High Permissiveness for Genetic Exchanges between Enteroviruses of Species A, including Enterovirus 71, Favors Evolution through Intertypic Recombination in Madagascar

Abstract

Human enteroviruses of species A (EV-A) are the leading cause of hand-foot-and-mouth disease (HFMD). EV-A71 is frequently implicated in HFMD outbreaks and can also cause severe neurological manifestations. We investigated the molecular epidemiological processes at work and the contribution of genetic recombination to the evolutionary history of EV-A in Madagascar, focusing on the recently described EV-A71 genogroup F in particular. Twenty-three EV-A isolates, collected mostly in 2011 from healthy children living in various districts of Madagascar, were characterized by whole-genome sequencing. Eight different types were identified, highlighting the local circulation and diversity of EV-A. Comparative genome analysis revealed evidence of frequent recent intra- and intertypic genetic exchanges between the noncapsid sequences of Madagascan EV-A isolates. The three EV-A71 isolates had different evolutionary histories in terms of recombination, with one isolate displaying a mosaic genome resulting from recent genetic exchanges with Madagascan coxsackieviruses A7 and possibly A5 and A10 or common ancestors. The engineering and characterization of recombinants generated from progenitors belonging to different EV-A types or EV-A71 genogroups with distantly related nonstructural sequences indicated a high level of permissiveness for intertypic genetic exchange in EV-A. This permissiveness suggests that the primary viral functions associated with the nonstructural sequences have been highly conserved through the diversification and evolution of the EV-A species. No outbreak of disease due to EV-A has yet been reported in Madagascar, but the diversity, circulation, and evolution of these viruses justify surveillance of EV-A circulation and HFMD cases to prevent possible outbreaks due to emerging strains.IMPORTANCE Human enteroviruses of species A (EV-A), including EV-A71, are the leading cause of hand-foot-and-mouth disease (HFMD) and may also cause severe neurological manifestations. We investigated the circulation and molecular evolution of EV-A in Madagascar, focusing particularly on the recently described EV-A71 genogroup F. Eight different types, collected mostly in 2011, were identified, highlighting the local circulation and diversity of EV-A. Comparative genome analysis revealed evidence of frequent genetic exchanges between the different types of isolates. The three EV-A71 isolates had different evolutionary histories in terms of recombination. The engineering and characterization of recombinants involving progenitors belonging to different EV-A types indicated a high degree of permissiveness for genetic exchange in EV-A. No outbreak of disease due to EV-A has yet been reported in Madagascar, but the diversity, circulation, and evolution of these viruses justify the surveillance of EV-A circulation to prevent possible HFMD outbreaks due to emerging strains.

Keywords: RNA virus; emergence; enterovirus; picornavirus; recombination; viral evolution.

FIG 1

Maximum likelihood analysis of the nucleotide sequences of different genomic regions of the enterovirus A species. A sequence alignment including the complete genomes of 153 virus strains from 21 types of enterovirus A was constructed and split into four partition data sets corresponding to the capsid-encoding region P1 (A), the 5′ untranslated region (5′ UTR) (B), and the nonstructural protein-encoding regions P2 (C) and P3 (D). Phylograms were reconstructed by the maximum likelihood (ML) method. Bootstrap values of ≥70% are indicated at key nodes. For clarity, consistent clusters of strains are boxed with the name of the corresponding EV-A type or EV-A71 genogroup. Isolate naming is according to the following scheme: EV type, laboratory name, country code (e.g., MDG), and the last two numbers of the year of isolation. Madagascan sequences are highlighted in red. Types of isolates are indicated on the right (colored vertical lines and type names). The two main clusters of P2 and P3 sequences are indicated as clades I and II. Scale bars indicate the ML of substitution per nucleotide position. Gen, genogroup; Sgen, subgenogroup.

FIG 2

Genomic structure of mosaic recombinant genomes of enterovirus A isolates. (A) Similarity plot analyses of different EV-A genomic sequences, with the genomic sequence of EV-A71 MAD7842-MDG10 as a reference. A scheme indicating the genomic organization of EV and the structural and nonstructural regions is shown. Similarity plot profiles are color coded according to EV type or the genogroup of EV-A71. Numbers of isolates for each type and genogroup are indicated in brackets. The nucleotide sequence identity threshold indicating evolution through genetic exchanges in the noncoding and nonstructural coding regions was set at 95%. Additional similarity plots with this and other EV-A genomes as references are shown Fig. S1 in the supplemental material. (B) Putative genomic crossovers with recombinant partners and positions in the nonstructural protein-coding regions of a set of EV-A genomes. The consistency of recombination events was assessed with seven different statistical methods implemented in a recombination detection program (RDP4) (44). Genomes for which significant evidence of recombination (P value of <0.001) was obtained with 7 methods (****), 6 methods (***), 5 methods (**), and 4 methods (*) are indicated. The nucleotide positions of the recombinant parts are indicated by dots if determined by the program (99% confidence interval) or by arrows if undetermined. Detailed values obtained with the program are presented in Table S2.

FIG 3

Evolutionary history of EV-A71 genogroup F, as assessed by maximum clade credibility phylogenetic analysis. (A) Patterns of evolution were reconstructed for EV-A71 from the nucleotide sequences encoding the capsid protein 1D^VP1 and for a segment of the P3 (nonstructural protein-encoding) region, as indicated on the schematic diagram of the EV genome. (B) Maximum clade credibility tree based on the capsid protein 1D^VP1 sequence, including EV-A71 of various geographic origins from all described EV-A71 (sub)genogroups. (C) Maximum clade credibility tree based on the nonstructural P3 region specifically including EV-A71 strains within genogroup F and EV-A sequences belonging to clade I (Fig. 1D). The maximum clade credibility trees were inferred by Bayesian Markov chain Monte Carlo analysis, with a discrete model, to determine the virus corresponding to tree branches. Tips are scaled according to the date of virus sampling (timescale on the x axis). The main lineages are color coded according to EV-A71 (sub)genogroup and EV-A type. Time to the most recent common ancestor (tMRCA) and highest posterior probability density (HPD) intervals (blue lines) are indicated for the main nodes. The size of the circles at the nodes is proportional to posterior probability (PP).

FIG 4

Comparative plaque phenotypes of the progenitor and chimeric viruses. Plaque phenotype (A to D) was assessed for the progenitor and chimeric viruses on Vero E6 cells or RD cells for viruses with the 5′ half of the CV-A10 genome. Negative and positive controls (coxsackievirus B3) are presented (E). Schematic diagrams of the progenitor and chimeric genomes are shown. Virus names and genomes are color coded according to genetic background and affiliation to the nonstructural CDS clade I (red) or clade II (blue) (Fig. 1C and D). Titers of viral suspensions are indicated.

FIG 5

Replication kinetics of the progenitor and chimeric viruses. Replication kinetics of the progenitor (black) and chimeric (gray and white) viruses were analyzed in triplicate following infection of Vero E6 cells or RD cells. RD cells were used for viruses carrying the 5′ half of the CV-A10 genome. Infected cells and supernatants were collected at 0, 6, 16, 24, and 48 h postinfection (hpi). Virus production is expressed as the number of genome copies and TCID₅₀, as indicated. The data are presented are mean values ± standard deviations. The occurrence of a cytopathic effect and complete cell lysis are indicated. Virus names and genomes are color coded according to genetic background and affiliation to the nonstructural CDS clade I (red) or clade II (blue), as indicated in the legend of Fig. 4.

FIG 6

Circulation of enterovirus A isolates in different geographic areas of Madagascar. The Madagascan isolates used in this study are color coded by type and shown on the map of the island (map from Wikimedia Commons Atlas of the World [https://commons.wikimedia.org/wiki/Atlas_of_the_world]).