E6/E7 Expression of Human Papillomavirus Type 20 (HPV-20) and HPV-27 Influences Proliferation and Differentiation of the Skin in UV-Irradiated SKH-hr1 Transgenic Mice

Abstract

The functional role of UV irradiation, in combination with the E6 and E7 proteins of the cutaneous human papillomavirus (HPV) types in the malignant conversion of benign papillomatous lesions, has not been elucidated. Transgenic SKH-hr1 hairless mice expressing HPV-20 and HPV-27 E6 and E7 proteins in the suprabasal compartment were generated and exposed to chronic UV irradiation. Histological and immunohistochemical examination of skin samples revealed enhanced proliferation of the epidermal layers and papilloma formation in both transgenic strains in comparison to what was observed with nontransgenic mice. Squamous cell carcinoma developed in the HPV-20 E6/E7 transgenic line as well as in the HPV-27 E6/E7 transgenic line. Several weeks after cessation of UV-B exposure, enhanced proliferation, as measured by BrdU incorporation, was maintained only in HPV-20 transgenic skin. Keratin 6 expression was increased in the transgenic mice throughout all cell layers. Expression of the differentiation markers involucrin and loricrin was reduced and disturbed. p63α expression was differentially regulated with high levels of cytoplasmic expression in clusters of cells in the granular layer of the skin in the transgenic lines 20 weeks after cessation of UV-B exposure, in contrast to uninterrupted staining in the nontransgenic lines. p53 was expressed in clusters of cells in nontransgenic and HPV-27 transgenic mice, in contrast to an even distribution in a higher number of cells in HPV-20 transgenic animals.

UV irradiation is a major etiological factor in the development of nonmelanoma skin cancer, one of the most frequent cancers occurring among the Caucasian population worldwide (6, 38). These basal cell and squamous cell carcinomas (SCC) occur primarily on sun-exposed sites. Wild-type p53 protein is expressed at increased levels after UV irradiation, thereby shutting off DNA synthesis and cell replication and allowing for DNA repair and maintenance of genetic integrity after DNA damage. Severe UV irradiation induces not only acute inflammatory reactions but also DNA damage, which is reflected in the induction of p53 mutations (5, 36, 39, 43, 57). These p53 mutations are distributed as scattered cells or in clusters of p53-immunoreactive keratinocytes (4, 35).

Papillomavirus infections have been implicated in the etiology of nonmelanoma skin cancer both in immunocompetent and in immunosuppressed individuals. This is evidenced in the malignant conversion of 30% to 60% of warts occurring on sun-exposed body sites in patients with the immune-impairing disease Epidermodysplasia verruciformis (37). Human papillomavirus (HPV) DNA has been demonstrated in the majority of premalignant lesions and SCC of the skin (12, 21, 23, 52). The mechanism by which the high-risk mucosal HPV types participate in anogenital carcinogenesis has been elucidated. The expression of the E6 and E7 genes of the high-risk mucosal HPV types stimulates proliferation and is a precondition for immortalization and malignant growth (58, 59). Very little is known about the molecular pathways involved after infection with cutaneous papillomavirus types. The proinflammatory cytokines (interleukin-1α [IL-1α], IL-1β, IL-6, and IL-17 and alpha, beta, and gamma interferons) induced by UV irradiation either induce or inhibit the promoters of a number of HPV types associated with cutaneous lesions (13, 14, 48). The E6 proteins of certain cutaneous HPV types are unable to promote p53 degradation in vitro (18, 51, 53) but effectively inhibit UV-induced apoptosis by attenuating the transcription of p53-regulated proapoptotic genes as well as stimulating the degradation of the proapoptotic protein Bak (26). The ability of the E7 protein of cutaneous HPV types to bind to the pocket protein Rb is not a determinant for oncogenicity for the cutaneous HPV types tested as it has been demonstrated for the high-risk mucosal HPV types (8, 58).

A number of in vivo models using transgenic mice have been reported in which the E6 and E7 properties of high-risk mucosal HPV-16 or HPV-18 have been studied. Tumors developed in different organs, independent of the promoter used for expression (1, 2, 7, 22, 29, 33). The in vivo properties of the E6 and E7 proteins of cutaneous papillomaviruses have received less attention. In one study in which transgenic mice harbored the early region of the HPV-8 genome driven by the K14 promoter and were backcrossed for six generations, 6% of mice developed nonmelanoma skin cancer (50). Another study described induction of papillomas, keratoacanthomas, and squamous cell carcinomas after a two-stage carcinogen treatment of HPV-38 E6/E7-expressing transgenic mice (15). We present here a comparative in vivo study to dissect the influence of chronic UV exposure in conjunction with the expression of the E6 and E7 genes of HPV types associated with either malignant or benign lesions. Transgenic mice were generated in which the human keratin 10 (K10) promoter directed expression of the respective E6 and E7 genes of HPV-20 and HPV-27 to the differentiating cells of the epithelium (3). HPV-20 was more prevalent in a number of squamous cell carcinomas of the skin (12), whereas HPV-27 was usually associated with common benign warts (10). An increasing proliferation and disrupting differentiation of the epidermis were observed in the transgenic animals, and they developed more papillomas than their nontransgenic littermates. The HPV-20 transgenic mice displayed a higher tendency for malignant progression, and a squamous cell carcinoma developed in these mice during the course of the study. Proliferation of the skin was enhanced in the transgenic mice, as measured by BrdU incorporation, and expression of the differentiation markers involucrin and loricrin was reduced and disturbed in comparison to that observed with the control nontransgenic mice. The expression of p63, a p53 family member involved in the differentiation of skin, was differentially regulated by chronic UV-B exposure only in the transgenic mice and remained disrupted 20 weeks after cessation of UV-B irradiation. p53 expression was in addition increased in the irradiated transgenic mice in comparison to that in the irradiated nontransgenic mice.

MATERIALS AND METHODS

Construction of transgenes and generation of transgenic mouse lineages.

The plasmids for the K10 HPV-20 E6/E7 and K10 HPV-27 E6/E7 transgenes contained the respective E6 and E7 open reading frames under the control of the keratin 10 promoter and were separated by an internal ribosomal entry site (IRES). The constructs were generated in successive steps. Each component was amplified by PCR and digested with restriction enzymes, as indicated and shown in Fig. 1A. The E6 genes of HPV-20 and HPV-27 were each ligated as SmaI-E6-BglII fragments to a full-length BglII-XbaI IRES fragment and cloned into puC18 (Roche, Mannheim, Germany). These E6-IRES fragments were then excised by SmaI (E6) and KpnI (IRES) digestion, truncating the IRES by 190 bp at the 3′ end. A fusion product of the IRES 3′ ATG (nucleotide [nt] 660 to nt 833) with the 5′ end of the respective E7 gene was generated. The HPV-20 IRES was amplified as a KpnI-BsrdI fragment (HPV-20 E7 with 28 bp, nt 697 to nt 724) and the HPV-27 IRES as a KpnI-BanI fragment (HPV-27 with 18 bp, nt 539 to nt 556). The E7 genes were then completed by amplification of the respective 5′-truncated fragments for HPV-20 E7 (nt 724 to nt 1005) as a BsrdI and EcoRI fragment and HPV-27 E7 (nt 556 to nt 814) as a BanI and EcoRI fragment. The HPV-20 E6/IRES/E7 (total length, 1,321 bp) and the HPV-27 E6/IRES/E7 (1,294 bp) cassettes were constructed by ligation of the respective fragments. The respective complete transgene cassettes were cloned as SmaI-EcoRI fragments into the plasmid pPOLYIIIkeratin10pA (2) (kindly provided by M. Tomassino) containing the bovine homologue of the human keratin 10 promoter sequence and a simian virus 40 polyadenylation signal (Fig. 1A). The K10E6/IRES/E7pA transgenes of HPV-20 (6,361 bp) and HPV-27 (6,334 bp) were released from the vector by NotI enzyme digestion. The DNA was cleaned by sterile filtration (Millipore-Millex-GV filter, 0.22 μm; Millipore, Eschborn, Germany), floated on a Millipore membrane (VMW PO2500/Vm, 0.05 μm; Millipore, Eschborn, Germany), and dialyzed against microinjection buffer (10 mM Tris-HCl, pH 7.5, 0.1 mM EDTA) prior to injection into the pronuclei of fertilized oocytes. The hairless and immunocompetent mouse strain SKH-hr1 (purchased from Charles River Wiga, Sulzfeld, Germany) was used as genetic background for the development of heterozygous transgenic mice. Genomic DNA was prepared from tail biopsy specimens of 4-week-old founder mice, and transgenic animals were identified by PCR amplification of the respective transgene cassette. All animal experiments were performed according to German federal law (Regierungspräsidium Karlsruhe no. A39/02 and G39/02).

FIG. 1.

(A) Schematic presentation of the transgene constructs. The E6 and E7 open reading frames of HPV-27 and HPV-20 were separated by an IRES and inserted into the plasmid pPOLYIIIkeratin10pA under the control of the K10 promoter, using the restriction enzymes as indicated (HPV-27 above and HPV-20 below). polyA, simian virus 40 polyadenylation signal. (B) Reverse transcription of E6 and E7 transgene expression in the skin of mice from line 7013 (HPV-20 transgene) and line 606 (HPV-27 transgene). Lanes 1 to 4 represent individual mice of the respective lines in comparison to nontransgenic mice (lanes 5 to 8). GAPDH was included as positive control for reverse transcription and amplification.

Amplification was performed in 35 cycles of 95°C for 30 seconds, 55°C for 60 seconds and elongation at 72°C for 60 seconds, each (unless specified differently).

The PCR primers and conditions used for amplification of the respective fragments (restriction sites in bold) are as follows: HPV20-E6 (522 bp), 5′ GAC GAA TTC CCC GGG ATG GCT ACA CCT CCT TCT TCA GAA G 3′ (forward; nt 200 to nt 224) and 5′ GAA CAG ATC TTT ATT GAA AAT GCT TAC ACA GCC TAC AGA TTC 3′ (reverse; nt 697 to nt 667); HPV27-E6 (501 bp), 5′ GAC GAA TTC CCC GGG ATG CGC ACA AGG GCA GGG ATG TCA G 3′ (forward; nt 99 to nt 124) and 5′ GAA CAG ATC TAT GTA ATG TCC GCG AGG CTG GGT CGG GTG 3′ (reverse; nt 575 to nt 546); IRES of the encephalomyocarditis virus (512 bp; kindly provided by H. Pöpperl), 5′ GAC AGA TCT ATG TTA TTT TCC ACC ATA TTG CCG T 3′ (forward; nt 332 to nt 357) and 5′ TGC TCT AGA ATT ATC ATC GTG TTT TTC AAA GGA A 3′ (reverse; nt 833 to nt 809); IRES/E7 fusion fragment of HPV-20 (201 bp), 5′ TCC TCA AGC GTA TTC AAC AAG G 3′ (forward; nt 660 to nt 681) and 5′ CTT GCA ATG TGA CCT CTT TAC CAA TCA TAT TAT CAT CGT G 3′ (reverse; E7, nt 724 to nt 697; IRES, nt 833 to nt 822 [in italics]) (annealing temperature of 55°C); IRES/E7 fusion fragment of HPV-27 (196 bp), 5′ TCC TCA AGC GTA TTC AAC AAG G 3′ (forward; nt 660 to nt 681) and 5′ GGG TCG GGT GCC GTG CAT ATT ATC ATC GTG 3′ (reverse; E7, nt 556 to nt 539; IRES, nt 833 to nt 822 [in italics]); truncated E7 of HPV-20 (295 bp), 5′ ATG ATT GGT AAA GAG GTC ACA TTG CAA GAT ATT GTG C 3′ (forward; nt 697 to nt 732) and 5′ GAC GAA TTC TTA GGA TCC GCC ATG TTT GCA GTT C 3′ (reverse; nt 1005 to nt 980) (annealing temperature of 57°C); truncated E7 of HPV-27 (281 bp), 5′ ATG CAC GGC ACC CGA CCC AGC CTC GCG GAC 3′ (forward; nt 539 to nt 568) and 5′ GAC GAA TTC CGC GCA GTG GGG GCA CAC TAG ATT C 3′ (reverse; nt 814 to nt 790) (annealing temperature of 57°C); HPV-20 E6/IRES/E7 transgene cassette (1,321 bp), 5′ ATG GCT ACA CCT CCT TCT TCA GAA G 3′ (forward; nt 200 to nt 224) and 5′ TTA GGA TCC GCC ATG TTT GCA GTT C 3′ (reverse; nt 1005 to nt 981) (annealing at 65°C and elongation at 72°C for 90 seconds per cycle); and HPV-27 E6/IRES/E7 transgene cassette (1,294 bp), 5′ ATG CGC ACA AGG GCA GGG ATG TCA G 3′ (forward; nt 99 to nt 123) and 5′ CGC GCA GTG GGG GCA CAC TAG ATT C 3′ (reverse; nt 814 to nt 790) (annealing with elongation at 72°C for 150 seconds).

Analysis of E6 and E7 transgene expression.

Transcription of the E6 and E7 open reading frames of HPV-20 and HPV-27 was monitored by reverse transcriptase (RT) PCR on total RNA isolated from skin samples of first-generation PCR-positive mice. Transgenic mice and wild-type controls were sacrificed, and skin samples were collected and shock frozen. The frozen samples were powderized, and guanidine isothiocyanate lysis buffer was added immediately. RNA was extracted and purified as described by Chomczynski and Sacchi (11). Contaminating DNA was removed by digestion with RNase-free DNase. Samples of 1 μg total cellular RNA were reverse transcribed with SensiScript RT (QIAGEN, Hilden, Germany) according to the manufacturer's instructions. Expression of each sample was determined relative to the level of expression of GAPDH (glyceraldehyde-3-phosphate dehydrogenase). All samples were run in duplicate. Three highly expressing founder lines were determined for each construct and were subsequently used for breeding of the different transgenic lineages (Fig. 1B).

The primers and conditions used for RT-PCR are as follows: HPV20-E6 (497 bp), 5′ ATG GCT ACA CCT CCT TCT TCA GAA G 3′ (forward; nt 200 to nt 224) and 5′ TTA TTG AAA ATG CTT ACA CAG CCT ACA G 3′ (reverse; nt 697 to nt 670) (95°C for 30 s, 65°C for 60 s, and 72°C for 30 s for 35 cycles).; HPV20-E7 (308 bp), 5′ ATG ATT GGT AAA GAG GTC ACA TTG C 3′ (forward; nt 697 to nt 721) and 5′ TTA GGA TCC GCC ATG TTT GCA GTT C 3′ (reverse; nt 1005 to nt 981) (95°C for 30 s, 55°C for 60 s, and 72°C for 30 s for 35 cycles); HPV27-E6 (476 bp), 5′ ATG CGC ACA AGG GCA GGG ATG TCA G 3′ (forward; nt 99 to nt 123) and 5′ ATG TAA TGT CCG CGA GGC TGG GTC G 3′ (reverse; nt 575 to nt 551) (95°C for 30 s and 72°C for 120 s for 35 cycles); HPV27-E7 (275 bp), 5′ ATG CAC GGC ACC CGA CCC AGC CTC G 3′ (forward; nt 539 to nt 563) and 5′ CGC GCA GTG GGG GCA CAC TAG ATT C 3′ (reverse; nt 814 to nt 790) (95°C for 30 s and 72°C for 120 s for 35 cycles); and GAPDH (460 bp), 5′ CTT CAT TGA CCT CAA CTA CAT GGT 3′ (forward) and 5′ GCC TTC TCC ATG GTG GTG AAG AC 3′ (reverse) (95°C for 30 s, 60°C for 30 s, and 72°C for 30 s for 35 cycles).

Chronic UV-B irradiation of transgenic mice.

UV-B irradiation was performed with a Bio-Spectra system (Vilber Lourmat, Marne La Vallee, France) at a wavelength of 312 nm. Irradiation was initiated at 6 weeks of age with an initial dose of 90 mJ/cm², followed by a weekly increase of 25% until the dose reached 200 mJ/cm² (20), to allow for epidermal thickening. Mice were immobilized in order to localize the UV-B irradiation to a restricted area on the dorsal surface and were exposed three times weekly for 15 weeks. Each animal was anethesized with 3% Sevorane (Abbott, Wiesbaden, Germany) in an inhalation anesthesizer (Provet, Lyssach, Switzerland) and placed in a covered compartment with an upper square opening (3 by 2 cm) and 40 cm below the UVB bulb. Groups of HPV-20 E6/E7 (11 male and 13 female) and HPV-27 E6/E7 (12 male and 13 female) transgenic mice, as well as nontransgenic SKH-hr1 mice (9 male and 9 female), were exposed to chronic UV-B irradiation. Control groups of nonirradiated mice consisted of HPV-20 E6/E7 transgenic (10 male and 11 female), HPV-27 E7/E7 transgenic (11 male and 12 female), and nontransgenic SKH-hr1 (8 male and 7 female) mice. An observation period of 20 weeks followed the UVB irradiation period. The tumor incidence and number were recorded once a week for each mouse. One male and one female from each group (selected because they had developed papillomas) were sacrificed 24 h after UV-B treatment at weeks 7, 11, and 15 and every fifth week during the follow-up period in order to obtain tissue samples (3 by 2 cm). Skin samples, taken from the defined exposed or nonexposed dorsal surface of the mice, were partially frozen in isopentane (−20°C) and liquid nitrogen or fixed in 4% neutral-buffered formalin and paraffin embedded for histological and immunohistochemical analyses.

Skin proliferation and BrdU incorporation.

Mice were injected intraperitoneally 22 h after UV-B exposure with a filter-sterilized solution of 10% 5-bromo-2-deoxyuridine (BrdU; Sigma-Aldrich, Taufkirchen, Germany) in phosphate-buffered saline (PBS) at 50 μg/g of body weight. Mice were killed 2 h later, and the skin from the dorsal surface was fixed in 4% neutral-buffered formalin and paraffin embedded. Sections (5 μm) were stained according to the protocol provided with the BrdU immunohistochemistry system (Oncogene, San Diego, CA). BrdU incorporation was measured by counting the BrdU-positive cells within an area of 400 cells in five histologically similar microscopic fields, with magnification at ×200 (Axioplan; Carl Zeiss, Oberkochen, Germany) on each section.

Statistical analysis.

Two animals of each group (irradiated and control animals) were sacrificed in weeks 7, 11, and 15 and weeks 5, 10, 15, and 20 after cessation of UV-B irradiation in order to monitor progression during the experiment. The irradiated animals were selected based on papilloma development. This left 12 animals in the HPV-20 E6/E7 group, 13 animals in the HPV-27 E6/E7 group, and 6 nontransgenic animals which had been exposed to chronic UV-B treatment for 15 weeks. The corresponding number of untreated animals in each group was 9 HPV-20 E6/E7 transgenic, 11 HPV-27 E6/E7 transgenic, and 3 nontransgenic mice. Statistical analysis was performed for all morphologically appearing papillomas of all of the mice of the three UV-treated genotypes which remained in the long-term UV study of the experiment until week 35. The total numbers of developed papillomas, of all regressed papillomas, and of papillomas present in week 35 of the experiment were included in the analyses.

Differences among the three irradiated genotypes were scored by the Kruskal-Wallis test, and pairwise comparisons of genotypes were performed using the Wilcoxon rank sum test.

For the statistical analysis, the proportion of BrdU-labeled cells as well as the proportion of p53-expressing cells among basal cells was arcsin transformed. To account for five repeated measurements from the same animal, a linear mixed model with factor genotype and compound symmetry covariance structure was used. Comparisons were made between HPV-20 and HPV-27 transgenic mice and between HPV-20 or HPV-27 transgenic and nontransgenic animals for each time point and status of UV-B irradiation. P values of ≤0.05 were regarded as significant.

Immunohistochemical analysis of skin.

Serial sections (5 μm) of paraffin-embedded skin samples were hematoxylin-eosin stained for histological examination or stained immunohistochemically for mouse keratin 6, mouse involucrin, mouse loricrin, mouse p53, and p63 expression. p63 immunohistochemistry was performed with a MOM immunodetection kit (Vector Laboratories) using a modified protocol. Briefly, sections were deparaffinized in xylene and rehydrated in aqueous solutions with decreasing alcohol content, followed by a wash in 1× PBS-0.5% Tween 20 (pH 7.4). Antigen retrieval for the cytokeratins was achieved by boiling the slides in Dako retrieval solution (DakoCytomation, Carpinteria, CA) for 20 min. p53 retrieval was performed by boiling the slides at 750 W in a standard microwave oven in 0.01 M citrate buffer (pH 6.0) three times for 10 min each time. Slides were cooled for 20 min and washed in H₂O and 1× PBS-0.5% Tween 20 (pH 7.4). Endogenous peroxidase activity was inactivated by treatment in 3% H₂O₂ in methanol for 15 min, followed by a wash in 1× PBS-0.5% Tween 20 (pH 7.4). The samples were blocked with blocking serum (Vectastain Elite ABC kit; Vector Laboratories) during incubation in a humid chamber for 20 min. The blocking serum was removed, and the respective first antibody (diluted in 1% bovine serum albumin in PBS) was added per slide, followed by overnight incubation in a humid chamber. These first antibodies included rabbit polyclonal anti-mouse keratin 6 (1:500), rabbit polyclonal anti-mouse involucrin (1:1,000), rabbit polyclonal anti-mouse loricrin (1:500) (all obtained from Hiss Diagnostics, Freiburg, Germany), and rabbit polyclonal anti-mouse p53 (CM5; Novocastra, Great Britain). After the slides were incubated with a biotin-labeled goat anti-rabbit secondary antibody for 30 min and with Vectastain ABC solutions (Vector Laboratories), they were exposed to 3,3′-diaminobenzidine as substrate. The slides were counterstained with hematoxylin (Merck, Darmstadt, Germany) for cytokeratin and with Meyers' hemalaun (AppliChem, Darmstadt, Germany) for p53.

p63 was visualized by incubation with the mouse monoclonal anti-p63 antibody (4A4; Santa Cruz). Slides were deparaffinized, rehydrated, and washed in PBS-0.5% Tween 20 (pH 7.4), and antigen retrieval was achieved as described for p53. Samples were blocked in MOM mouse immunoglobulin-blocking reagent (MOM immunodetection kit; Vector Laboratories) for 1 h and covered with MOM diluent (MOM immunodetection kit; Vector Laboratories) for 5 min. The solutions were removed, and the slides were incubated with the p63 antibody (1:500 dilution in 1% bovine serum albumin in PBS) in a humid chamber for 1 h. Subsequent steps were performed according to the MOM immunodetection kit protocol (Vector Laboratories). Immunohistochemical staining was developed by exposure to 3,3′-diaminobenzidine, and counterstaining was performed with Meyers' hemalaun (AppliChem, Darmstadt, Germany).

p53 incorporation was measured by counting the p53-positive cells within an area of 400 cells in five histologically similar microscopic fields, with magnification at ×400 (Axioplan; Carl Zeiss, Oberkochen, Germany) on each section.

RESULTS

Generation of HPV-20 E6/E7 and HPV-27 E6/E7 transgenic mice.

The multistage development of squamous cell carcinoma of the skin induced by UV irradiation has been extensively studied in the hairless SKH-1 mouse model (reviewed in reference 6). In humans, papillomavirus-induced lesions occurring on sun-exposed sites progress to squamous cell carcinomas (24). We generated a SKH-hr1 transgenic model to examine the influence of HPV-20 and HPV-27 E6 and E7 gene expression on the differentiating layers of the skin during and after chronic UV-B irradiation. Hairless SKH-hr1 mice were chosen for the study to ensure direct penetration during the UV-B irradiation and to minimize injury which could occur during the regular shaving of hair. Three heterozygous lineages were generated for each transgene. The respective papillomavirus genes were expressed under the control of the K10 promoter. The E6 and E7 genes were separated by an IRES in order to obtain full expression of both proteins (Fig. 1A). The expression of the viral E6 and E7 genes in the skin was analyzed by RT-PCR (Fig. 1B). Immunohistochemical staining using antibodies directed against the E6 and E7 proteins of HPV-20 and HPV-27, respectively, proved unsatisfactorily and therefore was not evaluated (data not shown). All organs were examined histologically, but no specific phenotype was observed for the transgenic founders in comparison to nontransgenic SKH-hr1 mice, nor did any spontaneous tumors develop during 2 years of observation of either the transgenic or the nontransgenic mice.

Increased tumor development after chronic exposure to UV-B irradiation.

Chronic irradiation on SKH-hr1 hairless mice has been shown to result in skin tumors after 8 weeks (42). We irradiated the two transgenic lines and the nontransgenic mice with UV-B three times weekly for 15 weeks and by increasing the dose weekly by 25%. Multiple sessile and exophytic papillomas developed in the irradiated animals of both transgenic mouse lineages as well as in nontransgenic SKH-hr1 mice. Twelve mice of each irradiated genotype were sacrificed during the course of the experiment in order to monitor the irradiated skin by immunohistochemical analysis. Tumor development in these mice could not be analyzed statistically because papilloma formation had been a selection criterion for sacrifice. Biological differences were nevertheless noted among the three genotypes regarding tumor development (Table 1). A total of 20 papillomas developed in six of the HPV-20 E6/E7 transgenic mice sacrificed during the course of the study compared to 16 papillomas in seven of the HPV-27 E6/E7 transgenic mice and 6 papillomas in four irradiated nontransgenic control animals. The first papilloma was observed in the HPV-27 E6/E7 transgenic mice during week 12 of chronic UV-B exposure, whereas papilloma development started only in week 16 (i.e., 1 week after cessation of UV-B exposure) in the other two groups. Two squamous cell carcinomas were observed in two HPV-20 E6/E7 transgenic mice (Fig. 2) and in one HPV-27 E6/E7 transgenic mouse. Interestingly, two fibropapillomas developed in the HPV-27 E6/E7 animals and a fibropapilloma as well as a rarely observed mast cell tumor developed in the HPV-20 E6/E7 transgenic mice.

TABLE 1.

Tumor development in transgenic and nontransgenic mice

Group	Sacrificed animals		Animals in long-term study
	No. of animals	Total no. of papillomas/no. of animals with papillomas	No. of animals	Total no. of papillomas/no. of animals with papillomas	Mean no. of papillomas/ animal (±SD)	Total no. of regressed papillomas	Mean no. of regressed papillomas/ animal (±SD)	Total no. of papillomas in 35th week	Mean no. of papillomas in 35th week (±SD)
Nontransgenic mice	12	6/4	6	8/5	1.33 (±1.03)	7	1.17 (±0.98)	1	0.17 (±0.41)
HPV-20 E6/E7 transgenic mice	12	20/6	12	19/5	1.58 (±2.27)	6	0.5 (±1.0)	13	1.08 (±1.62)
HPV-27 E6/E7 transgenic mice	12	16/7	13	10/3	0.77 (±1.79)	6	0.46 (±1.2)	4	0.3 (±1.63)

FIG. 2.

Histology of paraffin-embedded sections of tumors of HPV-20 E6/E7 and HPV-27 E6/E7 transgenic mice stained with hematoxylin and eosin (magnification, ×100). (a) Representative exophytic papilloma with strong epidermal thickening, hyperkeratosis, and highly accumulated perspiratory glands in an HPV-27 E6/E7 transgenic mouse (no. 518, week 30 of experimentation) (b) Papilloma with anaplastic foci, partial loss of discrimination of epithelial and collagenous structures as well as multiple mitosis, and keratotic follicular hyperplasia. (c) Invasive squamous cell carcinoma with multiple proliferating islands with central hornification as well as concentric lamination of collagenous parts and infiltrative growth into deep areas of the subcutis in HPV-20 E6/E7 transgenic mice (no. 7360, week 30; no. 393, week 25).

Papilloma incidence in the long-term study of all three genotypes.

Although a weak tendency for increased papilloma induction in the HPV transgenic lines was evident in the Kaplan-Meier plot (data not shown), the log rank test did not show statistical evidence (P = 0.274). The reason for this may be twofold. The animals sacrificed during the course of irradiation were probably more prone to tumor development because they had previously developed papillomas but were removed from the experiment. This left a small number of animals, developing papillomas at a later stage, completing the long-term UV study.

Comparative statistical analysis of all irradiated animals in the long-term UV study showed a difference in the number of regressed papillomas between the genotypes (P = 0.0388, Kruskal-Wallis test). A significant difference between the HPV-27 E6/E7 transgenic mice and nontransgenic mice was measured in the pairwise comparison (P = 0.0091, Wilcoxon rank sum test) when all three genotypes were compared.

Chronic UV-B exposure enhances proliferation in HPV-20 E6/E7 and HPV-27 E6/E7.

Expression of keratin 6 is associated with epidermal hyperproliferation (30). We performed immunohistochemical staining for keratin 6 on skin samples taken from the dorsal surface of irradiated and control mice throughout the study period. Differences in the expression patterns between the nontransgenic mice and both of the transgenic lines, HPV-20 E6/E7 and HPV-27 E6/E7, were observed at different time points during the chronic UV-B exposure as well as during the observation period after cessation of irradiation. Keratin 6 was highly expressed in all epidermal layers of the HPV-20 E6/E7 transgenic mice during week 7 of irradiation (Fig. 3), in comparison to a varying, patchy expression in the HPV-27 E6/E7 transgenic mice. The staining pattern in the nontransgenic mice was much weaker and more limited to the basal and suprabasal layers. Keratin 6 expression remained more pronounced in the transgenic lines during the subsequent weeks of irradiation, with a clear staining throughout the squamous cell and granular layers as well as in numerous basal cells. It also remained up-regulated throughout the 20-week period after cessation of irradiation, mainly in the basal and spinous layers. Expression of keratin 6 in the nontransgenic mice reverted gradually after cessation of irradiation. No differences were observed between the nontransgenic and the transgenic mice in the nonirradiated control groups.

FIG. 3.

Immunohistochemical staining of keratin 6 after (a to c) 7 weeks of UVB exposure (+UV) and (d to f) in the control nonirradiated animals (−UV). (a) Patches of positive-staining cells in the basal, spinous, and granular layers of the dorsal skin of nontransgenic mice, (b) an almost continuous positive staining throughout all layers in the HPV-20 E6/E7 transgenic mice, and (c) a more pronounced staining in the suprabasal cells in the HPV-27 E6/E7 transgenic mice were observed.

We further monitored the proliferation activity by scoring DNA synthesis by BrdU incorporation. Mice were pulse-labeled with BrdU for 24 h prior to sampling. The number of BrdU-positive cells in a field of 400 cells was determined in five representative skin locations per section from two irradiated and two nonirradiated animals of each experimental group at weeks 7 (Fig. 4A), 11, and 15 and week 10 after cessation of irradiation (Fig. 4B). The results are presented graphically in Fig. 4C. Differences were statistically significant for all three genotypes (P = 0.025) at week 7 of UV-B irradiation. A higher proliferation was measured in the HPV-20 transgenic mice than in the HPV-27 transgenic mice (P = 0.050) and nontransgenic mice (P = 0.059). These distinctions were not statistically significant during the period that followed. Significant differences between all three genotypes (P = 0.033) were again observed at 10 weeks after cessation of chronic UV-B exposure (25th week of experimentation). Here, the proliferation was again higher in the HPV-20 irradiated transgenic mice than in their nontransgenic (P = 0.052) and HPV-27 transgenic (P = 0.084) counterparts.

FIG. 4.

Proliferation activity as measured by BrdU incorporation in (a to c and g to i) mice after chronic UV-B exposure (+UV) and (d to f and j to l) nonirradiated mice (−UV). (A) DNA synthesis was increased during week 7 of exposure in (b) the HPV-20 E6/E7 mice compared to (a) the nontransgenic and (c) the HPV-27 E6/E7 transgenic mice (magnification, ×400). (B) After cessation of UVB irradiation (week 25), DNA synthesis diminished in (g) the nontransgenic and (i) HPV-27 E6/E7 mice but stayed up-regulated in (h) the HPV-20 E6/E7 transgenic mice. (C) Graphical presentation (box plots) of the quantification of BrdU-positive cells/400 cells counted in representative histological sites for each genotype and time point in the irradiated and nonirradiated groups. Statistical differences were analyzed in a linear mixed model.

Reduced skin differentiation in HPV-20 E6/E7 and HPV-27 E6/E7 transgenic mice.

On the whole, no obvious differences between the transgenic lines and nontransgenic mice were observed by histomorphological examination of the skin during the period of chronic UV-B exposure. The epidermal changes resembled those seen in actinically damaged skin in humans. Hyperplasia was observed starting at week 7 of chronic UV-B irradiation. Intraepidermal proliferation increased over time to reach an average of 10 to 15 cell layers in week 15 of UV-B irradiation, after which the epidermal thickness decreased again within the first 5 weeks after cessation of irradiation to an average of 6 to 8 cell layers. Hyperkeratosis was very distinct at this point, with single areas of hyperplasia, anaplastic basal cells, and multiple papillomatous foci. Differences between the transgenic and the nontransgenic mice became evident during the 20-week observation period after cessation of UV-B exposure. A final decrease to three to four cell layers was already observed in week 10 after cessation of irradiation in the nontransgenic mice, whereas this was seen only in week 20 after cessation in the transgenic mice. Larger areas of mild hyperproliferating epidermis (five to six cell layers) with moderate hyperkeratosis and subepithelial fibrosis were seen in the transgenic mice. These changes were not observed in the nontransgenic mice.

Both loricrin and involucrin are major components of the cornified cell envelope barrier structure of the epidermis (54). Involucrin is regularly expressed in the upper suprabasal cell layers of the stratum granulosum during early stages of epithelial differentiation, whereas loricrin expression is usually bound to the upper layers of the keratinized layer (54). We monitored the expression of these differentiation markers during the course of the study. Epithelial differentiation decreased in the transgenic lines during increased UV-B exposure, in contrast to that observed with the irradiated nontransgenic mice. Regular involucrin expression was observed mainly throughout the granular and keratinizing layers in the HPV-27 E6/E7 transgenic line and the nontransgenic mice, whereas staining of this protein extended down into the spinous cell layer in the HPV-20 E6/E7 transgenic mice at week 7 of irradiation. The strong involucrin staining in the nontransgenic mice was maintained throughout the period of irradiation (Fig. 5A). Involucrin expression was limited to single cells in the HPV-20 E6/E7 transgenic mice in week 15, as well as during the subsequent 8 weeks of irradiation, and was generally weaker than in the nontransgenic mice. This patchy pattern extended through the granular and squamous cell layers, reaching down in single cells of the basal cell layer (Fig. 5A). A similar single-cell distribution of involucrin-expressing cells was observed in the HPV-27 E6/E7 transgenic mice at week 15 of irradiation.

FIG. 5.

Disruption of the differentiation markers involucrin (A) and loricrin (B) in both transgenic lines upon chronic UVB exposure. After 15 weeks of irradiation, a regular strong expression of involucrin occurred in (a) the nontransgenic mice, in contrast to a patchy pattern of single cells extending through the granular and squamous cell layers in (b) the HPV-20 E6/E7 and (c) the HPV-27 E6/E7 transgenic mice. Involucrin was evenly expressed in (d to f) the keratinizing layers of the skin of the corresponding control mice. In week 15 of UV irradiation, loricrin expression (g) extended to the basal cells of the granular layer in nontransgenic mice and (h) occurred in a few cells with a tessellated character in the stratum corneum and stratum granulosum of the HPV-20 E6/E7 transgenic mice, and (i) a stronger expression was restricted to single scattered areas of the upper granular and keratinizing layers in the HPV-27 E6/E7 mice. Normal loricrin expression was observed in the corresponding control mice which did not receive any UV irradiation (j to l) (magnification, ×200).

Differences in loricrin expression were most pronounced in week 11 of UV irradiation. Loricrin expression extending to the basal cell layers of the stratum granulosum remained until week 15 of UV irradiation in the nontransgenic mice. In contrast, only a few single cells expressed loricrin with a tessellated character in the stratum corneum and stratum granulosum of the HPV-20 E6/E7 transgenic mice (Fig. 5B), and loricrin was almost lost or rarely found in the keratotic hyperplastic follicles (data not shown). Loricrin expression in the HPV-27 E6/E7 transgenic mice was significantly stronger than in the HPV-20 E6/E7 transgenic mice but was also restricted to single scattered areas of the upper cell layers of the stratum spinosum. This expression remained restricted to single cells extending into the upper granular and keratinizing layers.

p53 expression is increased in irradiated HPV-20 E6/E7 and HPV-27 E6/E7 transgenic mice.

Transient p53 accumulation is a physiological response to a single dose of UV-B irradiation (43). Chronic UV-B exposure, however, induces constitutive p53 alterations (4, 42, 46). We analyzed p53 expression in skin samples taken from all three groups of irradiated mice 24 h after UV-B irradiation in weeks 7, 11, and 15 of exposure as well as at four time points during the subsequent observation period. Sections were stained with the CM-5 antibody which detects wild-type as well as mutated p53. The number of cells staining positive for p53 was evenly spread and higher in the transgenic mice than in the nontransgenic mice during week 7 of irradiation. Significant differences between the transgenic lines and the nontransgenic irradiated mice were observed during week 11 of chronic UV exposure (P = 0.022) (Fig. 6A and B). An increased number of p53-expressing cells were already present in the basal cell layer at week 7 of irradiation in the HPV-20 E6/E7 transgenic mice. The basal cell layer almost uniformly expressed p53 in week 11, in addition to some p53-positive-staining clusters of cells in the spinous layer. This distinction reached significance compared to that observed with the nontransgenic mice (P = 0.023) and was higher than in the HPV-27 transgenic mice. This was followed by virtually every basal cell showing p53 expression in the transgenic mice at the end of UV-B irradiation at week 15, in addition to the numerous p53-expressing cell clusters reaching into the squamous cell layers. p53 expression in the HPV-27 E6/E7 transgenic mice showed differences as early as week 7 of irradiation. p53-expressing cell clusters were visible in addition to the positive-staining basal cells. This difference was significant compared to that observed with the nontransgenic mice (P = 0.028) (Fig. 6B). In week 11 of UV-B exposure, p53 expression appeared rather irregular, with clusters of intense positive-staining cells in the basal layers which extended into the upper part of the squamous cell layer (Fig. 6A). Numerous p53-expressing cells were present in the basal and suprabasal cell layers in week 15 of exposure. p53 expression in all three groups reverted after cessation of UV-B irradiation to a unified pattern of single scattered p53-positive-staining basal cells and single clusters of positive cells in the areas of moderate hyperproliferation. No differences in p53 expression were observed between control nonirradiated transgenic mice and nontransgenic animals (Fig. 6A).

FIG. 6.

(A) p53 expression during week 11 of UV irradiation. (a) Single p53-positive-staining clusters of cells were observed in the basal cell layer in the nontransgenic mice, and (b and c) almost every basal cell showed p53 expression, with clusters reaching into the squamous cell layers (arrows) in the transgenic mice. No obvious differences were observed between control nonirradiated transgenic (e and f) and nontransgenic (d) mice. (B) Graphical presentation (box plots) of the quantification of p53-positive cells/400 cells counted in representative histological sites for each genotype and time point in the irradiated and nonirradiated groups. Statistical differences were analyzed in a linear mixed model.

p63 expression is reduced in the HPV-20 E6/E7 transgenic mice in comparison to the HPV-27 E6/E7 transgenic mice.

p63, a p53 homologue, is essential for regenerative proliferation and epithelial stratification during development as well as for maintenance of the proliferative potential of the basal keratinocytes (32, 45). We used the 4A4 antibody for immunohistochemical staining, which recognizes all isoforms of p63. p63 was expressed in the nuclei of the basal and suprabasal cells of the epidermis of the nontransgenic lines as well as the two transgenic lines in week 15 of UV-B exposure (Fig. 7A). In addition, single patches of strong cytoplasmic staining for p63 were present in the upper layer of the stratum spinosum of the HPV-20 transgenic mice. In contrast, only the basal cells of the nonirradiated control mice expressed p63. Large differences between groups were also noted during the observation period after cessation of chronic UV-B irradiation (Fig. 7B). p63 expression was still present in the basal cell layer of the nontransgenic mice, but an additional marked cytoplasmic expression was observed in the stratum granulosum and stratum corneum, in contrast to a very patchy distribution of these cells in the epidermis of the HPV-20 transgenic mice. The pattern in the HPV-27 transgenic mice differed at this time point in that expression in the basal and suprabasal cells was more pronounced, with a lower number of cytoplasmic staining patches in the stratum corneum (data not shown).

FIG. 7.

(A) p63α was expressed as demonstrated by immunohistochemical staining in the nuclei of the basal and suprabasal cells of the epidermis of the (a) nontransgenic mice and (c) HPV-27 E6/E7 transgenic mice at week 15 of UV-B exposure. (b) p63α staining extended to single patches of strong cytoplasmic staining in the upper layer of the stratum spinosum in the HPV-20 transgenic mice (magnification, ×200). (B) Immunohistochemical staining of p63 at week 15 after cessation of UV treatment. (g) Nuclear expression of p63 in the basal cell layer, with cytoplasmic staining in the stratum granulosum and stratum corneum of the nontransgenic mice, and (h) a patchy distribution of positive-staining cells in the HPV-20 E6/E7 transgenic mice were observed. (i) More-pronounced p63 staining in the HPV-27 E6/E7 transgenic mice in the basal and suprabasal cells (magnification, ×400). Regular staining was observed in the corresponding nonirradiated mice (d to f and j to l).

DISCUSSION

The functional role of UV irradiation in combination with the E6 and E7 proteins of the cutaneous HPV types in the malignant conversion of benign papillomatous lesions has not been elucidated. We generated transgenic mice expressing the HPV-20 and HPV-27 E6 and E7 genes, respectively. The majority of reported studies using transgenic mice expressing the E6 and E7 proteins of the genital HPV types as well as the one study on HPV-8 transgenic mice made use of the FVB/N mouse strain. Malignant conversion of keratinocytes, as well as tumor formation, was more efficient in this strain than in mice with other genetic backgrounds (56). We chose the SKH-hr1 hairless mouse strain for the generation of our transgenic mice because it has in the past been used extensively as a model for squamous cell carcinoma of the skin by UVB irradiation (16, 20, 46). The use of the K10 promoter enabled us to investigate whether cells which have left the stem cell compartment could be influenced by expression of the E6 and E7 genes of HPV-20 and HPV-27. However, the genetic background of the SKH-hr1 mice may have contributed to our failure to generate transgenic lines using the E6/E7 genes under the influence of the K14 promoter as well as to the rather modest effects of the transgenes on malignant development.

Interaction of the viral proteins and chronic UV-B irradiation led to an enhanced proliferation of the epidermal layers of the skin of both transgenic mouse lines. This was reflected in the higher rate of papilloma formation and an earlier onset in increased DNA synthesis, as measured by BrdU incorporation in the transgenic mice, than in the irradiated nontransgenic mice. Although the statistical values on tumor formation did not reach significance due to the small numbers of animals completing the long-term UV study and the selection for papilloma-bearing animals being sacrificed during the course of UV-B irradiation, some correlation was evident. Malignant skin lesions were observed in both transgenic lines but not in the nontransgenic mice. Squamous cell carcinomas developed in two HPV-20 transgenic mice and one HPV-27 transgenic mouse. The higher numbers of papillomas per animal observed in the HPV-20 and HPV-27 transgenic lines than in the nontransgenic lines were also mirrored in the increased BrdU incorporation in both transgenic lines during the initial phase of the study. Interestingly, more papillomas regressed and at an earlier time point in the nontransgenic and the HPV-27 E6/E7 transgenic mice than in the HPV-20 E6/E7 transgenic mice, which is reflected in the recurrent increased BrdU incorporation 10 weeks after cessation of UV-B exposure (25th week) in the HPV-20 transgenic line in contrast to that observed in the nontransgenic and the HPV-27 transgenic line. One of the proinflammatory cytokines induced by UV-B irradiation in keratinocytes, IL-1, is known to induce keratin 6 expression (28) via the induction of p38, which in turn phosphorylates keratin 6 (55). IL-1 is up-regulated in HPV-infected warts (25). Keratin 6 was increased in all cell layers following UV-B exposure in both transgenic mouse strains but only in scattered cells in the nontransgenic mice. The expression of this hyperproliferation marker (30) remained up-regulated in the transgenic mice even after cessation of chronic UV-B exposure, similar to what has been reported for SCC of the skin (49). This indicates that the expressions of the E6 and E7 genes of these two HPV types are able to exert a prolonged influence on the proliferation of the differentiating cell layer.

The concerted action of viral gene expression and chronic UV-B exposure also affected differentiation of the epidermal cell layers in both transgenic mouse strains. Proliferation and differentiation are coupled and spatially segregated in the normal epidermis. A disturbance in this control may be correlated with an invasive phenotype (9). Expression of both involucrin and loricrin was reduced and disturbed by the expression of the viral E6 and E7 genes in the presence of UV-B irradiation. Loricrin was expressed at very low levels in a few cells in the HPV-20 E6/E7 mice. The same pattern was observed in the HPV-27 E6/E7 mice, although the level of expression was slightly higher. Involucrin and loricrin expression is modulated by the mitogen-activated protein kinase cascades, which may be induced by UV-B irradiation and which, in turn, regulate promoter activity through the transcription factors Sp-1 and AP-1 (17, 27). The E6 and E7 genes of both HPV-20 and HPV-27 seem to influence these pathways, as the expressions of these differentiation markers in the nontransgenic mice, which were irradiated in the same way, were not similarly modified. A link between both involucrin and loricrin expression and the late viral proteins of a number of HPV types has been reported (34, 40, 47), but the role of the early proteins of the papillomaviruses is unknown.

Another group of proteins which plays an important role in epithelial stratification during development, as well as in the maintenance of a proliferative potential of the basal keratinocytes in the mature epidermis, is p63, a p53 family member (31, 41). The two p63 isoforms, TAp63α and ΔNp63α, counterbalance each other. TAp63α is an inhibitor of keratinocyte differentiation, as is evidenced in transgenic mice, where induction of TAp63α expression, in the presence of reduced levels of ΔNp63α, resulted in epidermal fragility. Loricrin expression was also absent (31). The number of p63-expressing cells in squamous cell carcinomas is related to the degree of differentiation. They are more abundant, and their distribution is chaotic in poorly differentiated tumors (44). In our study, chronic UV-B exposure induced expression of p63 in basal and suprabasal cells of all three groups of mice, as observed in week 15 of irradiation. Faint cytoplasmic p63 staining was also noticed in single cells of both transgenic lines at this time. Interestingly, additional single patches of cells highly expressing p63 in the cytoplasm were observed at this time point in the stratum granulosum of the skin of the HPV-20 E6/E7 transgenic mice. A more pronounced difference in p63 expression between the three lines was seen 15 weeks after cessation of chronic UV-B exposure. High levels of cytoplasmic staining appeared in clusters of cells in the granular layer in the transgenic lines, in contrast to uninterrupted staining in the granular layer in the nontransgenic mice. p63 staining in basal and suprabasal cells was present in all three groups, although the number and intensity were more pronounced in the transgenic lines—even more so in the HPV-27 E6/E7 mice than in the HPV-20 E6/E7 mice. Unfortunately, the available antibody did not distinguish between the different isoforms of p63. In vitro studies have demonstrated a differentiated modulation of the promoters of these HPV types by both TAp63α and ΔNp63α. This modulation is directly influenced by the expressions of the respective E6 and E7 proteins (19; J.-W. Fei, H. Hohmann, and E.-M. de Villiers, unpublished data).

UV-B irradiation increases the number of cells and prolonged the period of immunoreactivity to p53 in SKH-1 mice (4, 5). The induction of p53 mutations by chronic UV-B exposure and its association with the development of malignant tumors have been demonstrated in these mice (4, 42, 46). In our study, p53 expression measured under chronic UV exposure in transgenic skin was reflected not only in the number of expressing cells but also in the distribution of these cells. The clustering of the p53-expressing cells in the nontransgenic and the HPV-27 transgenic mice could be indicative of UV-mutated p53 (4), although it contrasted the equal extended distribution of p53-expressing cells seen in the HPV-20 transgenic line which had received UV irradiation simultaneously. The HPV genes were expressed in the suprabasal layers, whereas the majority of cells with increased p53 expression were observed in the basal cell layer. p53 expression reverted to an equal expression after cessation of chronic UV-B exposure.

This in vivo study provides insight into the functions of the E6 and E7 proteins of both HPV-20 and HPV-27. These early proteins of both HPV types influenced proliferation as well as differentiation of the differentiating layers of the mouse epidermis when expressed under the influence of chronic UV-B exposure. The model of chronic UV irradiation used here resembles humans who are exposed to high doses of UV light early in life, with reduced UV-B levels later in life. Additional infection with papillomavirus results in the expression of the early viral proteins, which then leads to the development of nonmelanoma skin cancer in the predisposed cells after several decades. Our previous reports on the differential modulation of individual papillomavirus types by UV irradiation and proinflammatory cytokines (14, 48) were substantiated in the present study by the differences observed between the two transgenic mouse strains. Although these differences were in most instances subtle, they nevertheless provide a basis for future in vitro experiments to determine how infections with HPV-20 and HPV-27 influence the pathways involved in epidermal differentiation and malignant progression.