Characterization of a novel adenovirus isolated from a skunk$

Robert A. Kozak a, James G. Ackford a, Patrick Slaine a, Aimin Li b, Susy Carman c,Doug Campbell a, M. Katherine Welch c, Andrew M. Kropinski a, Éva Nagy a,n

a Department of Pathobiology, Ontario Veterinary College, University of Guelph, Guelph, Ontario, Canadab Public Health Ontario, Toronto, Ontario, Canadac Animal Health Laboratory, Laboratory Services Division, University of Guelph, Guelph, Ontario, Canada

a r t i c l e i n f o

Article history:Received 5 February 2015Returned to author for revisions17 May 2015Accepted 19 June 2015Available online 17 July 2015

Keywords:MastadenovirusSkunk adenovirusWhole genome sequenceLiver tropism

a b s t r a c t

Adenoviruses are a ubiquitous group of viruses that have been found in a wide range of hosts. A noveladenovirus from a skunk suffering from acute hepatitis was isolated and its DNA genome sequenced. Theanalysis revealed this virus to be a new member of the genus Mastadenovirus, with a genome of31,848 bp in length containing 30 genes predicted to encode proteins, and with a GþC content of 49.0%.Global genomic organization indicated SkAdV-1 was similar in organization to bat and canineadenoviruses, and phylogenetic comparison suggested these viruses shared a common ancestor.SkAdV-1 demonstrated an ability to replicate in several mammalian liver cell lines suggesting a potentialtropism for this virus.

& 2015 Elsevier Inc. All rights reserved.

Introduction

Members of the diverse family Adenoviridae are classified intofive genera, Mastadenovirus, Aviadenovirus, Atadenovirus, Siadeno-virus, and Ichtadenovirus (Harrach et al., 2011). Adenoviruses arenon-enveloped, icosahedral viruses that range in diameter from 70to 100 nm. Their double-stranded DNA genomes are linear, non-segmented, ranging in size from 26 to 48 kb, and typically containbetween 22 and 40 genes (Davison et al., 2003). To date, over 44species have been identified in the family Adenoviridae, and thereis a wide range of host species capable of sustaining these viruses(Hall et al., 2012; Marek et al., 2014; Park et al., 2012; Robinsonet al., 2013). Due to the relative speed and low cost of sequencingtechnologies, new isolates will continue to be added to each ofthese genera, and it is likely that the number of novel species willcontinue to grow.

Adenoviruses have been isolated from human and a wide range ofanimal species, including lizards (Wellehan et al., 2004), fowl (Griffinand Nagy, 2011; Ojkic and Nagy, 2000), dogs (Morrison et al., 1997),bats (Li et al., 2010), horses (Cavanagh et al., 2012), titi monkeys (genusCallicebus) (Chen et al., 2011), and sea lions (Cortes-Hinojosa et al.,2015). While numerous mammalian species harbor adenoviruses, nonovel viruses from this family have yet been reported in skunks(family Mephitidae), although these animals may serve as a spillover

host for canine adenoviruses (Alexander et al., 1972; Karstad et al.,1975). Skunks have traditionally been the focus of rabies virussurveillance, although they may also serve as hosts for parvoviruses(Allison et al., 2013), orthoreoviruses (Victoria et al., 2008), West Nilevirus (Anderson et al., 2001) and Powassan virus (El Khoury et al.,2013). Therefore it is evident that skunks may serve as an under-appreciated reservoir of viruses capable of infecting other mammals(Chen et al., 2011).

We have characterized a novel adenovirus isolated from a skunk,hereafter referred to as skunk adenovirus 1 (SkAdV-1), and sequencedits genome using Illumina MiSeq next-generation sequencing technol-ogy. Genomic analysis reveals similarities between SkAdV-1 and bothbat adenoviruses (BtAdV) and canine adenoviruses (CAdV). Phyloge-netic analysis suggests that SkAdV-1 may have evolved in skunks fromthe common ancestor of both bat and canine adenoviruses. This studyis the first in depth investigation into double-stranded DNA viruses,and more specifically adenoviruses, found in skunks. Based on ourresults we propose that this novel skunk adenovirus is a foundingmember of a new species, to be termed Skunk mastadenovirus A,within the genus Mastadenovirus.

Results

Gross pathology and virus isolation

A dead skunk (Mephitis mephitis) was originally presented to theCanadian Wildlife Health Cooperative, Department of Pathobiology,University of Guelph, for surveillance of wildlife rabies. However,the brain tissue was negative for rabies virus by indirect fluorescent

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/yviro

Virology

http://dx.doi.org/10.1016/j.virol.2015.06.0260042-6822/& 2015 Elsevier Inc. All rights reserved.

☆The GenBank accession number for the skunk adenovirus sequence isKP238322.

n Corresponding author. Tel.: þ1 519 824 4120x54783.E-mail address: enagy@uoguelph.ca (É. Nagy).

Virology 485 (2015) 16–24

antibody test. This juvenile female skunk was in good bodycondition, weighing 283 g. On post-mortem and histological eva-luation, there were lesions of acute hepatitis, and mild icterus wasobserved in a wide range of tissues. The lungs were edematous anda clear pericardial effusion was present. Spleen and mesentericlymph nodes were enlarged. The liver was bronze-colored, butotherwise unremarkable. Tissues were fixed in 10% formalin, sec-tioned in routine fashion and stained with hematoxylin and eosinfor histological examination. No significant changes were observedin brain, heart, kidney, adrenal glands, esophagus and trachea.Lymphoid tissue in spleen and lymph node were depleted. Therewas congestion, edema and interstitial inflammation in the lung. Inthe liver, there was diffuse hepatic necrosis, with large, amphophilicintranuclear inclusions present in affected areas. Collectively, thesechanges were interpreted to be typical of adenovirus infection.Frozen liver tissue was subsequently used for virus isolation andanalysis by transmission electron microscopy. Images displayedvirions that were icosahedral in shape (data not shown), with adiameter of approximately 80–100 nm, both of which are morpho-logical characteristics of adenoviruses (Pantelic et al., 2008). Virusisolation was attempted in different cell lines, and the mostsignificant cytopathic effect (CPE) was observed in Madin-Darbycanine kidney (MDCK) cells. The virus was plaque purified in MDCKcells and used for subsequent work. The CPE, manifested as cellrounding and detachment was seen as early as 24 h post-infection(h.p.i.) (Fig. 1A and B). Replication of the virus in MDCK cells wasevaluated through a single step growth curve (Fig. 1C). Productiveinfection in these cells yielded viral titers of 108 PFU/ml after threedays. Electron microscopic examination of the purified virus pre-paration showed typical adenovirus particles (Fig. 2).

Characteristics of SkAdV-1 genome

The genome of SkAdV-1 was sequenced using an IlluminaMiSeq, with paired-end libraries at 2�150 bp cycle run. Any reads

with a Q score below 30 were discarded, and 2857 times coverageof the complete viral genome was achieved. The genome length ofSkAdV-1 is 31,848 bp, with left and right inverted terminal repeatseach being 181 bp in length. The virus has a base composition of26.2% A, 24.8% C, 24.2% G, 24.8% T, and the GþC content was 49.0%.These and related data are presented in Table 1 and in comparisonto canine and bat adenoviruses.

A large number of potential genes were identified by in silicoanalysis using Kodon and comparative genomics with other sequencedadenoviral genomes. An open reading frame (ORF) was taken as apotential coding sequence if it contained more than 75 nucleotidesafter an ATG as possible start codon. The initial list was reduced to 30genes based on significant homology to known adenovirus genes,11 ofthese genes were on the complementary DNA strand. The overall

Fig. 1. Virus replication and cytopathic effect (CPE) in MDCK cells infected with SkAdV-1. Cells were infected with a multiplicity of infection (MOI) of 5 (A) or were leftuninfected (B), and were evaluated for CPE at 24 h post-infection (h.p.i.) and imaged using brightfield microscopy. (C) Cells were infected with SkAdV-1 at an MOI of 5, and atvarious h.p.i. the titers were determined in MDCK cells in order to establish a one-step growth curve. All experiments were performed in duplicate, and error bars representthe standard deviation.

Fig. 2. Electron micrograph of purified SkAdV-1 particles after negative staining.The bar represents 20 nm.

R.A. Kozak et al. / Virology 485 (2015) 16–24 17

number of predicted genes and their arrangement were similar toother adenoviruses, notably BtAdVs and CAdVs (Li et al., 2010;Morrison et al., 1997). Furthermore, SkAdV-1 showed significantnucleotide sequence similarity to BtAdV-2 (species Bat mastadenovirusB) and CAdV-1 (species Canine mastadenovirus A) with 64% and 64.2%pairwise identity, respectively. Global genomic analysis revealed theregion of greatest difference between SkAdV-1 and canine and batadenoviruses was at the left and right end regions. The globalorganization of the SkAdV-1 genome is illustrated in Fig. 3. Genespredicted to encode proteins included a number of core genescommon to adenoviruses including: DNA polymerase, pTP, 52K,penton base, hexon, protease and DNA binding protein. No evidenceof gene duplication was found. Additionally, we did not detect anyunique ORFs at either end of the SkAdV-1 genome, a region tradi-tionally associated with variation in adenovirus genomic content(Davison et al., 2003). Collectively, analysis of the genes of SkAdV-1indicated that themajority of themwere similar to those of canine and

bat adenoviruses, although the small T-antigen showed greatestsimilarity to one found in equine adenovirus 1 (EAdV-1). The completelist of predicted genes, including homologs with the greatest percentidentity is summarized in Table 2. Interestingly, similar to batadenoviruses, no gene predicted to code for VA RNA was detected inSkAdV-1 (Li et al., 2010).

Phylogenetic analysis of SkAdV-1

Phylogenetic analysis of sequences of four genes of the SkAdV-1 genome indicated that SkAdV-1 should be classified within thegenus Mastadenovirus. SkAdV-1 formed a separate branch, distinctfrom CAdVs and BtAdVs, although they likely share a commonancestor. Specifically, the amino acid sequences of the DNApolymerase, penton base, hexon and fiber were examined. Mole-cular typing and classification of adenoviruses are often done byanalysis of the hexon amino acid sequence (Lu and Erdman, 2006).The hexon of SkAdV-1 was similar in length to those of CAdVs andBtAdVs, and was shorter than the HAdV sequences examined. Theassessment of the DNA polymerase sequence found highestidentity (69%) to CAdV-2; although SkAdV-1 was arranged in itsown branch, separate from both the canine and bat adenoviruses(Fig. 4A). Phylogeny indicated the hexon of SkAdV-1 to be mostsimilar to those of CAdVs and BtAdVs, with highest identity (86%)to that of CAdV-2 (Fig. 4B). We next examined the predictedsequences of the fiber and penton base proteins. These proteinsinteract with cellular receptors, and act in concert to facilitate viralentry into the host cell (Russell, 2009). Both predicted proteinsshowed divergence from other adenoviruses, with highest identity(43% and 82%, respectively) to CAdV-1 (Fig. 4C and D). This

Table 1Characteristics of SkAdV-1 genome and comparison to canine and batmastadenoviruses.

Designation Genome size (bp) Percent GC CG dinucleotides % identitya

SkAdV-1 31,848 49.0 984 100CAdV-1 30,536 47.0 989 62.6CAdV-2 31,323 50.3 1064 64.2BtAdV-3 31,681 56.8 1895 63.5BtAdV-2 31,161 53.5 1432 64

a Percent nucleotide identity compared to SkAdV-1.

Fig. 3. Organization and global comparison of SkAdV-1 genome. The genome is represented by a thick black line, and predicted genes and ORFs and their directions aredenoted in purple arrows. The predicted organization of the SkAdV-1 indicates 30 predicted genes distributed over both strands. Below the genome map, global pairwisecomparison of homology to canine and bat adenoviruses is shown. Analysis was performed using mVISTA and regions of greater than 75% homology are colored red.

R.A. Kozak et al. / Virology 485 (2015) 16–2418

phylogenetic analysis reinforced the hypothesis that SkAdV-1 is amember of a novel mastadenovirus species. Additionally, weperformed multiple alignments of the predicted fiber and pentonbase proteins using ClustalX2 (Larkin et al., 2007). Comparison ofthe fiber sequences indicated that the SkAdV-1 fiber was approxi-mately 50 amino acids longer than that of BtAdV-3 or CAdV-2.However, there was significant similarity between the knobregions of SkAdV-1 fiber and those of CAdV-1, CAdV-2 andBtAdV-3 (Fig. 5B). Interestingly, examination of the penton basesequence revealed that it did not contain the RGD motif (Fig. 5A),which is a known cell adhesion motif, suggesting that SkAdV-1may enter cells through an RGD independent mechanism, similarto that observed with CAdV-2 (Soudais et al., 2000).

Virus replication in bat and canine cell lines

Our sequence analysis indicated a high probability that SkAdV-1 evolved from the common ancestor of BtAdVs. To investigatepotential cell tropisms, both bat (Tb-1 Lu) and canine (MDCK) celllines were infected with SkAdV-1 and virus titers were determinedat 72 h.p.i. (Fig. 6). Titers were three logs higher in MDCK cells(p¼0.01), and there was greater CPE, compared with Tb-1 Lu cells.Recent work by Li et al. (2010) demonstrated that while Tb-1 Lucells were not susceptible to BtAdV-3, the virus could infect MDCKcells (Li et al., 2010). This similar phenotype was noted withSkAdV-1, although when the number of CG nucleotides wasanalyzed using EMBOSS 6.3.1 geecee (⟨http://mobyle.pasteur.fr⟩),SkAdV-1 had fewer GC dinucleotides than any BtAdV species(Table 1). It has been noted that reduction in the number of CpGdinucleotides is often linked to viral adaptation and stabilitywithin a particular host species, suggesting SkAdV-1 may beadapting to the skunk host (Karlin et al., 1994; Upadhyay et al.,2013).

Virus replication in mammalian liver cell lines

The fact that the virus was isolated from an acute case ofhepatitis in a skunk, suggested that SkAdV-1 would replicate inother mammalian liver cells. Moreover, the liver tropism of somemammalian AdVs is well established (Einfeld et al., 2001;Nakamura et al., 2003). Therefore, we examined CPE and virustiters, at 72 h.p.i., in three mammalian (Huh-7, Huh-7.5 andHepG2) and one non-mammalian (avian CH-SAH) liver cell linesinfected with SkAdV-1. The cells were infected at a multiplicity ofinfection (MOI) of 3 and extensive CPE was seen in all cells exceptfor CH-SAH (Fig. 7). Moreover, the virus titers were 1�107 PFU/mlin Huh-7 cells, 1�108 PFU/ml in Huh-7.5 cells and 2�107 PFU/mlin HepG2 cells. By contrast, in the chicken derived CH-SAH cellsthe virus titer was 9�104 PFU/ml.

Discussion

Massive parallel sequencing technology has expanded ourknowledge of the viruses present in both animals and humans(Ge et al., 2012; Lysholm et al., 2012; Ng et al., 2012; Wevers et al.,2010). Here we have isolated a novel adenovirus from a skunkexhibiting acute hepatitis, and sequenced and characterized itsgenome. Our investigation showed that the SkAdV-1 genome issimilar in size and arrangement to the genomes of four types ofthree species within the genus, notably bat and canine adeno-viruses (Kohl et al., 2012; Li et al., 2010; Morrison et al., 1997).However, phylogenetic placement of this virus indicates thatSkAdV-1 is a member of its own distinct species though withinthe genus Mastadenovirus. We failed to find any unique ORFs, norany gene duplications. The analyses of the DNA polymerase,penton base, hexon, and fiber predicted amino acid sequencessuggest that SkAdV-1 and both CAdVs and BtAdVs may have

Table 2Predicted gene products of SkAdV-1.

Gene Location (bp) Strand Length (aa) Description Closest viral homolog (GenBank accession no.) Identity (%)

E1A 28.6K 462–1305 þ 260 Early E1A protein BtAdV-2 (YP_004782095.1) 53E1B 20.8K 1413–1961 þ 182 E1B protein, small T-antigen EAdV-1 (AEP16406.1) 54E1B 50.5K 1775–3136 þ 453 E1B protein, large T-antigen BtAdV-2 (YP_004782097.1) 48IX 3213–3512 þ 99 protein XI/hexon associated CAdV-2 (AP_000611.1) 63IVa2 3531–5149 – 449 IVa2 multifunctional protein CAdV-2 (AP_000612.1) 70pol 4622–12599 – 1143 DNA polymerase CAdV-2 (AP_000613.1) 69pTP 7900–12599 – 610 precursor terminal protein BtAdV-3 (YP_005271183.1) 7952K 9775–10965 þ 396 late encapsidation protein BtAdV-2 (YP_004782102.1) 69pIIIa 10832–12577 þ 581 IIIa protein precursor CAdV-1 (NP_044192.1) 79III 12625–14061 þ 478 penton base protein CAdV-1 (NP_044193.1) 82pVII 14091–14525 þ 144 core protein BtAdV-2 (YP_004782105.1) 72V 14620–15924 þ 434 protein V, minor core protein CAdV-2 (AP_000619.1) 62pX 15953–16165 þ 70 Mu peptide precursor BtAdV-2 (YP_004782107.1) 83pVI 16235–17056 þ 273 minor capsid protein precursor CAdV-2 (NP_044197.1) 66hexon 17132–19855 þ 907 capsid protein CAdV-2 (ACF19805.1) 86protease 19868–20488 þ 206 23K endopeptidase BtAdV-2 (YP_004782110.1) 84DBP 20533–21846 – 437 DNA binding protein BtAdV-3 (YP_005271193.1) 60100K 21859–23952 þ 697 100 kDa hexon assembly protein CAdV-2 (AP_000625.1) 7322K 23804–24310 þ 168 22 kDa encapsidation protein BtAdV-3 (YP_005271196.1) 56pVIII 24452–25120 þ 222 hexon associated protein BtAdV-3 (YP_005271197.1) 80E3 12.5K 25107–25454 þ 115 E3 12.5K control protein CAdV-1 (NP_044204.1) 41E3 ORFA 25521–26726 þ 401 E3 ORFA BtAdV-3 (YP_005271198.1) 33U exon 26743–26910 – 55 U exon BtAdV-3 (YP_005271199.1) 74fiber 26909–28729 þ 606 fiber protein CAdV-1 (ACT22747.1) 43E4 ORF6/7 28759–29763 – 89 E4 ORF6/7 control protein BtAdV-3 (YP_005271201.1) 48E4 34K 28990–29763 – 257 E4 34 kDa early protein CAdV-2 (AP_000634.1) 52E4 ORFD 29771–30160 – 129 E4 ORFD EAdV-1 (AEP16435.1) 30E4 ORFC 30160–30534 – 124 E4 ORFC BtAdV-3 (YP_005271204.1) 31E4 ORFB 30567–30926 – 119 E4 ORFB BtAdV-3 (YP_005271205.1) 61E4 ORFA 31008–31403 – 131 E4 ORFA CAdV-2 (AP_000638.1) 45

BtAdV-2¼bat adenovirus 2; EAdV-1¼equine adenovirus 1; CAdV-2¼canine adenovirus 2; BtAdV-3¼bat adenovirus 3; CAdV-1¼canine adenovirus 1.

R.A. Kozak et al. / Virology 485 (2015) 16–24 19

evolved from a common ancestor. It is also possible that a batadenovirus may have made a species jump into a skunk, andevolved within that host. This theory is supported by the observa-tion that SkAdV-1 showed reduced titers in the Tb-1 Lu fruit batcell line. However, unlike BtAdV-3, SkAdV-1 did not replicate wellin Vero cells (data not shown), potentially indicating further viraladaptation. Moreover, the report of adenovirus transmissionbetween bats, coupled with the detection of high virus titers inthe intestines of infected bats suggests a gastrointestinal diseasethat could also spread to other animals (Sonntag et al., 2009).

Studies examining the epidemiology of CAdV-2 have demon-strated a wide distribution across a number of mammals, andpostulate that this virus may have evolved from bat adenovirus(Buonavoglia and Martella, 2007; Kohl et al., 2012). Notably, therehave been reports of disease similar to infectious canine hepatitis,suggesting SkAdV-1 infections may be widespread but underreported (Alexander et al., 1972; Karstad et al., 1975). However,these studies failed to determine if the viral infection was in factdue to SkAdV-1 or due to CAdV-1, the etiological agent of thesimilar disease in dogs. Therefore, it remains to be determined

Fig. 4. Phylogenetic comparison of SkAdV-1 genes. Phylogenetic trees were based on amino acid sequences of the (A) DNA polymerase, (B) hexon, (C) fiber, and (D) pentonbase. Sequences were aligned using ClustalW (BLOSUM cost matrix), and phylogenic analysis was in Geneious Version 8.05 using the PhyML function with the unweightedpair group method with arithmetic mean (UPGMA) under default settings (Bootstrap resampling method, 1000 replicates).

R.A. Kozak et al. / Virology 485 (2015) 16–2420

whether skunks act as a reservoir, an amplifier host, a spilloverevent for SkAdV-1, or if in fact the virus evolved from a pre-decessor in this animal. It is possible that this virus has a broaderhost range, similar to other adenoviruses, and further investigationis necessary (Buonavoglia and Martella, 2007; Chen et al., 2011; Yuet al., 2013). Furthermore, it must be noted that although SkAdV-1was isolated from a diseased animal, Koch's postulates have notbeen fulfilled to link this virus to the documented pathology.Therefore, additional work with animal models will be required toformally demonstrate it.

The pathology documented in this case is in agreement withthat associated with adenoviruses (Morrison et al., 1997). Hepatitisis a known sequela of adenoviral infection, as evidenced by arecent report showing TMAdV causes hepatitis in new worldmonkeys (Chen et al., 2011). Furthermore, this new skunk adeno-virus is capable of infecting and causing CPE in a number ofmammalian liver cell lines, specifically Huh-7, Huh-7.5 and HepG2,suggesting SkAdV-1 may display a liver tropism commonly foundamong other adenoviruses (Einfeld et al., 2001; Nakamura et al.,2003). The higher titers obtained in Huh-7, Huh-7.5 and HepG2cell lines compared to CH-SAH could suggest the virus replicatesor is internalized more efficiently in mammalian liver cells. It hasbeen reported that CAR and integrin are highly expressed in livercells (Fechner et al., 1999), however expression levels could besignificantly lower in avian cells, thereby reducing viral uptake.Nevertheless, it is also possible that virus internalization or the

later steps of viral replication are inferior in avian cells, and furtherstudies are required to elucidate on these. The absence of otherreceptors, presence of mitigating factors (eg. Homologs of mouseCAR-like soluble protein) (Kawabata et al., 2007), or differences inreceptor structure must also be considered. Adenovirus attach-ment is facilitated by fiber protein interacting with CAR, followedby an interaction of the RGD motif on the penton base with αVβ3or αVβ5 integrins to allow internalization (Bergelson et al., 1997;Zhang and Bergelson, 2005). Assessment of the SkAdV-1 pentonbase sequence did not reveal an RGD, RGE or other known integrinbinding motif, suggesting that viral entry into cells may bethrough an unknown pathway (Belin and Boulanger, 1993;Bergelson et al., 1997; Soudais et al., 2000). CAdV-2 enters thecell through a CAR-independent pathway, and as this virus likelyevolved from the same common ancestor as SkAdV-1, it is likelythey share similar mechanisms of infection (Soudais et al., 2000).Furthermore, other mammalian adenoviruses have been isolatedthat lack this motif, a recent example is a novel adenovirus inspecies Human mastadenovirus D, which was found to contain YGDon the penton base (Robinson et al., 2013).

The varied levels of virus replication may also be due todifferences in host immune factors and the ability of SkAdV-1 tomodulate the innate immune response. The E3 region is involvedin suppression of host cytokines, chemokines, MHC-I, and apop-tosis (Friedman and Horwitz, 2002; Lesokhin et al., 2002; Woldand Gooding, 1991). This region is dispensable for replication, and

Fig. 5. Multiple alignments of the penton base and fiber knob region sequences of selected mammalian adenoviruses. (A) The sequence alignment of the penton base fromCAdV-1, CAdV-2, BtAdV-3, SkAdV-1, and HAdV-3 was performed using ClustalX2. The RGD motif in HAdV-3 is highlighted (\widehat\widehat\widehat), and absent from theremaining adenovirus sequences. (B) Multiple alignments of the knob region of the fiber sequences of selected mammalian adenoviruses. The sequence alignments wereperformed using ClustalX2.

R.A. Kozak et al. / Virology 485 (2015) 16–24 21

variations in it may alter viral pathogenesis (Marinheiro et al.,2011). Interestingly our elucidation of the SkAdV-1 genome pre-dicted the presence of the E3 12.5K followed by E3 ORFA, which is

a similar organization to canine adenoviruses (Morrison et al.,1997).

Conclusions

Currently, there is a paucity of studies on the virome of skunks,and the main focus of many skunk studies is on rabies. Thecharacterization of a novel skunk adenovirus not only broadens ourunderstanding of the evolution of adenoviruses, but also suggests thatthese animals may be an underappreciated reservoir of viruses. Thedata here present a strong case for proposing SkAdV-1 as the firstrepresentative of a novel species in genus Mastadenovirus, as it meetsmultiple criteria laid out by the International Committee on theTaxonomy of Viruses; importantly the phylogenetic distance of theDNA polymerase is more than 15%. Further studies are required notonly to characterize the epidemiology of this adenovirus in the skunkpopulation, but to also determine its role in causing disease, and itshost range.

Materials and methods

Virus isolation and propagation

To isolate virus, tissue homogenate of a skunk liver was filteredthrough a 0.45 mm syringe top filter (BD Biosciences, Mississauga,Ontario) before being added to MDCK, Vero, A-72, Crandall-Reesfeline kidney and mink lung epithelial cell cultures. The cell lineswere evaluated daily for CPE, and based on the extent of CPE theMDCK cell line was selected for further work. Plaque purification,virus propagation and titrations were performed in MDCK cells asdescribed previously (Griffin and Nagy, 2011; Ojkic and Nagy,2000).

Fig. 6. Viral titers and CPE in Tb-1 Lu and MDCK cells. Cells were infected at an MOIof 3, and (A) the virus titers were determined at 72 h.p.i. by titration in MDCK cells.Experiments were performed in triplicate, and error bars represent the standarderror of the mean. Samples were compared using Students t-test with significantdifferences indicated by * (pr0.05). (B) Less CPE was seen in Tb-1 Lu cellscompared to MDCK cells at this time point. Experiments were performed intriplicate and one representative sample is shown.

Fig. 7. SkAdV-1 replication in mammalian and non-mammalian liver cell lines. Mammalian (Huh-7, Huh-7.5 and HepG2) and avian (CH-SAH) liver cell lines were infected atan MOI of 3 and CPE was evaluated at 72 h.p.i. by brightfield microscopy. Experiments were performed in triplicate, and one representative example is shown.

R.A. Kozak et al. / Virology 485 (2015) 16–2422

DNA extraction and sequencing

SkAdV-1 was concentrated and the genomic DNAwas extractedas described previously (Ojkic and Nagy, 2001). Concentration andquality of the genomic DNA were determined using Qubit dsDNAAssay kit (Life Technology Inc., Oakville, Canada). Adenoviralgenomic DNA was acoustically fragmented using the CovarisM220 ultrasonicator and viral libraries were prepared usingIllumina Truseq DNA Library Preparation kit (Illumina, California,USA). Indexed libraries were evaluated using Agilent High Sensi-tivity DNA chip on an Agilent 2100 Bioanalyser system (AgilentTechnologies Canada Ltd, Mississauga, Canada) according to man-ufacturer's specifications.

Electron microscopy

Dilutions of the concentrated and purified virus sample wereadhered to a copper grid for 10 s, stained with 1% uranyl acetatefor 7 s, then washed once with double distilled H2O, and allowedto air dry. Grids were imaged using the FEI Tecnai G2 F20 operatedat 120 kV. The transmission electron microscopy was performed atthe University of Guelph.

Cell lines and virus replication

The mammalian cell lines were maintained in Dulbecco'sModified Eagle Medium (DMEM) (Thermo Fischer Scientific,Whitby, Ontario) supplemented with 10% fetal bovine serum(FBS) (Sigma, Oakville, Ontario), 2 mM L-glutamine, and penicillin(100 IU/ml)/streptomycin (100 ug/ml) (Sigma, Oakville, Ontario).The cell lines used include MDCK, HepG2, Huh-7, Huh-7.5, andTb-1 Lu. The chicken hepatoma cells (CH-SAH) were maintained inDMEM/F12 (Sigma) supplemented with 10% FBS (Sigma, Oakville,Ontario), 2 mM L-glutamine and penicillin (100 IU/ml)/streptomy-cin (100 mg/ml) (Sigma, Oakville, Ontario) (Griffin and Nagy, 2011).To assess viral replication cell monolayers were grown to 80%confluency prior to infection, and were infected as describedpreviously (Ojkic and Nagy, 2001).

Genomic and phylogenetic analysis

The primary deep sequence data, representing 2857-fold cover-age of the viral genome, was de novo assembled using SeqManNGen12 (DNASTAR, Madison, WI). The ORFs were identified inKodon (Applied Maths, Austin, TX) with respect to other adeno-viruses, with homologs identified on the basis of BLASTP analyses.Splice sites were calculated using alternative splice site predictor(ASSP) (Wang and Silverman, 2006), and cross-referenced with theannotated genomes of BtAdV-2 and BtAdV-3 (Accession no.JN252129, GU226970). Amino acid sequences for the DNA poly-merase, penton base, hexon and fiber were downloaded fromGenBank, and aligned using ClustalW (BLOSUM cost matrix).Phylogenetic analysis was performed in Geneious Version 8.05(Biomatters, Auckland, New Zealand) using the PhyML functionwith the unweighted pair group method with arithmetic mean(UPGMA) under default settings (Bootstrap resampling methodwith 1000 replicates). Multiple alignments of the penton base andfiber sequences of selected mammalian adenoviruses were per-formed using ClustalX2 (Larkin et al., 2007). The number of CGdinucleotides were enumerated using EMBOSS 6.3.1 geecee(⟨http://genome.lbl.gov/vista/index.shtml⟩). Evaluation of globalpairwise identity of the SkAdV-1 whole genome to other adeno-viral sequences was by using mVISTA (⟨http://genome.lbl.gov/vista/mvista⟩) (Brudno et al., 2003).

Nucleotide accession numbers

The GenBank accession number of the skunk adenovirus 1(SkAdV-1) genome sequence reported in this paper is KP238322.

In order to perform phylogenetic comparisons, the followingadenovirus genomes were retrieved from GenBank: bovine ade-novirus 3 (BAdV-3, AF030154), bovine adenovirus 1 (BAdV-1,AC_000191.1), bovine adenovirus 2 (BAdV-2 AF252854), bovineadenovirus 10 (BAdV-10, AF027599), canine adenovirus 1 (CAdV-1,Y07760), canine adenovirus 2 (CAdV-2, U77082), equine adeno-virus 1 (EAdV-1, L79955), equine adenovirus 2 (EAdV-2, L80007),human adenovirus 12 (HAdV-12, X73487), human adenovirus 1(HAdV-1, AF534906), human adenovirus 8 (HAdV-8, AB448767),human adenovirus 4 (HAdV-4, AY487947), human adenovirus 3(HAdV-3, NC_011203), human adenovirus 12 (HAdV-12,AB330093), simian adenovirus 22 (SAdV-22, AY530876), humanadenovirus 40 (HAdV-40, L19443), human adenovirus 52 (HAdV-52, DQ923122), murine adenovirus 1 (MAdV-1, NC_000942),murine adenovirus 2 (MAdV-2, NC_014899), murine adenovirus3 (MAdV-3, EU835513), porcine adenovirus 5 (PAdV-5, AF289262),tree shrew adenovirus 1 (TSAdV-1, AF258784), titi monkey ade-novirus (TMAdV, NC_020487), bat adenovirus 2 (BtAdV-2,FJ983127), bat adenovirus 3 (BtAdV-3, GU226970).

Statistical analyses

Figures were generated using GraphPad Prism 6.0 software(GraphPad Software, Inc., La Jolla, CA). Statistical analyses werealso performed using this software. Differences between meanswere evaluated using the Student's t-test and were deemedsignificant at p-values r0.05.

Acknowledgments

The authors would like to acknowledge Susan Emery, Betty-AnneMcBey and David Leishman for their excellent technical assistance.This research was supported by the Natural Sciences and EngineeringResearch Council of Canada (NSERC), RGPIN 41625-2013.

The authors have no conflicts of interest to declare.

References

Alexander, A.D., Flyger, V., Herman, Y.F., McConnell, S.J., Rothstein, N., Yager, R.H.,1972. Survey of wild mammals in a Chesapeake Bay area for selected zoonoses.J. Wildl. Dis. 8, 119–126.

Allison, A.B., Kohler, D.J., Fox, K.A., Brown, J.D., Gerhold, R.W., Shearn-Bochsler, V.I.,Dubovi, E.J., Parrish, C.R., Holmes, E.C., 2013. Frequent cross-species transmis-sion of parvoviruses among diverse carnivore hosts. J. Virol. 87, 2342–2347.

Anderson, J.F., Vossbrinck, C.R., Andreadis, T.G., Iton, A., Beckwith 3rd, W.H., Mayo,D.R., 2001. Characterization of West Nile virus from five species of mosquitoes,nine species of birds, and one mammal. Ann. N.Y. Acad. Sci. 951, 328–331.

Belin, M.T., Boulanger, P., 1993. Involvement of cellular adhesion sequences in theattachment of adenovirus to the HeLa cell surface. J. Gen. Virol. 74, 1485–1497.

Bergelson, J.M., Cunningham, J.A., Droguett, G., Kurt-Jones, E.A., Krithivas, A., Hong,J.S., Horwitz, M.S., Crowell, R.L., Finberg, R.W., 1997. Isolation of a commonreceptor for Coxsackie B viruses and adenoviruses 2 and 5. Science 275,1320–1323.

Brudno, M., Do, C.B., Cooper, G.M., Kim, M.F., Davydov, E., Program, N.C.S., Green, E.D.,Sidow, A., Batzoglou, S., 2003. LAGAN and Multi-LAGAN: efficient tools for large-scale multiple alignment of genomic DNA. Genome Res. 13, 721–731.

Buonavoglia, C., Martella, V., 2007. Canine respiratory viruses. Vet. Res. 38,355–373.

Cavanagh, H.M., Mahony, T.J., Vanniasinkam, T., 2012. Genetic characterization ofequine adenovirus type 1. Vet. Microbiol. 155, 33–37.

Chen, E.C., Yagi, S., Kelly, K.R., Mendoza, S.P., Tarara, R.P., Canfield, D.R., Maninger, N.,Rosenthal, A., Spinner, A., Bales, K.L., Schnurr, D.P., Lerche, N.W., Chiu, C.Y., 2011.Cross-species transmission of a novel adenovirus associated with a fulminantpneumonia outbreak in a new world monkey colony. PLoS Pathog. 7, e1002155.

Cortes-Hinojosa, G., Gulland, F.M.D., Goldstein, T., Venn-Watson, S., Rivera, R.,Waltzek, T.B., Salemi, M., Wellehan Jr., J.F.X., 2015. Phylogenomic characteriza-tion of California sea lion adenovirus-1. Infect. Genet. Evol. 31, 270–276.

R.A. Kozak et al. / Virology 485 (2015) 16–24 23

Davison, A.J., Benko, M., Harrach, B., 2003. Genetic content and evolution ofadenoviruses. J. Gen. Virol. 84, 2895–2908.

Einfeld, D.A., Schroeder, R., Roelvink, P.W., Lizonova, A., King, C.R., Kovesdi, I.,Wickham, T.J., 2001. Reducing the native tropism of adenovirus vectors requiresremoval of both CAR and integrin interactions. J. Virol. 75, 11284–11291.

El Khoury, M.Y., Camargo, J.F., White, J.L., Backenson, B.P., Dupuis 2nd, A.P., Escuyer,K.L., Kramer St, L., George, K., Chatterjee, D., Prusinski, M., Wormser, G.P., Wong,S.J., 2013. Potential role of deer tick virus in Powassan encephalitis cases inLyme disease-endemic areas of New York, USA. Emerg. Infect. Dis. 19,1926–1933.

Fechner, H., Haack, A., Wang, H., Wang, X., Eizema, K., Pauschinger, M., Schoemaker,R., Veghel, R., Houtsmuller, A., Schultheiss, H.P., Lamers, J., Poller, W., 1999.Expression of coxsackie adenovirus receptor and alphav-integrin does notcorrelate with adenovector targeting in vivo indicating anatomical vectorbarriers. Gene Ther. 6, 1520–1535.

Friedman, J.M., Horwitz, M.S., 2002. Inhibition of tumor necrosis factor alpha-induced NF-kappa B activation by the adenovirus E3-10.4/14.5K complex.J. Virol. 76, 5515–5521.

Ge, X., Li, Y., Yang, X., Zhang, H., Zhou, P., Zhang, Y., Shi, Z., 2012. Metagenomicanalysis of viruses from bat fecal samples reveals many novel viruses ininsectivorous bats in China. J. Virol. 86, 4620–4630.

Griffin, B.D., Nagy, E., 2011. Coding potential and transcript analysis of fowladenovirus 4: insight into upstream ORFs as common sequence features inadenoviral transcripts. J. Gen. Virol. 92, 1260–1272.

Hall, N.H., Archer, L.L., Childress, A.L., Wellehan Jr., J.F., 2012. Identification of anovel adenovirus in a cotton-top tamarin (Saguinus oedipus). J. Vet. Diagn.Investig. 24, 359–363.

Harrach, B., Benkő, M., Both, G.W., Brown, M., Davison, A.J., Echavarría, M., Hess, M.,Jones, M.S., Kajon, A., 2011. Virus Taxonomy: Classification and Nomenclature ofViruses: Ninth Report of the International Committee on Taxonomy of Viruses.In: King, A.M.Q., Adams, M.J., Carstens, E.B., Lefkowitz, E.J. (Eds.), Elsevier, SanDiego, pp. 125–141.

Karlin, S., Doerfler, W., Cardon, L.R., 1994. Why is CpG suppressed in the genomes ofvirtually all small eukaryotic viruses but not in those of large eukaryoticviruses? J. Virol. 68, 2889–2897.

Karstad, L., Ramsden, R., Berry, T.J., Binn, L.N., 1975. Hepatitis in skunks caused bythe virus of infectious canine hepatitis. J. Wildl. Dis. 11, 494–496.

Kawabata, K., Tashiro, K., Sakurai, F., Osada, N., Kusuda, J., Hayakawa, T., Yamanishi,K., Mizuguchi, H., 2007. Positive and negative regulation of adenovirus infec-tion by CAR-like soluble protein, CLSP. Gene Ther. 14, 1199–1207.

Kohl, C., Vidovszky, M.Z., Muhldorfer, K., Dabrowski, P.W., Radonic, A., Nitsche, A.,Wibbelt, G., Kurth, A., Harrach, B., 2012. Genome analysis of bat adenovirus 2:indications of interspecies transmission. J. Virol. 86, 1888–1892.

Larkin, M.A., Blackshields, G., Brown, N.P., Chenna, R., McGettigan, P.A., McWilliam,H., Valentin, F., Wallace, I.M., Wilm, A., Lopez, R., Thompson, J.D., Gibson, T.J.,Higgins, D.G., 2007. Clustal W and Clustal X version 2.0. Bioinformatics 23,2947–2948.

Lesokhin, A.M., Delgado-Lopez, F., Horwitz, M.S., 2002. Inhibition of chemokineexpression by adenovirus early region three (E3) genes. J. Virol. 76, 8236–8243.

Li, Y., Ge, X., Zhang, H., Zhou, P., Zhu, Y., Zhang, Y., Yuan, J., Wang, L.F., Shi, Z., 2010.Host range, prevalence, and genetic diversity of adenoviruses in bats. J. Virol.84, 3889–3897.

Lu, X., Erdman, D.D., 2006. Molecular typing of human adenoviruses by PCR andsequencing of a partial region of the hexon gene. Arch. Virol. 151, 1587–1602.

Lysholm, F., Wetterbom, A., Lindau, C., Darban, H., Bjerkner, A., Fahlander, K.,Lindberg, A.M., Persson, B., Allander, T., Andersson, B., 2012. Characterization ofthe viral microbiome in patients with severe lower respiratory tract infections,using metagenomic sequencing. PloS ONE 7, e30875.

Marek, A., Ballmann, M.Z., Kosiol, C., Harrach, B., Schlotterer, C., Hess, M., 2014.Whole-genome sequences of two turkey adenovirus types reveal the existence

of two unknown lineages that merit the establishment of novel species withinthe genus Aviadenovirus. J. Gen. Virol. 95, 156–170.

Marinheiro, J.C., Dos Santos, T.G., Siqueira-Silva, J., Lu, X., Carvalho, D., da Camara, A.A.,Arruda, E., Arruda, K., Erdman, D.D., Harsi, C.M., 2011. A naturally occurringhuman adenovirus type 7 variant with a 1743 bp deletion in the E3 cassette.J. Gen. Virol. 92, 2399–2404.

Morrison, M.D., Onions, D.E., Nicolson, L., 1997. Complete DNA sequence of canineadenovirus type 1. J. Gen. Virol. 78, 873–878.

Nakamura, T., Sato, K., Hamada, H., 2003. Reduction of natural adenovirus tropismto the liver by both ablation of fiber-coxsackievirus and adenovirus receptorinteraction and use of replaceable short fiber. J. Virol. 77, 2512–2521.

Ng, T.F., Marine, R., Wang, C., Simmonds, P., Kapusinszky, B., Bodhidatta, L.,Oderinde, B.S., Wommack, K.E., Delwart, E., 2012. High variety of known andnew RNA and DNA viruses of diverse origins in untreated sewage. J. Virol. 86,12161–12175.

Ojkic, D., Nagy, E., 2000. The complete nucleotide sequence of fowl adenovirus type8. J. Gen. Virol. 81, 1833–1837.

Ojkic, D., Nagy, E., 2001. The long repeat region is dispensable for fowl adenovirusreplication in vitro. Virology 283, 197–206.

Pantelic, R.S., Lockett, L.J., Rothnagel, R., Hankamer, B., Both, G.W., 2008. Cryoelec-tron microscopy map of Atadenovirus reveals cross-genus structural differ-ences from human adenovirus. J. Virol. 82, 7346–7356.

Park, Y.M., Kim, J.H., Gu, S.H., Lee, S.Y., Lee, M.G., Kang, Y.K., Kang, S.H., Kim, H.J.,Song, J.W., 2012. Full genome analysis of a novel adenovirus from the SouthPolar skua (Catharacta maccormicki) in Antarctica. Virology 422, 144–150.

Robinson, C.M., Zhou, X., Rajaiya, J., Yousuf, M.A., Singh, G., DeSerres, J.J., Walsh, M.P.,Wong, S., Seto, D., Dyer, D.W., Chodosh, J., Jones, M.S., 2013. Predicting the nexteye pathogen: analysis of a novel adenovirus. mBio 4 (2) e00595-12.

Russell, W.C., 2009. Adenoviruses: update on structure and function. J. Gen. Virol.90, 1–20.

Sonntag, M., Muhldorfer, K., Speck, S., Wibbelt, G., Kurth, A., 2009. New adenovirusin bats, Germany. Emerg. Infect. Dis. 15, 2052–2055.

Soudais, C., Boutin, S., Hong, S.S., Chillon, M., Danos, O., Bergelson, J.M., Boulanger,P., Kremer, E.J., 2000. Canine adenovirus type 2 attachment and internalization:coxsackievirus-adenovirus receptor, alternative receptors, and an RGD-independent pathway. J. Virol. 74, 10639–10649.

Upadhyay, M., Samal, J., Kandpal, M., Vasaikar, S., Biswas, B., Gomes, J.,Vivekanandan, P., 2013. CpG dinucleotide frequencies reveal the role of hostmethylation capabilities in parvovirus evolution. J. Virol. 87, 13816–13824.

Victoria, J.G., Kapoor, A., Dupuis, K., Schnurr, D.P., Delwart, E.L., 2008. Rapididentification of known and new RNA viruses from animal tissues. PLoS Pathog.4, e1000163.

Wang, Y., Silverman, S.K., 2006. Experimental tests of two proofreading mechan-isms for 50-splice site selection. ACS Chem. Biol. 1, 316–324.

Wellehan, J.F., Johnson, A.J., Harrach, B., Benko, M., Pessier, A.P., Johnson, C.M.,Garner, M.M., Childress, A., Jacobson, E.R., 2004. Detection and analysis of sixlizard adenoviruses by consensus primer PCR provides further evidence of areptilian origin for the atadenoviruses. J. Virol. 78, 13366–13369.

Wevers, D., Leendertz, F.H., Scuda, N., Boesch, C., Robbins, M.M., Head, J., Ludwig, C.,Kuhn, J., Ehlers, B., 2010. A novel adenovirus of Western lowland gorillas(Gorilla gorilla gorilla). Virol. J. 7, 303.

Wold, W.S., Gooding, L.R., 1991. Region E3 of adenovirus: a cassette of genesinvolved in host immunosurveillance and virus-cell interactions. Virology 184,1–8.

Yu, G., Yagi, S., Carrion Jr., R., Chen, E.C., Liu, M., Brasky, K.M., Lanford, R.E., Kelly, K.R.,Bales, K.L., Schnurr, D.P., Canfield, D.R., Patterson, J.L., Chiu, C.Y., 2013. Experi-mental cross-species infection of common marmosets by titi monkey adeno-virus. PloS ONE 8, e68558.

Zhang, Y., Bergelson, J.M., 2005. Adenovirus receptors. J. Virol. 79, 12125–12131.

R.A. Kozak et al. / Virology 485 (2015) 16–2424