Decoding rRNA sequences for improved metagenomics of sylvatic mosquito species

Abstract

RNAseq-based metagenomics on mosquitoes are frequently performed for surveillance or microbiome discovery studies to understand the disease ecology of arboviruses. A major hurdle to overcome in these studies is the depletion of abundant rRNA, which is commonly achieved using oligo-based protocols. The lack of complete reference rRNA sequences in sequence databases for most mosquito vectors makes such studies feasible for only a narrow range of species, biasing the knowledge on mosquitoes and their viruses. We performed total RNAseq and mitochondrial cytochrome c oxidase I gene sequencing on specimens of 29 mosquito species to compare the functionality of rRNA sequences as an RNA barcode for taxonomy and phylogeny. Using a score-based strategy to bioinformatically filter out contaminating rRNA reads, we successfully assembled full-length 28S and 18S rRNA sequences for all specimens. Phylogenetic reconstructions revealed a high degree of congruity between DNA and RNA barcodes. This assembly strategy presented here and the expansion of the rRNA reference library in public databases by 232 new complete 28S and 18S rRNA sequences provide new tools to improve surveillance and microbiome discovery for a more holistic understanding of arbovirus-transmission by mosquitoes in designing preemptive measures.