Eqin

R.

G.MM.a Dipb Dip

Boloc Cen

1. I

divcov7 taniDN

Veterinary Microbiology 170 (2014) 213–223

A R

Artic

Rece

Rece

Acce

Keyw

Equ

Asin

Evol

Gen

Skin

Vira

*

Univ

Tel.:

http

037

uus asinus Papillomavirus (EaPV1) provides new insightsto equine papillomavirus diversity

Lecis a, G. Tore a, A. Scagliarini b, E. Antuofermo a, C. Dedola a, C. Cacciotto a,. Dore a, E. Coradduzza a, L. Gallina a, M. Battilani b, A.G. Anfossi a,

Muzzeddu c, B. Chessa a, M. Pittau a, A. Alberti a,*

artimento di Medicina Veterinaria, Universita degli Studi di Sassari, via Vienna 2, 07100 11 Sassari, Italy

artimento di Scienze Mediche Veterinarie, Alma Mater Studiorum Universita di Bologna, via Tolara di sopra 50, 40064 Ozzano Emilia,

gna, Italy

tro Fauna Bonassai, Olmedo, Sassari, Italy

ntroduction

The Papillomaviridae family consists of a large anderse group of viruses characterised by a double strand,alently linked circular genome typically ranging fromo 8 kb in size, and causing proliferative lesions inmals and human (Howley and Lowy, 2007). Based onA sequence homologies, about 200 different human

papillomavirus types (HPVs) have been identified andclassified into a number of genera (Bernard et al., 2010;Rector and Van Ranst, 2013). According to their associationwith cancer, HPVs of the alpha genus (a-HPVs) areclassified as low-risk, such as types 6 and 11, and ashigh-risk, such as types 16 and 18 (Munoz et al., 2003;Howie et al., 2011). While low-risk a-HPVs have beenassociated to benign proliferative lesions, high-risk a-HPVs can cause cervical, anogenital, and head and neckcancer (zur Hausen, 1999; Cogliano et al., 2005). Lately,HPVs of the beta genus (b-HPVs) are being increasinglyinvestigated due to a possible role in development ofnonmelanoma skin cancer (Howley and Lowy, 2007;Haedicke and Iftner, 2013). As a general principle, it is

T I C L E I N F O

le history:

ived 16 November 2013

ived in revised form 28 January 2014

pted 4 February 2014

ords:

ine papillomaviruses

ara white donkey

ution

ome

l diversity

A B S T R A C T

We detected a novel papillomavirus (EaPV1) from healthy skin and from sun associated

cutaneous lesions of an Asinara (Sardinia, Italy) white donkey reared in captivity in a

wildlife recovery centre. The entire genome of EaPV1 was cloned, sequenced, and

characterised. Genome is 7467 bp long, and shows some characteristic elements of horse

papillomaviruses, including a small untranslated region between the early and late regions

and the lack of the retinoblastoma tumour suppressor binding domain LXCXE in E7.

Additionally, a typical E6 ORF is missing. EaPV1 DNA was detected in low copies in normal

skin of white and grey donkeys of the Asinara Island, and does not transform rodent

fibroblasts in standard transformation assays. Pairwise nucleotide alignments and

phylogenetic analyses based on concatenated E1-E2-L1 amino acid sequences revealed

the highest similarity with the Equine papillomavirus type 1. The discovery of EaPV1, the

prototype of a novel genus and the first papillomavirus isolated in donkeys, confirms a

broad diversity in Equidae papillomaviruses. Taken together, data suggest that EaPV1 is a

non-malignant papillomavirus adapted to healthy skin of donkeys.

� 2014 Elsevier B.V. All rights reserved.

Corresponding author at: Dipartimento di Medicina Veterinaria,

ersita degli Studi di Sassari, Via Vienna 2, 07100 Sassari, Italy.

+39 079 229448; fax: +39 079 229451.

E-mail address: alberti@uniss.it (A. Alberti).

Contents lists available at ScienceDirect

Veterinary Microbiology

jo u rn al ho m epag e: ww w.els evier .c o m/lo cat e/vetmic

://dx.doi.org/10.1016/j.vetmic.2014.02.016

8-1135/� 2014 Elsevier B.V. All rights reserved.

R. Lecis et al. / Veterinary Microbiology 170 (2014) 213–223214

accepted that PVs sharing less than 60% nucleotide identityin the L1 ORF represent different genera. Similarly,nucleotide identities below 70% and 90% define differentspecies and types, respectively (De Villiers et al., 2004;Bernard et al., 2010). Non-human PVs have been detectedin normal skin and in both benign and malignantproliferative lesions of 54 different animal species (Teraiet al., 2002; Alberti et al., 2010; Rector and Van Ranst,2013; Scagliarini et al., 2013). In single animal species agenotype diversity comparable to that found in humanshas not yet been identified, even if there was anexponential increase in the number of animal papilloma-viruses identified over the last decade (Rector and VanRanst, 2013). The association of animal papillomavirustypes to defined degrees of pathogenicity needs to behighlighted, the suspected great diversity of animal PVsstill remaining uncovered. Indeed, each single animalspecies potentially hosts a number of PV types comparableto the one found in human.

Based on the number of types identified in singleanimal species, the most widely studied animal PVinfections are those of domestic animals (Lange et al.,2013b). According to the PapillomaVirus Episteme (PaVE,http://pave.niaid.nih.gov/#home), 13 viruses have beenfound and fully sequenced in dogs, 12 in cattle (Zhu et al.,2012), 7 in horses (Lange et al., 2013b), 3 in sheep (Albertiet al., 2010) and cats (Munday et al., 2013).

Horse PVs are often quoted as an example of exceptionto Clay’s rule, according to which related PVs infectphylogenetically related host species (Gottschling et al.,2011). Indeed, horses are commonly infected by bovinePVs type 1 and 2 (BPV1, BPV2), which are the causativeagents of equine sarcoid (Nasir and Campo, 2008). Apartfrom BPV1 and BPV2, the seven horse papillomaviruses(EcPV1 to 7) therefore represent the PV diversity so fardescribed in Equidae.

Equidae PVs seem to be associated to distinct clinicalconditions (Lange et al., 2013a,b). EcPV1 has beenisolated from benign proliferative lesions of young horses(Postey et al., 2007). EcPV2, EcPV3, EcPV4, EcPV5, EcPV6,and EcPV7 were found both in benign and malignantgenital neoplasia, in aural plaques, in semen and inapparently healthy skin, some of them often in the samehost (Lange et al., 2011, 2013a,b; Kainzbauer et al., 2012;Sykora et al., 2012). The absence of closely related virusescomplicates the taxonomic and phylogenetic allocationof Equidae PVs. In a recent work conducted to quantifythe phylodynamic forces driving PVs evolution, Gottschl-ing et al. (2011) obtained significant associations ofDelta + Zeta-PVs with Perissodactyla and Ruminantia, withthe two horse PVs EcPV1 and EcPV2 at the root of theselineages.

Aim of this study was to investigate papillomavirusdiversity in Equidae and to establish the presence of PV inspecies related to horse. We identified EaPV1, the firstpapillomavirus found in donkey. By means of molecularand bioinformatics tools we fully characterised EaPV1genome, described its main features at the nucleotide andamino acid levels, and investigated its phylogeny andevolution. The ability of the EaPV1 early region totransform rat fibroblasts in standard transformation assay

and the viral loads in clinically healthy skin of a number ofanimals living in close contact were also investigated.

2. Material and methods

2.1. Animals and samples

During a survey conducted to identify cutaneous PVs inwild and domestic animals, swabs were collected fromnormal skin and from skin presenting changes compatiblewith solar damage (erythema, scales, and dark pigmentedmacules) of an Asinara white donkey identified as EA01M(Fig. 1A), and hosted in the Wildlife Recovery Centre ofBonassai (Sardinia, Italy). The Asinara donkey (Equus asinus

var. albina) is a rare breed of feral white donkey indigenousto the island of Asinara, which lies off the Northwest coastof Sardinia, Italy. It is one of the ‘seven indigenous donkeybreeds of limited distribution’ listed by the Italian breedersassociation. A population of around 120 animals is actuallyfree-ranging on the Asinara island National Park, while asmaller group of around 30 animals lives in the Porto ConteRegional Park (Northern Sardinia). All these animals arecharacterised by a white coat, pink or light blue eyes, dueto incomplete albinism, and a small size (around 1 mheight).

Full-thickness skin specimens were obtained from boththe right and left pinna of donkey EA01M by punch biopsy(Fig. 1B). Tissues were fixed in 10% neutral-bufferedformalin, embedded in paraffin, and sectioned to 4–6 mmslices. Adjacent slices were mounted onto glass slides andstained with haematoxylin and eosin (H&E) for lightmicroscopy. A second set of tissues was kept at �80 8Cuntil nucleic acids extraction. Cutaneous swabs were alsosampled from 10 white and 8 grey free-living donkeys in theAsinara Island. All donkeys were subjected to dermatolo-gical and physical examination.

2.2. Nucleic acid extraction and PV genome investigation

DNA was extracted from punch biopsies and skin swabsrespectively using the DNeasy Blood & Tissue Kit and theQIAamp DNA Mini Kit (Qiagen, Italy), following manufac-turer recommendations. To identify PV genomes, therolling circle amplification (RCA) technique combinedwith restriction enzyme digestion was performed. RCA wascarried out by using the TempliPhi 100 Amplification Kit(GE Healthcare, Italy), following the protocols previouslydescribed by Rector et al. (2004) and modified in Albertiet al. (2010). Briefly, 5 ml of DNA extractions were mixedwith 10 ml of sample buffer and subsequently heated for3 min at 95 8C, then transferred on ice. Ten microliters ofTempliPhi reaction buffer, 0.4 ml of TempliPhi enzyme mixcontaining phi 29 DNA polymerase, random hexamers in50% glycerol, and 0.4 ml of 10 mM dNTPs per sample weremixed and added to the cooled sample. The final reactionwas subsequently incubated for 16 h at 30 8C. Phi 29polymerase was eventually inactivated at 65 8C for 10 min.RCA products were digested with restriction enzymesEcoRI, BamHI and HindIII and run on a 0.8% agarose gel tovisualize the presence of a DNA profile consistent with thelength of a papillomaviral genome, or of multiple bands with

sizeparseqPCRGGTAGEco

primandwefragprecomapptionafte(Invleasclon

2.3.

of

punandGCTAGof t

Fig.

pigm

app

R. Lecis et al. / Veterinary Microbiology 170 (2014) 213–223 215

s adding up to this length. An EcoRI band representingt of a PV genome was cloned into pUC19 and fullyuenced. To achieve EaPV1 genome full sequencing, a

was performed using primers EAPV1/PCR/F (50-TGCGTGAATGCTTTTATG-30) and EAPV1/PCR/R (50-GGAT-

CAAACTGGCTATG-30), designed on the sequence of theRI RCA product.To confirm the integrity of the PV circular genome,

ers EAPV1/RCA/F1 (50-AGACACACTACCTAGTGTGG-30) EAPV1/RCA/R1 (50-TCTAATTTTGGGAGGGTCGC-30)

re designed based on the sequence of the PCR-generatedment, and used to amplify the EcoRI fragment

viously obtained by RCA. PCRs were performed bybining 100 ng of DNA with 10 mM of each of the

ropriate primers and following Qiagen recommenda-s for Taq polymerase. All amplicons were sequencedr cloning into pCR4 with the TOPO1 TA Cloning1 Kititrogen, Italy), following vendor recommendation. Att three clones were automatically sequenced for eaching (BMR genomics, Padova, Italy).

Traditional and real-time PCR

In order to investigate the presence of EaPV1 in a panelfree-living donkeys, DNA extractions obtained fromch biopsies and swabs were tested by traditional

real time PCR with primers EasL1_298F (50-CTGCGAGCTATTGAGGT-30) and EasL1_473R (50-

AATTTGCACCTGCTTTGG-30), designed to target 176 bphe L1 gene of EAPV1. Traditional PCR amplifications

were performed in a 50 ml volume containing approxi-mately 150 ng of DNA, and by following Qiagen recom-mendations for Taq polymerase. PCR profile included aninitial denaturation step at 95 8C for 2 min, followed by 35cycles of denaturation (20 s), annealing at 60 8C (20 s) andextension at 72 8C (20 s). Amplicons were routinely clonedinto pCR4 using the TOPO1 TA Cloning1 Kit (Invitrogen,Italy), following vendor recommendations. Three cloneswere selected after each transformation of Escherichia coli

DH5 alpha and sequenced for comparison. Nucleic acidquantification was obtained by Real time PCR, performedwith Rotor gene 3000 system (Corbett research, Australia),using the Rotor Gene SYBR Green PCR kit (Qiagen). Thereaction was carried out in a final volume of 25 mlcontaining 12.5 ml of 2� SYBR Green PCR Master Mix,1 mM of EasL1_298F and EasL1_473R primers, 7.5 ml ofwater and 2.5 ml of each DNA sample. The reactionincluded a denaturation step at 95 8C for 5 min followedby 40 cycles at 95 8C for 5 s and annealing/extension at60 8C for 15 s. The absolute quantification of the viral DNAwas obtained by plotting 10-fold dilutions of 1.2–1.2 � 105 copies/ml of pUC19 plasmid containing the4400 bp EcoRI fragment against the corresponding thresh-old cycle value. Only donkeys positive to traditional PCRwere tested by real time PCR.

2.4. EAPV1 genome sequencing and characterisation

The RCA-generated EcoRI fragment and the PCR-gener-ated fragment containing EaPV1 genome were cloned into

1. Macroscopic and microscopic features of typical dark pigmented areas found on the skin of Asinara white donkeys. A: donkey EA01M. B: Dark

ented areas in both the right and left pinna. C, D: Histological examination of punch skin biopsies derived from donkey EA01M. Koilocytes can be

reciated in the squamous epithelium.

R. Lecis et al. / Veterinary Microbiology 170 (2014) 213–223216

pUC19 (Invitrogen, Italy) to generate pUC19/EaPV1/4400and pUC19/EaPV1/3000. Briefly, both the RCA and PCRproducts were digested with 100 units of EcoRI overnightand run on a 0.8% agarose gel. The appropriate fragmentswere extracted using the QIAquick Gel Extraction Kit(Qiagen, Italy). Fragments were then ligated into thepUC19 vector (previously cut with EcoRI and depho-sphorylated), by using the Rapid DNA Dephos & LigationKit (Roche, Italy). Ligation products were used to transformOne Shot MAX Efficiency DH5AlphaTM-T1R competent cells(Invitrogen, Italy). Bacteria were incubated for blue-whitecolony screening on agar plates containing X-gal and100 mg/ml ampicillin, and white colonies were checkedby EcoRI digestion of miniprep DNA. Three clones for eachtransformation were selected alternatively containingplasmids pUC19/EaPV1/4400 or pUC19/EaPV1/3000. Thecomplete genome of the Equus asinus Papillomavirus type 1(EaPV1) was determined by primer-walking sequencing ofthe cloned DNA fragments, starting from the universalprimers sites in the multiple cloning site of the pUC19plasmid. Sequencing was performed on an ABI Prism 3100Genetic Analyzer (Perkin-Elmer Applied Biosystems, FosterCity, CA, USA). To exclude the presence of two or moreadjacent EcoRI sites in the EaPV1 genome, potentiallygenerating small DNA fragments not detectable in agarosegel electrophoresis after RCA digestion, and thereforehampering the correct reconstruction of the EaPV1 genome,sequences overlapping regions of the RCA- and PCR-generated fragments were aligned, and checked for fullhomology. The nucleotide sequence of EaPV1 genome wasdeposited in GenBank using the National Center forBiotechnology Information (NCBI, Bethesda, MD) BankItv3.0 submission tool (http://www3.ncbi.nlm.nih.gov/BankIt/) under accession number KF741371. Prediction ofopen reading frames (ORFs) was performed using the ORFFinder tool on the NCBI server of the National Institutes ofHealth (http://www.ncbi.nlm.nih.gov/gorf.html). Molecu-lar weight of the putative proteins was calculated by usingthe ExPASy (Expert Protein Analysis System) Compute pI/MW tool (http://www.expasy.org/tools/pi_tool.html). Pair-wise sequence alignments and sequences similarities werecalculated using the ClustalW (Thompson et al., 1994) andthe identity matrix options in Bioedit (Hall, 1999),respectively.

2.5. Phylogenetic analyses

To perform multiple nucleotide sequence alignments,the sequences of EaPV1 and 94 other PVs (Table S1) wereimported in DAMBE version 4.2.7 (Xia and Xie, 2001) andaligned at the amino acid level with ClustalW. According toGottschling et al. (2011), this was done separately for thedifferent ORFs, and the E1, E2, and L1 ORFs were pastedtogether in one compiled alignment. The highly variable E4gene region nested within the E2 gene was excluded frommolecular analyses. The evolutionary history of EaPV1 andthe 94 PVs considered in this study was inferred by usingthe Maximum Likelihood method based on the WhelanAnd Goldman model (Whelan and Goldman, 2001),identified as the best-suited evolutionary model for ourdata. Initial tree(s) for the heuristic search were obtained

by applying the Neighbor-Joining method to a matrix ofpairwise distances estimated using a JTT model. A discreteGamma distribution was used to model evolutionary ratedifferences among sites (five categories (+G, para-meter = 1.1232)). The rate variation model allowed forsome sites to be evolutionarily invariable ([+I], 6.7697%sites). There were a total of 1015 positions in the finaldataset. Evolutionary analyses were conducted in MEGA5(Tamura et al., 2011). Statistical support for internalbranches of the trees was evaluated by bootstrapping with1000 iterations (Felsenstein, 1985). Maximum parsimony(MP) trees and consensus values were generated using thesame software. Trees were edited with NJplot (Perriere andGouy, 1996) and Treeview v. 1.5.2 (Page, 1996).

2.6. Standard transformation assays

A fragment of 1091 bp including the LCR region and thefull-length E7 gene was amplified with primers EAPV_LCR_-BAMHI/F (50-CACGGATCCACATGCGTGCGAAGACATTTGG-30) and EAPV_LCR_ECORI/R (CGCGAATTCTCAGAATC-CATTTCTGTTCACGC). After double digestion with BamHIand EcoRI, the fragment was cloned into pcDNA3.1(+)(Invitrogen, Italy) pre-digested with the same enzymes.Plasmid pcDNA3.1(+)/EAPV/LCR-E7 was therefore gener-ated. Plasmid pCMV/RAS, constitutively expressing thewild-type (wt) Ha-ras protein, was purchased by Clontech(Italy). 208F rat cells (Quade, 1979) were grown inDulbecco’s modified Eagle medium (DMEM) supplementedwith 10% newborn calf serum, at 37 8C with 5% CO2 and 95%humidity. Standard transformation assay was conducted aspreviously described (Alberti et al., 2002). Briefly, one daybefore transfection, 208F cells were trypsinized and countedwith a ScepterTM Automated Cell Counter (Millipore, Italy).About 5 � 105 cells were seeded in 6-cm-diameter dishes.Cells were transfected alternatively with 3.75 mg ofpcDNA3.1(+)/EAPV/LCR-E7 alone or in association with3.75 mg of pCMV/RAS (Clontech) using the CalPhos Mam-malian Transfection kit (Clontech), following vendorsrecommendations. Control transfections were also per-formed by transfecting 208F cells with empty pcDNA 3.1(+),and with pCMV/RAS. One day after transfection, cells weretrypsinized and split in five dishes. After the cells reachedconfluence, medium was replaced with DMEM supplemen-ted with 5% newborn calf serum and 1 mM dexamethasoneevery 3 days. Foci of transformed cells were counted 1month after transfection and expressed as mean number offoci per dish. Data from three independent experiments(means � standard deviations) were considered when com-paring transformation activity of different plasmids.

3. Results

On general physical examination, white and greydonkeys free-living on the Asinara Island did not showany abnormalities. Dermatological examination revealedthe presence of white/brownish adherent scales, alopecia,and erythema on the external aspect of both ears’s pinna inall white donkey examined, including EA01M. Few smallmultifocal dark pigmented macules were also detected(Fig. 1B). Similar but milder lesions were observed on the

dorDerthebioma(Figandtionobs201dispkoi

fragthecircwaRCAunccan

Fig.

divi

Nuc

R. Lecis et al. / Veterinary Microbiology 170 (2014) 213–223 217

sal aspect of the nose and in the periocular area.matological lesions did not significantly differ among

subjects examined. Histological examination of skinpsies collected from both the pigmented and erythe-tous area on the ear pinna of the white donkey EA01M. 1C and D), revealed moderate epidermal hyperplasia hyperkeratosis, mild acanthosis, melanin pigmenta-

and multifocal vacuolization in the basal layer. Aserved in other animal PVs infections (Alberti et al.,0), keratinocytes with eccentric hyperchromatic nucleilaced by large perinuclear vacuoles compatible with

locytes were abundant in the squamous epithelium.RCA combined to EcoRI digestion generated twoments of approximately 4400 and 3000 bp in size,

refore consistent with the presence of a papillomaviralular genome (data not shown). The 4400 bp fragment

s successfully cloned into pUC19. Sequencing of this fragment revealed the following genes of a novel,lassified PV genome, designed EaPV1: partial L2, L1, theonical non coding region (NCR), E7, and the 50- end of

E1. EaPV1 full genome sequencing was accomplished bystandard PCR amplifying a fragment of about 3000 bp (seeSection 2), which included partial E1, E2, a second NCR, andpartial L2 genes. Additional PCRs conducted in order toconfirm genome integrity gave evidence of a covalentlylinked circular genome adding up to 7467 bp.

The complete nucleotide sequence of E. asinus PV type 1(EaPV1, GenBank accession number KF741371) has a GCcontent of 50.1% and includes the classical PV major ORFsE7, E1, E2, L2, L1. Moreover, the open reading frame E4 wasidentified as a result of a spliced message unifying the firstfew codons of E1 (nt 1063–1084) with a downstream ORFin the +1 frame of the E2 ORF (nt 3807–4132). Notably,EaPV1 lacks the E6 oncogene, and similarly to whatobserved in all other equine PVs, it shows an additionalNCR located between the early and late regions (Fig. 2). Asmall (165bp-long) ORF, ending about 50 bp upstream theE7 gene, was identified in the canonical NCR. No putativeconserved domains or significant similarity to PV E6 wereobserved on Protein BLAST search for this ORF.

2. Genomes of Equidae papillomaviruses. The eight Equidae papillomavirus genomes and their open reading frames (ORFs) are presented. Genomes are

ded into sections: early genes, late genes and non-coding regions (respectively Early, Late, and NCRs). Numbers indicate nucleotide positions.

leotide position number one is defined as the first following the stop codon of the L1 ORFs.

R. Lecis et al. / Veterinary Microbiology 170 (2014) 213–223218

Nucleotide and amino acid alignments defined EaPV1similarity with the other Equidae papillomaviruses (Tables1 and 2). Overall, EaPV1 shared the highest nucleotide (nt)and amino acid (aa) identities with EcPV1. With referenceto the single ORFs, the highest similarity was found withEcPV6 L1, with 56% and 47% identity on the nt and aa levels,respectively.

According to the predicted amino acid and nucleotidefeatures summarised in Table 2, the main differences foundbetween EaPV1 and the other equine PVs consisted in theabsence of E6 and consequently of its metal-binding motifs(CX2CX28-30CX2C), and in the lack of the AP1 (ActivatorProtein-1) binding sites (TGANTC). A TATA box wasidentified in the LCR at nucleotide 697 (TATAAA). The E1

binding site (A (A/T)GATTGTTGTTAACAAT) shows an A atposition two as it has been found in EcPV1, EcPV2, EcPV4,and EcPV6. The pRB-binding domain (L-X-C-X-E) wasabsent in the putative E7 protein as in all equine- andDelta-papillomaviruses (Narechania et al., 2004) and theputative E4 protein showed a typical high proline content(about 10%).

Phylogenetic trees were obtained from a concatenatedE1/E2/L1 nucleotide sequence alignment of EaPV1 and 94PV-types representative of the different PV genera andspecies (Fig. 3). Amino acid alignments were constructedseparately for the different ORFs and then included in thefinal dataset in one combined alignment of 1015 positions.The resulting maximum likelihood phylogenetic trees

Table 1

Nucleotide (and amino acid) identities among Equidae papillomaviruses.

EaPV1 EcPV1 EcPV2 EcPV3 EcPV4 EcPV5 EcPV6 EcPV7

E7/11.1 kDa 42 (24) 37 (23) 38 (17) 42 (20) 39 (22) 41 (19) 39 (19)

E1/70.6 kDa 54 (46) 54 (46) 53 (43) 51 (42) 53 (41) 52 (43) 52 (43)

E2/45.8 kDa 49 (33) 48 (31) 48 (34) 49 (32) 47 (33) 47 (31) 48 (31)

L2/59.7 kDa 44 (28) 44 (31) 41 (27) 44 (25) 46 (28) 44 (25) 43 (27)

L1/57.4 kDa 55 (45) 54 (46) 55 (46) 54 (45) 50 (43) 56 (47) 55 (47)

Fig. 3. Well-resolved maximum likelihood tree including EaPV1 and 94 PV-types representative of the different papillomavirus genera and species. ML tree

of the 95 PVs was inferred by using a combined E1–E2–L1 amino acid sequence analysis. PV genera are indicated in Greek letters and according to the PAVE

website (http://pave.niaid.nih.gov). Branch lengths are drawn to scale. Numbers on branches are ML bootstrap support values. Only values above 70 are

shown.

Table 2

Genomic nucleotide and amino acid features of Equidae PVs.

Predicted features* EaPV1 (IT-M)

KF741371

EcPV1

NC_003748jEcPV2

HM461973

EcPV3

NC_017862

EcPV4

NC_020085

EcPV5

NC_020084

EcPV6

NC_020500

EcPV7

NC_020501

Predicted nt-feature

Genome size (bp) 7467 7610 7803 7582 7554 7519 7551 7619

GC content (%) 50.1 53.0 56.0 50.9 54.9 50.2 51.3 52.5

E2 binding site

(ACC-N5-7-GGT)

118, 180, 230, 288, 387,

449, 480, 580, 596, 708,

1009, 1669, 3204, 3311,

3425, 3736, 4011

417, 650, 781, 998,

3910, 5050,

6179, 6270, 7409

418, 587, 948,

5458, 5823

202, 218, 338, 432,

524, 551, 682, 1154,

1446, 1944,

5022, 5797

1455, 2102, 3663 88, 93, 223, 267,

611, 5060, 6241

164, 191, 211, 506,

660, 2372, 4360,

4791, 5463, 5637

156, 179, 202,

291, 398, 419,

513, 668, 1956,

2012, 4376,

5050, 6331, 6617,

6847, 7007

E1 binding site

A(A/T)GATTGTTGTTAACAAT

630

AAGATTGTTGTT-

ATAAT

822

AAGCTTGTTGTTG-

TTGTCTAC

771

AAGATTGTTGT-

GGACCAC

717

TAGATTGTTGTTG-

TTAAC

562

AAGATTGTTATT-

AGGAAT

566

ATGATTGTGGTT-

AACAAC

695

AACATTGTTGTT-

GTCAACT

703

CGATTGTTGTT-

GTTCAC

Polyadenylation sites

ATAAA

59a, 3109, 5956a 154a, 3137, 4631b 43a 22, 3141, 7544a 22, 4581a, 6706a 45 23a, 3110, 3173,

6514, 7126a

24a, 3374, 7047a

Sp1 binding sites

GGCGGG

1476, 1483, 4015, 4342,

5536, 5591, 6773

624 (LCR); 4697,

4879, 5440, 6859

615 (LCR); 5337,

5559, 5810, 7353

181 (LCR); 2711,

5068, 5107, 5594

635 (LCR); 1721,

2706, 5593, 5996,

6812

399 (LCR); 1806,

3946, 4171

1936, 4607, 5577 815 (LCR); 1970,

4449, 5359

NF1 binding sites

CGGAA

327, 2283, 3154,

3313

1365, 4211 3536, 3934, 4051,

4205, 4525, 6065

1157, 1483, 3742,

6387

196, 893,1002,

2183, 3821

2038, 3831, 4300

4854, 6316, 7047

1235,1299, 3636,

4682, 5620

380, 2009, 4042,

4703, 6646, 7030

AP1 binding sites

TGANTCA

– 4272, 5784 727, 673 – – 524 247, 651, 375,

1429

342, 1457

-(25-35) tata signals in

LCR- TATAAA or

TATA(A/T)A(A/T)

697 872 – 41 – – – 117

Predicted aa-feature

Retinoblastoma tumour

suppressor binding

domain in E7 (LXCXE)

– – – – – – – –

ATP-dependent helicase

motive in E1

GPSDTGKS

1342 (448)

GPPNTGKS

1327 (443)

GPPNTGKS

1363 (455)

GPPNTGKS

1327(443)

GPPNTGKT

1357(453)

GPPDTGKS

1372(458)

GPPDTGKS

1315(439)

GPPDTGKS

1330(444)

Cyclin interaxion RXL

motive in E1, KRRLF

325 (109) 328 (110) 331 (110) 328(110) 313(105) 334(112) 304(102) 310(104)

Metal-binding motifs in E6

(CX2CX28-30CX2C)

– 34(12)-144(48)

85(253)-122(366)

256(86)-366(122)

481(169)-591(197)

61(21)-171(57)

280(94)-390-(130)

28(10)-138(46)

250(84)-360(120)

28(10)-138(46)

250(84)-360(120)

55(19)-165(55)

274(92)-384(128)

67(23)-177(59)

286(96)-396(132)

Metal-binding motifs in E7

(CX2CX28-30CX2C)

169(57)-277 (93) – 208(70)-318(106) 139(47)-249(83) 118(40)-228(76) 115(39)-225(75) 133(45)-243(81) 133(45)-243(81)

Nuclear localization signal

in L2

KRRRR

1585(529)

KQRRK

1498(500)

RKRKRR

1465(489)

RRRRKR

1441(481)

RRKRR

1426(476)

RRKRR

1414(472)

RRKRKR

1453(485)

RRRRKR

1417(473)

Nuclear localization signal

in L1

KKRR

1516(506)

RRRKRR

1396(466)

KKRK

1483(495)

KRRRK

1480(494)

KRRKR

1471(491)

KRRRK

1537(513)

KRRRK

1480(494)

KRKRK

1489(497)

a Is followed by a CA dinucleotide and a G/T cluster.b Is a double polyadenilation signal.

* For all sequences nucleotide positions are numbered starting from the first nucleotide following the stop codon of the L1 ORFs.

R.

Lecis et

al.

/ V

eterina

ry M

icrob

iolo

gy

17

0 (2

01

4)

21

3–

22

3

21

9

R. Lecis et al. / Veterinary Microbiology 170 (2014) 213–223220

clustered the PVs in their respective genera, according tothe new PV classification (Bernard et al., 2010). Also, themain genera associations observed by Gottschling et al.(2011) were maintained. In particular, the Artiodactyla –infecting Delta and Epsilon viruses grouped together withthe PVs infecting Equidae (Zeta, Dyoiota, Dyorho). EaPV1originates as a separate branch at the root of theDelta + Epsilon papillomavirus lineage, related to butdistinct from the two equine clades including respectivelythe Dyoiota (EcPV2, EcPV4, EcPV5) and Dyorho (EcPV3,EcPV6, EcPV7) papillomaviruses. EcPV1 (Zeta) seems to bepart of the Dyoiota clade, and does not cluster with theDyorho viruses, as previously observed by Lange et al.(2013b). However, this later observation is not supportedby bootstrap values, which are under 50%. Therefore, thehypothesis of a monophyletic origin for EcPV1 + DyorhoPVs is not statistically relevant.

Ten out of 18 white and brown Asinara donkeys (55%)were positive to EaPV1 when skin DNA extractions weretested by traditional PCR. The viral loads found in skinsamples of PCR-positive donkeys, evaluated by quantita-tive real time PCR, ranged from 3.46 to 270 copies/ml(Table 3).

Standard transformation assays (Fig. 4) showed thatthere was no statistically significant difference in thenumber of foci counted in transformed 208F cells whenplates transfected with pcDNA3.1(+)/EAPV/LCR-E7 werecompared to plates transfected with pcDNA3.1(+) only,used as control plasmid. Also, the expression of EaPV1 LCR-E7 did not seem to enhance the transforming activity of thewild-type Ha-ras when 208F were cotransfected withpcDNA3.1(+)/EAPV/LCR-E7 and pCMV/RAS. However,

foci of transformed cells obtained by transfecting rat208F cells with pcDNA3.1(+)/EAPV/LCR-E7 seemed to bemore differentiated when compared to foci in cellstransfected with pcDNA3.1(+) only, in which fewer andmore dispersed transformed cells were present.

4. Discussion

According to the Papillomavirus Episteme (PAVE),155 different human papillomaviruses (HPV) have been

Table 3

Prevalence and quantification of EaPV1 in skin of Asinara donkeys.

Donkey ID Breed Sex Traditional PCR Real time PCRa

1 White M + 1.47E+01

2 White F + 4.21E+01

3 Grey F + 3.46E+00

4 Grey M � ntb

5 White F � nt

6 White M � nt

7 White F + 1.39E+01

8 Grey M + 2.70E+02

9 White M � 1.12E+02

10 White F � 1.75E+01

11 White F � nt

12 Grey F + 3.15E+01

13 Grey F + 2.69E+01

14 Grey F + 2.44E+02

15 Grey F � nt

16 Grey M � nt

17 White F + 1.87E+01

18 White F + 4.69E+01

a Number of genome copies/ml DNA.b Not tested.

Fig. 4. Transformation assays in 208F cells. 208F cells were transfected with pcDNA3.1(+)/EAPV/LCR-E7 alone or in combination with pCMV/RAS as

described in the text. Micrographs show the typical appearance of transformed foci after 30 days.

idehavpoiPVsgen

potdiv

Fig.

clos

the

R. Lecis et al. / Veterinary Microbiology 170 (2014) 213–223 221

ntified to date. On the other hand, only 106 animal PVse been rescued from 52 different Vertebrate species. Asnted out by Rector and Van Ranst (2013), non-human

are distributed over 32 different genera, whether theus Gamma, Mu and Nu exclusively contain HPV types.Considering these numbers, it becomes clear thatentially a great proportion of animal papillomavirusersity has yet to be uncovered. Intensive investigation

of individual animal species is needed to improve ourknowledge about PVs evolution, which will not progresswithout a more systematic sampling of PVs diversity(Gottschling et al., 2011). Improving the knowledge aboutepitheliotropic animal PVs will help addressing manyquestions in human and veterinary medicine, such as therole played by PVs in skin cancer development. Uncoveringanimal PVs diffusion and distribution could also allow the

5. Maximum likelihood tree showing the Equidae papillomavirus diversity. Equidae PVs group in four major paraphyletic lineages (in red): the Dyorho

e – to-root lineage, the Zeta and Dyoiota lineages, and the putative novel genus EaPV1, basal to the Delta + Epsilon artiodactyl PVs.(For interpretation of

references to color in this figure legend, the reader is referred to the web version of the article.)

R. Lecis et al. / Veterinary Microbiology 170 (2014) 213–223222

investigation of animal models for the study of malignanttransformation.

In this paper we report the identification of EaPV1, thefirst papillomavirus found in donkey. Following the criteriaestablished for the classification of PVs (De Villiers et al.,2004; Bernard et al., 2010), according to which members ofthe same genus share at least 60% nucleotide sequenceidentity in the L1 ORF, EaPV1 represents the prototype of anovel PV genus. Indeed, EaPV1 L1 ORF shares 54–56%identity with any other papillomavirus of Equidae (Table 1).Cloning and full sequencing of the EaPV1 genome revealed astructure congruent with the other Equidae PVs (Fig. 2)characterised by the presence of an additional smalluntraslated region (NCR2) placed between the early andlate regions, and of the core PV ORFs E1, E2, L2, L1. On theother hand, the absence of a canonical E6 ORF represents aunique feature of EaPV1 among Equidae PVs. This featurewas also reported in bovine papillomaviruses of theXipapillomavirus genus (Rector and Van Ranst, 2013),clustering together in a lineage weakly related to EaPV1,and in general to equine PVs (Fig. 3).

Phylogenetic analysis of EaPV1 and 94 other PVs (TableS1, Fig. 3) was consistent with previous observations andconfirms the presence of four major strongly supportedPVs crown groups, as previously described (Gottschlinget al., 2011). Since several new animal PV types notconsidered by Gottschling and coworkers were included inour phylogenetic analyses, the crown groups were updatedbased on significant associations observed among Alpha,Omega, Dyodelta, Psi, Omicron, Upsilon (Alpha + Omicroncrown); Beta, Dyoeta, Dyoxy, Gamma, Tau, Pi, Phi, Xi(Beta + Xi crown); Delta, Epsilon, Zeta, Dyoiota, Dyorho(Delta + Zeta crown); Kappa, Mu, Lambda (Lambda + Mucrown).

EaPV1 confirms that equine PVs, which are paraphyleticand distributed in four main lineages (Fig. 5), arecharacterised by a broad genetic diversity. EaPV1 segre-gates in an independent branch close to the root ofDelta + Epsilon Artiodactyl PVs, this endorsing a highdegree of differentiation and the inclusion of this virusin a new genus. The high prevalence (10/18) and the lowviral loads of EaPV1 in healthy skin of donkeys sampled inthe Asinara Island (Table 3) might suggest that thisPapillomavirus does not represent a risk factor for theprogression of sun-related lesions into skin carcinoma (seeSection 3). However, further evidence is needed to supportthis hypothesis. Notably, veterinarian practitioners and/oranimal keepers have never observed benign or malignantproliferative lesions in the Asinara white donkeys’ skin(personal communication), although the breed is appar-ently prone to skin disease due to its incomplete albinism.The inability of EaPV1 LCR-E7 to transform rat fibroblastsin standard transformation assays and to enhance thetransforming activity of Ha-ras, supports that EaPV1 is alow-risk epithelial virus. The low pathogenicity of EaPV1 isalso supported by the lack of any other viral oncogene,including E6. This is particularly relevant both as anexample and model of an apparently non-pathogenicanimal papillomavirus adapted to healthy skin, and also forthe sanitary and management implications of Asinara

breeds of limited distribution’ listed by the Italianbreeder’s association, endemic to the island of Sardiniaand legally protected.

In this work, different molecular techniques were usedto identify EaPV1, the prototype of a novel genus of Equidae

PVs. The identification of EaPV1 confirms a broad diversityin Equidae PVs and suggests, similarly to what observed inhumans, a high prevalence of non-pathogenic PVs presentin small quantities on donkeys skin. Further investigationutilizing more sequence data from a larger number ofEquidae viruses and extensive epidemiological studiesincluding different species are required to fill in theexisting gaps in the PV phylogenetic tree, and to betterunderstand the coevolution of papillomaviruses withintheir hosts. Further research could also elucidate thepresence and distribution of this novel virus EaPV1 in otherdonkey breeds and populations over the Mediterraneanarea.

Appendix A. Supplementary data

Supplementary data associated with this article can befound, in the online version, at http://dx.doi.org/10.1016/j.vetmic.2014.02.016.

References

Alberti, A., Murgia, C., Liu, S.L., Mura, M., Cousens, C., Sharp, M., Miller,A.D., Palmarini, M., 2002. Envelope-induced cell transformation byovine betaretroviruses. J. Virol. 76, 5387–5394.

Alberti, A., Pirino, S., Pintore, F., Addis, M.F., Chessa, B., Cacciotto, C.,Cubeddu, T., Anfossi, A., Benenati, G., Coradduzza, E., Lecis, R., Antuo-fermo, E., Carcangiu, L., Pittau, M., 2010. Ovis aries Papillomavirus 3: aprototype of a novel genus in the family Papillomaviridae associatedwith ovine squamous cell carcinoma. Virology 407, 352–359.

Bernard, H.U., Burk, R.D., Chen, Z., Van Doorslaer, K., Hausen, H., De Villiers,E.M., 2010. Classification of papillomaviruses (PVs) based on 189 PVtypes and proposal of taxonomic amendments. Virology 401, 70–79.

Cogliano, V., Baan, R., Straif, K., Grosse, Y., Secretan, B., El Ghisassi, F., 2005.Carcinogenicity of human papillomaviruses. Lancet Oncol. 6, 204.

De Villiers, E.M., Fauquet, C., Broker, T.R., Bernard, H.U., Zur Hausen, H.,2004. Classification of papillomaviruses. Virology 324, 17–27.

Felsenstein, J., 1985. Confidence limits on phylogenies: an approach usingthe bootstrap. Evolution 39, 783–791.

Gottschling, M., Goker, M., Stamatakis, A., Bininda-Emonds, O.R., Nindl, I.,Bravo, I.G., 2011. Quantifying the phylodynamic forces driving papil-lomavirus evolution. Mol. Biol. Evol. 28, 2013–2101.

Haedicke, J., Iftner, T., 2013. Human papillomaviruses and cancer. Radio-ther. Oncol. 108, 397–402.

Hall, T.A., 1999. BioEdit: a user-friendly biological sequence alignmenteditor and analysis program for Windows 95/98/NT. Nucl. AcidsSymp. Ser. 41, 95–98.

Howie, H.L., Koop, J.I., Weese, J., Robinson, K., Wipf, G., Kim, L., Galloway,D.A., 2011. Beta-HPV 5 and 8 E6 promote p300 degradation byblocking AKT/p300 association. PLoS Pathog., http://dx.doi.org/10.1371/journal.ppat.1002211.

Howley, P.M., Lowy, D.R., 2007. Papillomaviruses. In: Knipe, D.M., Howley,P.M. (Eds.), Fields Virology. Lippincott Williams & Wikins,Philadelphia, pp. 2299–2354.

Kainzbauer, C., Rushton, J., Tober, R., Scase, T., Nell, B., Sykora, S., Brandt, S.,2012. Bovine papillomavirus type 1 and Equus caballus papilloma-virus 2 in equine squamous cell carcinoma of the head and neck in aConnemara mare. Equine Vet. J. 44, 112–115.

Lange, C.E., Tobler, K., Ackermann, M., Favrot, C., 2011. Identification oftwo novel equine papillomavirus sequences suggests three genera inone cluster. Vet. Microbiol. 21 (149) 85–90.

Lange, C.E., Tobler, K., Lehner, A., Grest, P., Welle, M.M., Schwarzwald, C.C.,Favrot, C., 2013a. EcPV2 DNA in equine papillomas and in situ and

invasive squamous cell carcinomas supports papillomavirus etiology.Vet. Pathol. 50, 686–692. white donkey breed, one of the ‘seven indigenous donkey

Lang

Mun

Mun

Nare

Nas

Page

Perr

Post

Qua

Rect

R. Lecis et al. / Veterinary Microbiology 170 (2014) 213–223 223

e, C.E., Vetsch, E., Ackermann, M., Favrot, C., Tobler, K., 2013b. Fournovel papillomavirus sequences support a broad diversity amongequine papillomaviruses. J. Gen. Virol. 94, 1365–1372.day, J.S., Dunowska, M., Hills, S.F., Laurie, R.E., 2013. Genomic char-

acterization of Felis catus papillomavirus-3: a novel papillomavirusdetected in a feline Bowenoid in situ carcinoma. Vet. Microbiol. 165,319–325.oz, N., Bosch, F.X., de Sanjose, S., Herrero, R., Castellsague, X., Shah,

K.V., Snijders, P.J., Meijer, C.J., 2003. Epidemiologic classification ofhuman papillomavirus types associated with cervical cancer. N. Engl.J. Med. 348, 518–527.chania, A., Terai, M., Chen, C.Z., DeSalle, R., Burk, R.D., 2004. Lack of

the canonical pRB-binding domain in the E7 ORF of artiodactylpapillomaviruses is associated with the development of fibropapil-lomas. J. Gen. Virol. 85, 1243–1250.ir, L., Campo, M.S., 2008. Bovine papillomaviruses: their role in theaetiology of cutaneous tumors of bovids and equids. Vet. Dermatol.19, 243–254., R.D.M., 1996. TREEVIEW: an application to display phylogenetic

trees on personal computers. Comput. Appl. Biosci. 12, 357–358.iere, G., Gouy, M., 1996. WWW-query: an on-line retrieval system forbiological sequence banks. Biochimie 78, 364–369.ey, R.C., Appleyard, G.D., Kidney, B.A., 2007. Evaluation of equinepapillomas, aural plaques, and sarcoids for the presence of Equinepapillomavirus DNA and Papillomavirus antigen. Can. J. Vet. Res. 71,28–33.de, K., 1979. Transformation of mammalian cells by avian myelocyto-matosis virus and avian erythroblastosis virus. Virology 98, 461–465.or, A., Bossart, G.D., Ghim, S.J., Sundberg, J.P., Jenson, A.B., Van Ranst,M., 2004. Characterization of a novel close-to-root papillomavirusfrom a Florida manatee by using multiply primed rolling-circleamplification: Trichechus manatus latirostris papillomavirus type 1.J. Virol. 78 (22) 12698–12702.

Rector, A., Van Ranst, M., 2013. Animal papillomaviruses. Virology 445,213–223.

Scagliarini, A., Gallina, L., Battilani, M., Turrini, F., Savini, F., Lavazza, A.,Chiari, M., Coradduzza, E., Peli, A., Erdelyi, K., Alberti, A., 2013. Cervuselaphus papillomavirus (CePV1): new insights on viral evolution indeer. Vet. Microbiol. 165, 252–259.

Sykora, S., Samek, L., Schonthaler, K., Palm, F., Borzacchiello, G., Aurich, C.,Brandt, S., 2012. EcPV-2 is transcriptionally active in equine SCC butonly rarely detectable in swabs and semen from healthy horses. Vet.Microbiol. 158, 194–198.

Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., Kumar, S., 2011.MEGA5: molecular evolutionary genetics analysis using maximumlikelihood, evolutionary distance, and maximum parsimony methods.Mol. Biol. Evol. 28, 2731–2739.

Terai, M., DeSalle, R., Burk, R.D., 2002. Lack of canonical E6 and E7 openreading frames in bird papillomaviruses: Fringilla coelebs papilloma-virus and Psittacus erithacus timneh papillomavirus. J. Virol. 76,10020–10023.

Thompson, J.D., Higgins, D.G., Gibson, T.J., 1994. CLUSTAL W: improvingthe sensitivity of progressive multiple sequence alignment throughsequence weighting, positions-specific gap penalties and weightmatrix choice. Nucl. Acids Res. 22, 4673–4680.

Whelan, S., Goldman, N., 2001. A general empirical model of proteinevolution derived from multiple protein families using a maximum-likelihood approach. Mol. Biol. Evol. 18, 691–699.

Xia, X., Xie, Z., 2001. DAMBE: data analysis in molecular biology andevolution. J. Hered. 92, 371–373.

Zhu, W., Dong, J., Shimizu, E., Hatama, S., Kadota, K., Goto, Y., Haga, T.,2012. Characterization of novel bovine papillomavirus type 12 (BPV-12) causing epithelial papilloma. Arch. Virol. 157, 85–91.

zur Hausen, H., 1999. Papillomaviruses in human cancers. Proc. Assoc.Am. Phys. 111, 581–587.