Abstract
BACKGROUND: Epidemic arbovirus transmission occurs among humans by mosquito bites and the sylvatic transmission cycles involving non-human primates (NHPs) still exists. However, limited data are available on the extent in NHPs infections and their role. In this study, we have developed and validated a high-throughput serological screening tool to study the circulation of multiple arboviruses that represent a significant threat to human health, in NHPs in Central Africa. METHODOLOGY/PRINCIPAL FINDINGS: Recombinant proteins NS1, envelope domain-3 (DIII) for the dengue (DENV), yellow fever (YFV), usutu (USUV), west nile (WNV) and zika (ZIKV) and envelope 2 for the chikungunya (CHIKV) and o'nyong-nyong (ONNV) were coupled to Luminex beads to detect IgG directed against these viruses. Evaluation of test performance was made using 161 human sera of known arboviral status (66 negative and 95 positive). The sensitivity and specificity of each antigen were determined by statistical methods and ROC curves (except for ONNV and USUV). All NS1 antigens (except NS1-YFV), CHIKV-E2 and WNV-DIII had sensitivities and specificities > 95%. For the other DIII antigens, the sensitivity was low, limiting the interest of their use for seroprevalence studies. Few simultaneous reactions were observed between the CHIKV+ samples and the NS1 antigens to the non-CHIKV arboviruses. On the other hand, the DENV+ samples crossed-reacted with NS1 of all the DENV serotypes (1 to 4), as well as with ZIKV, USUV and to a lesser extent with YFV. A total of 3,518 samples of 29 species of NHPs from Cameroon and the Democratic Republic of Congo (DRC) were tested against NS1 (except YFV), E2 (CHIKV/ONNV) and DIII (WNV) antigens. In monkeys (n = 2,100), the global prevalence varied between 2 and 5% for the ten antigens tested. When we stratified by monkey's biotope, the arboreal species showed the highest reactivity. In monkeys from Cameroon, the highest IgG prevalence were observed against ONNV-E2 and DENV2-NS1 with 3.95% and 3.40% respectively and in DRC, ONNV-E2 (6.63%) and WNV-NS1 (4.42%). Overall prevalence was low in apes (n = 1,418): ranging from 0% for USUV-NS1 to 2.6% for CHIKV-E2. However, a very large disparity was observed among collection site and ape species, e.g. 18% (9/40) and 8.2% (4/49) of gorillas were reactive with CHIKV-E2 or WNV-NS1, respectively in two different sites in Cameroon. CONCLUSIONS/SIGNIFICANCE: We have developed a serological assay based on Luminex technology, with high specificity and sensitivity for simultaneous detection of antibodies to 10 antigens from 6 different arboviruses. This is the first study that evaluated on a large scale the presence of antibodies to arboviruses in NHPs to evaluate their role in sylvatic cycles. The overall low prevalence (<5%) in more than 3,500 NHPs samples from Cameroon and the DRC does not allow us to affirm that NHP are reservoirs, but rather, intermediate hosts of these viruses.