First Evidence of Genotypes Ad3a16 and Ad3a18 in North America, Obtained by Genetic Analysis of Infectious Human Adenovirus from Wastewaters of Two Urban Communities in Canada

Abstract

A 1-year study found seven infectious human adenovirus serotypes, including Ad3, Ad31, Ad46, Ad27, Ad22, Ad51, and clinical clone PB3, in wastewaters of two major cities in Canada. Comparative genomic analysis revealed the existence of the reportedly highly contagious Ad3a16/18 genotypes. This is the first report of Ad3a16/18 genotypes in North America.

TEXT

Human adenoviruses (HAdVs) are a ubiquitous and diverse group of nonenveloped icosahedral viruses comprised of 53 serotypes that are further segregated into seven clades or subgroups (A to G) based on their hemagglutination properties and DNA homology (2). This group of viruses is responsible for a broad spectrum of illnesses affecting the central nervous system (11), the lungs (17), the gastrointestinal tract (8), and the eyes (8).

Adenovirus serotype 3 can cause large outbreaks (4, 7, 10, 12, 13, 14) resulting in severe acute respiratory diseases (1, 22). Results from the Seoul National University Children's Hospital performing viral surveillance for childhood lower respiratory tract infections (LRTIs) indicated that Ad3 and Ad7 serotypes accounted for more than half of adenoviral LRTIs, including a large nationwide epidemic of Ad7-associated pneumonia as well as annual outbreaks of LRTIs associated with Ad3 (3). Two newly identified genotypes, Ad3a16 and Ad3a18, were found to have replaced the previously predominating Ad3 types and became the major genome types during the Ad3 outbreak in Korea in 1998 to 1999.

In this study we recovered and characterized HAdVs from wastewaters collected over a 1-year period in two major cities in Ontario, Canada. By sequencing the most informative fragment of the hexon gene, which contains hypervariable region 7 (HVR-7), it was found that the homogenous population of Ad3 isolates from this study possess three unique amino acid substitutions in loop 2 which distinguish reportedly highly virulent Ad3a16 and Ad3a18 (3) from all other Ad3 genome types.

Thirty-six wastewater samples, including those from the primary, secondary, and final stages of treatment, were taken from two very large municipal wastewater treatment plants located in Ontario, Canada, from April 2008 to May 2009. The samples were delivered to the laboratory within a day of collection and immediately processed. The samples were concentrated using highly electropositive 0.2-μm-pore-size filters that combine nano alumina fibers on a microglass and cellulose scaffold (NanoCeram; Argonide, Sanford, FL). Viral particles were eluted by purging the filter unit with ∼400 ml of sterile 1.5% beef extract with 0.25% (wt/vol) glycine adjusted to pH 9.3. Virus particles were then precipitated from the eluted liquid using an equal volume of 16% polyethylene glycol 8000 (Sigma Chemical Co.) and incubated for 1 h on ice before being centrifuged at 15,000 × g for 30 min. After precipitation, the resulting pellets were resuspended in ∼10 ml of sterile RNase-free water and filtered using a 0.22-μm-pore-size filter. The recovery efficiency of the concentration method was 19.53% ± 2%, as determined on three separate occasions using human adenovirus serotype 5 and the 293T cell line. The final filtrate was stored at −80°C for cell culture infectivity assay.

Infectious HAdVs were isolated using established cell line human hepatocellular carcinoma (PLC/PRF/5). A seed culture of PLC/PRF/5 cells (obtained from Martha Brown, University of Toronto, Canada) at passage level 8 was used in this study between passages 10 to 26. The infectivity assay was executed using the total culturable method described elsewhere (5). The cells were reincubated at 37°C for 14 days, here referred to as first passage, under an atmosphere supplemented with 5% CO₂. To extend the time for virus replication, the cell cultures from the first passage were frozen and thawed three times and the supernatant was subjected to a second passage in corresponding monolayers and incubated for an additional 14 days. Cells were examined daily for cytopathic effects.

Viral genomic DNA was extracted from the second-passage cell cultures showing cytopathogenic symptoms as well as randomly chosen cell monolayers with no sign of cytopathology using the QIAamp DNA minikit (Qiagen, Mississauga, Canada). An ∼620-bp DNA fragment containing an intact hypervariable region 7 of the hexon gene was amplified by PCR. Further, this ∼620-bp amplicon was used for phylogenetic analysis, as it could detect all known human adenovirus serotypes through analysis of the nucleotide sequence (16). All PCR amplifications were performed using Phusion high-fidelity DNA polymerase (New England BioLabs, Pickering, Ontario, Canada). The PCR was performed as previously described by Sarantis et al. (16). Both strands of the amplicon were sequenced using an ABI BigDye Terminator v. 3.1 cycle sequencing kit (Applied Biosystems, Foster City, CA) and an ABI 3100 automated sequencer (PE Biosystems). Sequencing was done at the Ontario Genomics Innovation Centre, Ottawa Hospital Research Institute. Sequence editing and alignment were done using ClustalW (20). The consensus DNA sequences were compared with all sequences in the GenBank database using the PubMed National Center for Biotechnology Information BLAST program. The genetic relatedness of HAdV serotypes was inferred using the neighbor-joining method (15). The evolutionary distances were computed using the maximum composite likelihood method (19) and are expressed as the number of base substitutions per site. Phylogenetic analyses were executed in MEGA4 (18).

Of 360 subsamples of monolayers of the PLC/PRF/5 cell line that were inoculated and observed for cytopathology, we found 217 cytopathic effects (CPE). However, these samples were likely to contain numerous cytopathogenic viruses other than adenoviruses. When we performed integrated cell culture-PCR (ICC-PCR), it was observed that 45% (35 of 78) of the tested samples were positive for the presence of infectious HAdVs. No obvious cytotoxicity from the sample concentrates was observed in any of the inoculated cells.

Amplicons that indicated the presence of HAdVs in cell cultures were subtyped by sequencing the ∼620-bp DNA fragment of the hexon gene which contains an intact HVR-7 region. Thirty samples were successfully sequenced. The BLAST analysis suggested that all amplicons from the cell cultures belonged to HAdVs and further revealed the existence of seven different HAdV serotypes, including serotypes 3, 31, 46, 22, 51, and 27 and clinical isolate clone PB3. A 100% agreement was obtained with existing GenBank sequences for serotypes 46 and 3; an agreement of 99% was obtained for serotypes 31, 51, 22, and 46 and clinical clone PB3 (Table 1).

Table 1.

Percentages of hexon gene sequence homology of HAdV isolates from this study with existing GenBank sequences

Isolate ID	% nt homologya	Serotype	Clade
97-PLC	100	3	B
107-PLC	99	31	A
108-PLC	100	3	B
118-PLC	100	3	B
119-PLC	100	3	B
120-PLC	100	3	B
45-PLC	100	46	D
46-PLC	99	46	D
47-PLC	89	22	D
48-PLC	90	22	D
49-PLC	90	22	D
59-PLC	98	31	A
60-PLC	91	31	A
61-PLC	87	31	A
63-PLC	99	31	A
64-PLC	99	31	A
65-PLC	99	31	A
69-MA	91	27	D
70-PLC	91	27	D
71-PLC	91	27	D
72-PLC	91	27	D
73-PLC	99	Clone PB3
74-PLC	99	Clone PB3
85-PLC	99	51	D
86-PLC	99	51	D
87-PLC	99	51	D
89-PLC	99	22	D
90-PLC	99	22	D
92-PLC	97	22	D

^ant, nucleotide.

Each of the HAdV serotypes was clearly distinct and further grouped into one of three major clusters, A, B, and D, as shown by a phylogenetic tree (Fig. 1). Each of these three major clusters represents a clade, which accurately classified different HAdV serotypes based on their genome homology. Most HAdV serotypes were grouped into clade D, followed by clades B and A. It is noteworthy that 73-PLC and 74-PLC isolates grouped together with a bootstrap support of 100% with the reference clinical isolate clone PB3, forming an outlier of clade B (Fig. 1), and furthermore indicating the close genetic relatedness of these two HAdV isolates to clade B.

Fig. 1.

Phylogenetic analysis of the hexon nucleic acid sequences of infectious HAdV isolates from this study and a set of sequences downloaded from GenBank used as references for different serotypes. The reference sequences are indicated in bold and labeled with their GenBank accession number, strain name, and serotype designation. The phylogenetic tree was constructed using the neighbor-joining method. The bootstrap values were generated with 1,000 pseudoreplicates. Scale bar indicates number of substitutions per site.

The results of DNA sequencing of HVR-7 revealed a high degree of heterogeneity among the infectious HAdV population. Studies of adenoviruses frequently show several to be cocirculating (3, 6, 9, 12), though occasionally one can predominate (6, 9, 12); the period of turnover can be quite rapid (9). Although the main characteristic of infectious HAdVs in this study was a rapid serotype turnover, it was observed that the HAdV serotype 3 isolated during the first sampling period (May and June 2008) was identical in the two urban centers, indicating its presence throughout a wider population in Canada. Yeung et al. (21), performing a surveillance program in southern Ontario during the period September 2007 to June 2008, reported that Ad3 was the most commonly identified serotype in respiratory specimens throughout the study; it was found in 37% of the patients aged 4 years or less but 71% of those older than 4 years.

In performing BLAST analysis on our samples, it was found that the Ad3 isolates from Ontario shared 100% homology of hexon gene sequences with Ad3a16 and Ad3a18 genotypes isolated during the 1998 to 1999 Ad3 outbreak in Korea (3, 7) and the 2003 to 2005 outbreak in Japan (6). The hexon gene sequences of both Ad3a16 and Ad3a18 genotypes are identical to those of three unique amino acid substitutions in loop 2, which segregates them from all other Ad3 genotypes into a distinct lineage (3). Multiple alignment of all Ad3 isolates from this study with a reference Ad3a16/18 genotype, targeting the loop 2 region, showed three identical amino acid substitutions at positions 417, 429, and 439 (Fig. 2), further indicating the presence of Ad3a16/18 genotype in the sampled region of Canada. These three amino acid substitutions in loop 2 affect the predicted hydrophobicity in this region, and it is believed (3) that these mutations contributed to the emergence of the new genotypes, Ad3a16 and Ad3a18, and an Ad3-associated epidemic of childhood pneumonia. Choi and colleagues (3) found that Ad3a16 and Ad3a18 genotypes emerged in 1998 to 1999 and, due to an apparently greater fitness than other Ad3 genotypes, very soon replaced the previously predominating types during the Ad3 outbreak in Korea, which may explain the genetic homogeneity of Ad3 isolates found in Ontario. The same virus genotype was also dominant between 2003 and 2005 in Japan (6). However, there have been no reports in the English language of these novel genotypes in other parts of the world.

Fig. 2.

Multiple alignment of loop 2 region of hexon protein, including the Ad3 isolates obtained in this study with a reference sequence of Ad3a16 and/or Ad3a18 genotypes. Three amino acid substitutions, unique for Ad3a16 and Ad3a18 genotypes, are highlighted at 417, 429, and 439 amino acid positions of hexon protein.

In conclusion, this study reports the recovery of a range of adenovirus serotypes from wastewater samples, with the first appearance of the unique Ad3a16 and Ad3a18 in North America, providing additional evidence of the widespread distribution of these two recently emerged Ad3 genome types. Taking into account the reported high virulence of these genome types in South Korea and Japan and their apparent superior fitness over other Ad3 genome types, it may be important to monitor their distribution and associated disease from clinical and epidemiological points of view.

Nucleotide sequence accession numbers.

Nucleotide sequences were deposited in the GenBank database, and their accession numbers ranged from HM449007 to HM449018.