Highly divergent hepaciviruses from African cattle

Abstract

The hepatitis C virus (HCV; genus Hepacivirus) is a highly relevant human pathogen. Unique hepaciviruses (HV) were discovered recently in animal hosts. The direct ancestor of HCV has not been found, but the genetically most closely related animal HVs exist in horses. To investigate whether other peridomestic animals also carry HVs, we analyzed sera from Ghanaian cattle for HVs by reverse transcription-PCR (RT-PCR). Nine of 106 specimens from different sampling sites contained HV RNA (8.5%) at median viral loads of 1.6 × 10(5) copies/ml. Infection seemed unrelated to cattle age and gender. Near-full-genome sequencing of five representative viruses confirmed taxonomic classifications. Cattle HVs formed two distinct phylogenetic lineages that differed by up to 17.7% on the nucleotide level in the polyprotein-encoding region, suggesting cocirculation of different virus subtypes. A conserved microRNA122-binding site in the 5' internal ribosomal entry site suggested liver tropism of cattle HVs. Phylogenetic analyses suggested the circulation of HVs in cattle for several centuries. Cattle HVs were genetically highly divergent from all other HVs, including HCV. HVs from genetically related equine and bovine hosts were not monophyletic, corroborating host shifts during the evolution of the genus Hepacivirus. Similar to equine HVs, the genetic diversity of cattle HVs was low compared to that of HCV genotypes. This suggests an influence of the human-modified ecology of peridomestic animals on virus diversity. Further studies should investigate the occurrence of cattle HVs in other geographic areas and breeds, virus pathogenicity in cattle, and the potential exposure of human risk groups, such as farmers, butchers, and abattoir workers.

Importance: HCV (genus Hepacivirus) is a major human pathogen, causing liver failure and cancer. Unique hepaciviruses (HVs) were discovered over the last few years in animals, but the direct ancestor of HCV has not been found. The animal HV most closely related to HCV so far originated from horses, suggesting that other livestock animals also harbor HVs. Therefore, we investigated African cattle and discovered previously unknown HVs at high prevalence and viral loads. Because of the agricultural importance of cattle, it may be relevant to investigate HV pathogenicity. The frequent exposure of humans to cattle also may warrant investigations of the zoonotic potential of these viruses. Evolutionary analyses suggested that cattle HVs have existed for centuries. Despite the genetic relatedness of their animal hosts, HVs from cattle and horses were not phylogenetically related, corroborating frequent host shifts during the evolution of the genus Hepacivirus.

FIG 1

Sampling sites. (A) Geographic location of Kumasi, central Ghana. (B) Twelve sampling sites in the metropolitan area of Kumasi were marked with dots and numbered from west to east. Gray background, Kumasi city limits. Positive sampling sites are given in red. The maps were created using QGIS (www.qgis.org) and data freely available under an OpenStreetMap license (www.openstreetmap.org).

FIG 2

Phylogenetic characterization of cattle hepaciviruses. (A) Neighbor-joining phylogenetic reconstruction of a 3,715-nt fragment encompassing the complete hepacivirus (HV) structural genome region using a complete deletion option. Numbers at nodes indicate support of grouping from 1,000 bootstrap replicates. Only values above 75% are shown. Corresponding sampling sites are given to the right. (B) Bayesian phylogeny of the complete HV polyprotein done in MrBayes using a WAG amino acid substitution model and 2,000,000 generations sampled every 100 steps with 25% discarded as burn-in. The branch leading to the bovine viral diarrhea virus 1 (BVDV) outgroup was truncated for graphical reasons. Filled circles, posterior probability support of 1.0. Hosts are depicted to the right of taxa, which are colored accordingly. The GenBank accession number of viruses used in this study were the following: NZP-1, JQ434001; A6-006, JQ434003; G5-077, JQ434006; G10-73, JQ434002; H3-011, JQ434008; F80-68, JQ434005; B10-022, JQ434004; H10-094, JQ434007; CHV, JQ434007; HCV1a, NC_004102; HCV2a, AB047639; HCV3a, X76918; HCV4a, Y11604; HCV5a, Y13184; HCV6a, AY859526; HCV7, EF108306; PDB829, KC796074; PDB491, KC796078; PDB452, KC796090; PDB445, KC796091; RMU10-3382, KC411777; NLR07-oct70, KC411784; NLR08-365, KC411796; SAR-3, KC411806; SAR-46, KC411807; RHV089, KC815312; NrHV-1, KJ950938; NrHV-2, KJ950939; RHV339, NC_021153; BWC08, KC551800; BWC05, KC551801; BWC04, KC551802; GBV-B, NC_001655; PDB112, KC796077. Branches leading to novel cattle HVs are in red.

FIG 3

Genomic characterization of cattle hepaciviruses. (A) Genome organization of cattle HVs. Black arrows on the top indicate predicted signal peptidase cleavage sites; red arrows indicate N-linked glycosylation sites. (B) Comparison of amino acid sequence identity within and between HV polyproteins, including five novel cattle HVs, one representative each of HCV genotypes 1 to 7, and all available equine HVs (see the legend to Fig. 2 for GenBank accession numbers), calculated using SSE V1.1 with a sliding window of 400 and a step size of 200 residues. (C) Mean minimum folding energy differences (MFED), calculated using SSE V1.1, comparing the original sequences of the five full cattle HV polyprotein genes against 50 replicates scrambled according to dinucleotide content and coding sequence (CDLR option) with a sliding window of 250 and a step size of 30 nucleotides.

FIG 4

Cattle hepacivirus genome ends. (A) 5′-End secondary structure of cattle HVs (represented by virus GHC25) with stem-loops numbered next to structures. The microRNA122-binding site is given in blue. PK, pseudoknot. (B) 3′-End secondary structure of cattle HVs. Kissing loops are highlighted. Differences between the primary sequences of the 5′ and 3′ genome ends are depicted in black next to structures; covariant mutations are in green.