Background
Plasmodium vivax is the most widely distributed human malaria parasite and responsible for 100-300 million clinical cases each year including severe disease and death [
1]. Studies on naturally acquired immune responses against different
P. vivax antigens, including, among others, merozoite surface antigens, Duffy-binding protein, circumsporozoite surface protein, apical membrane 1, have demonstrated their immunogenicity in natural infections [
2‐
6]. In addition, some of these studies have also demonstrated differences in acquisitiuon of IgG subclasses against particular antigens. The association of clinical protection with specific antigens of
P. vivax, however, has been reported in only two prospective cohort longitudinal studies [
7,
8]. In the first, clinical protection was associated with IgG3 antibodies against the N-terminus of the merozoite surface protein 1 (PvMSP1) in residents of the Brazilian Amazon region of Portuchuelo [
7]. In the second, clinical protection was reported in children from Papua New Guinea [
8]. This latter study used a functional assay on naturally-acquired binding inhibitory antibodies against the Duffy-binding protein region II (PvDBPII) to demonstrate the association of these antibodies with protection against
P. vivax infection.
Antibodies in these studies were measured using the ELISA assay [
9], a technique requiring high amounts of coating-antigens and immune sera when large numbers of samples must be screened. To overcome these limitations, suspension array technologies with high-throughput capacity to simultaneously analyse several proteins with minimal amount of immune sera have been developed [
10]. Recently, these methodologies have been reported to measure antibodies to multiple malaria vaccine candidate antigens [
11], to measure simultaneously antibody recognition of twenty eight
Plasmodium falciparum erythrocyte membrane protein 1 (PfEMP1) domains [
12], and to develop a high throughput functional assay to study binding of ICAM1 to the 3D7 strain PfEMP1 repertoire [
13].
The aim of this study was to measure naturally acquired IgG anti-PvMSP1 antibodies in human patients infected with P. vivax from different endemic regions of Brazil and Papua New Guinea through implementation and validation of the BioPlex suspension array system.
Discussion
The use of multiplex assays has not been reported for
P. vivax antigens and, therefore, a comparison of the BioPlex assay values with those obtained by ELISA was initially made to determine the concordance and correlation of these techniques. To do so, two different statistics, the Kappa statistics and the Spearman's rank correlation test, were used. Kappa Statistics has been proposed as the most appropriate method when the same parameter with two similar techniques is compared [
16]. The coefficient correlation tests, such as Spearman's rank correlation or Pearson correlation, has been used to validate multiplex assays in
P. falciparum and other diseases [
11,
12,
17]. Using Kappa statistics, a good agreement, with significant values, was obtained for both recombinant proteins with the samples from PNG. In contrast, even if the agreement of the results obtained from Brazil was above 70%, Kappa values were not significant. Most of the discrepancy, however, was the result of samples slightly positive by BioPlex and negative by ELISA. In addition, the Spearman's rank correlation test demonstrated significant correlations for all comparisons (Figure
2). As this correlation has been used to validate the use of multiplex assays in
P. falciparum and other diseases, the data from Brazil remain meaningful; yet, it also argues that further research into different statistical analyses is desirable before claiming conclusive results exclusively based on multiplex assays.
Naturally-acquired IgG responses against PvMSP1 have demonstrated that a large proportion of individuals expose to
P. vivax contain IgG antibodies against different regions of the molecule [
15,
18,
19]. The results obtained with the BioPlex assay with subjects from Brazilian adults and PNG children confirm these findings even though a lower overall response to PvMSP1 was observed in PNG children (Table
1). Interestingly, non-infected individuals (no symptoms, negative microscopy and/or negative PCR) from both regions contained antibodies against PvMSP1. This data indicate that anti-PvMSP1 antibodies are long-lived and/or that these individuals presented a sub-patent parasitaemia, not detected by microscopy or PCR, representing undetected carriers. Both explanations have supporting evidence as anti-PvMSP1 antibodies have been reported to last up-to 30 years [
20] and direct mosquito feeding of patients with sub-patent vivax infections have successfully transmitted
P. vivax to mosquitoes [
21]. Determining their frequencies will have implications for malaria control programmes in these regions.
Studies on naturally acquired immunity in
P. vivax in PNG have shown that it is acquired in children below nine years of age [
22], while acquired immunity in Brazil is only reached later, when closer to 15-20 years old in native populations [
7] and also in migrant adult working populations [
18]. Interestingly, age was significantly associated with the response to PvMSP1-C and not to PvMSP1-N in both regions; yet, in PNG this association was mainly observed in younger children (positive individuals mean age = 27 months, negative individuals mean age = 37 months) whereas in Brazil was observed in older populations (positive individuals mean age = 32 years, negative individuals mean age = 23 years). Further prospective studies with larger sample sizes are required to determine the role, if any, of PvMSP1 in age-dependent acquisition of immunity.
This is the first study that uses the BioPlex system to measure IgG subclasses in malaria. To validate its use, sera from asymptomatic patients from Brazil were chosen as previous studies had demonstrated that these patients contained predominantly IgG3 antibodies against PvMSP1-N [
7]. As expected, IgG3 anti-PvMSP1-N antibodies were the predominant isotype in asymptomatic patients from Rio Machado. Remaining IgG profiles were very much in agreement with previous studies using ELISA in which IgG1 and IgG3 cytophilic isotypes are the predominant isotypes against PvMSP1 with low or undetectable levels of IgG2 and IgG4 [
19,
23]. Interestingly, IgG isotype profiles of PNG and Brazil were different with a predominance of IgG1 antibodies against PvMSP1 in PNG and a predominance of IgG3 antibodies against PvMSP1 in Brazil. Here again, whether this result is due to the difference in the age groups, remains to be determined in larger surveys from these regions including all age-groups.
Conclusions
Several studies on naturally acquired IgG responses against different P. vivax antigens have been described using the ELISA assay. The BioPlex assay described and validated here is the first to measure naturally acquired humoral IgG antibody responses against a P. vivax antigen and the first one to study IgG subclasses in malaria. These studies compared the naturally acquired humoral IgG responses against PvMSP1 in different endemic regions of Brazil and PNG. Results demonstrated differences in percentage of responders, differences between the age correlating to responses and differences on isotype profiles. Whether these differences are due to the limited power of this study, which was mainly designed to validate the use of the multiplex assay for PvMSP1, or to real differences in the immunoepidemiology of P. vivax in these endemic regions, is presently under investigation. Most relevant, prospective comparative longitudinal cohort studies in endemic different regions using the BioPlex assay can now be envisaged to look for associations against infection and clinical protection having clear implications in antigen discovery and vaccine development for P. vivax.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
CFB produced recombinant proteins and implemented the BioPlex assay. MB performed ELISA assays. SS and FA performed the statistical analysis. EPC & FA provided samples from Rio Machado. DS provided samples from PNG. DS, FA, PLA and IM critically read the manuscript. CFB and HAP conceived this study and drafted the manuscript. All authors' read and approved the final manuscript.