Changes in diversification patterns and signatures of selection during the evolution of murinae-associated hantaviruses

Abstract

In the last 50 years, hantaviruses have significantly affected public health worldwide, but the exact extent of the distribution of hantavirus diseases, species and lineages and the risk of their emergence into new geographic areas are still poorly known. In particular, the determinants of molecular evolution of hantaviruses circulating in different geographical areas or different host species are poorly documented. Yet, this understanding is essential for the establishment of more accurate scenarios of hantavirus emergence under different climatic and environmental constraints. In this study, we focused on Murinae-associated hantaviruses (mainly Seoul Dobrava and Hantaan virus) using sequences available in GenBank and conducted several complementary phylogenetic inferences. We sought for signatures of selection and changes in patterns and rates of diversification in order to characterize hantaviruses' molecular evolution at different geographical scales (global and local). We then investigated whether these events were localized in particular geographic areas. Our phylogenetic analyses supported the assumption that RNA virus molecular variations were under strong evolutionary constraints and revealed changes in patterns of diversification during the evolutionary history of hantaviruses. These analyses provide new knowledge on the molecular evolution of hantaviruses at different scales of time and space.

Figure 1

Tanglegram of the two phylogenetic trees based on analyses of segments M and S. Grey lines are used to illustrate the segments that were isolated from the same virus strains. Hence, a similar topology for both segments should result in horizontal grey line. By contrast, oblique lines indicate dissimilarities of strain position, which most probably results from reassortments of viral segments between differentiated strains. Note that this pattern of dissimilarities seems to be much more frequent in the HNTV clade and the SEOV-GOUV-JURV clade than in the three other clades. ∗ indicate bootstrap values >80%.

Figure 2

M and S segments of murine-associated hantaviruses BEAST time calibrated (ultrametric) trees. Root age was arbitrarily assigned to 1. A Generalized mixed Yule coalescent (GMYC) model was used to determine thresholds such that nodes before the threshold are considered as species diversification events, whereas branches crossing the thresholds define clusters following a coalescent process (in red). Shifts in diversification rates determined using MEDUSA algorithm are represented by numbered blue circles. Increases in lineage diversification rates among the tree compared to the reminder tree determined by the PRC Parametric Rate Comparison) test implemented in the iterates package are represented by orange arrows. Inserts illustrate the Lineage-through-time (LTT) plots based on the BEAST time calibrated (ultrametric) trees for the M and S segment of murine-associated hantaviruses in which the occurrence and the position of the switch between two evolution patterns (red line) was determined by the method of Pons et al. implemented in the package SPLITS [54].

Figure 3

Detection of phylotypes (subsets of taxa with close phylogenetic relationships and common geographic trait values) among the phylogenies of M (a) and S, for the small dataset (b) and for the large dataset (c), segments of murine-associated hantaviruses by Phylotype [55]. Results are presented on the ML phylogenetic trees, in background and in which the selected phylotypes are represented by coloured regions. Phylotype identifiers are indicated in circles and refer to Pi in Table S2.