Virology 401 (2010) 314–321

Contents lists available at ScienceDirect

Virology

j ourna l homepage: www.e lsev ie r.com/ locate /yv i ro

Phenotypic effects of HPV-16 E2 protein expression in human keratinocytes

Julie E. Burns 1, Hannah F. Walker, Christian Schmitz 2, Norman J. Maitland ⁎YCR Cancer Research Unit, Department of Biology (Area 13), University of York, PO Box 373, YORK, YO10 5YW, UK

⁎ Corresponding author. Fax: +44 1904 328710.E-mail addresses: j.e.burns@leeds.ac.uk (J.E. Burns),

(N.J. Maitland).1 Present address: Section of Experimental Oncology

Medicine, St James's University Hospital, Beckett Street2 Present address: cliMECS GmbH, IZFG, Inselstraße 2

0042-6822/$ – see front matter © 2010 Elsevier Inc. Adoi:10.1016/j.virol.2010.03.002

a b s t r a c t

a r t i c l e i n f o

Article history:Received 11 May 2009Returned to author for revision14 July 2009Accepted 1 March 2010Available online 27 March 2010

Keywords:PapillomavirusHPV-16E2 proteinHaCaTDifferentiation

Expression of the HPV E2 open reading frame in cervical cancer cells has been shown to affect the expressionof both viral and cellular genes. We have examined the phenotypic effects of the expression of humanpapillomavirus 16 E2 open reading frame in the human keratinocyte cell line HaCaT. Increased levels ofapoptotic cell death were seen within 24 h of the transfection of HPV-16 E2 expression constructs. However,in those cells which survived selection and retained the intact E2 ORF, long-term stable expression of E2, asdetected by RT-PCR, produced cells which developed phenotypes typical of terminally differentiated cells.These included characteristic morphological changes and expression of involucrin, filaggrin and senescencemarkers. This provides the first evidence of a role for E2 in stimulation of the normal epithelial differentiationprogramme, which would promote the progression of the HPV life cycle.

njm9@york.ac.uk

, Leeds Institute of Molecular, Leeds, LS9 7TF, UK.2, D-04103 Leipzig, Germany.

ll rights reserved.

© 2010 Elsevier Inc. All rights reserved.

Introduction

Infection of human mucosal or squamous epithelium with papillo-maviruses results inwart-like lesions (zurHausen andDeVilliers, 1994)in which the papillomavirus (HPV) life cycle is intimately linked to thedifferentiation state of the host epithelial cells. The virus enters theepithelium throughmicroscopic lesions and infects basal keratinocytes,where the viral genome is uncoated and early functions are expressed.The “early” gene products interact with cellular factors to transcribe,maintain and replicate the viral genome in the proliferating basal layer.As the host cell differentiates, the viral genome is amplified, encapsi-dated by viral structural proteins until mature virus particles are shedwith squames from the epithelial surface.

The E2 proteins (reviewed in Burns et al., 2003) encoded bypapillomaviruses have been shown to play an essential role in thecontrol of viral replication and transcription. The E2 open readingframes (ORFs) encode nuclear proteins that bind in a sequence-specificmanner to the consensus sequence ACCGN4CGGT, which occurs severaltimes in the virus genome (Androphy et al., 1987; Dostatni et al., 1988).The E2 protein is a dimer with a tripartite structure, comprising anN-terminal transactivation module (TAD) of about 200 amino acids,which includes the transcriptional activation and replication func-

tions, a C-terminal module (CT) of about 90 amino acids containingDNA-binding and protein dimerization domains, and a flexible linkerregion of about 80 amino acids which separates the N- and C-terminalmodules (Giri and Yaniv, 1988; Gauthier et al., 1991). The amino acidsequences of N- and C-terminal modules are highly conserved betweenthe E2 proteins of different papillomaviruses; whereas the linker regionis more variable, both in sequence and length.

Throughout the vegetative replication cycle, the HPV genome ismaintained as an episome, and an intact E2 ORF is required for episomalmaintenance and high efficiency viral DNA replication (Ustav andStenlund, 1996; Piirsoo et al., 1996). However, integration into host cellDNA may occur, particularly with “high risk” HPV types such as 16 and18 (Durst et al., 1985). Integration generally disrupts the E1/E2 ORFs(Durst et al., 1985; Daniel et al., 1995) leading to dysregulatedexpression of the viral oncoproteins E6 and E7 and progression tomalignancy. Immunostaining of cervical lesions reveals that HPV-16 E2is expressed in CIN I, CIN II and koilocytes (believed to be sites of viralreplication) but expression is lost in later stages of tumour development(Maitland et al., 1998), consistent with a role for E2 in repressing HPVgene expression through the p97 promoter.

Studies examining the role of E2 in mammalian cells have generallyused transiently transfected cancer cell lines (reviewed in Burns et al.(2003)), where ectopic expression of E2 in cervical carcinoma cell linesresulted in growth suppression of lines containing HPV DNA, probablydue to repression of endogenous E6/E7 expression (Hwang et al., 1993;Dowhanick et al., 1995; Goodwin et al., 1998; Naeger et al., 1999; Moonet al., 2001). Studies in HeLa cells (which contain integrated copies ofHPV-18) indicated that transient overexpression of E2 inducesapoptosis (Desaintes et al., 1997, 1999), which was confirmed forother cell lines, both HPV-positive and -negative (Webster et al., 2000),

315J.E. Burns et al. / Virology 401 (2010) 314–321

and was shown to be both transactivation independent and mediatedthrough activation of caspase 8 (Demeret et al., 2003; Thierry andDemeret, 2008).

Subsequent studies indicated that prolonged growth arrest due tosustained expression of E2 in HeLa cells induces senescence, which hasbeen attributed to loss of E7 expression and re-activation of the Rbpathway (Goodwin et al., 2000; Wells et al., 2000; Kang et al., 2004;

Fig. 1. Cell death 24 h after transfection with various pZeoSV constructs. (A) HaCaT, HeLa (poconstructs as described in Materials and methods and unfixed cultures examined after 24 hoptics. Both EGFP-E2 and EGFP-E2 CT were exclusively expressed in the nuclei of transconcentrated in a perinuclear ring. Arrows show dead and dying green cells. All images w(B) pZeoSV-E2 transfected HaCaT cells were stained 24 h after transfection by addition of proby phase contrast and fluorescent microscopy. PI-positive staining was mostly in floatintransactivation as measured by EGFP expression 45 h after co-transfection with pZeoSV conwas used as a positive control. Bars show mean and standard deviation of three experimen

Johung et al., 2007). However, since HeLa cells are incapable of terminaldifferentiation, it is possible that irreversible senescence may be adefault pathway for cells which exit the cell cycle. The object of thepresent study was to stably transfect a p53-negative, HPV-negative,differentiation-competent, keratinocyte cell line (HaCaT) with HPV-16E2 and examine the effects of E2 expression on cell growth andmorphology.

sitive control) and LNCaP (negative control) cells were transfected with EGFP-E2 fusionby inverted microscopy. Images show same cells under fluorescence and phase contrastfected cells (visible in phase contrast images). EGFP-E2 in HaCaT and HeLa becameere captured using the same magnification and exposure settings. Scale bar=40 μm.pidium iodide (PI) directly into the medium and unfixed cultures viewed within 15 ming cells (not visible) or detaching, dying cells. Scale bar=200 μm. (C) E2-mediatedstructs and E2-response plasmid, normalized to pZeoSV:βgal level=1. pLNCX-FLAG-E2ts.

316 J.E. Burns et al. / Virology 401 (2010) 314–321

Results

HaCaT cells were transfected with an efficiency of 10–15% (asassessed by EGFP expression from pCMV-EGFP 24 h after transfec-tion). Before selection was initiated, approximately 5-fold higherlevels of cell death were seen in cultures transfected with full-lengthand C-terminal E2 vectors compared to N-terminal E2 and emptyvector or β-galactosidase controls (Fig. 1). Similar results were foundusing DAPI permeability and annexin V binding assays (data notshown) and propidium iodide staining of unfixed cells (Fig. 1B).Transient transactivation assays indicated expression of full-length E2protein at levels sufficient to transactivate and E2 responsivepromoter 2–3-fold (Fig. 1C). Due to the low transfection efficiencyof HaCaT cells, nuclear changes and cell death were monitored inindividual cells using EGFP-E2 fusion constructs (Fig. 1A); however,these constructs were not used for stable transfections, since over-expression of EGFP alone can cause apoptosis in a significantpercentage of transfected HaCaT cells (data not shown) thuspotentially confounding any effects of E2, even though EGFP-E2fusion proteins have been reported not to have enhanced stability(Bellanger et al., 2001).

Following zeocin selection for stable transfectants, it was observedthat cells transduced with E2 constructs showed evidence of a

Fig. 2. (A) Representative phase contrast images of unfixed transfected cell cultures (passadensity cultures, right image shows confluent culture. E2-transfected cultures (zeoE2FL and zto empty vector (zeo) which had more typical basal epithelial cell morphology. Scale bars=identified by bright field microscopy. Images show the same field under bright and phase cocontrol (zeo) and E2 transfectant cultures (2 FL and 3 CT clones). Graph shows mean and s

differentiated morphology, including elongated cells, squame-likecells and multi-layered colonies, even though the cells had beenmaintained in calcium-free medium (Fig. 2A). These morphologieswere not seen in empty vector or LacZ transfectants and were similarin appearance to HaCaT cells stimulated to differentiate by calcium(Boukamp et al., 1988; Fig. S1). The altered morphology was morepronounced in clonal cultures and was seen in cells with or withoutcontinued zeocin selection, therefore the selection agent was notresponsible. Cells were examined by immunocytochemistry forexpression of markers of epidermal differentiation. Involucrin,filaggrin and cytokeratins 1 and 10 were all expressed at high levelsin E2-transfected cells, particularly in morphologically differentiatedareas (Fig. 3). Expression of all markers was highest in areas of multi-layered, differentiated morphology, particularly in the larger,squame-like cells. Cells transfected with empty vector or LacZ showedlittle evidence of multi-layering or differentiation, lower saturationdensity and only low levels of expression of differentiationmarkers, asis normal for HaCaT cells in the absence of exogenous differentiationstimuli (Fig. 3). Since E2 is reported to induce senescence in HeLa cells(Goodwin et al., 2000) and our transfected cells showed morpholog-ical similarities with senescent cells, we looked at expression ofsenescence-activated β-galactosidase (SA-βgal) (Dimri et al., 1995).Both E2 FL and E2CT transfectants showed areas of increased

ges 6 (E2 FL), 7 (E2 CT), or 9 (zeo)). Left image of each pair shows morphology of loweoE2CT) showed higher numbers of elongated cells and stratified areas when compared200 μm. (B) Staining for SA-β-gal. Blue-stained senescent cells (passages 8–10) werentrast illumination. Scale bars=100 μm. (C) Comparison of SA-β-gal positive cells in 1tandard deviation of positive staining cells in random fields.

Fig. 3. Immunodetection of epidermal differentiation markers, filaggrin, involucrin, and cytokeratins 1 and 10 (CK1, CK10) in stably transfected and HaCaT parental cells. Imagesshow the same field with fluorescent and phase contrast illumination. Expression of all markers was highest in E2-transfected cells, in areas of multi-layered, differentiatedmorphology, particularly in large, squame-like cells. Empty vector (zeo) transfectants showed very low levels of all markers even when cells were densely packed. For each antibody,all images were captured at the same exposure settings. Scale bars 40 μm.

317J.E. Burns et al. / Virology 401 (2010) 314–321

expression of SA-β-gal when compared to cells transfected withempty vector or parental HaCaT under similar culture conditions(Fig. 2B). SA-βgal-positive cells tended to be cells with elongated

morphology or in differentiated areas. Overall levels of senescent cellswere similar in E2FL and E2CT transfected cells (Fig. 2C) but varied incultures from a few positive cells per field to over 40%.

Fig. 4. (A) PCR detection of E2 ORF in genomic DNA of transfectants (clones and pools)at various passage numbers (shown above each lane). M=size marker lane. E2 NT:246 bp E2 N-terminal fragment amplified by primers specific for SV40 promoter ofpZeoSV and bp 2840–2819 of HPV-16 genome. E2 CT: 190 bp E2 C-terminal fragment,amplified using primers E2MC and E2C, corresponding to HPV-16 bp 3529–3719. HPRT:267 bp fragment amplified by human HPRT gene-specific primers, demonstrating thepresence of intact human DNA. (B) RT-PCR detection of E2mRNA in late passage (6–10)transfectants. +/−=with or without reverse transcriptase in first-strand reaction.E2 CT: PCR for E2 C-terminal fragment, as panel A above, using 50% of first-strand cDNAproduct as template. G3PDH: PCR for 693 bp glyceraldehyde 3-phosphate dehydroge-nase exon 2–8 fragment (positive control for cDNA synthesis) using 10% of first-strandcDNA template. Gel images were captured using GeneSnap (Syngene) software andmounted with Adobe Photoshop 6.

318 J.E. Burns et al. / Virology 401 (2010) 314–321

During the transition to malignancy, the HPV-16 genomefrequently becomes integrated into the host cell chromatin withdisruption of the E2 ORF. To check that the E2 ORFs were still intact instable transfectants, PCR was performed on genomic DNA. As shownin Fig. 4, the appropriate E2 ORF fragments were obtained fromtransfected cells up to at least passage 8, with no evidence ofdisruption. Cells remained resistant to zeocin for at least six passageswithout selection, suggesting that the zeocin resistance gene had alsoremained intact. Cells at various passages were screened forexpression of E2 by western blotting, probing with anti-HPV-16 E2SCT rabbit polyclonal antiserum (Stevenson et al., 2000). No evidencefor E2 protein expression was seen by this method. This was notunexpected given that E2 expression in human cells at levels detectablebywestern blotting is usually highly toxic to cells and indirect assays areusually required. However, reverse transcriptase-PCR (RT-PCR) indi-cated that the E2 ORF was actively transcribed (Fig. 4), although wewere unable to detect positive transactivation of an E2-responsivepromoter by transactivation assays of the stable transfectants aswe haddone in the transient phase of transfection (24–72 h). This may havebeen because the transfection efficiency, by the reporter plasmid, of thestable E2 transfectants was very low (≤3%) and no significantexpression of EGFP above background could be detected.

Discussion

The aim of the present study was to study the outcome ofexpression of levels of HPV-16 E2 protein that permit cell growth inthe natural host cell for viral infection. This was achieved by

generation of stably transfected keratinocyte cell lines whichexpressed HPV-16 E2 using the SV40 early promoter to driveexpression. The human keratinocyte cell line HaCaT was chosen forthis study because it is immortal, non-tumorigenic, contains noendogenous HPV sequences and is able to grow and differentiatephenotypically as normal keratinocytes in the presence of 1.4 mMcalcium (Boukamp et al., 1988). For growth and selection, cells weregrown in low calciummedium, where the only calcium was from FCS,which maintains HaCaT cells in an undifferentiated state.

The results showed that low levels of E2 expression were capableof triggering keratinocytes to terminally differentiate. In the contextof the virus life cycle, this was not unexpected since, although thevirus requires the cell to remain cycling in order establish an infectionand amplify its genome, it also needs the host cell to re-enter thedifferentiation process in order for the life cycle to be completed.Hence expression of E2 may not only influence expression of E6 andE7 and thus affect cell growth, but also stimulate differentiation. SinceHaCaT cells contain no endogenous HPV (Boukamp et al., 1988), itcannot be that the effect of E2 was simply due to suppression of thegrowth-stimulatory action of E6 and E7 as in HeLa cells (Goodwin andDiMaio, 2000) rather it is likely that this is a direct effect by E2 oncellular gene expression. Indeed, the HPV-8 E2 protein has beenshown to directly suppress the cellular β4-integrin promoter inprimary keratinocytes (Oldak et al., 2004). HPV-8 E2 also interactswith C/EBP to synergistically activate the involucrin promoter inkeratinocytes (Hadaschik et al., 2003) independently of the presenceof specific E2BS or p53, while E2, C/EBP and p300 are known to co-operate during terminal differentiation of keratinocytes (Müller et al.,2002). The C/EBP-α and -β interaction domains were mapped to theC-terminus of E2, consistent with our results where the C-terminalfragment of E2 was sufficient for induction of differentiation.

Overexpression of E2 in HPV-18 positive HeLa cells leads to earlyapoptosis of a fraction of cells and irreversible senescence withinthree days, directly linked to repression of E6 and E7 and consequentreactivation of the p53 and pRB pathways (Desaintes et al., 1999;Goodwin and DiMaio, 2000; Wells et al., 2003). In HPV-16-infectedcervical carcinoma cells with intact p53 and Rb pathways, Moon et al.(2001) found no evidence of E2-induced apoptosis but ‘enlarged andflattened’ cells. In HPV-16 positive SiHa cells, expression of HPV-16 E2induced early apoptosis (Sanchez-Perez et al., 1997). In the presentstudy, increased expression of SA-β-gal was seen in areas ofdifferentiated morphology but this may not be a direct consequenceof E2. Normal terminal differentiation of keratinocytes does notinvolve senescence (Gandarillas, 2000); however, HaCaT cells cannotcompletely differentiate in monolayer culture (Boukamp et al., 1988)and may thus be expected to senesce under these conditions. Thiscontrasts with the widespread rapid senescence seen when E2FL wasexpressed after infection of HeLa, CaSki and HT-3 cells with andwithout Ad5-E2 expression vector (Goodwin et al., 2000). This effectwas attributed to loss of E7 expression (Kang et al., 2004). Since HeLacells are incapable of terminal differentiation, irreversible senescencemay be a default pathway for cells which exit the cell cycle. It is alsolikely that the E2 expression levels using adenovirus vectors weretransient and considerably higher than in the present study.

The levels of cell death seen in E2 transfected cells 24–48 h post-transfection, i.e. the transient phase before transgene integration,indicated that HPV-16 E2 was also capable of inducing apoptosis inHaCaT cells. Although it was previously shown (Webster et al., 2000)that E2-induced apoptosis in HPV-ve cells was p53-dependent, Parishet al. (2006) showed that HPV-16 E2 was able to induce apoptosis inHPV-ve, p53-null Saos-2 cells in the presence of exogenouslyexpressed mutant p53s now able to interact with E2, but unable tobind DNA or activate transcription. Although considered to be p53negative, HaCaT carries two p53 mutations (H179Y and R282W) onseparate alleles (Lehman et al., 1993). In a yeast functional assay,HaCaT p53 was found to fall into the heterozygous wild type/mutant

319J.E. Burns et al. / Virology 401 (2010) 314–321

range (Jia et al., 1997). Chylicki et al. (2000) showed that p53 induceddifferent pathways for apoptosis and differentiation, with differentia-tion requiring lower levels of p53 and possibly being transcription-independent. Certainly HaCaT cells are capable of p53-independentapoptosis, since irradiation in thepresence or absenceof E6andE7 led toapoptotic cell death without any change in p53 or p21 levels (Magal etal., 1998). However, blocking HaCaT p53 can partially block apoptosis(Henseleit et al., 1997), suggesting that HaCaT p53 retains some activity,but that both p53-independent and -dependent apoptosis pathways arepresent. Our data indicate that induction of apoptosis and thendifferentiationwas resident in the C-terminus of theHPV-16 E2 protein,the region that interacts with p53 (Parish et al., 2006). Overexpressionof HPV-18 E2 in HeLa cells was shown to trigger apoptosis via theextrinsic pathway by induction of caspase 8 (Desaintes et al., 1999;Demeret et al., 2003), but thiswas a functionof nuclear expressionof theTAD, which was lacking in our C-terminal constructs.

Therefore we propose that HPV-16 E2 has multi-functionalproperties, which can influence the normal viral life cycle inkeratinocytes. When expressed at levels that allow for cell growth/survival in a cell capable of terminal differentiation, it acts not only topromote genome expression and replication but also to promote thedifferentiation programme which is essential for the completion ofthe HPV life cycle, and ultimately the assembly of viral particles.

Materials and methods

Cell culture

HaCaT cells were used throughout this study. This immortalizedcell line was derived from skin keratinocytes, contains no endogenousHPV sequences and is capable of differentiation (Boukamp et al.,1988). Cells were routinely maintained in a calcium-free formulationof 1:1 Dulbecco's modified Eagle medium: Ham's F12 (DMEM:F12)[custom-made formulation of Cat. no. 21331, Invitrogen] supplemen-ted with 10% foetal calf serum (FCS; PAA) and split weekly when justconfluent at a ratio of 1:20. HeLa cells were grown in DMEMsupplemented with 10% FCS. LNCaP cells (derived from a humanprostate carcinoma) weremaintained in RPMI containing 10% FCS andsplit 1:5 when around 80% confluent. For calcium-induced differen-tiation cells were transferred to medium containing 1.4 mM (normalDMEM:Ham's F12) or 5 mM Ca2+.

Both cell lines and stable transfectants were checked on a regularbasis for mycoplasma infection using the Stratagene mycoplasmadetection kit and found to be consistently negative. For transfection,approximately 1.5×105 cells were plated per well in 6-well dishes andtransfected with 1 μg plasmid DNA when 70–80% confluent (24–48 h)using either Fugene6 (Roche) or ESCORT (Sigma) according tomanufacturers’ protocols. To obtain stable transfectants, cells werepassaged24–48 hafter transfection into2×10 cmdishes, selectionwithzeocin (125 μg/ml, Invitrogen) was started 24 h later and mediumreplaced twice weekly. Stably transfected cultures were maintainedeither aspools or ring-cloned after 2–3 weeksof selection. Inmost cases,zeocin selection was maintained through all subsequent passages,although some established cultures were removed from selection andmaintained in zeocin-freemedium. Transfected cellswere split at a ratioof 1:10 when 80–95% confluent.

Plasmids

The vector pZeoSV (Invitrogen) allows constitutive expression ofcloned genes in mammalian cells under the control of the SV40promoter, whilst expressing resistance to the antibiotic zeocin underCMV promoter control. The ORFs encoding full-length E2 (E2FL,aa 1–365) or E2 C-terminus (E2CT, aa 251–365) were inserted intothe SpeI site of the vector pZeoSV (Invitrogen). pZeoSVlacZ, whichencodes β-galactosidase, or the empty pZeoSV vector, was used as a

negative control. Transient transfection efficiency was monitoredby transfection of cells with plasmid pCMV:EGFP which expressesenhanced green fluorescent protein (EGFP) under the control of theCMV promoter (modified from pEGFP-1 (Clontech) by insertion ofthe CMV promoter from pCR3 (Invitrogen)) (Schmitz, 2000).

To generate EGFP-E2 fusion constructs under the control of the CMVpromoter, the open reading frame (ORF) encoding green fluorescentprotein (EGFP) was PCR-amplified from pEGFP-1 (Clontech) withoutincluding the TAA stop codon and inserted into the retroviral expressionvector pLNCX (Clontech) to generate pLNCX-EGFP. The full-length,wild-type HPV-16 E2 ORFwas inserted in framedownstreamof EGFP togenerate pLNCX-EGFP-E2. pCMV-EGFP-E2CT was generated by insert-ing the truncated E2 C-terminal ORF (aa 277–365) downstream, and in-frame with an EGFP ORF lacking a STOP codon in pCMV-EGFP.

PCR

DNA was extracted from stably transfected cell pellets by lysis in10 mM Tris–HCl pH 8, 100 mM EDTA, 20 μg/ml RNAase A, 0.5% SDSfollowed byovernight digestion at 55 °Cwith proteinase K (0.1 mg/ml),phenol/chloroform extraction and ethanol precipitation. Interferingagents were removed by washing in Microcon 30 (Amicon) concen-trators. One hundred nanograms of genomic DNA was used as atemplate for PCR using Immolase Taq polymerase (BioLine) andprimers specific for the E2FL and E2CT ORFs, located both within theORF andwithin the upstream SV40 sequence. Primer sequences wereE2C: 5′-CAGAGACTCAGTGGACAGTG-3′ (HPV-16 bp 3529–3548);E2MC: 5′-CCAATGCCATGTAGACG-3′ (HPV-16 bp 3718–3702); E2F:5′-GGTCACGTAGGTCTGTACTATC-3′ (HPV-16 bp 2840–2819); SV40promoter: 5′-GCCCAGTTCCGCCCATTCTCC-3′ (SV40 promoter(pZeoSV). The integrity of genomic DNA was checked by amplifica-tion of a 267 bp fragment of human hypoxanthine guanine phos-phoribosyl transferase (HPRT) with the primers described inMaitland et al. (1998). Cycling conditions were 94 °C (10 min)followed by 35 cycles of 94 °C (1 min), 55 °C (1 min), 72 °C (2 min)and final extension of 72 °C (7 min). PCR products were examined byeither agarose or Visigel (Stratagene) electrophoresis.

Total RNA was prepared using RNA-bee (AMS Biotechnology),treated with RNase-free DNase (Promega) and 1 μg reverse-tran-scribed using oligo dT primer and Superscript II reverse transcriptase(Invitrogen). Second-strand cDNA synthesis was performed using 10%of the reverse transcribed product and products analyzed by agaroseor Visigel electrophoresis. Amplification of exons 2–8 of humanglyceraldehyde-3 phosphate dehydrogenase (G3PDH) was used as apositive control for RNA amplification, using primers 5′-AAGGT-GAAGGTCGGAGTCAA-3′ and 5′-GGACACGGAAGGCCATGCCA-3′.

For quantitative RT-PCR, total RNA was extracted using theRNAeasy method (Qiagen) and 1 μg total RNA was reverse-transcribed using random hexamer primers. Second-strand synthesiswas carried out in triplicate in an ABI 7000 thermal cycler using30 ng of the first-strand product, Power SYBR Green Master Mix(Applied Biosystems) and 10 μM of each forward and reverseprimers. Amplification of HPRT was used as an endogenous control.Primers for involucrin (exon 1) were 5′-TCCTCCTCCAGTCAATACCC-3′and 5′-GCTGATCCCTTTGTGTT-3′. Primers for CK10 (exons 3/4–5)were 5′-ACGAGGAGGAAATGAAAGAC-3′ and 5′-GGACTGTAGTTC-TATCTCCAG-3′. HPRT primers were 5'-GATGATGAACCAGGTTAT-GACC-3' and 5'-CCAAATCCTCAGCATAATGATTAGG-3'.

Cell death

To identify potentially apoptotic cells, unfixed cells were stainedwith 4′,6-diamidine-2-phenylindole-dihydrochloride (DAPI; Sigma)or propidium iodide (Sigma) by adding 1 μl stain solution directly tothe medium, and examined by fluorescence microscopy within15 min (Waller et al., 2000). To label with annexinV-ALEXA568

320 J.E. Burns et al. / Virology 401 (2010) 314–321

(Roche), cells were rinsed twice with PBS, and incubatedwith a 1/200dilution of annexin in 10 mM HEPES, 140 mM NaCl, 5 mM CaCl2. Tento 30 min later, cells were examined on an inverted UV fluorescentmicroscope for staining andmorphology. In order to compare levels ofcell death, random fields were captured and the percentage of labeledcells was counted.

To directly monitor cellular and nuclear morphology following E2transfection, cells were transfected with EGFP-E2 fusion constructs asdescribed above and examined by fluorescence microscopy.

Detection of senescent cells

Senescent cells were detected by staining for acidic β-galactosi-dase according to the method of (Dimri et al., 1995). Cells were fixedwith 3% formaldehyde in PBS, washed with PBS and stained overnightat 37 °C in 40 mM citric acid/sodium phosphate buffer, pH 6.0, 5 mMpotassium ferricyanide, 5 mM potassium ferrocyanide, 150 mMsodium chloride, 2 mM magnesium chloride, 1 mg/ml X-gal. Blue-stained (positive) cells were identified by bright field microscopy.Senescent cells were quantitated by capturing 4 random fields perculture and counting the numbers of blue and non-blue cells.

Immunohistochemistry

For immunohistochemistry, cells were grown on chamber slides,fixed with 4% paraformaldehyde in PBS and hybridized with antiseraagainst involucrin (Sigma), CK10 (AbCam) or filaggrin (Biogenesis)overnight at 4 °C followed by washing and labeling with fluorescentlylabeled secondary antibody (anti-mouse ALEXA488 (Molecular Probes)).For labeling with CK1 (AbCam), antigen retrieval was performedaccording to manufacturer's recommendations prior to hybridization.Cells were mounted in Vectashield and examined by fluorescencemicroscopy.

Transactivation assay

Transactivation assays were carried out as previously described(Hernandez-Ramon et al., 2008). Cells were co-transfected intriplicate in 6-well plates with 850 ng pZeoSV:E2FL, pZeoSV:E2CT orpZeoSVlacZ and 850 ng per well of E2 response plasmid p2x2E2BS-SV40min-EGFP. Cells were harvested 45 h after transfection and EGFPexpression was measured by FACS analysis.

Acknowledgments

We thank Yorkshire Cancer Research for funding this research andN. Fusenig, German Cancer Research Centre, Heidelberg, Germany, forproviding the HaCaT cell line.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.virol.2010.03.002.

References

Androphy, E.J., Lowy, D.R., Schiller, J.T., 1987. Bovine papillomavirus E2 trans-activatinggene product binds to specific sites in papillomavirus DNA. Nature 325, 70–73.

Bellanger, S., Demeret, C., Goyat, S., Thierry, F., 2001. Stability of the human papillomavirustype 18 E2 protein is regulated by a proteasome degradation pathway through itsamino-terminal transactivation domain. J. Virol. 75, 7244–7251.

Boukamp, P., Petrussevska, R.T., Breitkreutz, D., Hornung, J., Markham, A., Fusenig, N.E.,1988. Normal keratinization in a spontaneously immortalized aneuploid humankeratinocyte cell line. J. Cell Biol. 106, 761–771.

Burns, J.E., Burn, C.L., Hernandez-Ramon, E.E., Schmitz, C., Lloansi Vila, A., Maitland, N.J.,2003. Forms and functions of papillomavirus E2 proteins. Recent Res. Dev. Biochem.4, 795–823.

Chylicki, K., Ehinger, M., Svedberg, H., Gullberg, U., 2000. Characterization of themolecular mechanisms for p53-mediated differentiation. Cell Growth Differ. 11,561–571.

Daniel, B.,Mukherjee, G., Seshadri, L., Vallikad, E., Krishna, S., 1995. Changes in the physicalstate and expression of human papillomavirus type 16 in the progression of cervicalintraepithelial neoplasia lesions analysed by PCR. J. Gen. Virol. 76, 2589–2593.

Demeret, C., Garcia-Carranca, A., Thierry, F., 2003. Transcription-independent triggeringof the extrinsic pathway of apoptosis by human papillomavirus 18 E2 protein.Oncogene 22, 168–175.

Desaintes, C., Demeret, C., Goyat, S., Yaniv, M., Thierry, F., 1997. Expression of thepapillomavirus E2 protein in HeLa cells leads to apoptosis. EMBO J. 16, 504–514.

Desaintes, C., Goyat, S., Garbay, S., Yaniv, M., Thierry, F., 1999. Papillomavirus E2 inducesp53-independent apoptosis in HeLa cells. Oncogene 18, 4538–4545.

Dimri, C.P., Lee, X., Basile, G., Acosta, M., Roskelley, C., Medrano, E.E., Linskens, M.,Rubelj, I., Pereira-Smith, O., Peacocke, M., Campisi, J., 1995. A biomarker thatidentifies senescent human cells in culture and in aging skin in vivo. Proc. Natl.Acad. Sci. U. S. A. 92, 9363–9367.

Dostatni, N., Thierry, F., Yaniv, M., 1988. A dimer of BPV-1 E2 containing a proteaseresistant core interacts with its DNA target. EMBO J. 7, 3807–3816.

Dowhanick, J.J., McBride, A.A., Howley, P.M., 1995. Suppression of cellular proliferationby the papillomavirus E2 protein. J. Virol. 69, 7791–7799.

Durst, M., Kleinheinz, A., Hotz, M., Gissmann, L., 1985. The physical state of humanpapillomavirus type 16 DNA in benign and malignant genital tumours. J. Gen. Virol.66, 1515–1522.

Gandarillas, A., 2000. Epidermal differentiation, apoptosis and senescence: commonpathways? Exp. Gerontol. 35, 53–62.

Gauthier, J.M., Dillner, J., Yaniv, M., 1991. Structural analysis of the humanpapillomavirus type 16 E2 transactivator with antipeptide antibodies reveals ahigh mobility region linking the transactivation and the DNA-binding domains.Nucleic Acids Res. 19, 7073–7079.

Giri, I., Yaniv, M., 1988. Structural and mutational analysis of E2 trans-activatingproteins of papillomaviruses reveals three distinct functional domains. EMBO J. 7,2823–2829.

Goodwin, E.C., DiMaio, D., 2000. Repression of human papillomavirus oncogenes inHeLa cervical carcinoma cells causes the orderly reactivation of dormant tumorsuppressor pathways. Proc. Natl. Acad. Sci. U. S. A. 97, 12513–12518.

Goodwin, E.C., Naeger, L.K., Breiding,D.E., Androphy, E.J., DiMaio, D., 1998. Transactivation-competent bovine papillomavirus E2 protein is specifically required for efficientrepression of human papillomavirus oncogene expression and for acute growthinhibition of cervical carcinoma cell lines. J. Virol. 72, 3925–3934.

Goodwin, E.C., Yang, E., Lee, H.-W., DiMaio, D., Hwang, E.-S., 2000. Rapid induction ofsenescence in human cervical carcinoma cells. Proc. Natl. Acad. Sci. U. S. A. 97,10978–10983.

Hadaschik, D., Hinterkeuser, K., Oldak, M., Pfister, H.J., Smola-Hess, S., 2003. Thepapillomavirus E2 protein binds to and synergizes with C/EBP factors involved inkeratinocyte differentiation. J. Virol. 77, 5253–5265.

Henseleit, U., Zhang, J., Wanner, R., Haase, I., Kolde, G., Rosenbach, T., 1997. Role of p53in UVB-induced apoptosis in human HaCaT keratinocytes. J. Invest. Dermatol. 109,722–727.

Hernandez-Ramon, E.E., Burns, J.E., Zhang,W.,Walker,H.F., Allen, S., Antson,A.A.,Maitland,N.J., 2008. Dimerization of the human papillomavirus type 16 E2N terminus results inDNA looping within the upstream regulatory region. J. Virol. 82, 4853–4861.

Hwang, E.-S., Riese II, D.J., Settleman, J., Nilson, L.A., Honig, J., Flynn, S., DiMaio, D., 1993.Inhibition of cervical carcinoma cell line proliferation by the introduction of abovine papillomavirus regulatory gene. J. Virol. 67, 3720–3729.

Jia, L.Q., Osada, M., Ishioka, C., Gamo, M., Ikawa, S., Suzuki, T., Shimodaira, H., Niitani, T.,Kudo, T., Akiyama, M., Kimura, N., Matsuo, M., Mizusawa, H., Tanaka, N., Koyama, H.,Namba, M., Kanamaru, R., Kuroki, T., 1997. Screening the p53 status of human celllines using a yeast functional assay. Mol. Carcinog. 19, 243–253.

Johung, K., Goodwin, E.C., DiMaio, D., 2007. Human papillomavirus E7 repression incervical carcinoma cells initiates a transcriptional cascade driven by theretinoblastoma family, resulting in senescence. J. Virol. 81, 2102–2116.

Kang, H.T., Lee, C.J., Seo, E.J., Bahn, Y.J., Kim, H.J., Hwang, E.S., 2004. Transition to anirreversible state of senescence in HeLa cells arrested by repression of HPV E6 andE7 genes. Mech. Ageing Dev. 125, 31–40.

Lehman, T.A., Modali, R., Boukamp, P., Stanek, J., Bennett, W.P., Welsh, J.A., Metcalf, R.A.,Stampfer, M.R., Fusenig, N., Rogan, E.M., Harris, C.C., 1993. p53 mutations in humanimmortalized epithelial cell lines. Carcinogenesis 14, 833–839.

Magal, S.S., Jackman, A., Pei, X.F., Schlegel, R., Sherman, L., 1998. Induction of apoptosisin human keratinocytes containing mutated p53 alleles and its inhibition by boththe E6 and E7 oncoproteins. Int. J. Cancer 75, 96–104.

Maitland, N.J., Conway, S., Wilkinson, N.S., Ramsdale, J., Morris, J.R., Sanders, C.M., Burns,J.E., Stern, P.L., Wells, M., 1998. Expression patterns of the human papillomavirustype 16 transcription factor E2 in low- and high-grade cervical intraepithelialneoplasia. J. Pathol. 186, 275–280.

Moon, M.S., Lee, C.J., Um, S.J., Park, J.S., Yang, J.M., Hwang, E.S., 2001. Effect of BPV1 E2-mediated inhibition of E6/E7 expression in HPV16-positive cervical carcinomacells. Gynecol. Oncol. 80, 168–175.

Müller, A., Ritzkowsky, A., Steger, G., 2002. Cooperative activation of humanpapillomavirus type 8 gene expression by the E2 protein and the cellularcoactivator p300. J. Virol. 76, 11042–11053.

Naeger, L.K., Goodwin, E.C., Hwang, E.S., DeFilippis, R.A., Zhang, H., DiMaio, D., 1999.Bovine papillomavirus E2 protein activates a complex growth- inhibitory programin p53-negative HT-3 cervical carcinoma cells that includes repression of cyclin Aand cdc25A phosphatase genes and accumulation of hypophosphorylatedretinoblastoma protein. Cell Growth Differ. 10, 413–422.

321J.E. Burns et al. / Virology 401 (2010) 314–321

Oldak, M., Smola, H., Aumailley, M., Rivero, F., Pfister, H., Smola-Hess, S., 2004. Thehuman papillomavirus type 8 E2 protein suppresses β4-integrin expression inprimary human keratinocytes. J. Virol. 78, 10738–10746.

Parish, J.L., Kowalczyk, A., Chen, H.T., Roeder, G.E., Sessions, R., Buckle, M., Gaston, K.,2006. E2 proteins from high- and low-risk human papillomavirus types differ intheir ability to bind p53 and induce apoptotic cell death. J. Virol. 80, 4580–4590.

Piirsoo, M., Ustav, E., Mandel, T., Stenlund, A., Ustav, M., 1996. Cis and transrequirements for stable episomal maintenance of the BPV-1 replicator. EMBO J.15, 1–11.

Sanchez-Perez, A.-M., Soriano, S., Clarke, A.R., Gaston, K., 1997. Disruption of the humanpapillomavirus type 16 E2 gene protects cervical carcinoma cells from E2F-inducedapoptosis. J. Gen. Virol. 78, 3009–3018.

Schmitz, C., 2000. Transient and stable expression of the human papillomavirus type 16(HPV-16) early protein 2 (E2) in human keratinocytes. DPhil Thesis, University ofYork, UK.

Stevenson, M., Hudson, L.C., Burns, J.E., Stewart, R.L., Wells, M., Maitland, N.J., 2000.Inverse relationship between the expression of the human papillomavirus type 16transcription factor E2 and virus DNA copy number during the progression ofcervical intraepithelial neoplasia. J. Gen. Virol. 81, 1825–1832.

Thierry, F., Demeret, C., 2008. Direct activation of caspase 8 by the proapoptotic E2protein of HPV18 independent of adaptor proteins. Cell Death Differ. 15,1356–1363.

Ustav, M., Stenlund, A., 1991. Transient replication of BPV-1 requires two viralpolypeptides encoded by the E1 and E2 open reading frames. EMBO J. 10, 449–457.

Waller, A.S., Sharrard, R.M., Berthon, P., Maitland, N.J., 2000. Androgen receptorlocalisation and turnover in human prostate epithelium treated with theanitandrogen, Casodex. J. Mol. Endocrinol. 24, 339–351.

Webster, K., Parish, J., Pandya, M., Stern, P.L., Clarke, A.R., Gaston, K., 2000. The humanpapillomavirus (HPV) 16 E2 protein induces apoptosis in the absence of other HPVproteins and via a p53-dependent pathway. J. Biol. Chem. 275, 87–94.

Wells, S.I., Francis, D.A., Karpova, A.Y., Dowhanick, J.J., End, J.D.B., Benson, J.D., Howley, P.M., 2000. Papillomavirus E2 induces senescence in HPV-positive cells via pRB- andp21CIP-dependent pathways. EMBO J. 19, 5762–5771.

Wells, S.I., Aronow, B.J., Wise, T.M., Williams, S.S., Couget, J.A., Howley, P.M., 2003.Transcriptome signature of irreversible senescence in human papillomavirus-positive cervical cancer cells. Proc. Natl. Acad. Sci. U. S. A. 100, 7093–7098.

zur Hausen, H., De Villiers, E.-M., 1994. Human papillomaviruses. Annu. Rev. Microbiol.48, 427–447.