Differential Infection Patterns and Recent Evolutionary Origins of Equine Hepaciviruses in Donkeys

Abstract

The hepatitis C virus (HCV) is a major human pathogen. Genetically related viruses in animals suggest a zoonotic origin of HCV. The closest relative of HCV is found in horses (termed equine hepacivirus [EqHV]). However, low EqHV genetic diversity implies relatively recent acquisition of EqHV by horses, making a derivation of HCV from EqHV unlikely. To unravel the EqHV evolutionary history within equid sister species, we analyzed 829 donkeys and 53 mules sampled in nine European, Asian, African, and American countries by molecular and serologic tools for EqHV infection. Antibodies were found in 278 animals (31.5%), and viral RNA was found in 3 animals (0.3%), all of which were simultaneously seropositive. A low RNA prevalence in spite of high seroprevalence suggests a predominance of acute infection, a possible difference from the mostly chronic hepacivirus infection pattern seen in horses and humans. Limitation of transmission due to short courses of infection may explain the existence of entirely seronegative groups of animals. Donkey and horse EqHV strains were paraphyletic and 97.5 to 98.2% identical in their translated polyprotein sequences, making virus/host cospeciation unlikely. Evolutionary reconstructions supported host switches of EqHV between horses and donkeys without the involvement of adaptive evolution. Global admixture of donkey and horse hepaciviruses was compatible with anthropogenic alterations of EqHV ecology. In summary, our findings do not support EqHV as the origin of the significantly more diversified HCV. Identification of a host system with predominantly acute hepacivirus infection may enable new insights into the chronic infection pattern associated with HCV.

Importance: The evolutionary origins of the human hepatitis C virus (HCV) are unclear. The closest animal-associated relative of HCV occurs in horses (equine hepacivirus [EqHV]). The low EqHV genetic diversity implies a relatively recent acquisition of EqHV by horses, limiting the time span for potential horse-to-human infections in the past. Horses are genetically related to donkeys, and EqHV may have cospeciated with these host species. Here, we investigated a large panel of donkeys from various countries using serologic and molecular tools. We found EqHV to be globally widespread in donkeys and identify potential differences in EqHV infection patterns, with donkeys potentially showing enhanced EqHV clearance compared to horses. We provide strong evidence against EqHV cospeciation and for its capability to switch hosts among equines. Differential hepacivirus infection patterns in horses and donkeys may enable new insights into the chronic infection pattern associated with HCV.

Keywords: donkey; equine hepacivirus; evolution; hepatitis C virus; pathogenesis.

FIG 1

EqHV infection patterns. (A) Anti-EqHV antibody (ab) detection depicted in pie charts (red = positive). Asterisks, origin of the EqHV-RNA positive animals. (B) LIPS ratios of control sera form horses and donkeys; Bulgaria includes as well 53 sera from mules. EqHV-RNA positive donkey and mule sera are indicated in red. All three positive sera form France originate from one animal; no seroprevalence rate for this country is indicated due to the low sample size. Dotted line, cutoff (16,249.2 RLU). (C) Seroprevalence rates in different age groups. (D) Aspartate aminotransferase (AST), gamma-glutamyl transferase (γGGT), and glutamate dehydrogenase (GLDH) levels were determined in the sera of Bulgarian donkeys. Sera are shown according to their LIPS status, and RNA-positive samples are given in orange and blue.

FIG 2

Phylogenetic relationships of EqHV, including the novel donkey hepaciviruses. (A) Maximum-likelihood (ML) phylogeny based on the nucleotide sequences encoding for the complete EqHV polyprotein, including the newly described donkey EqHV strains (orange). Bootstrap values larger than 75% are depicted as filled circles. Taxon designations indicate GenBank accession numbers, country and year of sampling. (B to D) ML phylogenies based on the complete NS3 (1,872 nucleotides), partial NS3 (293 nucleotides), and partial NS5B (261 nucleotides), respectively. Cyan, noncontemporary strains from two horses. Partial NS3 sequences for which fewer than 200 nucleotides were characterized were not included in the analysis shown in panel C to avoid further loss of genomic information and robustness of phylogenetic reconstruction.

FIG 3

EqHV evolutionary patterns. (A) Analysis of EqHV polyprotein sequences from three consecutive samples of one EqHV RNA-positive donkey sampled in France. Gray bars, synonymous substitutions; black bars, nonsynonymous substitutions. (B) On the right are the locations of nonsynonymous mutations in the E2 genes of EqHV strains infecting horses. On the left is indicated the ML phylogeny of the translated sequences as before. (C) Mean folding energy differences (MFED) for complete polyprotein sequences of EqHV strains representing both donkey EqHV lineages and all available EqHV polyprotein sequences shown in Fig. 2A. (D) Complete polyprotein ML phylogenies with branch lengths reestimated using either nonsynonymous or synonymous substitutions. Bootstrap values larger than 75% are depicted as filled circles. (E) Root-to-tip divergence plots based on ML trees shown in panel C and Fig. 2A.