Mesoniviruses are mosquito-specific viruses with extensive geographic distribution and host range

Abstract

Background: The family Mesoniviridae (order Nidovirales) comprises of a group of positive-sense, single-stranded RNA ([+]ssRNA) viruses isolated from mosquitoes.

Findings: Thirteen novel insect-specific virus isolates were obtained from mosquitoes collected in Indonesia, Thailand and the USA. By electron microscopy, the virions appeared as spherical particles with a diameter of ~50 nm. Their 20,129 nt to 20,777 nt genomes consist of positive-sense, single-stranded RNA with a poly-A tail. Four isolates from Houston, Texas, and one isolate from Java, Indonesia, were identified as variants of the species Alphamesonivirus-1 which also includes Nam Dinh virus (NDiV) from Vietnam and Cavally virus (CavV) from Côte d'Ivoire. The eight other isolates were identified as variants of three new mesoniviruses, based on genome organization and pairwise evolutionary distances: Karang Sari virus (KSaV) from Java, Bontag Baru virus (BBaV) from Java and Kalimantan, and Kamphaeng Phet virus (KPhV) from Thailand. In comparison with NDiV, the three new mesoniviruses each contained a long insertion (180 - 588 nt) of unknown function in the 5' region of ORF1a, which accounted for much of the difference in genome size. The insertions contained various short imperfect repeats and may have arisen by recombination or sequence duplication.

Conclusions: In summary, based on their genome organizations and phylogenetic relationships, thirteen new viruses were identified as members of the family Mesoniviridae, order Nidovirales. Species demarcation criteria employed previously for mesoniviruses would place five of these isolates in the same species as NDiV and CavV (Alphamesonivirus-1) and the other eight isolates would represent three new mesonivirus species (Alphamesonivirus-5, Alphamesonivirus-6 and Alphamesonivirus-7). The observed spatiotemporal distribution over widespread geographic regions and broad species host range in mosquitoes suggests that mesoniviruses may be common in mosquito populations worldwide.

Figure 1

Ultrastructure of mesoniviruses in C6/36 cells. Bars = 100 nm. (A) virus particles of Nam Dinh strain V3872 inside a cytoplasmic vacuole; (B) virions of Bontag Baru JKT-9876 strain at the cell surface (thin arrow) and inside a cytoplasmic vacuole (thick arrow); (C) virions of Bontag Baru strain JKT-9876 free in the cytosol (thin arrows). Thick arrow indicates a portion of a paracrystalline agglomerate of empty virus particles in the host cell cytosol; (D) virions of Bontag Baru strain JKT-9853 at the surface of the cell (thin arrow) and inside a cytoplasmic vacuole (thick arrow); (E) virus particles of Karang Sari (JKT-10701) inside an expanded cistern of granular endoplasmic reticulum (thick arrow); (F) virus particles of Bontag Baru JKT-7774 inside a cytoplasmic vacuole; (G and H) virus particles of Kamphaeng Phet strain KP84-0156 inside cytoplasmic vacuoles (thick arrows) and free in cytosol (thin arrow in H).

Figure 2

Genome organisation of the mesoniviruses. In ORF1a and ORF1b the putative ribosomal frame-shift (RFS) element and the locations of conserved functional domains, including the 3C-like serine protease (3CLP), RNA-dependent RNA polymerase (RdRp), zinc-binding (Z), helicase (Hel), exoribonuclease (ExoN), N7-methyltransferase (NMT) and 2’-O-methyltransferase (OMT) domains, are shown. Sequence insertions in ORF1a are shown as black blocks. In ORF2a encoding the S glycoprotein the sites of proteolytic cleavage to generate S1, S2 and the uncharacterised N-terminal fragment are shown as black inverted triangles. Putative ORF4 that is relatively conserved in the 3’terminal region of the genome is shown as black rectangles and other small ORFs >180 nt that occur variously in the 3’-terminal domains and elsewhere in the genome are shown as grey rectangles.

Figure 3

Phylogenetic analysis of the mesonivirus proteins. (A). Maximum-likelihood phylogeny of the full-length proteins encoded in ORF2a (unprocessed S proteins). (B). Maximum-likelihood analysis of conserved protein domains of ORF1ab (3CL^pro, RdRp, HEL1). Scale bars indicate amino acid substitutions/site.

Figure 4

Polypeptides encoded in the ORF3a/ORF3b region of mesonivirus genomes. (A). Clustal X alignment of the sequences of polypeptides encoded in ORF3a illustrating the predicted signal peptide, exodomain, transmembrane domain and cytoplasmic domain, two domains of highly conserved sequence (boxed), conserved cysteine residues (black shading) and a potential N-glycosylation site (underlined). (B). Clustal X alignment of the sequences of polypeptides encoded in ORF3b illustrating the cytoplasmic domain, transmembrane domain and ectodomain and a potential N-glycosylation site (underlined). (C). Clustal X alignment of the sequences of polypeptides that would be generated by -1 ribosomal frame-shift at the predicted ‘slippery’ sequence (CACUUUU) in the overlap region. (D). Schematic representation of the predicted membrane topology of the putative 3a, 3b and 3ab glycoproteins using SignalP and TMHMM and NetNGlyc (http://www.expasy.org).

Figure 5

Predicted structures of the mesonivirus ORF1a/ORF1b ribosomal frame-shift element (RFS) element. (A). Stem-loop structure predicted previously for NDiV [3] using pseudoknotsRG software, illustrating nucleotides that are substituted in various viruses (red circles). Nucleotide substitutions (which are primarily non-compensatory) are shown in the boxes. The corresponding sequence alignment is shown in Additional file 4: Figure S4. (B). An alternative conserved pseudoknot structure predicted from the multiple sequence alignment using ipknot software. The putative ‘slippery’ sequence (GGAUUUU) at the ribosomal frame-shift site is shaded in grey.