J. Comp. Path. 2010, Vol. 142, 9e18 Available online at www.sciencedirect.com

Cor

002

doi

www.elsevier.com/locate/jcpa

Ferritin Heavy Chain (FHC) is Up-regulatedin Papillomavirus-Associated Urothelial Tumours

of the Urinary Bladder in Cattle

S. Roperto*, G. Borzacchiello*, R. Brun*, F. Costanzo†, M. C. Faniello†,C. Raso†, A. Rosati‡, V. Russo*, L. Leonardix, D. Saracinok, M. C. Turco‡,

C. Urraro* and F. Roperto*

*Department of Veterinary Pathology and Animal Health, Naples University Federico II, Naples, †Department of

Experimental Medicine and Clinics, Magna Græcia University, Catanzaro, ‡Department of Pharmaceutical Sciences,

University of Salerno, Fisciano (SA), xDepartment of Biopathological Sciences and Hygiene of Animal and Food Production,

Perugia University, Perugia and kAzienda Sanitaria Locale SA/3, Vallo della Lucania, Salerno, Italy

resp

1-99

:10.1

Summary

The up-regulation of ferritin heavy chain (FHC) is reported in six papillary and in four invasive urothelial tu-mours of the urinary bladder of cattle grazing on mountain pastures rich in bracken fern. All tumours con-tained sequence of bovine papillomavirus type-2 (BPV-2) as determined by polymerase chain reaction(PCR) analyses and validated by direct sequencing of the amplified products. The oncoprotein E5 was alsodetected in these tumours by immunoprecipitation and by immunofluorescence and laser scanning confocalmicroscopy. Expression of FHC was evaluated by western blot analysis, reverse transcriptase (RT) PCR,real-time RT-PCR and immunohistochemistry. The oligonucleotide sequence of the bovine ferritin ampliconswas identical to that of human ferritin. Nuclear overexpression of p65, an important component of nuclearfactor kB (NF-kB) transcription factors, was also observed. These findings suggest that FHC up-regulationmay be mediated by activation of NF-kB and that in turn this may be related to the resistance of bovinepapillomavirus type-2 (BPV-2) infected urothelial cells to apoptosis.

� 2009 Elsevier Ltd. All rights reserved.

Keywords: cattle; ferritin heavy chain; papillomavirus; urinary bladder; urothelial tumour

Introduction

Iron is essential to many fundamental metabolic pro-cesses such as DNA synthesis and cell cycleprogression (Richardson et al., 2008). However,iron-mediated cellular toxicity is also known to occurthrough oxidative stress, resulting in the production ofreactive oxygen species (ROS) (De Domenico et al.,2008). The generation of ROS plays a crucial rolein cell signalling, but increased formation of ROScan harm cellular integrity (Galaris et al., 2008).Therefore, iron homeostasis must be strictly regulatedin order to prevent imbalanced cellular metabolismand pathological change (Gu et al., 2008).

ondence to: S. Roperto (e-mail: sante.roperto@unina.it).

75/$ - see front matter

016/j.jcpa.2009.05.009

There is emerging evidence that iron plays a role inthe development of cancer (Brookes et al., 2006). Thecarcinogenic role of iron has been attributed to its po-tential to induce ROS, DNA adducts, abnormal cel-lular proliferation and to repress cellular adhesion(Boult et al., 2008). It is believed that the accumula-tion of cellular iron may be a feature of all neoplastictransformation (Boult et al., 2008) and that increasedintracellular iron may trigger the common oncogenicWnt signalling pathway (Brookes et al., 2008; Klausand Birchmeier, 2008). Therefore, it is not surprisingthat some new therapeutic strategies for tumour ther-apy are based on the development of novel potent ironchelators (Richardson et al., 2008).

Numerous proteins are involved in iron metabo-lism and iron-binding ferritin plays a crucial role

� 2009 Elsevier Ltd. All rights reserved.

10 S. Roperto et al.

in coordinating intracellular iron homeostasis(Sammarco et al., 2008). Ferritin is a ubiquitous andhighly conserved protein and is the main molecule in-volved in intracellular iron storage in all organisms(Li et al., 2006; Zandman-Goddard and Shoenfeld,2007; Recalcati et al., 2008). The molecule is com-posed of ferritin heavy chains (FHC) and ferritin lightchains (FLC) that assemble to form a 24-subunit pro-tein (Torti and Torti, 2002; Li et al., 2006). FHC andFLC are encoded by distinct genes that are highlyconserved in all species and are differentially regu-lated at the post-transcriptional level (Sammarcoet al., 2008). The FHC has ferroxidase activity thatoxidizes ferrous iron to ferric iron, while the FLC lacksferroxidase activity but cooperates with the FHC toincrease iron uptake (Lawson et al., 1989; Torti andTorti, 2002; Orino et al., 2004). The ferroxidaseactivity is associated with the presence of seven highlyconserved amino acids (Harrison and Arosio, 1996;Orino and Watanabe, 2008). Furthermore, the ironresponsive element (IRE), a regulatory structure offerritin mRNA, is particularly conserved amongspecies (Torti and Torti, 2002).

FHC is increasingly recognized as having a role innumerous pathological processes including inflamma-tion (Kwak et al., 1995; Pham et al., 2004), autoimmu-nity (Zandman-Goddard and Shoenfeld, 2007;Sammarco et al., 2008), neurodegenerative disorders(Reis et al., 2006) and cancer (Aung et al., 2007; Boultet al., 2008; Brookes et al., 2008). New evidence isemerging that suggests an important oncogenic rolefor FHC (Brookes et al., 2006; Boult et al., 2008), butthere are still few data providing molecular explana-tions for ferritin dysregulation and these often lead toconflicting conclusions (Kwok and Richardson, 2002;Torti and Torti, 2002; Tsuji, 2005).

Disturbances in ferritinmetabolismhavenot been in-vestigated in comparative oncology. The aim of thepresent study was to determine whether up-regulationofFHCoccurs inurothelial tumours of theurinaryblad-der of cattle associatedwith infection by bovine papillo-mavirus type-2 (BPV-2). Such tumours commonlyoccur in cattle grazing mountain pastures rich inbracken fern and bovine BPV-2 is known to play a cru-cial role in oncogenesis (Pamukcu et al., 1976; Campoet al., 1992; Campo, 2002; Borzacchiello et al., 2003;Borzacchiello and Roperto, 2008; Roperto et al., 2008).

Materials and Methods

Samples

Ten samples of neoplastic urothelium were collectedat public slaughterhouses from cows aged 4e24 yearsthat had suffered from chronic enzootic haematuria.All animals were raised as mountain cattle in the

south of Italy (Roperto et al., 2005) and were knownto have grazed pastures rich in bracken fern. Normal(control) bladder mucosa was obtained from fivehealthy cows aged 4e14 years that had grazedlowland pastures without bracken.

Papillary neoplastic proliferations were detected inthe bladder tissue from six animals; whilst polypoidand sessile infiltrative lesions were present in samplesfrom four cows. Neoplastic lesions were classified ac-cording to the criteria given in the World Health Or-ganization (WHO) Blue Book on the pathology andgenetics of tumours of the human urinary systemand male genitalia (Lopez-Beltran et al., 2004; Sauteret al., 2004). This histopathological classification en-compasses all of the microscopical patterns of bovineurinary bladder tumours (Roperto et al., 2007).

Histopathology, Immunofluorescence and

Immunohistochemistry

Tissues were fixed in 10% neutral buffered formalin,processed routinely and embedded in paraffin wax.Sections (5 mm) were stained with haematoxylinand eosin (HE) for histopathological assessment.

For immunofluorescence labelling, sections weredewaxed, rehydrated and heated in a microwaveoven (twice for 5 min at 750 W) to allow antigen un-masking. Sections were then incubated with a 1 in 50dilution of rabbit antiserum specific for the E5oncoprotein (kindly provided by Dr. R. Schlegel,Georgetown University, USA) and thereafter withFITC-conjugated secondary antibody (Chemicon,California, USA). Immunofluorescence was detectedwith a Zeiss LSM 510 laser scanning confocal micro-scope (Carl Zeiss GmbH, Jena, Germany).

For immunohistochemistry (IHC), dewaxed sec-tions were first blocked for endogenous peroxidaseactivity by incubation in H202 0.3% in methanol for20 min. Antigen retrieval was performed by micro-wave heating as above. The primary antiserum wasspecific for ferritin heavy chain (1 in 200; cloneH-53, sc-25617, Santa Cruz Biotechnology, USA).A commercial kit (Dako, LSABK680; Dako Cytoma-tion, Burlingame, California) including peroxidase-conjugated streptavidin and a mixture of biotinylatedanti-rabbit and anti-mouse immunoglobulins wasused as the secondary reagent. ‘‘Visualization’’ waswith H202 and 3,30-diaminobenzidine followed bycounterstaining with haematoxylin. Control sectionswere treated with phosphate-buffered saline (PBS) in-stead of the primary antiserum.

Virus Analysis

A small fragment of frozen tissue from each urothelialtumour was digested by Proteinase-K in lysis buffer

Bovine Papillomavirus-Associated Urothelial Tumours 11

(50 mM KCl, 10 mM TriseHCl, pH 8.3, 2.5 mMMgCl2, 100 mg/ml gelatin, 0.45% NP-40 and 0.45%Tween-20) to recover DNA. Samples (10 ml) wereamplified by PCR utilizing one unit of Taq polymer-ase (Finenzymes, Espoo, Finland) in buffer providedby the manufacturer with 1.5 mM MgCl2. Thereaction was carried out in an iCycler (Bio-Rad Lab-oratories, Milan, Italy) with a forward primer of50-TTGCTGCAATGCAACTGCTG-30 and a reverseprimer of 50-TCATAGGCACTGGCACGTT-30 thatamplify a fragment encompassing part of the E5and L2 open reading frame (ORF) of BPV-1(311 bp, from nucleotide 3915 to 4226) and BPV-2(306 bp, from nucleotide 3919 to 4225) (Otten et al.,1993). PCR conditions were as follows: denaturationfor 3 min at 95�C followed by 35 cycles of denatur-ation at 95�C for 45 s, annealing at 50�C for 45 sand extension at 72�C for 1 min. The final PCR prod-ucts were separated by electrophoresis in an agarosegel (2%) and visualized by ethidium bromide stain-ing. In each experiment, a blank sample consistingof reaction mixture without DNA and a positive sam-ple consisting of BPV-2 cloned DNA (kindly providedby Dr. M.S. Campo, University of Glasgow, Scot-land) were included. A band corresponding to theamplified sequences of BPV-2 was detected in eachof the tumour samples. To confirm these PCR data,amplified bands were excised from the gel and puri-fied through silica gel membranes by the QIAquickPCR quantification kit, according to the manufactur-er’s instructions (QIAgen, Milan, Italy). AmplifiedDNA was subject to direct automated sequencing(Biogen, Rome, Italy).

Immunoprecipitation of E5 Protein

Tissues were lysed in ice-cold buffer containing50 mM TriseHCl (pH 7.5), 1% (v/v) Triton X-100,150 mM NaCl, 2 mM PMSF, 1.7 mg/ml aprotinin,50 mM NaF and 1 mM sodium orthovanadate.Lysates were clarified by centrifugation at 12,300gfor 30 min. Supernatants were collected and proteinconcentrations were determined by a modifiedBradford assay (Bio-Rad). One mg of protein fromeach sample was immunoprecipitated with 2 mg ofanti-E5 antibody (kindly provided by Dr. M.S.Campo, University of Glasgow, Scotland) and 30 mlof G-sepharose (GE Healthcare, Piscataway, NJ).Immunoprecipitates were washed four times in com-plete lysis buffer and then heated in LDS loadingbuffer 4� (Invitrogen, Carlsbad, California) at 70�Cfor 10 min according to the manufacturer’s protocol.Immunoprecipitates were separated in 4e12% poly-acrylamide gels and transferred to nitrocellulosemem-branes (Bio-Rad). Membranes were blocked in 5%

non-fat dried milk, and sequentially incubated withprimary and secondary antibodies. Detection of bind-ing was by use of an enhanced chemiluminescencesystem (Amersham Biosciences, Piscataway, NJ).

Western Blotting

Samples of neoplastic and control tissue were immedi-ately fixed in liquid nitrogen and stored at�80�C un-til tested. The samples were homogenized in lysisbuffer containing 50 mM TriseHCl pH 7.5,150 mM NaCl and 1% Triton X-100 to which1 mM DTT, 2 mM PMSF, 1.7 mg/ml aprotinin,25 mM NaF and 1 mM Na3VO4 (SigmaeAldrich,St Louis, Missouri) were added at the time of use.The proteins were concentrated by centrifugation at20,000g for 30 min at 4�C. The protein concentrationwas measured using the Bradford method (Bio-Rad).

Total extracts (30 mg) from these bladder tissueswere separated by 10% sodium dodecyl sulphatepolyacrylamide gel electrophoresis (SDS-PAGE),electrophoretically transferred onto a nitrocellulosemembrane and incubated with rabbit anti-humanH-ferritin (sc-25617; 1 in 1000; Santa Cruz Biotech-nology, Santa Cruz, CA) or rabbit anti-human g-tu-bulin horseradish peroxidase (HRP)-conjugated (1 in1000; Santa Cruz Biotechnology) primary antibodies.To detect protein levels of p65 (Rel/A; one of five sub-units of the NF-kB), an anti-human p65 polyclonalantibody (C-20; 1 in 5000) was employed (SantaCruz Biotechnology). An anti-human b-actin mono-clonal antibody (1 in 5000; Sigma) was used in paral-lel to determine equivalent loading conditions.Immunoreactivity was detected by sequential incuba-tion with HRP-conjugated secondary antibody andenhanced chemiluminescence reagents following thestandard manufacturer’s protocols (GEHealthCare).

The intensity of the bands was evaluated densito-metrically and expressed as the ratio of FHC to g-tu-bulin and of p65 levels to b-actin. Scanningdensitometry of the immunoblots was performedwith an Image Scan (SnapScan 1212; Agfa-GevaertN.V., Mortsel, Belgium). The area under the curverelated to each protein was determined using Gimp2 software (http://gimp-win.sourceforge.net/index.html). Background was subtracted from these calcu-lated values. Significance between the two groupswas calculated by the Student T-test (Ryder andRobakiewicz, 1998).

ReverseTranscriptase-Polymerase Chain Reaction (RT-PCR)

RT-PCR was performed in order to evaluatetranscription of the gene encoding the FHC. Thiswas determined in semi-quantitative fashion by

Fig. 1. (A) Detection of BPV-2 DNA sequences in samples fromanimals with tumours (lanes 2e11). The size of the E5 am-plified products was 125 bp. Lane 1, no cDNA; lane 12,positive control; lane M, DNA molecular weight markerVIII. (B) E5 oncoprotein detected by immunoprecipita-tion in tumour samples. (C) Immunofluorescence labellingof E5 (arrows) in the cytoplasm of some basal and supraba-silar tumour cells detected by confocal microscopy. Bar,10 mm.

Fig. 3. Nuclear expression pattern of FHC in bladder carcinomasas determined by western blotting, densitometry and anal-ysis by the Scion Image�Program. Significancewas deter-mined by the Student T-test (P< 0.04).

12 S. Roperto et al.

comparison with expression of the housekeeper geneencoding glyceraldehyde-3-phosphate dehydroge-nase (GAPDH). A total of 2 mg of RNA were usedto generate complementary DNA (cDNA) with ran-dom primers by use of the SuperScript II� reversetranscriptase (Invitrogen) as recommended by themanufacturer. PCR amplification was carried out un-der standard conditions with Taq DNA polymerase

Fig. 2. (A)Western blot analysis showing the expression pattern ofFHC in neoplastic tissues. Positive control was an extract ofthe HeLa cell line and the negative control comprised anequivalent quantity of normal tissue. (B) Histogramswere produced following analysis of densitometric traceswith the Scion Image� Program and significance deter-mined by the Student T-test (P< 0.002).

(Promega Biosciences, San Luis Obispo, CA). Thenucleotide sequences of the PCR primers were 50-CATCAACCGCCAGATCAAC-30 (forward) and50-GATGGCTTTCACCTGCTCAT-30 (reverse)for FHC; 50 GAGTCCACTGGGGTCTTCAC-30

(forward) and 50-CTTCCGCGTCCCCAGAGTC-30 (reverse) for GAPDH. The PCR products wereseparated in an agarose gel 1.5% containing 0.04%ethidium bromide. Direct sequence analysis wasperformed on PCR fragments after purification withMicrocon�PCRcolumns (MilliporeCorp., Billerica,MD). Fragments were sequenced in both directionsusing ABI PRISM Big-Dye Terminator Kit� (Ap-plied Biosystems) with the ABI-3100 genetic Analyzer(Applied Biosystems).

Real-Time RT-PCR

Total RNA was extracted from tissues using Trizolreagent (Invitrogen). 1 mg of RNA was reverse-transcribed by use of the QuantiTect ReverseTranscription Kit� (Qiagen Inc., Valencia, CA) ac-cording to the manufacturer’s protocol. Quantitativereal-time RT-PCR was performed using QuantiTectSYBR Green PCR Kit� (Qiagen Inc.) with thefollowing primer set for FHC: 50-GCAGGTGCGC-CAGAACTAC-30 (forward) and 50-CACATCATCGCGGTCAAAGT-30 (reverse). The primer set foramplification of the b-actin gene was: 50-AGCAAGCGAGAGTACGATGAGT-30 (forward) and 50-ATCCAACCGACTGCTGTCA-30 (reverse). Eachsample was run in triplicate. The amount of FHCmRNA was expressed as the ratio to b-actin

Fig. 4. (A) Strong intracytoplasmic immunoreactivity for FHC isevident inmany neoplastic urothelial cells. Some nuclei arealso positively labelled. Bar, 10 mm. (B) Weak intracyto-plasmic immunoreactivity for FHC is present in control ur-othelium and fewnuclei are positively labelled. Bar, 10 mm.

Bovine Papillomavirus-Associated Urothelial Tumours 13

mRNA. Relative gene expression was calculatedusing the 2�DDCt method (Livak and Schmittgen,2001).

Nuclear and Cytoplasmic Fractions

Tissues were cut and weighed and a tenfold excess (tothe volume of the tissue) of solution A (10 mMHepespH 7.5, 10 mM KCl, 1.5 mM MgCl2, 1 mM EDTA,1 mM EGTA, 10 mM sodium pyrophosphate, 0.5%Triton, 2 mM DTT, 1 mM PMSF, 2 mM Na3VO4,50 mMNaF, 5 mg/ml aprotinin and 5 mg/ml leupep-tin) was added to each sample. The samples werehomogenized on ice and subsequently centrifuged at700 g for 3 min at 4�C. Pellets were resuspended in so-lution B (20 mM Hepes, 400 mM NaCl, 1.5 mMMgCl2, 0.1 mM EDTA and 20% glycerol, to which2 mM DTT, 1 mM PMSF, 2 mM Na3VO4, 50 mMNaF, 5 mg/ml aprotinin and 5 mg/ml leupeptinswere added at the time of use). The nuclear extractswere separated by centrifugation at 15,000g for10 min at 4�C.

Fig. 5. (A) RT-PCR analysis for FHC gene expression in normal and nas housekeeper. (B) Quantitative densitometric analysis of theanalyzed by Student T-test (P¼ 0.001).

Electrophoresis Mobility Shift Assay (EMSA)

10 mg nuclear extracts were analyzed for their abilityto bind to 32P-labelled NF-kB oligonucleotidesspanning the kB motif: (sense: 50-CAACGGCAGGGGAATCTCCCTCTCCTT-30 and antisense:50-GTTGCCGTCCCCTTAGAGGGAGAGGAA-30)in an EMSA.DNA-protein complexes were separatedin a 6% DNA retardation polyacrylamide gel. Bandspecificity was assessed by use of 50� specific ornon-specific (Oct-1 consensus oligo; Promega BioSci-ences) cold competitors.

Results

Histopathological Characterization

The ten tumours investigated here comprised fourlow-grade urothelial carcinomas, two high-grade ur-othelial carcinomas, three low-grade invasive carci-nomas and a single high-grade invasive carcinoma.

Virus Analysis and Immunofluorescence

PCRanalysis, validated by direct sequencing of the am-plified products, confirmed the presence of true BPV-2sequences in all tumour samples (Fig. 1A). Sequenceanalysis was performed by use of the ClustalW (http://www.ebi.ac.uk/clustalw/) BLAST (http://www.ncbi.nlm.nih.gov/BLAST) programmes. These analyses re-vealed the presence of DNA sequences matching theBPV-2 complete genome from the GenBank DataBase (GenBank NC-001521). E5 oncoprotein was de-tected in all tumours by immunoprecipitation(Fig. 1B) and immunofluorescence. Some basal andsuprabasal neoplastic urothelial cells had strong cyto-plasmic E5 immunoreactivity (Fig. 1C). Neither BPV-2 sequences nor E5 protein expression was detected inbladder samples from healthy cattle.

eoplastic bovine bladder tissue. GAPDH gene expression was usedse gels was performed with the Scion Image� Program and data

Fig. 6. The top line represents the sequence of bovine H-ferritin detected in samples of tumour tissue. This sequence is compared with thenucleotide sequences of bovine and human H-ferritin extracted from GenBank.

14 S. Roperto et al.

Western Blot Analysis and Immunohistochemistry

Protein expression levels as evaluated by western blotanalysis revealed that H-ferritin concentration was

increased in tumour tissue. Positive control extractsfrom the HeLa cell line (known to contain a largeamount of FHC), and negative control extractsfrom healthy bladder were included (Fig. 2A). The

Bovine Papillomavirus-Associated Urothelial Tumours 15

related intensity of the signal was quantified by densi-tometric scanning of the filter (Fig. 2B). Nuclear pro-teins from diseased and healthy tissues were used toperform SDS-PAGE. FHC concentration was higherin tumour samples (Fig. 3).

IHC confirmed that FHC was overexpressed in thecytoplasm of neoplastic urothelial cells; however,many nuclei were also positively labelled (Fig. 4A).In normal urothelium there was homogeneous andweak expression of FHC within the cytoplasm andin occasional nuclei (Fig. 4B).

RT-PCR

The amount of mRNA encoding FHC was increasedin samples of tumour tissue when compared withtissue from healthy cattle, but the expression of thegene encoding GAPDH was equivalent (Fig. 5).The oligonucleotide sequence of the amplicons wasidentical to that of human ferritin (Fig. 6).

Real-Time RT-PCR

The expression of the gene encoding FHC was signifi-cantly greater (P¼ 0.006) in neoplastic comparedwith healthy tissues (Fig. 7).

Western Blotting and EMSA for NF-kB

Nuclear p65 expression was found to be significantlyincreased in neoplastic urothelial cells relative to

Fig. 7. RT-PCR analysis of FHC gene expression. Total RNAfrom samples of normal bovine bladder andbovinebladdercarcinoma were reverse-transcribed and analyzed by RT-PCR. FHCmRNA levels are expressed as a ratio to b-actinmRNA (*P¼ 0.006).

healthy tissue (P¼ 0.047; Fig. 8A,B). The activity ofNF-kB within nuclear extracts as determined byEMSA is summarized in Fig. 8C.

Discussion

Disturbances in iron homeostasis are increasinglyemerging as important elements in the pathogenesisof a range of diseases including cancer (Torti andTorti, 2002; Tsuji, 2005). In particular, FHC appearsto be involved in some molecular pathways impli-cated in tumourigenesis. The concentration of serumferritin increases in animals and human beings withcancer (Newlands et al., 1994; Torti and Torti,2002) and this is believed to be a marker of malig-nancy (Kwok and Richardson, 2002; Orino andWatanabe, 2008).

The present study has demonstrated up-regulationof p65, a component of NF-kB transcription factors,and of FHC in urothelial tumours of the urinary blad-der of cattle infected by BPV-2. These observationsare consistent with the suggestion that high risk hu-man papillomavirus (HPV) 16 E6 activates NF-kBand protects against apoptosis in airway epithelialcells (James et al., 2006). Similarly, there is a positiverelationship between the level of HPV and the expres-sion of p65 in human laryngeal squamous cell carci-noma (Du et al., 2003).

It has been shown that the activation of NF-kB/p65is sufficient to up-regulate FHC expression (Phamet al., 2004; Reis et al., 2006). In turn, FHC is respon-sible for iron sequestration, resulting in suppression ofapoptosis by the inhibition of ROS (Pham et al., 2004;Bubici et al., 2006; Aung et al., 2007). It is thereforelikely that FHC up-regulation, mediated by NF-kB,may be one of the molecular keys whereby BPV-2-in-fected urothelial cells are protected from apoptosis orprogrammed cell death (PCD). FHC expression wasalso found to be up-regulated in some urothelial cellnuclei. It is generally believed that nuclear ferritinacts as a DNA protectant, suggesting a potentialrole of nuclear ferritin in the growth of tumour cells(Thompson et al., 2002).

Bovine bladder tumours are characterized by a rel-atively low metastatic capacity, with regional lymphnode metastases found in approximately 10% of cases(Pamukcu et al., 1976; Roperto et al., 2005). It hasbeen suggested that high levels of GM3 gangliosidemay be involved in the down-regulation of the meta-static potential of bovine urothelial tumours (Ropertoet al., 2007). Furthermore, it is also possible that theup-regulation of FHC observed in the present studymay play a role in the biological behaviour of bovineurothelial tumours, since overexpression of FHCinhibits CXCR4-mediated chemotaxis and cell

Fig. 8. (A) Nuclear extracts were obtained from samples of normal bovine bladder and bovine bladder carcinoma. p65 levels were ana-lyzed by western blotting and expressed as the optical density ratio to b-actin protein levels. Significance between the two groupswas evaluated by the StudentT-test (*P¼ 0.047). (B)Western blotting analysis of p65 levels in nuclear extracts. A and B are rep-resentative results derived from analysis of control bladder samples, while C and D are representative results derived from tumourtissues. (C) Samples shown in (B) were also analyzed for NF-kB activity by EMSA. Arrows indicate the position of NF-kB dimers.These bandsweremore intense in neoplastic as comparedwith control samples. Band specificitywas assessed by using 100� specific(NF-kB) or non-specific (Oct-1) cold competitors.

16 S. Roperto et al.

migration (Li et al., 2006). Overexpression of FHCwas also detected in the cancer line RBT 157, derivedfrom a rat bladder tumour, which phenotypically andcytogenetically resembles human superficial bladdercancer. RBT 157 is characterized by a very low met-astatic capacity and a well-differentiated phenotype(Vet et al., 1997).

The present study is the first to demonstrate thatFHC expression is up-regulated in urothelial neopla-sia. In parallel studies, we have found that tumours ofthe bovine urinary bladder express hepcidin (data notshown), an antimicrobial protein synthesized by theliver and known to play a central role in iron homeo-stasis. This observation further suggests that ironmight play a role in the mechanisms underlyingdevelopment of BPV-2-associated urothelial tumours.It is of note that the BPV-2 E5 oncoprotein sharesmany biological features with the viral Nef proteinknown to be involved in iron metabolism via HFE,a non-classical class I molecule of the major histocom-patibility complex (Drakesmith et al., 2005; Tsirimo-naki et al., 2006). Recently, it has been shown thatHFE plays a role in papillomavirus-associated cervi-cal cancer (Cardoso et al., 2006).

Ultimately, an in-depth understanding of themolecular mechanisms by which iron metabolism isinvolved in disease pathogenesis, including carcino-genesis, warrants further investigation. It has been

suggested that iron may be a candidate for targetedapproaches to treatment of neoplastic and non-neoplastic diseases in which there is dysregulatediron metabolism (Richardson et al., 2008). In partic-ular, new evidence appears to suggest that virusesare able to specifically target and regulate proteinsinvolved in iron homeostasis (Laham and Ehrlich,2004). Therefore, it has been proposed that futureresearch should investigate viruseiron interactionsat a molecular level, thus elucidating more specificareas for therapeutic intervention (Drakesmith andPrentice, 2008). For example, it has been shownthat iron-reduction therapy by the use of chelatorcompounds can influence the pathogenesis of viraldiseases, including those progressing to neoplasia(Drakesmith and Prentice, 2008). The bovine modelmight therefore be very useful for evaluating suchnovel therapeutic strategies.

Acknowledgments

This work was supported in part by grants fromthe Italian Ministry of University and Scientific Re-search, Law number 5 of Regione Campania, theAssessorato alla Sanita of Regione Basilicata and Re-gione Campania and the Assessorato all’Agricoltura,Foreste, Forestazione, Caccia e Pesca of Regione Cala-bria. The authors wish to thank Drs G. Di Domenico

Bovine Papillomavirus-Associated Urothelial Tumours 17

and G. Marino of Azienda Sanitaria Locale SA/3, DrsE. Massari, G. Toma, and A. Russillo of AziendaSanitaria Locale PZ/2, Dr G. Salvatore from RegioneBasilicata and Dr F. Logozzo from Regione Calabriafor their technical help.

References

Aung W, Hasegawa S, Furukawa T, Saga T (2007) Poten-tial role of ferritin heavy chain in oxidative stress and ap-optosis in human mesothelial and mesothelioma cells:implications for asbestos-induced oncogenesis. Carcino-genesis, 28, 2047e2052.

Borzacchiello G, Iovane G, Marcante ML, Poggiali F,Roperto F et al. (2003) Presence of bovine papillomavi-rus type 2 DNA and expression of the viral oncoproteinE5 in naturally occurring urinary bladder tumours incows. Journal of General Virology, 84, 2921e2926.

Borzacchiello G, Roperto F (2008) Bovine papillomavi-ruses, papillomas and cancer in cattle.Veterinary Research,39, 45e63.

Boult J, Roberts K, Brookes MJ, Hughes S, Bury JP et al.(2008) Overexpression of cellular iron import proteinsis associated with malignant progression of esophagealadenocarcinoma. Clinical Cancer Research, 14, 379e387.

Brookes MJ, Boult J, Roberts K, Cooper BT, Hotchin NAet al. (2008) A role for iron in Wnt signalling. Oncogene,27, 966e975.

BrookesMJ, Hughes S, Turner FE, Reynolds G, Sharma Net al. (2006)Modulation of iron transport proteins in hu-man colorectal carcinogenesis. Gut, 55, 1449e1460.

Bubici C, Papa S, Pham CG, Zazzeroni F, Franzoso G(2006) The NF-kB-mediated control of ROS and JNKsignaling. Histology and Histopathology, 21, 69e80.

Campo MS, Jarrett WFH, Barron RJ, O’Neil BW,Smith KT (1992) Association of bovine papillomavirustype 2 and bracken fern with bladder cancer in cattle.Cancer Research, 52, 6898e6904.

Campo MS (2002) Animal models of papillomavirus path-ogenesis. Virus Research, 89, 249e261.

Cardoso CS, Araujo HC, Cruz E, Afonso A,Mascarenhas C et al. (2006) Haemochromatosis gene(HFE) mutations in viral-associated neoplasia: linkageto cervical cancer. Biochemical and Biophysical Research

Communications, 341, 232e238.De Domenico I, McVeyWard D, Kaplan J (2008) Regula-

tion of iron acquisition and storage: consequences foriron-linked disorders. Nature Reviews. Molecular Cell Biol-

ogy, 9, 72e81.Drakesmith H, Chen N, Ledermann H, Screaton G,

Townsend A et al. (2005) HIV1 Nef downregulates thehemochromatosis protein HFE, manipulating cellularhomeostasis. Proceedings of the National Academy Sciences

the United States of America, 102, 11017e11022.Drakesmith H, Prentice A (2008) Viral infection and iron

metabolism. Nature Reviews. Microbiology, 6, 541e552.Du J, Chen GG, Vlantis AC, Xu H, Tsang RKY et al.

(2003) The nuclear localization of NF-kB and p53 is

positively correlated with HPV16 E7 level in laryngealsquamous cell carcinoma. Journal of Histochemistry and Cy-

tochemistry, 53, 533e539.Galaris D, Skiada V, Barbouti A (2008) Redox signaling

and cancer: the role of ‘‘labile’’ iron. Cancer Letters,266, 21e29.

Gu JM, Lim SO, Oh SJ, Yoon SM, Seong JK et al. (2008)HBx modulates iron regulatory protein 1-mediated ironmetabolism via reactive oxygen species. Virus Research,133, 167e177.

Harrison PM, Arosio P (1996) The ferritins: molecularproperties, iron storage function and cellular regulation.Biochimica et Biophysica Acta, 1275, 161e203.

James MA, Lee JH, Klingelhutz AJ (2006) Human papil-lomavirus type 16 E6 activates NF-kB, induces cIAP-2expression, and protects against apoptosis in a PDZbinding motif-dependent manner. Journal of Virology,80, 5301e5307.

Klaus A, Birchmeier W (2008) Wnt signaling and its im-pact on development and cancer. Nature Reviews. Cancer,8, 387e398.

Kwak EL, Larochelle DA, Beaumont C, Torti SV,Torti FM (1995) Role for NF-kappa B in the regulationof ferritin H by tumour necrosis factor-alpha. Journal ofBiological Chemistry, 270, 15285e15293.

Kwok JC, Richardson DR (2002) The iron metabolism ofneoplastic cells: alterations that facilitate proliferation?Critical Reviews in Oncology/Hematology, 42, 65e78.

Laham N, Ehrlich R (2004) Manipulation of iron to deter-mine survival e competition between host and patho-gen. Immunology Research, 30, 15e28.

Lawson DM, Treffry A, Artymiuk PJ, Harrison PM,Yewdall SJ et al. (1989) Identification of the ferroxidasecentre in ferritin. FEBS Letters, 254, 207e210.

Li R, Luo C, Mines M, Zhang J, Fan G-H (2006) Chemo-kine CXCL12 induces binding of ferritin heavy chain tothe chemokine receptor CXCR4, alters CXCR4 signal-ing, and induces phosphorylation and nuclear transloca-tion of ferritin heavy chain. Journal of Biological Chemistry,281, 37616e37727.

Livak KJ, Schmittgen TD (2001) Analysis of relative geneexpression data using real-time quantitative PCR andthe2(�DeltaDeltaC(T))method.Methods,25, 402e408.

Lopez-Beltran A, Sauter G, Gasser T, Hartmann A,Schmitz-Drager BJ et al. (2004) Infiltrating urothelialcarcinoma. In: Pathology and Genetics of Tumours of the Uri-

nary System and Male Genital Organs, JN Eble, G Sauter,JL Epstein, IA Sesterhenn, Eds, IARC Press, Lyon,pp. 93e109.

Newlands C,HoustonDM,Vasconcelos DY (1994)Hyper-ferritinemia associated with malignant histiocytosis ina dog. Journal of the American VeterinaryMedical Association,205, 849e851.

Orino K, Harada S, Natsuhori M, Takehara K,Watanabe K (2004) Kinetic analysis of bovine spleenapoferritin and recombinant H and L chain homopoly-mers: iron uptake suggests early stage H chain ferroxi-dase activity and second stage L chain cooperation.Biometals, 17, 29e134.

18 S. Roperto et al.

Orino K, Watanabe K (2008) Molecular, physiologicaland clinical aspects of the iron storage protein ferritin.Veterinary Journal, 178, 191e201.

Otten N, von Tscharner C, Lazary S, Antczak DF,Gerber H (1993) DNA of bovine papillomavirus type1 and 2 in equine sarcoids: PCR detection and direct se-quencing. Archives of Virology, 132, 121e131.

Pamukcu AM, Price JM, Bryan GT (1976) Naturally oc-curring and bracken fern-induced bovine urinary blad-der tumours. Veterinary Pathology, 13, 110e122.

Pham CG, Bubici C, Zazzeroni F, Papa S, Jones J et al.(2004) Ferritin heavy chain upregulation by NF-kB in-hibits TNFa-induced apoptosis by suppressing reactiveoxygen species. Cell, 119, 529e542.

Recalcati S, Invernizzi P, Arosio P, Cairo G (2008) Newfunctions for an iron storage protein: the role of ferritinin immunity and autoimmunity. Journal of Autoimmunity,30, 84e89.

Reis K, Halldin J, Fernaeus S, Pettersson C, Land T (2006)NADPH oxidase inhibitor diphenyliodonium abolisheslipopolysaccharide-induced down-regulation of trans-ferring receptor expression in N2a and BV-2 cells. Jour-nal of Neuroscience Research, 84, 1047e1052.

Richardson DR, Kalinowski DS, Jansson PJ, Lovejoy DB(2008) Cancer cell ironmetabolism and the developmentof potent iron chelators as anti-tumour agents. Biochimicaet Biophysica Acta, doi:10.1016/jbbagen.2008.04.003.

Roperto S, Ambrosio V, Borzacchiello G, Galati P,Russo V et al. (2005) Bovine papillomavirus type-2(BPV-2) infection and expression of uroplakin IIIb,a novel urothelial cell marker, in urinary bladder tu-mours of cows. Veterinary Pathology, 42, 812e818.

Roperto S, Borzacchiello G, Casellato R, Galati P, Russo Vet al. (2007) Sialic acid and GM3 ganglioside expressionin papillomavirus-associated urinary bladder tumoursof cattle with chronic enzootic haematuria. Journal ofComparative Pathology, 132, 87e93.

Roperto S, Brun R, Paolini F, Urraro C, Russo V et al.(2008) Detection of bovine papillomavirus type 2(BPV-2) in the peripheral blood of cattle with urinary

bladder tumours: possible biological role. Journal of Gen-eral Virology, 89, 3027e3033.

Ryder EF, Robakiewicz P (1998) Statistics for the molecu-lar biologist: group comparisons. Current Protocols in Mo-

lecular Biology, A. 3I(Suppl. 67), 1e22.Sammarco MC, Ditch S, Banerjee A, Grabczyk E (2008)

Ferritin L and H subunits are differentially regulatedon a post-transcriptional level. Journal of Biological Chem-istry, 283, 4578e4587.

Sauter G, Algaba F, Amin MB, Busch C, Cheville J et al.(2004) Non-invasive urothelial tumours. In: Pathologyand Genetics of Tumours of the Urinary System and Male

Genital Organs, JN Eble, G Sauter, JI Epstein,IA Sesterhenn, Eds, IARC Press, Lyon, pp. 110e112.

Thompson KJ, Fried MG, Ye Z, Boyer P, Connor RJ(2002) Regulation, mechanisms and proposed functionof ferritin translocation to cell nuclei. Journal of Cell Sci-ence, 115, 2165e2177.

Torti FM, Torti SV (2002) Regulation of ferritin genes andprotein. Blood, 99, 3505e3516.

Tsirimonaki E, Ullah R, Marchetti B, Ashrafi GH,McGarry L et al. (2006) Similarities and differences be-tween the E5 oncoproteins of bovine papillomavirusestype 1 and type 4: cytoskeleton, motility and invasive-ness in E5-transformed bovine and mouse cells. Virus Re-search, 115, 158e168.

Tsuji Y (2005) JunD activates transcription of the humanferritin H gene through an antioxidant response elementduring oxidative stress. Oncogene, 24, 7567e7578.

Vet JAM, van Moorselaar RJA, Debruyne FMJ,Schalken JA (1997) Differential expression of ferritinheavy chain in a rat transitional carcinoma progressionmodel. Biochimica et Biophysica Acta, 1360, 39e44.

Zandman-Goddard G, Shoenfeld Y (2007) Ferritin in au-toimmune diseases. Autoimmunity Reviews, 6, 457e463.

½ RA

eceived, November 24th, 2008

ccepted, May 29th, 2009

�