Molecular comparison and evolutionary analyses of VP1 nucleotide sequences of new African human enterovirus 71 isolates reveal a wide genetic diversity

Abstract

Most circulating strains of Human enterovirus 71 (EV-A71) have been classified primarily into three genogroups (A to C) on the basis of genetic divergence between the 1D gene, which encodes the VP1 capsid protein. The aim of the present study was to provide further insights into the diversity of the EV-A71 genogroups following the recent description of highly divergent isolates, in particular those from African countries, including Madagascar. We classified recent EV-A71 isolates by a large comparison of 3,346 VP1 nucleotidic sequences collected from GenBank. Analysis of genetic distances and phylogenetic investigations indicated that some recently-reported isolates did not fall into the genogroups A-C and clustered into three additional genogroups, including one Indian genogroup (genogroup D) and 2 African ones (E and F). Our Bayesian phylogenetic analysis provided consistent data showing that the genogroup D isolates share a recent common ancestor with the members of genogroup E, while the isolates of genogroup F evolved from a recent common ancestor shared with the members of the genogroup B. Our results reveal the wide diversity that exists among EV-A71 isolates and suggest that the number of circulating genogroups is probably underestimated, particularly in developing countries where EV-A71 epidemiology has been poorly studied.

Figure 1. Phylogenetic relationships between EV-A71 full-length VP1 sequences inferred using the Neighbor-Joining method.

The evolutionary distances were computed using the Kimura 2-parameter method. The length of the branches is proportional to the number of nt divergence. The percent of bootstrap replicates is indicated for the main nodes.

Figure 2. Molecular differentiation of the EV-A71 genogroups.

The percentages of nucleotide and amino acid divergences were calculated in pairwise comparisons of 59 sequences selected as representative of the EV-A71 genogroups and subgenogroups. A scatter plot was produced with the divergence values to determine a threshold between intra-and inter-genogroup differences.

Figure 3. Bayesian maximum clade credibility phylogram (A) and chronogram (B) inferred with the 1D gene encoding the VP1 capsid protein, depicting the phylogeny of EV-A71.

The trees were estimated with general time reversible nucleotide substitution model, an uncorrelated exponential clock model, and the Bayesian skyline as a tree prior. The posterior probability of the eight nodes used to compare different evolutionary models (see also Figures 4 and 3) are indicated. The posterior probability (pp) are indicated for the main nodes.

Figure 4. Time to the most recent common ancestor (TMRCA) calculated with different models for major nodes of interest within the EV-A71 phylogeny (see also Figure 3 for the location of these nodes).

The phylogenetic reconstructions were done with the general time reversible (squares) or SRD06 (triangles) nucleotide substitution models, a relaxed clock model (uncorrelated exponential distribution of rates, solid symbols; lognormal distribution of rates, open symbols), and the Bayesian skyline as a demographical tree prior. The bars indicate the Bayesian credibility intervals (95% highest posterior density, 95% HPD) intervals estimated from the posterior distribution obtained through Markov chain Monte Carlo analyses.