Phylodynamic Characterization of an Ocular-Tropism Coxsackievirus A24 Variant

Abstract

Recent phylodynamic studies have focused on using tree topology patterns to elucidate interactions among the epidemiological, evolutionary, and demographic characteristics of infectious agents. However, because studies of viral phylodynamics tend to focus on epidemic outbreaks, tree topology signatures of tissue-tropism pathogens might not be clearly identified. Therefore, this study used a novel Bayesian evolutionary approach to analyze the A24 variant of coxsackievirus (CV-A24v), an ocular-tropism agent of acute hemorrhagic conjunctivitis. Analyses of the 915-nucleotide VP1 and 690-nt 3Dpol regions of 21 strains isolated in Taiwan and worldwide during 1985-2010 revealed a clear chronological trend in both the VP1 and 3Dpol phylogenetic trees: the emergence of a single dominant cluster in each outbreak. The VP1 sequences included three genotypes: GI (prototype), GIII (isolated 1985-1999), and GIV (isolated after 2000); no VP1 sequences from GII strains have been deposited in GenBank. Another five genotypes identified in the 3Dpol region had support values >0.9. Geographic and demographic transitions among CV-A24v clusters were clearly identified by Bayes algorithm. The transmission route was mapped from India to China and then to Taiwan, and each prevalent viral population declined before new clusters emerged. Notably, the VP1 and 3Dpol genes had high nucleotide sequence similarities (94.1% and 95.2%, respectively). The lack of co-circulating lineages and narrow tissue tropism affected the CV-A24v gene pool.

Fig 1

Comparison of coxsackievirus A24v strains in (A–C) 305 codon sites of 111 VP1 sequences and in (D–F) 230 codon sites of 44 3Dpol sequences. (A, D) Secondary structure guide (PDB ID codes 4Q4V for VP1, 2IJD for 3Dpol products). (B, E) Consensus residues graphically depicted by Weblogo. (C, F) Entropy-based variability plotted by DEMBA.

Fig 2. Maximum clade credibility phylogeny of 111 VP1 sequences of coxsackievirus A24v without outgroup.

Blue bars at nodes indicate 95% HPDs of TMRCAs. Branch color indicates the likely location, and branch thickness indicates posterior probability from Bayesian inference (PP-BMCMC). Support values are also shown for major nodes and are indicated as BS-NJ/BS-ML/PP-BMCMC, where BS is the bootstrap value, NJ indicates the neighbor-joining method, and ML indicates the maximum likelihood method. The VP1 genotypes are shown on the right and are distinguished by color. Within-genotype nucleotide and amino acid similarities are also shown on the right. For comparison, the 3D^pol genotypes are distinguished by shading (GA: purple, GB: green, GC: yellow, GD: orange, and GE: blue). The branch length indicates the evolution time, and the scale bar at the bottom indicates the calendar time.

Fig 3

Bayesian skyline plots of (A) 111 VP1 sequences, (B) 11 VP1 sequences of genotype III strains, and (C) 99 VP1 sequences genotype IV strains. The x-axis is the time scale (years), and the y-axis is the logarithmic Neτ scale (where Ne is the effective population size and τ is the generation time). The thick solid line indicates the median estimates, and the shaded area indicates the 95% highest posterior density.

Fig 4. Maximum clade credibility phylogeny of 111 VP1 sequences of coxsackievirus A24v with outgroups.

Branch thickness indicates support value (posterior probability, PP), and branch color indicates the most probable location. Support values for major nodes are also shown. The right side of the figure shows within-genotype nucleotide/amino acid similarities. The VP1 genotypes for each strain are distinguished by color (Genotype I: purple, Genotype III: orange, and Genotype IV: blue). The 3D^pol genotypes for each strain are distinguished by shading. Evolution time is indicated by branch length (see scale bar).