A Preliminary Study of Viral Metagenomics of French Bat Species in Contact with Humans: Identification of New Mammalian Viruses

The prediction of viral zoonosis epidemics has become a major public health issue. A profound understanding of the viral population in key animal species acting as reservoirs represents an important step towards this goal. Bats harbor diverse viruses, some of which are of particular interest because they cause severe human diseases. However, little is known about the diversity of the global population of viruses found in bats (virome). We determined the viral diversity of five different French insectivorous bat species (nine specimens in total) in close contact with humans. Sequence-independent amplification, high-throughput sequencing with Illumina technology and a dedicated bioinformatics analysis pipeline were used on pooled tissues (brain, liver and lungs). Comparisons of the sequences of contigs and unassembled reads provided a global taxonomic distribution of virus-related sequences for each sample, highlighting differences both within and between bat species. Many viral families were present in these viromes, including viruses known to infect bacteria, plants/fungi, insects or vertebrates, the most relevant being those infecting mammals (Retroviridae, Herpesviridae, Bunyaviridae, Poxviridae, Flaviviridae, Reoviridae, Bornaviridae, Picobirnaviridae). In particular, we detected several new mammalian viruses, including rotaviruses, gammaretroviruses, bornaviruses and bunyaviruses with the identification of the first bat nairovirus. These observations demonstrate that bats naturally harbor viruses from many different families, most of which infect mammals. They may therefore constitute a major reservoir of viral diversity that should be analyzed carefully, to determine the role played by bats in the spread of zoonotic viral infections.

The family Dicistroviridae encompasses small single-stranded positive-sense RNA viruses infecting invertebrates, which can be distinguished from members of the "picornavirus-like superfamily" taxon on the basis of the location of the non-structural protein genes (ORF-1) at the 5' end of the genome rather than at the 3' end. Dicistroviruses have been isolated from six invertebrate orders from the Insecta and decapod crustaceans [4]. In this study, multiple contig sequences from bat specimen b2 (Pipistrellus pipistrellus) displayed various HSPs related to the family Dicistroviridae, matching structural and nonstructural viral proteins in BLASTx analysis (Table S1). The low degree of nucleotide sequence identity observed, with overlapping regions for some of these HSP sequences, suggested that different dicistroviruses were present (data not shown). Based on the longest HSP sequence (1,819 nt) matching the ORF1 gene, a specific set of primers was designed and used to confirm the presence of dicistrovirus sequences in the lungs of specimen b2 (Table S1 and Fig. S1A).
Phylogenetic analysis demonstrated that the related virus, Paris dicistrovirus, most closely matched the 2/4 nonstructural polyprotein region of the Cripavirus genus. However, it was different from the other dicistroviruses described (Fig. S1B).
The family Nodaviridae encompasses non enveloped isometric viruses with two positive sense-strand RNA molecules, RNA-1 and RNA-2. The RNA-1 segment encodes protein A (ORF1a), a catalytic subunit of the RNA-dependent RNA polymerase (RdRp), and protein B (ORF1b), whereas the second segment, RNA-2, encodes the coat precursor protein alpha [5]. This family contains two genera, Alphanodavirus and Betanodavirus, which principally infect insects and fish [6]. We identified multiple read and contig sequences from pooled tissues of bat specimen b7 (Eptesicus serotinus) that matched nodavirus sequences in BLASTn and BLASTx analysis (Table S1). The presence of nodavirus sequences was further confirmed by PCR on the brain sample of this specimen (Table S1 and Fig. S2A). Phylogenetic analysis performed after translation to obtain the amino-acid sequence encoded by this contig demonstrated that this virus, the Sers nodavirus, belonged to the genus Alphanodavirus, and was closely related to Boolarra virus (Fig. S2B).  (Table S1 and Fig. S3A).

3/4
Phylogenetic analysis, based on the translated HSP, confirmed that this virus, tentatively named Bordeaux luteovirus, belonged to the family Luteoviridae, and demonstrated that it was a putative new viral genus closely related to the genus Enamovirus (Fig. S3B).
The genus Sobemovirus encompasses plant RNA viruses with icosahedral virions containing a positive-sense single-stranded RNA genome of almost 4 kb in length containing three or four ORFs encoding various proteins, including the coat protein (ORF4). The host range of each sobemovirus species is narrow and confined to a few plant species from Poaceae or Fabaceae. In this study, we identified one HSP derived from a single contig sequence from the same bat specimen b3, which was found to be closely related to the coat protein of sobemoviruses in BLASTx analysis (Table S1 and Phage-related sequences corresponded mostly to enterobacterial phages, which probably infected the intestinal bacterial population. However, the length of contig sequences obtained after de novo assembly was not sufficient for subsequent relevant sequence comparisons or phylogenetic analysis.