Characterisation of the faecal virome of captive and wild Tasmanian devils using virus-like particles metagenomics and meta-transcriptomics

Rowena Chong,Mang Shi,Catherine E. Grueber,Edward C. Holmes,Carolyn J. Hogg,Katherine Belov,Vanessa R. Barrs
DOI: https://doi.org/10.1101/443457
2018-01-01
bioRxiv
Abstract:Background: The Tasmanian devil is an endangered carnivorous marsupial threatened by devil facial tumour disease (DFTD). While research on DFTD has been extensive, little is known about the viruses present in devils, and whether any of these are of potential conservation relevance for this endangered species. Methods: Using both metagenomics based on virus-like particle (VLP) enrichment and sequence-independent amplification (VLP metagenomics), and meta-transcriptomics based on bulk RNA sequencing, we characterised and compared the faecal viromes of captive and wild Tasmanian devils. Results: A total of 54 devil faecal samples collected from captive (n = 2) and wild (n = 4) populations were processed for virome characterisation using both approaches. We detected many novel, highly divergent viruses, including vertebrate viruses, bacteriophage and other dietary associated plant and insect viruses. In total, 18 new vertebrate viruses, including novel sapelovirus, astroviruses, bocaviruses, papillomaviruses and gammaherpesvirus were identified, as well as known mammalian pathogens including rabbit haemorrhagic disease virus 2 (RHDV2). Captive devils showed significantly lower levels of viral diversity than wild devils. Comparison of the two methodological approaches revealed substantial differences in the number and types of viruses detected, with meta-transcriptomics mainly identifying RNA viruses, and VLP metagenomics largely identifying DNA viruses. Conclusion: This study has greatly expanded our knowledge of eukaryotic viruses in the Tasmanian devil and provides important baseline information that will contribute to the conservation and captive management of this endangered species. In addition, our results showed that a combination of VLP metagenomics and meta-transcriptomics may be a more comprehensive approach to virome characterisation than either method alone.
What problem does this paper attempt to address?