Detection and Analysis of Six Lizard Adenoviruses by Consensus Primer PCR Provides Further Evidence of a Reptilian Origin for the Atadenoviruses

Abstract

A consensus nested-PCR method was designed for investigation of the DNA polymerase gene of adenoviruses. Gene fragments were amplified and sequenced from six novel adenoviruses from seven lizard species, including four species from which adenoviruses had not previously been reported. Host species included Gila monster, leopard gecko, fat-tail gecko, blue-tongued skink, Tokay gecko, bearded dragon, and mountain chameleon. This is the first sequence information from lizard adenoviruses. Phylogenetic analysis indicated that these viruses belong to the genus Atadenovirus, supporting the reptilian origin of atadenoviruses. This PCR method may be useful for obtaining templates for initial sequencing of novel adenoviruses.

Adenoviruses (AdVs) are categorized into four genera: Mastadenovirus (from mammals) (7, 42), Aviadenovirus (from birds) (19), and two recently accepted genera, Atadenovirus (5, 9) and Siadenovirus (13). Atadenoviruses from ruminants (3, 8, 11, 32, 33), birds (21, 23), snakes (4, 15, 34), and a marsupial (44) were isolated and studied. The two siadenoviruses were from a frog (14) and poultry (39). The mixed-host origin of these genera suggests that host switches may have occurred. Prior to the availability of reptile AdV phylogenetic information, a reptilian origin of atadenoviruses was proposed on the basis of comparison of phylogenetic trees of the adenoviruses and host rRNA (18). A fifth genus is proposed for a sturgeon adenovirus (4).

AdV-like particles have been identified in many reptile species, including 10 snake species (22, 28, 36, 38, 40, 43, 47), 4 lizard species (24, 27, 29, 30), and 1 crocodilian species (25). Lesions in reptiles associated with AdV-like agents include hepatitis (25, 27, 29, 43), enteritis (22, 30, 47), esophagitis (24, 40), splenitis (22), and encephalopathy (41). The only reptile adenovirus previously further classified was a corn snake (Elaphe guttata) isolate (15) which was consistent with the genus Atadenovirus. A Boa constrictor isolate was identical to the corn snake isolate (34).

Methods previously used for diagnosis of AdV infection in reptiles include virus isolation (26), electron microscopy (22), DNA in situ hybridization (ISH) (38), and plaque reduction neutralization (PRN) (34). Virus isolation requires further diagnostics for speciation. Electron microscopy and available ISH protocols do not speciate reptile adenoviruses (38). The cross-reactivity of neutralizing antibodies to reptile adenoviruses in PRN is not known. PRN also requires that a virus had been previously cultured, making this a poor method for novel virus discovery. Consensus PCR is a rapid way to obtain a sequencing template from clinical samples (45). A PCR protocol used for the snake atadenovirus (15) did not work with gecko samples; a technique usable for diverse novel adenoviruses was needed. The protocol described here has been used to amplify these atadenoviruses as well as a mastadenovirus and an aviadenovirus (J. F. X. Wellehan, unpublished data).

Samples.

The eublepharid gecko sample was obtained from a disease outbreak. Fat-tail geckos (Hemitheconyx caudicinctus) were wasting, with a high mortality rate. Histologically, lymphoplasmacytic enteritis with intranuclear inclusion bodies and multifocal lymphocytic pancreatitis were seen. Leopard geckos (Eublepharis macularius) in this collection did not have a high mortality rate, but a few did suffer weight loss and death. In these leopard geckos, both eublepharid gecko adenovirus and Cryptosporidium spp. were found.

The Tokay gecko (Gekko gecko) sample was collected from a cloacal wash of an animal with marked weight loss and regurgitation. Histologically, there was severe proliferative gastritis with myriad intralesional protozoa consistent with Cryptosporidium spp. No inclusion bodies were observed in tissue sections.

The Gila monster (Heloderma suspectum) adenoviral sample was collected from a cloacal wash of an animal with signs of regurgitation.

The blue-tongued skink (Tiliqua scincoides intermedia) sample was extracted from paraffin-embedded tissue of an asymtomatic neonatal animal euthanized for a spinal deformity. Histologically, the small intestine had numerous intranuclear basophilic inclusions. Transmission electron microscopy demonstrated adenovirus-like particles.

The bearded dragon (Pogona vitticeps) sample was extracted from paraffin-embedded tissue. Histologic examination revealed that the intestinal mucosa, hepatocytes, and bile ducts contained intranuclear inclusions.

The mountain chameleon (Chameleo montium) adenoviral sample was extracted from paraffin-embedded tissue from a previously described infection (30).

PCR and sequencing.

DNA was extracted with a DNEasy kit (QIAGEN, Valencia, Calif.). Nested-PCR amplification of a partial sequence of the adenoviral DNA polymerase gene was performed. For the first amplification, the 20-μl reaction mixture contained 2 μl of extracted DNA, 5% dimethyl sulfoxide, 1 μM concentrations for each primer (forward primer, polFouter [5′-TNMGNGGNGGNMGNTGYTAYCC-3′, where Y = C or T, N = A, C, G, or T, and M = A or C]; reverse primer, polRouter [5′-GTDGCRAANSHNCCRTABARNGMRTT-3′, where R = A or G, M = A or C, D = A, G, or T, S = G or C, H = A, T, or C, and B = G, T, or C]), 200 μM (each) for dATP, dCTP, dGTP, and dTTP, 2.5 U of Pwo DNA polymerase (Thermo Hybaid, Franklin, Mass.), and PCR buffer (Thermo Hybaid). The mixtures were amplified with an initial denaturation at 94°C for 5 min followed by 45 cycles at 94°C for 30 s, 46°C for 60 s, and 72°C for 60 s. There was a final extension at 72°C for 7 min. For the second round, 2 μl of product from the above-described reaction mixture was used with forward primer polFinner (5′-GTNTWYGAYATHTGYGGHATGTAYGC-3′, where W = A or T) and reverse primer polRinner (5′-CCANCCBCDRTTRTGNARNGTRA-3′) and amplified under the same conditions. Purified products were sequenced directly using a Big-Dye terminator kit (Perkin-Elmer, Branchburg, N.J.) and analyzed on ABI 377 automated DNA sequencers.

PCR amplification resulted in products that consisted of 318 to 324 bp before primer sequences were edited out. Sequences from the fat-tail gecko and leopard gecko samples were identical and were considered to represent the same virus. On the basis of naming conventions, these adenoviruses were named eublepharid adenovirus 1 (EuAdV-1, from the leopard and fat-tail geckos), gekkonid adenovirus 1 (GeAdV-1, from the Tokay gecko), agamid adenovirus 1 (AAdV-1, from the bearded dragon), helodermatid adenovirus 1 (HeAdV-1, from the Gila monster), chameleonid adenovirus 1 (ChAdV-1, from the mountain chameleon), and scincid adenovirus 1 (ScAdV-1, from the blue-tongued skink) (7).

Phylogenetic calculations.

The sequences were compared to known sequences in GenBank (National Center for Biotechnology Information, Bethesda, Md.), EMBL (Cambridge, United Kingdom), and Data Bank of Japan (Mishima, Shiuoka, Japan) databases by use of TBLASTX (2). All sequences except for that of EuAdV-1 showed the highest score with duck adenovirus 1 (DAdV-1) DNA polymerase (GenBank accession no. BK000404.1) (12). EuAdV-1 showed the highest score with bovine adenovirus 4 (BAdV-4) DNA polymerase (GenBank accession no. AF036092) (11). DAdV-1 and BAdV-4 are both in the genus Atadenovirus.

An alignment of predicted homologous sequences is shown (Fig. 1). Phylogenetic analyses of the predicted alignment were performed with PHYLIP (Phylogeny Inference Package; version 3.572c) program software (17) as described earlier (20). The phylogenetic tree (Fig. 2) shows the presently accepted species clustering. Similar AdV serotypes are classified as members of a species (7). While the two snake isolates studied so far are identical (34), these lizard AdV sequences are adequately different (less than 90% sequence identity) to be considered distinct adenovirus species. The tree also shows that these lizard AdVs group with the genus Atadenovirus. Further characterization would help confirm this. The classification of multiple species of reptile adenoviruses, all in the genus Atadenovirus, supports a reptilian origin for Atadenovirus.

FIG. 1.

Alignment of predicted homologous 89- to 91-amino-acid sequences from GenBank corresponding to the complement of bases 6458 to 6732 of human adenovirus 1 (GenBank accession number AF534906) or from our own studies. Sequences were aligned using MultAlin (10). Members of the same species are shown with black characters on a gray-shaded background. Genera are separated by lines. Viruses new to this study are in bold.

FIG. 2.

Phylogenetic tree of partial adenoviral DNA polymerase amino acid sequences. Protdist (Dayhoff PAM 001 matrix) followed by Fitch (global rearrangements) programs were used. The validity of the tree topology obtained was tested by using bootstrap analysis (16) with 100 resamplings from the aligned sequences followed by distance matrix calculations, and the most probable (consensus) tree was calculated. The tree was displayed using TreeView 1.5.2 (37). Novel lizard adenoviruses are in bold. Other sequences were retrieved from GenBank. The following serotypes comprising a viral species are shown in black characters on a gray background: Human adenovirus B (HAdV-B), HAdV-C, HAdV-E, Canine adenovirus (CAdV-A), and Bovine adenovirus D (BAdV-D). Sturgeon adenovirus 1 was used as an outgroup of the unrooted tree. Bootstrap values are shown. Branchings with bootstrap values less than 50 are not shown, and areas where these branchings occurred are indicated with checkered boxes.

The lizard sequences show a balanced G+C content (43.75 to 58.09% over the region sequenced) in contrast to results seen with the ruminant, marsupial, and bird atadenoviruses, so named because of their biased genome (32.35 to 47.43% G+C over the comparable region). Following a host switch, viral genes would undergo rapid evolution as they adapt to a new host. DNA containing CG dinucleotides is recognized by the innate immune system (1). In the absence of a previous host-adapted function of a sequence, there could be strong selective pressure away from CG dinucleotides. An A+T bias in genome composition could signify that a virus had evolved in a secondary host (6).

Phylogenetic analyses of mammalian herpesviruses suggest that the branching patterns of Herpesviridae are often congruent with those of host species (35). While reptilian herpesviruses fit well with herpesvirus phylogeny (46), the herpesviruses of amphibians and fish are highly divergent (31) and phylogenetic comparison with other herpesviruses is challenging. The fish and amphibian “herpesviruses” may have diverged long before the divergence of their hosts. In contrast, the adenoviruses are more clearly of a continuous lineage (6), providing the possibility to study coevolution of viruses through all vertebrate classes. The low resolution in this study emphasizes the need for additional sequences from more hosts.

Nucleotide sequence accession numbers.

Sequence data were submitted to GenBank; the accession numbers are AY576677 to AY576682.