Metaviromics Reveals Unknown Viral Diversity in the Biting Midge Culicoides impunctatus

Abstract

Biting midges (Culicoides species) are vectors of arboviruses and were responsible for the emergence and spread of Schmallenberg virus (SBV) in Europe in 2011 and are likely to be involved in the emergence of other arboviruses in Europe. Improved surveillance and better understanding of risks require a better understanding of the circulating viral diversity in these biting insects. In this study, we expand the sequence space of RNA viruses by identifying a number of novel RNA viruses from Culicoides impunctatus (biting midge) using a meta-transcriptomic approach. A novel metaviromic pipeline called MetaViC was developed specifically to identify novel virus sequence signatures from high throughput sequencing (HTS) datasets in the absence of a known host genome. MetaViC is a protein centric pipeline that looks for specific protein signatures in the reads and contigs generated as part of the pipeline. Several novel viruses, including an alphanodavirus with both segments, a novel relative of the Hubei sobemo-like virus 49, two rhabdo-like viruses and a chuvirus, were identified in the Scottish midge samples. The newly identified viruses were found to be phylogenetically distinct to those previous known. These findings expand our current knowledge of viral diversity in arthropods and especially in these understudied disease vectors.

Keywords: Culicoides impunctatus; RNA viruses; metaviromics.

Figure 1

MetaViC workflow: (A) pre-processing step to remove unwanted sequences (B) de novo assembly and contig classification.

Figure 2

An overview of the virus hits found in all midge pools. (A) A scatter plot showing contigs lengths (x-axis) and the corresponding percent identity (y-axis) for the top BLAST hit. The colour of the circle represents the midge pools and the size of the circle represent the number of mapped reads on a log10 scale. (B) A bar chart describing different virus groups found in the midge pools where each pool is represented in a different colour. (C) A bar chart showing different proteins found in the midge pools characterised based on the nearest BLAST homologue. The virus classification groups and protein categories were derived based on the closest viral sequence match. (D) Distribution of the closest viral homologs found in the midge pools. Krona plots show the nearest virus hit found in all three midge pools, M1, M2 and M3 respectively. Colours in the Krona charts represent the distinct taxonomic groups.

Figure 3

Novel nodavirus segments identified in the sample. Depth of coverage (orange) across the RNA1 (A) and RNA2 (B) segments with the coding regions for RNA dependent RNA polymerase (RdRp), Protein B2 and Capsid proteins shown in blue. (C) Phylogeny of the RdRp protein from the RNA1 segment and (D) phylogeny of the coat protein from the RNA2 generated using RAxML and showing the classification of the novel midge nodavirus RdRp and coat proteins with respect to existing protein sequences available in Genbank.

Figure 4

Phylogenetic tree showing the relationship of two novel midge rhabdo-like viruses identified in this study with currently known sequences from Genbank.

Figure 5

(A) RdRp based phylogeny describing clustering of a novel chuvirus species. Viruses that have two segments are highlighted with red tip labels. Blue tip labels highlight viruses that have one segment. Chuviruses found in the Scottish midge samples are highlighted in yellow. (B) Protein sequence similarity between Chuvirus Mos8Chu0 RdRp and midge chuvirus sequences with RdRp hits along with the sequence coordinates and coloured according to percentage identity.