Evolutionary Dynamics of Oropouche Virus in South America

Abstract

The Amazon basin is home to numerous arthropod-borne viral pathogens that cause febrile disease in humans. Among these, Oropouche orthobunyavirus (OROV) is a relatively understudied member of the genus Orthobunyavirus, family Peribunyaviridae, that causes periodic outbreaks in human populations in Brazil and other South American countries. Although several studies have described the genetic diversity of the virus, the evolutionary processes that shape the OROV genome remain poorly understood. Here, we present a comprehensive study of the genomic dynamics of OROV that encompasses phylogenetic analysis, evolutionary rate estimates, inference of natural selective pressures, recombination and reassortment, and structural analysis of OROV variants. Our study includes all available published sequences, as well as a set of new OROV genome sequences obtained from patients in Ecuador, representing the first set of genomes from this country. Our results show differing evolutionary processes on the three segments that comprise the viral genome. We infer differing times of the most recent common ancestors of the genome segments and propose that this can be explained by cryptic reassortment. We also present the discovery of previously unobserved putative N-linked glycosylation sites, as well as codons that evolve under positive selection on the viral surface proteins, and discuss the potential role of these features in the evolution of OROV through a combined phylogenetic and structural approach.IMPORTANCE The emergence and reemergence of pathogens such as Zika virus, chikungunya virus, and yellow fever virus have drawn attention toward other cocirculating arboviruses in South America. Oropouche virus (OROV) is a poorly studied pathogen responsible for over a dozen outbreaks since the early 1960s and represents a public health burden to countries such as Brazil, Panama, and Peru. OROV is likely underreported since its symptomatology can be easily confounded with other febrile illnesses (e.g., dengue fever and leptospirosis) and point-of-care testing for the virus is still uncommon. With limited data, there is a need to optimize the information currently available. Analysis of OROV genomes can help us understand how the virus circulates in nature and can reveal the evolutionary forces that shape the genetic diversity of the virus, which has implications for molecular diagnostics and the design of potential vaccines.

Keywords: Oropouche virus; arbovirus; bunyavirus; emerging infectious diseases; evolutionary biology; phylogenetic analysis.

FIG 1

Maximum-likelihood trees for the three OROV genome segments. (a) Small (S) segment phylogeny. (b) Medium (M) segment phylogeny. (c) Large (L) segment phylogeny. Branches are color coded by the country of origin for each sample, with gray samples indicating sequences obtained from nonhuman hosts and vectors. (d) A scaled comparison of the three segment phylogenies highlights the different node depths between all trees.

FIG 2

Time-calibrated and evolutionary rate analysis of the OROV genome. (a) A comparison of the three segments reveals an older common ancestor for the M segment, resulting in the divergence of two lineages that appear to diversify in the early to mid-1900s (lineage 1 and lineage 2 are indicated as L1 and L2 in the figure). (b) Estimated evolutionary rates of the segments. (c) Sliding window estimation of the relative genetic distance of each of the two M segment lineages (shown in orange and blue) to Iquitos virus (IQTV), used as an outgroup. Points of the sequences where the closest genotype to the outgroup changes could be interpreted as recombination events with unsampled sequences. (d) An analysis of the younger segments, S and L, reveals emergence date estimates for the virus in the early to middle 20th century, and shows relatively consistent estimates for the TMRCA of the new Ecuadorian sequences in the early 2010s. The posterior probabilities (PP) of each node is shown based on a gray-scale color scheme (0.0 PP = black to 1.0 PP = white).

FIG 3

Structure-based mapping of sites identified under selection and previously unreported N-linked glycosylation sites onto the Gc protein. (a) Schematic of the OROV M segment polyprotein, which includes the Gn and Gc components of the viral membrane proteins. The Gc protein is subdivided into a head domain (blue), two stalk domains (green and orange), and a C-terminal region, which encodes a putative fusion loop (dark gray). Sites identified as being under positive selection using the MEME approach are shown in red (lighter colored arrows indicate sites falling outside the Gc head and stem domains). Site 269 was identified as being under positive selection through multiple approaches and is marked with an asterisk. All N-linked glycosylation sites are shown in white, and the two previously undescribed sites are highlighted in blue and pink. (b) Estimated dN/dS ratio of each domain in Gc, obtained for the whole M segment, and for each M segment lineage. (c) Crystallographically observed trimeric spike structure of the Gc head domain, with sites under positive selection shown in red. The cartoon representation of each chain is embedded in a transparent surface representation (gray). Glycosylation sites are shown as in panel a. (d) Sites under positive selection and residues corresponding to N-linked glycosylation sites mapped to a composite model generated from previously reported OROV Gc (head domain, PDB ID 6H3X) and SBV Gc (stem domains, PDB ID 6H3S) crystal structures. The OROV head domain is shown in blue, with the C terminus loop residues shown in light green (taken from the SBV head domain) and SBV stem domains I and II shown in green and orange, respectively. (e) Putative N-linked glycosylation sites identified in this study are mapped onto the phylogeny of the M segment, showing the independent presence the two sites in the two M segment lineages.

FIG 4

Model of OROV genome evolution and reassortment that can resolve the inconsistent TMRCA estimates of the L and M segments. The upper panel depicts a simplified ancestral graph that describes the joint ancestry of the L (orange) and M (teal) segments. Sampled sequences are represented by dark and light gray circles at the tree tips. These are grouped into lineages 1 and 2, which correspond to the two main lineages in the M and L phylogenies (see Fig. 2). The light-gray branches represent hypothesized, highly diverse, unsampled OROV lineages. The graph contains one reassortment event (black diamond), resulting in the ancestor of lineage 2 in the M segment being descended from a divergent, unsampled lineage (alternatively, the reassortment might have occurred on the branch ancestral to lineage 1). As a result, the TMRCA of the L segment phylogeny (lower right panel; orange circle) exists substantially earlier of the M segment (lower left panel; teal circle). For simplicity, other more recent reassortment events between lineages 1 and 2 are not shown (i.e., the 2009 Mazagão outbreak clade).