Impact of Epstein-Barr virus co-infection on natural acquired Plasmodium vivax antibody response

Abstract

Background: The simultaneous infection of Plasmodium falciparum and Epstein-Barr virus (EBV) could promote the development of the aggressive endemic Burkitt's Lymphoma (eBL) in children living in P. falciparum holoendemic areas. While it is well-established that eBL is not related to other human malaria parasites, the impact of EBV infection on the generation of human malaria immunity remains largely unexplored. Considering that this highly prevalent herpesvirus establishes a lifelong persistent infection on B-cells with possible influence on malaria immunity, we hypothesized that EBV co-infection could have impact on the naturally acquired antibody responses to P. vivax, the most widespread human malaria parasite.

Methodology/principal findings: The study design involved three cross-sectional surveys at six-month intervals (baseline, 6 and 12 months) among long-term P. vivax exposed individuals living in the Amazon rainforest. The approach focused on a group of malaria-exposed individuals whose EBV-DNA (amplification of balf-5 gene) was persistently detected in the peripheral blood (PersVDNA, n = 27), and an age-matched malaria-exposed group whose EBV-DNA could never be detected during the follow-up (NegVDNA, n = 29). During the follow-up period, the serological detection of EBV antibodies to lytic/ latent viral antigens showed that IgG antibodies to viral capsid antigen (VCA-p18) were significantly different between groups (PersVDNA > NegVDNA). A panel of blood-stage P. vivax antigens covering a wide range of immunogenicity confirmed that in general PersVDNA group showed low levels of antibodies as compared with NegVDNA. Interestingly, more significant differences were observed to a novel DBPII immunogen, named DEKnull-2, which has been associated with long-term neutralizing antibody response. Differences between groups were less pronounced with blood-stage antigens (such as MSP1-19) whose levels can fluctuate according to malaria transmission.

Conclusions/significance: In a proof-of-concept study we provide evidence that a persistent detection of EBV-DNA in peripheral blood of adults in a P. vivax semi-immune population may impact the long-term immune response to major malaria vaccine candidates.

Fig 1. Profile of IgG antibody response against EBV capsid antigen p18 (VCA-p18) in individuals whose EBV-DNA could be detected (PersV_DNA) or not (NegV_DNA) over the follow-up period.

For each graphic, individual datapoints were expressed as ELISA absorbance (OD) and shown here as a scatter dot plots with lines showing the median with interquartile range (IQR). Numbers represented in the top and bottom of graphics represent the proportion of responders (%) and median (OD) with IQR values, respectively; p-values for significant difference between groups were included and calculated as described in methods. The proportion of responders was determined by considering an OD > 0.38 as ELISA-positive response (S2 Fig). All raw data are available in the S1 Table.

Fig 2. Pairwise correlations between EBV antibody levels in individuals whose EBV-DNA could be detected (PersV_DNA) or not (NegV_DNA) over the follow-up period.

Clustering was based on the Spearman correlation coefficient for assays measuring anti-EBV antibodies in plasma samples [IgM (VCA-p18, Zebra and EAd-p45/52) and IgG (VCA-p18 and EBNA-1)]. Matrix heatmaps were shown for each cross-sectional survey (baseline, 6- and 12-months), with top and bottom panels representing PersV_DNA and NegV_DNA groups, respectively. Positive correlations shown in blue and negative correlations shown in red, with numbers in bold statistically significant differences.

Fig 3. Profile of IgG antibody response against the surface-engineered DEKnull-2 vaccine of P. vivax in individuals with (PersV_DNA) or without (NegV_DNA) persistent viral DNA over the follow-up period.

In A, results are shown by cross-sectional surveys (baseline, 6- and 12-month), with individual data points expressed as ELISA Reactivity Index (RI) and shown here as a scatter dot plots with lines showing the median with interquartile range (IQR). In B, medians of RIs overtime for each group. Numbers in the top and bottom of each graphic represent the proportion of responders (%) and median (RI) with IQR values, respectively; p-values for significant difference between groups were included and calculated as described in methods. Reactivity Index (RI) was calculated as described in methods and RI > 1 corresponded to an ELISA-positive response.

Fig 4. Profile of IgG antibody response against the 19-kDa C-terminal region of the Merozoite Surface Protein-1 (MSP1-19) in individuals with (PersV_DNA) or without (NegV_DNA) persistent viral DNA over the follow-up period.

In A, results are shown by cross-sectional surveys (baseline, 6- and 12-month), with individual data points expressed as ELISA Reactivity Index (RI) and shown here as a scatter dot plots with lines showing the median with interquartile range (IQR). In B, medians of RIs overtime for each group. Numbers in the top and bottom of each graphic represent the proportion of responders (%) and median (RI) with IQR values, respectively; p-values for significant difference between groups were included and calculated as described in methods. Reactivity Index (RI) was calculated as described in methods and RI > 1 corresponded to an ELISA-positive response.