Molecular epidemiology of human coronavirus OC43 reveals evolution of different genotypes over time and recent emergence of a novel genotype due to natural recombination

Abstract

Although human coronavirus OC43-OC43 (HCoV-OC43) is the coronavirus most commonly associated with human infections, little is known about its molecular epidemiology and evolution. We conducted a molecular epidemiology study to investigate different genotypes and potential recombination in HCoV-OC43. Twenty-nine HCoV-OC43 strains from nasopharyngeal aspirates, collected from 2004 to 2011, were subjected to RNA-dependent RNA polymerase (RdRp), spike, and nucleocapsid gene analysis. Phylogenetic analysis showed at least three distinct clusters of HCoV-OC43, although 10 unusual strains displayed incongruent phylogenetic positions between RdRp and spike genes. This suggested the presence of four HCoV-OC43 genotypes (A to D), with genotype D most likely arising from recombination. The complete genome sequencing of two genotype C and D strains and bootscan analysis showed recombination events between genotypes B and C in the generation of genotype D. Of the 29 strains, none belonged to the more ancient genotype A, 5 from 2004 belonged to genotype B, 15 from 2004 to 2006 belonged to genotype C, and 1 from 2004 and all 8 from 2008 to 2011 belonged to the recombinant genotype D. Molecular clock analysis using spike and nucleocapsid genes dated the most recent common ancestor of all genotypes to the 1950s, genotype B and C to the 1980s, genotype B to the 1990s, and genotype C to the late 1990s to early 2000s, while the recombinant genotype D strains were detected as early as 2004. This represents the first study to describe natural recombination in HCoV-OC43 and the evolution of different genotypes over time, leading to the emergence of novel genotype D, which is associated with pneumonia in our elderly population.

Fig. 1.

Phylogenetic analysis of the complete RdRp, S, and N genes of the 21 HCoV-OC43 strains. The trees were constructed by the maximum-likelihood method, and bootstrap values were calculated from 1,000 trees. A total of 2,783, 4,089, and 1,347 nucleotide positions in RdRp, S, and N, respectively, were included in the analysis. The scale bar indicates the estimated number of substitutions per 2,000, 1,000, or 20 nucleotides as indicated. A, genotype/clade A; B, genotype/clade B; C, genotype/clade C; D, genotype D. Genotype D strains displaying incongruent phylogenetic relationships are in boldface.

Fig. 2.

Genome organization and bootscan analysis of the HCoV-OC43 genomes. Bootscanning was conducted with Simplot version 3.5.1 (F84 model; window size, 1,000 bp; step, 200 bp) on a gapless nucleotide alignment, which was generated with ClustalX with the genome sequences of strains HK04-02 (upper) and BE04 (lower) as the query sequences.

Fig. 3.

Phylogenetic analysis of nsp1 to nsp16, NS2a, HE, S, NS5a, E, M, and N genes of seven HCoV-OC43 genomes. The trees were constructed by the neighbor-joining method using Kimura's two-parameter correction, and bootstrap values were calculated from 1,000 trees. A total of 738, 1,815, 5,697, 1,488, 909, 861, 267, 591, 330, 411, 46, 2,784, 1,809, 1,563, 1,125, 900, 837, 1,244, 4,092, 330, 248, 678, and 1,347 nucleotide positions in nsp1, nsp2, nsp3, nsp4, nsp5, nsp6, nsp7, nsp8, nsp9, nsp10, nsp12, nsp13, nsp14, nsp15, nsp16, NS2a, HE, S, NS5a, E, M, and N, respectively, were included in the analysis. The scale bar indicates the estimated number of substitutions per 50 or 100 nucleotides as indicated. The corresponding nucleotide sequences of HCoV-HKU1 were used as the outgroups.

Fig. 4.

Alignment of the leader sequence and the homologous sequence upstream of the NS5a ORF. Sequence homology between the sequences is marked by asterisks. The putative TRS is highlighted in gray. The TRS suggested by St.-Jean et al. (39) are in boldface, and the corresponding deletions in BE03, BE04, HK04-01, and HK04-02 are indicated by the box. The initiation codons of NS5a are underlined.

Fig. 5.

Estimation of the time to the most recent common ancestor for HCoV-OC43 genotypes. The time-scaled phylogeny was summarized from all MCMC phylogenies of the S and N gene data set, which were analyzed under the relaxed-clock model with an uncorrelated exponential distribution in BEAST (version 1.6.1). A, genotype A; B, genotype B; C, genotype C. Genotype D strains are in boldface.