Pediatric Infections by Human mastadenovirus C Types 2, 89, and a Recombinant Type Detected in Japan between 2011 and 2018

Abstract

Between 2011 and 2018, 518 respiratory adenovirus infections were diagnosed in a pediatric clinic in Shizuoka, Japan. Detection and typing were performed by partial sequencing of both hexon- and fiber-coding regions which identified: adenovirus type 1 (Ad-1, n = 85), Ad-2 (n = 160), Ad-3 (n = 193), Ad-4 (n = 18), Ad-5 (n = 27), Ad-11 (n = 2), Ad-54 (n = 3), and Ad-56 (n = 1). Considering previous reports of the circulation of an endemic recombinant Ad-2, e.g., Ad-89, 100 samples typed as Ad-2 were randomly selected for further molecular typing by sequencing the penton base-coding region. Despite the high nucleotide sequence conservation in the penton base- coding region, 27 samples showed 98% identity to Ad-2. Furthermore, 14 samples showed 97.7% identity to Ad-2 and 99.8% identity to Ad-89, while the remaining 13 samples showed an average 98% pairwise identity to other Ad-C types and clustered with Ad-5. The samples typed as Ad-89 (n = 14) and as a recombinant Ad type (P5H2F2) (n = 13) represented 27% of cases originally diagnosed as Ad-2, and were detected sporadically. Therefore, two previously uncharacterized types in Japan, Ad-89 and a recombinant Ad-C, were shown to circulate in children. This study creates a precedent to evaluate the epidemiology and divergence among Ad-C types by comprehensively considering the type classification of adenoviruses.

Keywords: Human mastadenovirus C; adenovirus typing; molecular epidemiology; pediatric infections; recombination; respiratory infections.

Figure 1

Adenoviruses (Ads) detected from respiratory patients during 2012 to 2018. The vertical axis shows the year, and the horizontal axis shows the number of Ads detected. Data from 2011 were not shown because only three strains were detected in 2011, Ad-1 (n = 2) and Ad-3 (n = 1).

Figure 2

Phylogenetic trees of (A) the penton base RGD loop and (B) hexon loop 1. The posterior probability supporting the branching is shown at the tree nodes. Tip names are colored black for sequences obtained in the present study, blue for Ad-C reference sequences, and red for two samples used for whole-genome sequencing.

Figure 3

Phylogenetic trees for Ad-C types. The phylogenetic trees were inferred for (A) the complete genome sequences, and (B) penton base-, (C) hexon-, and (D) fiber-coding regions. The Bayesian posterior probability supporting the branching is shown adjacent to the nodes. Tip names of the recombinant genome sequences from the present study are colored in red.

Figure 4

Recombination analysis by sliding window in Simplot. Horizontal axes correspond to the relative genomic positions in K19-85-2012. (A) BootScan analysis shows the clustering of K19-85-2012 with other Ad-C types. The vertical axis shows the percentage of trees supporting the clustering. Series are colored according to the legend on top of the panel. (B) Graphical representation of the recombinant origin of the genomic segments colored as described above. (C) Genome annotation of K19-85-2012. (D) Sliding window analyses of the pairwise similarity per window of K19-85-2012 and the sequence described at the top of each chart. The vertical axis of each chart shows percentage similarity.

Figure 5

Phylogenetic trees considering the effects of detected recombination events. The Bayesian posterior probability supporting the branching is shown adjacent to the nodes. (A) Phylogenetic tree considering the section of the genome in K19-85-2012 detected as not recombinant with a length of 27,064 bp. Other phylogenetic trees correspond to sections of the genome for (B) event 1: 2527 bp, (C) event 2: 3685 bp, (D) event 3: 1527 bp, and (E) event 4: 1131 bp. See Table 2 and Figure 4B. Tip names of the recombinant genome sequences from the present study are colored in red.