Enhanced Arbovirus Surveillance with High-Throughput Metatranscriptomic Processing of Field-Collected Mosquitoes

Abstract

Surveillance programs are essential for the prevention and control of mosquito-borne arboviruses that cause serious human and animal diseases. Viral metatranscriptomic sequencing can enhance surveillance by enabling untargeted, high-throughput arbovirus detection. We used metatranscriptomic sequencing to screen field-collected mosquitoes for arboviruses to better understand how metatranscriptomics can be utilised in routine surveillance. Following a significant flood event in 2016, more than 56,000 mosquitoes were collected over seven weeks from field traps set up in Victoria, Australia. The traps were split into samples of 1000 mosquitoes or less and sequenced on the Illumina HiSeq. Five arboviruses relevant to public health (Ross River virus, Sindbis virus, Trubanaman virus, Umatilla virus, and Wongorr virus) were detected a total of 33 times in the metatranscriptomic data, with 94% confirmed using reverse transcription quantitative PCR (RT-qPCR). Analysis of metatranscriptomic cytochrome oxidase I (COI) sequences enabled the detection of 12 mosquito and two biting midge species. Screening of the same traps by an established public health arbovirus surveillance program corroborated the metatranscriptomic arbovirus and mosquito species detections. Assembly of genome sequences from the metatranscriptomic data also led to the detection of 51 insect-specific viruses, both known and previously undescribed, and allowed phylogenetic comparison to past strains. We have demonstrated how metatranscriptomics can enhance surveillance by enabling untargeted arbovirus detection, providing genomic epidemiological data, and simultaneously identifying vector species from large, unsorted mosquito traps.