Viral metagenomics reveals the presence of novel Zika virus variants in Aedes mosquitoes from Barbados.

Background: The Zika virus (ZIKV) epidemic of 2015/2016 spread throughout numerous countries. It emerged in mainland Latin America and spread to neighboring islands, including the Caribbean island of Barbados. Recent studies have indicated that the virus must have already been circulating in local mosquito populations in Brazil for almost 2 years before it was identified by the World Health Organization in 2015. Metagenomic detection assays have the potential to detect emerging pathogens without prior knowledge of their genomic nucleic acid sequence. Yet their applicability as vector surveillance tools has been widely limited by the complexity of DNA populations from field-collected mosquito preparations. The aim of this study was to investigate local vector biology and characterize metagenomic arbovirus diversity in Aedes mosquitoes during the ongoing 2015/2016 ZIKV epidemic.

Methods: We performed a short-term vector screening study on the island of Barbados during the ongoing 2015/2016 ZIKV epidemic, where we sampled local Aedes mosquitoes. We reanalyzed mosquito viral microbiome data derived from standard Illumina MiSeq sequencing to detect arbovirus sequences. Additionally, we employed deep sequencing techniques (Illumina HiSeq) and designed a novel bait capture enrichment assay to increase sequencing efficiency for arbovirus sequences from complex DNA samples.

Results: We found that Aedes aegypti seemed to be the most likely vector of ZIKV, although it prevailed at a low density during the observed time period. The number of detected viruses increased with sequencing depth. Arbovirus sequence enrichment of metagenomic DNA preparations allowed the detection of arbovirus sequences of two different ZIKV genotypes, including a novel one. To our knowledge, this is the first report of the S3116W mutation in the NS5 gene region of ZIKV polyprotein.

Conclusions: The metagenomic arbovirus detection approach presented here may serve as a useful tool for the identification of epidemic-causing arboviruses with the additional benefit of enabling the collection of phylogenetic information on the source. Apart from detecting more than 88 viruses using this approach, we also found evidence of novel ZIKV variants circulating in the local mosquito population during the observed time period.