Evidence for an endogenous papillomavirus-like element in the platypus genome

Abstract

Papillomaviruses (PVs) infect a wide range of vertebrates and have diversified into multiple genetic types, some of which have serious consequences for human health. Although PVs have to date only been characterized as exogenous viral forms, here we report the observation of an endogenous viral element (EPVLoa) in the genome of the platypus (Ornithorhynchus anatinus) that is related to PVs. Further data mining for endogenous PV-like elements is therefore warranted.

Papillomaviruses (PVs; family Papillomaviridae) are small, circular DNA viruses with a dsDNA genome approximately 8000 bp in length. PVs have been identified in a wide range of vertebrate species, particularly mammals (de Villiers et al., 2004; Antonsson & McMillan, 2006; Herbst et al., 2009; Lange et al., 2011), and >30 genera and 189 genetically distinct viral types have been described to date (de Villiers et al., 2004; Bernard et al., 2010). Depending on the viral type in question, human infection by PVs can be either asymptomatic or ultimately result in cancerous tumours (Bernard et al., 2006; Muñoz et al., 2006; Munday & Kiupel, 2010). Importantly, PVs possess a dsDNA genome, enter the cell nucleus and have a number of characteristics that might facilitate endogenization (Holmes, 2011). In particular, although the replication of PVs does not involve integration into the host genome, many genomic-integration events have been characterized in mammalian cells (Wentzensen et al., 2004).

We employed a genomic mining of 74 chordate genomes (Table S1, available in JGV Online) available at the NCBI database (http://www.ncbi.nlm.nih.gov/sutils/genom_table.cgi) to screen for endogenous PVs. We used L1 (major capsid) protein sequences of various vertebrate PVs as queries in a genomic blast analysis, employing a cut-off value of 10e⁻¹⁰ to signify a positive hit. The following representative PV sequences, which cover the full phylogenetic diversity of PVs (Bernard et al., 2010), were used as queries: bovine PV 8 (GenBank accession no. NC_009752, host = cow), canine PV 2 (NC_006564, dog), Caretta caretta PV 1 (NC_011530, loggerhead sea turtle), common chimpanzee PV 1 (NC_001838, chimpanzee), human PV 1 (NC_001356, human), human PV type 6b (NC_001355, human), Mastomys coucha PV 2 (NC_008519, Southern multimammate mouse), Phocoena spinipinnis PV (NC_003348, Burmeister’s porpoise), Psittacus erithacus timneh PV (NC_003973, African grey parrot), Rousettus aegyptiacus PV type 1 (NC_008298, Egyptian fruit bat) and Tursiops truncatus PV 2 (NC_008184, bottlenose dolphin).

Surprisingly, all of these queries resulted in strongly positive hits to sequences in the genome of the platypus (Ornithorhynchus anatinus); contig 159295.1, e-value 1e⁻²⁵, sequence identity = 93/325 (29 %), using the query with Caretta caretta PV 1 as an example. We term this sequence ‘endogenous PV-like element of O. anatinus’ (EPVLoa), although we only observed two copies with a strong match to L1 in the platypus genome: contig 159295.1, 1465 bp in length, containing one premature stop codon, and contig 9789.3, 20 736 bp in length, containing six premature stop codons. In addition, only partial L1 sequences were recovered. A reciprocal blast using the EPVLoa L1 sequence as the query confirmed its relationship to exogenous PVs, in this case exhibiting the closest match to human PV type 45 (e-value 5e⁻³⁵, sequence identity = 91/286 = 32 %; Fig. 1). A blast analysis using other PV proteins revealed no positive hits, as did an equivalent analysis using Polyomaviridae, the viral family related most closely to the Papillomaviridae (Woolford et al., 2007). That EPVLoa is very rare and that we were unable to identify a complete PV genome may be a function of the relatively low quality of the platypus genome, which is currently only available at sixfold sequencing coverage and therefore of uncertain nature (Lewin et al., 2009). Alternatively, it may be that these endogenization events are extremely rare, and/or that integration events only involve L1 sequences. With respect to the latter, it is notable that integration events involving partial PV genomes, as well as single PV genes, have been documented (zur Hausen, 2002). Finally, that EPVLoa is divergent from any known extant PV, and contains multiple premature stop codons, argues strongly against contamination by exogenous viruses.

Fig. 1.

Sequence alignment of partial L1 protein sequences from EPVLoa and exogenous papillomaviruses. Asterisks above the alignment denote those amino acids shared among all sequences; ⧫ represent sites in which EPVLoa possesses an amino acid residue shared with ≥80 % of the exogenous PVs in the alignment. Those amino acids shared by EPVLoa and HPV45 (GenBank accession no. ABP99855) are highlighted by double underlining at the bottom of the alignment. It is important to note that this is not the alignment used in the phylogenetic analysis (Fig. 2); in this case, all highly divergent regions, including insertions and deletions, were removed using Gblocks (Talavera & Castresana, 2007; Fig. S1). Virus abbreviations are given in Table 1.

To determine the evolutionary relationships between EPVLoa and exogenous PVs, we collected representatives of the full phylogenetic diversity of exogenous PVs from GenBank (n = 44; Table 1). We then aligned these sequences with EPVLoa using clustal_x (Larkin et al., 2007), with a subsequent manual adjustment undertaken using Se-Al (http://tree.bio.ed.ac.uk/software/seal/). This resulted in an L1 protein alignment of 312 aa in length, of which 15 % of amino acid sites (47/312) were conserved among all sequences including EPVLoa (Fig. 1). Although EPVLoa is clearly divergent from the exogenous PVs, all of these sequences share a number of relatively conserved regions (such as residues 47–52, 129–133, 140–147 and 154–174; Fig. 1). Next, we used the Gblocks program (Talavera & Castresana, 2007) to remove the divergent and ambiguously aligned regions, including all those containing insertions and deletions (Fig. S1). This resulted in a final sequence alignment of 196 aa (including 19 invariant amino acid residues) from which evolutionary relationships could be inferred. Phylogenetic analysis of this 196 residue alignment was performed using the maximum-likelihood method available in PhyML 3.0 (Guindon et al., 2010), incorporating the WAG+Γ model of amino acid substitution, with the robustness of each node determined using 1000 bootstrap replicates. The resulting phylogenetic tree placed EPVLoa as more divergent than all known exogenous PVs (Fig. 2), indicative of an ancient divergence event, and hence our designation that it is derived from a ‘PV-like’ virus.

Table 1.

GenBank accession numbers of L1 protein sequences of exogenous PVs used in this analysis

Virus	Abbreviation	GenBank accession no.
Common chimpanzee papillomavirus 1	CCPV	NP_045018
Human papillomavirus type 6b	HPV6b	NP_040304
Rhesus monkey papillomavirus	RMPV	NP_043338
Colobus guereza papillomavirus type 2	CGPV2	YP_004646343
Human papillomavirus type 92	HPV92	NP_775311
Macaca fascicularis papillomavirus type 2	MFPV2	YP_004646337
Canine papillomavirus 5	CPV5	YP_003204674
Canine papillomavirus 4	CPV4	YP_001648805
Capreolus capreolus papillomavirus 1	CaCPV1	YP_002004574
Deer papillomavirus	DPV	NP_041300
European elk papillomavirus	EElPV	NP_041313
Ovine papillomavirus 1	OPV1	NP_044438
Sus scrofa papillomavirus type 1	SSPV1	YP_002235542
Francolinus leucoscepus papillomavirus 1	FLPV1	YP_003104804
Erinaceus europaeus papillomavirus	EEuPV	YP_002427696
Equine papillomavirus 2	EPV2	YP_002635574
Caretta caretta papillomavirus 1	CCPV1	YP_002308363
Chelonia mydas papillomavirus 1	CMPV1	YP_002308370
Bovine papillomavirus 8	BPV8	YP_001429551
Fringilla coelebs papillomavirus	FCPV	NP_663767
Human papillomavirus 116	HPV116	YP_003084352
Mastomys natalensis papillomavirus	MNPV	NP_042019
Cottontail rabbit papillomavirus	CRPV	NP_077113
Bovine papillomavirus 1	BPV1	NP_056744
Canine oral papillomavirus	COPV	NP_056819
Felis domesticus papillomavirus type 1	FDPV1	NP_848025
Procyon lotor papillomavirus 1	PLPV1	YP_249604
Human papillomavirus 1	HPV1	NP_040309
Human papillomavirus type 41	HPV41	NP_040294
Ursus maritimus papillomavirus 1	UMPV1	YP_001931973
Phocoena spinipinnis papillomavirus	PSPV	NP_542623
Capra hircus papillomavirus type 1	CHPV1	YP_610959
Mastomys coucha papillomavirus 2	MCPV2	YP_803393
Mus musculus papillomavirus type 1	MMPV1	YP_003778198
Old World harvest mouse papillomavirus	WHMPV	YP_873945
Rattus norvegicus papillomavirus 1	RNPV1	YP_003169705
Rousettus aegyptiacus papillomavirus type 1	RAPV1	YP_717913
Trichechus manatus latirostris papillomavirus 1	TMLPV1	YP_164627
Erethizon dorsatum papillomavirus type 1	EDPV1	YP_224227
Canine papillomavirus 2	CPV2	YP_164635
Psittacus erithacus timneh papillomavirus	PETPV	NP_647590
Tursiops truncatus papillomavirus 1	TTPV1	YP_002117846
Bovine papillomavirus 3	BPV3	NP_694451
Equinus papillomavirus	EPV	NP_694429

Fig. 2.

Phylogenetic relationships of EPVLoa and exogenous PVs. Bootstrap values (>70 %) are shown for key nodes. The tree is midpoint-rooted for purposes of clarity only. Host species information is shown in parentheses; the PV genus of each sequence is shown in square brackets. Bar, 0.3 amino acid substitutions per site.

Despite the apparent rarity of EPVLoa in the platypus genome, its discovery is important for two reasons: not only is this the first observation of an endogenous PV-like element, but it also means that PVs, or viruses very closely related to PVs, must be capable of infecting germ-line cells. As a consequence, we suggest that more attention is given to the possibility that the endogenization of viruses of this kind has occurred during vertebrate evolution.