PDF, English Restricted access: Repository staff only until 26 April 2025. Login+Download (14MB) | Terms of use

Abstract

The African swine fever virus (ASFV) causes a highly lethal, haemorrhagic fever in both wild and domestic pigs and is currently panzootic with outbreaks in Africa, Asia and Europe. It poses a tremendous threat to livestock, as infection of a single individual leads to slaughter of the entire pig herd. ASFV is the only official member of the virus family Asfarviridae within the order Asfuvirales, and the closest relative to date is a mollusc virus, the Abalone asfa-like virus (AbALV), which also causes a high mortality disease. The aim of this work was to identify unknown asfarviruses from unprocessed sequencing data, which will help to elucidate the origin of ASFV and assess the risk of the emergence of other severe ASFV-like viruses. Using a highly streamlined profile Hidden Markov Model (HMM)-based approach, 320,000 unprocessed DNA and RNA sequencing datasets from the Sequence Read Archive (SRA) were screened for novel members of the Asfuvirales, and those datasets found positive were de-novo assembled and searched for viral genomes and hallmark genes. From a sequencing experiment of a marine shell-less mollusc, a circular asfarvirus genome named after the host Elysia marginata asfarvirus (EMAV) could have been assembled and comparative genomics was conducted to analyse the protein repertoire, orthologous genes and taxonomy with related viruses. Novel asfarvirus markers were identified in six other molluscan hosts, including a close relative of AbALV from a different abalone species, in two transcriptomic datasets of vertebrates, horse and cow, and in thirteen DNA and RNA datasets of protists. The phylogenetic tree reconstruction revealed 22 new virus sequences in the Asfuvirales clade and out of these, 14 sequences clustered in the Asfarviridae family. The novel asfarvirus sequences originated from hosts such as protists, molluscs and mammals. The subtree of the Asfarviridae is divided into three clades, showcased by ASFV, AbALV and EMAV, respectively. The closely related protist viruses and the intermixed hosts within that clade could indicate a high likelihood of an ASFV-like virus that has jumped from a protist to an intermediate (mollusc) or mammalian host. Finally, a second, non-canonical major capsid protein (MCP) was detected in five asfarviruses and protein sequence and structure analyses suggested an independent, monophyletic MCP lineage gained in an ancient asfarvirus. My results shed new light on the evolution, possible origins and host reservoirs of ASFV and in the future, it will be crucial to elucidate the pathogenic and zoonotic potential of EMAV and the other novel ASFV-like virus candidates.