, we determined the HLA antigen distribution among the study population. The observation of 13 allelic groups in the cohort suggests that the study population is heterogeneous, presenting a large variety of allelic groups. It was expected in our study mainly because Brazilian populations have peculiar features of a tri-hybrid populations formed with contribution of Caucasian, African, and native Amerindian origin, in which the phenotypic characteristics of each original population have been highly mixed. However the observation of high frequency of HLA-DR4 and HLA-DQ3 indicates that in this population the Amerindian HLA genotype is conserved [39
]. Therefore, previous works already show the association with IgG responders to Plasmodium
antigens and the HLA-DRB04 in this population [40
], indeed studies with HLA polymorphism observed in several populations have been attributed to a pathogen induced selection [42
The enrollment of populations with high HLA polymorphism but with a related degree of conservation and association with immune response with Plasmodium
antigens is ideal to validate the prediction algorithm using PvMSP9. The lack of association with the frequency of IFN-γ and IL-4 cellular responses and the number of ELISpots generated after stimulation with the five PvMSP-9 predicted epitopes supports the recall cellular immune response reported in our previous study [14
] and the promiscuous properties of the PvMSP9 derived peptides: pE, pH, pK, pJ and pL.
Several studies have suggested that single-epitope-based vaccines are not potent enough to induce full protection [30
]. However, the identification of immunogenic and promiscuous epitopes within a vaccine candidate antigen is extremely important, since it is possible to formulate a vaccine composed of relevant epitopes from different antigens. Additionally, the combination of multiple B cell and T cell epitopes was shown to increase immunogenicity [46
]. In conclusion the HLA-DR heterogeneity of the responding subjects and the prediction analysis using the ProPred server strongly suggest that these peptides was presented to T cells promiscuously. Thus the overall results suggest that HLA restriction will not be a problem if these peptides are used in a vaccine candidate.