06-3344 1626.1635

The Human Papillomavirus E6 Oncogene Dysregulates the CellCycle and Contributes to Cervical Carcinogenesis throughTwo Independent Activities Anny Shai,1 Tiffany Brake,1 Chamorro Somoza,2 and Paul F. Lambert1 1McArdle Laboratory for Cancer Research, University of Wisconsin School of Medicine and Public Health,Madison, Wisconsin and 2Arbor Vita Corporation, Sunnyvale, California we generated K14E6WT and K14E7WT transgenic mice expressingeither the HPV16 E6 or the HPV16 E7 oncogene, respectively.
Cervical cancer is a leading cause of death due to cancer The human keratin-14 promoter was used to direct transgene among women worldwide. Using transgenic mice to dissect the expression to the basal layer of the stratified squamous epithelium contributions of the human papillomavirus (HPV) 16 E6 and lining the skin, oral cavity, and reproductive tract (5–7). K14E6WT E7 oncogenes in cervical cancer, E7 was identified previously and K14E7WT transgenic mice display many of the known activities to be the dominant oncogene. Specifically, when treated with of each oncogene identified in tissue culture, including the ability exogenous estrogen for 6 months, E7 transgenic mice of E6 and E7 to inactivate p53 and pRb, respectively. Furthermore, developed cancer throughout the reproductive tract, but E6 these HPV16 transgenic mice develop tumors in the skin either transgenic mice did not. E6 contributed to carcinogenesis of spontaneously or with increased efficiency when induced chemi- the reproductive tract, as E6/E7 double transgenic mice cally with the carcinogens, 7,12-dimethylbenz( treated for 6 months with estrogen developed larger cancers than E7 transgenic mice. In the current study, we investigated O-tetradecanoylphorbol-13-acetate (5, 6, 8).
Prior studies showed that both the K14E6WT/K14E7WT trans- whether the E6 oncogene alone could cooperate with estrogento induce cervical cancer after an extended estrogen treat- genic mice and the K14E7WT singly transgenic mice developed ment period of 9 months. We found that the E6 oncogene cervical cancer following estrogen treatment for 6 months.
synergizes with estrogen to induce cervical cancer after However, similarly treated K14E6WT transgenic mice developed 9 months, indicating that E6 has a weaker but detectable only low-grade dysplasia (7). Estrogen is a cofactor in cervical oncogenic potential in the reproductive tract compared with carcinogenesis in this mouse model, as untreated K14E6WT/ the E7 oncogene. Using transgenic mice that express mutant K14E7WT or K14E7WT mice did not develop cancer. Subsequent forms of HPV16 E6, we determined that the interactions of E6 studies indicated that estrogen is required for multiple stages of with cellular A-helix and PDZ partners correlate with its cervical carcinogenesis (9). Reproductive tumors arising in the ability to induce cervical carcinogenesis. In analyzing the K14E6WT/K14E7WT transgenic mice were more aggressive than tumors arising in E6 transgenic mice, we learned that E6 those arising in the K14E7WT transgenic mice, indicating that the induces expression of the E2F-responsive genes, Mcm7 and E6 oncogene contributed to the malignant progression.
cyclin E, in the absence of the E7 oncogene. E6 also prevented In the current study, we investigated whether the E6 oncogene the expression of p16 in tumors of the reproductive tract could cooperate with estrogen to induce cervical cancer given an through a mechanism mediated by the interaction of E6 with extended (9 months) treatment period. To examine the mecha- A-helix partners. [Cancer Res 2007;67(4):1626–35] nism(s) by which the E6 oncogene contributes to cervical can-cer, we monitored cervical carcinogenesis in K14E6I128T andK14E6D146-151 mice, which express mutant forms of the HPV16 E6 oncogene (10, 11). K14E6I128T transgenic mice express a mutant Cervical cancer is the second most common type of cancer form of E6 greatly reduced in its ability to bind a-helix partners.
among women, with high mortality rates worldwide, despite Specifically, E6I128T protein binds the a-helix partners, E6AP and increased screening efforts (1). Human papillomavirus (HPV) E6BP, at 1% to 5% the levels of wild-type (WT) E6 protein (12).
infection contributes to nearly all cases of cervical cancer based E6AP or UBE3A belongs to the HECT family of E3 ubiquitin ligases on the observed presence of HPV DNA within these cancers (2) and (13) and is normally associated with the human neurologic more than half of the HPV-associated cervical cancers are disorder, Angelman's syndrome (14, 15). E6AP is thought to be attributed to infection with HPV16 (2, 3).
the primary cellular factor mediating the degradation of p53 by E6 Two viral genes, E6 and E7, are commonly expressed in cervical (16) and is thus a potentially important partner in mediating the cancer. In tissue culture, E6 and E7 display properties of oncogenic activities of E6. Correspondingly, K14E6I128T transgenic oncogenes, including the ability to immortalize and transform mice are defective for inactivating p53 (17). K14E6D146-151 trans- cells (4). To assess the oncogenic properties of these genes in vivo, genic mice encode a mutant E6 protein defective for interactingwith PDZ partners (18), such as DLG and Scribble, two genesknown to be tumor suppressors in Drosophila. We have previouslyused K14E6I128T and K14E6D146-151 transgenic mice to show a role Requests for reprints: Paul F. Lambert, McArdle Laboratory for Cancer Research, of the a-helix and PDZ domain partners of E6, respectively, in University of Wisconsin School of Medicine and Public Health, Madison, WI 53706.
mediating the oncogenic potential of E6 in the skin (10, 17, 19).
Phone: 608-262-8533; Fax: 608-262-2824; E-mail: [email protected].
These studies provide a framework for our studies of cervical I2007 American Association for Cancer Research.
doi:10.1158/0008-5472.CAN-06-3344 carcinogenesis studies reported herein.
Cancer Res 2007; 67: (4). February 15, 2007 Multiple Properties of E6 Contribute to Cervical Cancer We found that the E6 oncogene synergizes with estrogen to induce lesion scored as the final diagnosis. The tumors were measured with the cervical cancer after an extended estrogen treatment. Specifically, Zeiss Axiovision (version 3.1) program (Zeiss, Thorwood, NY). Any tumor K14E6WT mice, treated with estrogen for 9 months, developed with an area >2,000 Am2 was classified as a large cancer.
cervical cancers at an increased frequency compared with non- Quantification of bromodeoxyuridine. To quantify the amount of basal DNA synthesis, the total number of bromodeoxyuridine (BrdUrd)– transgenic mice. Compared with K14E6WT mice, K14E6I128T and positive basal cells was counted and divided by the total number of basal K14E6D146-151 mice in the absence and/or the presence of E7 cells and multiplied by 100 to determine the percentage. To quantify the displayed reduced oncogenic potential in the reproductive tract.
amount of epithelial hyperplasia, the total number of suprabasal BrdUrd- We also evaluated mouse reproductive tracts and their positive cells were counted and divided by the total number of cells and associated tumors for biomarkers commonly used for the multiplied by 100 to determine the percentage. BrdUrd was counted in diagnosis of human cervical cancers. Biomarkers included the eight, !40 microscopic fields per mouse, with a total of at least three or E2F-responsive genes Mcm7, involved in DNA replication (20), more mice per genotype group.
and cyclin E, involved in the G Statistical analysis. The Fisher's exact test was used to determine the 1-S transition (21, 22). We also monitored expression of the cyclin kinase inhibitor, p16. p16 is a significance in tumor incidence. The Wilcoxon rank-sum test was used to biomarker for HPV-associated cervical lesions and cancers (23).
determine the significance in BrdUrd quantification and in measurementsof tumor size and area. Statistical analysis was carried out using the MSTAT Finally, tumors were evaluated for p53 expression. In most cases, biomarker expression in the lesions and tumors from our Immunohistochemistry. Sections were prepared for immunohisto- nontransgenic and transgenic mice mirrored results observed in chemical analysis as described previously (7). For BrdUrd, cyclin E, pRb, and human cervical samples. Of particular interest, lesions from our p16 analysis, the slides were immersed in 2N HCl for 20 min to unmask K14E6WT and K14E6mutant (refers to both K14E6I128T and further. Primary antibody was applied to the sections at 1:100 for BrdUrd K14E6D146-151 herein) mice showed an up-regulation of the (Calbiochem, Darmstadt, Germany); 1:200 to 1:500 in blocking solution for E2F-responsive genes, Mcm7 and cyclin E, even in the absence p53 (CM5; Novocastra, Norwell, PA); and 1:50 for MCM7 (NeoMarkers, of E7, but at lower levels than in tumors arising in E7-expressing Fremont, CA), cyclin E (M-20; Santa Cruz Biotechnology, Santa Cruz, CA), mice. In contrast, we observed an inverse relationship between pRb (BD Biosciences, San Jose, CA), and p16 (M156; Santa Cruz the expression of p16 and pRb in the reproductive tumors from Biotechnology) overnight at 4jC. A universal biotinylated secondaryantibody was applied and developed.
K14E6WT versus K14E7WT mice. Biomarker expression in tumorsarising in K14E6WT/K14E7WT mice were similar to tumors arisingin K14E7WT mice, showing that E7 is the dominant oncogene in deregulating the p16/pRb pathway during cervical carcinogenesis.
K14E6WT and K14E6mutant transgenic mice express physio- In summary, this study shows that two properties of E6 logic levels of E6 protein in the epithelia of the skin and contribute to the development of cervical cancer. These reproductive tract. Prior detection of HPV16 E6 protein was contributions lead to a distinct pattern of dysregulation of cell difficult due to an absence of adequately sensitive antibodies. A cycle regulatory genes compared with that seen in E7-expressing recently generated HPV16 E6–specific antibody (24) allowed us to detect and compare levels of E6 protein expressed in ourtransgenic mice with cell lines derived from human precancerous Materials and Methods cervical intraepithelial neoplasia (CIN) lesions and cervical cancers(Fig. 1). HPV16 E6 protein was detected in the HPV16-positive SiHa Mouse lines and estrogen treatment. The K14E6WT (5), K14E6I128T and Caski but not in the HPV18-positive HeLa and HPV-negative (17), K14E6D146-151 (10), and K14E7WT (6) lines were all maintained on aheterozygous FVB background. K14E6WT, K14E6I128T, and K14E6D146-151 C33a cervical cancer cells (Fig. 1A). E6 protein was also detected in mice were mated to K14E7WT mice to generate double transgenic mice.
W12 clonal cell lines derived from a HPV16-positive CIN1 lesion.
Female mice were treated with 17-h estradiol as described previously (7) Dorsal skin from 9-day-old WT and mutant E6 transgenic mice for 9 months. Untreated control mice were held for the same period. All expressed E6 protein at levels slightly higher than SiHa cells, mice were bred and maintained in the American Association for whereas expression in adult mice were lower (Fig. 1B; data not Accreditation of Laboratory Animal Care–approved McArdle Laboratory shown). This result is consistent with our prior observations that Animal Care Facility in accordance with an institutionally approved K14-directed transgene expression is maximal in young pups and animal protocol.
wanes in adults (25). In homozygous K14E6WT transgenic mice, the Quantification of E6 levels. Five-week-old female mice were treated level of E6 protein was approximately half of the amount of E6 with estrogen, and reproductive tracts were harvested after 6 weeks of expressed in Caski cells (Fig. 1B). E6 protein was detected in the treatment to obtain a state of constant estrus in all of the mice. Dorsal skinwas also harvested at the time of sacrifice in addition to skin from 9-day-old reproductive tracts of all adult K14E6WT and K14E6mutant mice, at mice, as transgene expression is highest during this period in the skin. The levels lower than in both SiHa and Caski cells (Fig. 1C). Given that tissue was placed in cold HNTG lysis buffer [50 mmol/L HEPES (pH 7.5), <10% of the total protein from the harvested reproductive tract 150 mmol/L NaCl, 1.1% Triton X-100, 1 mmol/L EGTA, 10% glycerol, tissues comes from the stratified epithelium, we conclude that our 1 mmol/L phenylmethylsulfonyl fluoride, 1! PIC] and homogenized.
K14E6WT transgenic mice express E6 protein at levels similar to Protein lysates (100–200 Ag) were separated by SDS-PAGE, transferred, that observed in human cervical cancer. K14E6D146-151 mice and immunoblotted with a monoclonal HPV16 E6 antibody (24) at 5 Ag/mL.
expressed f1.5 times the amount of E6 protein relative to A mouse IgG secondary conjugated to horseradish peroxidase (Jackson K14E6WT mice in the reproductive epithelium (Fig. 1C). K14E6I128T Immunoresearch, West Grove, PA) was used at 1:10,000. Detection of E6 was transgenic mice (data not shown) expressed mutant E6 protein achieved by using the Enhanced Chemiluminescence Plus Western roughly equal to that in K14E6WT mice.
Detection kit (Amersham, Piscataway, NJ).
Analysis of reproductive tracts. Reproductive tracts were harvested after 9 months of estrogen treatment and analyzed as described previously(7). The fixed tissue was histologically sectioned and every tenth 5-Amsection was stained with H&E and pathologically examined, with the worst Cancer Res 2007; 67: (4). February 15, 2007

E6 synergizes with estrogen in the absence of E7 to form reproductive tumors after estrogen treatment for 9 months.
Whereas K14E6WT transgenic mice did not develop cancer after6 months of estrogen treatment, E6 contributed to the severity oftumors arising in K14E6WT/K14E7WT mice when treated withestrogen for either 6 or 9 months (7, 9). In the current study,K14E6WT mice were treated with estrogen for an extended 9-monthperiod to investigate whether E6 could induce cervical cancer inthe absence of E7. As expected, none of the untreated mice,regardless of genotype, developed cancer (data not shown). After9 months of estrogen treatment, 41% of K14E6WT mice developedtumors in the reproductive tract compared with 6.7% fornontransgenic mice (Table 1). This difference was statisticallysignificant (P = 0.02). In contrast, 100% of K14E7WT transgenic micetreated for 9 months developed cancer (9). Thus, E6, in the absenceof E7, can induce cervical cancers in cooperation with exogenousestrogen, albeit less robustly than E7.
An E6 mutant reduced in binding A-helix partners has a lower incidence of cancer and develops smaller tumors. E6binds to numerous cellular proteins. One subset (e.g., the E3ubiquitin ligase E6AP) binds to E6 via a leucine-rich a-helical(a-helix) motif, whereas another (e.g., Dlg and Scribble) bindsthrough PDZ domains. We used two lines of E6 mutant mice,K14E6I128T and K14E6D146-151, defective in binding a-helix and PDZpartners, respectively (10, 17), to examine the importance of theinteraction of E6 with each subset of partners in mediatingthe oncogenic properties of E6 in the reproductive tract. After a9-month treatment period with estrogen, K14E6I128T mice had amarginally significant reduction in tumor incidence relative toK14E6WT mice (19.4% versus 41%; P = 0.058; Table 1). In contrast,the K14E6D146-151 transgenic mice had similar rates of cancerincidence as the K14E6WT mice (P = 0.25). Tumors from thereproductive tract of K14E6WT and K14E6mutant mice were variablein size (Fig. 2A). Nonetheless, tumors arising in the K14E6I128Ttransgenic mice were generally smaller relative to tumors fromeither K14E6WT or K14E6D146-151 mice (Table 1). The largest tumorsize on average for K14E6I128T mice (1.38 mm2 cross-sectional area)was significantly smaller (P = 0.041) than that for K14E6WT mice(5.22 mm2). In contrast, there was no significant difference in thelargest tumor size for the K14E6D146-151 mice compared with the Figure 1. Comparison of E6 expression in K14E6 transgenic mice andhuman cervical cancer cell lines. A, levels of E6 in HPV-positive andHPV-negative cervical cell lines. Immunoblots probed with antibodies specificfor either E6 or glyceraldehyde-3-phosphate dehydrogenase (GAPDH).
The GAPDH-specific immunoblot was done to confirm equivalence in loading Table 1. Comparison of tumors from the reproductive and was done for all experiments displayed in this figure, although shownonly for the top-most one. For each sample, 200 Ag of total soluble protein tract of nontransgenic, K14E6WT and K14E6mutant were analyzed. The top blot was loaded with samples from HPV-negative C33A, HPV18-positive HeLa, and HPV16-positive SiHa and Caski humancervical cancer–derived cell lines. The second blot was loaded with samplesfrom various clonal populations of W12E (20850 and 20963) and W12I (20861, 20822, 20862, and 201402) cell lines derived from a HPV16-positive incidence (%) multiplicity tumor (mm2) CINI lesion. W12E clones harbor the viral genome in the extrachromosomalstate. W12I clones harbor the genome in a chromosomally integrated state.
B, levels of E6 in the dorsal skin of K14E6WT and K14E6mutant transgenic mice at age postnatal day 9 or 10. In this immunoblot, 200 Ag (top) or 250 Ag (bottom) of total cellular protein from each mouse tissue sample and K14E6I128T (n = 36) 100 or 125 Ag of SiHa and Caski extracts were analyzed. Extracts from K14E6D146-151 (n = 28) different animals (A and B) of the same genotype were loaded to assessreproducibility of findings. Bottom, levels of E6 protein in E6 homozygoustransgenic mice in both skin and ear. C, levels of E6 in the reproductive tractof K14E6WT and K14E6mutant transgenic mice. Top, 200 Ag of total cellular Abbreviation: NTG, nontransgenic.
protein from each mouse tissue sample were analyzed. As above, extracts *Cancer incidence in K14E6I128T was marginally significant compared from different animals (A, B , and C ) of the same genotype were loaded toassess reproducibility of findings. In the bottom blot assessing relative with K14E6WT, P = 0.058, Fisher's exact test.
amounts of E6 protein in K14E6WT and K14E6D146-151 mice, the amount of cK14E6I128T tumors were significantly different than K14E6WT protein loaded is indicated in each lane. Mice were treated with estrogen to tumors, P = 0.04, Wilcoxon rank-sum test.
synchronize them in estrus, thereby eliminating variability in cervical epithelialthickness.
Cancer Res 2007; 67: (4). February 15, 2007 Multiple Properties of E6 Contribute to Cervical Cancer Figure 2. Characterization of reproductive tumors and the proliferative index of the cervix. A, comparison of tumor sizes between K14E6WT, K14E6I128T, andK14E6D146-151 transgenic mice. B, comparison of tumor sizes between K14E7WT, K14E6WT/K14E7WT, and K14E6mutant/K14E7WT transgenic mice. C, classification ofreproductive tumors by location. Middle, a cartoon representation of the murine reproductive tract and identifies the approximate borders for determining tumor locationused in histopathologic diagnosis. Left (for K14E6WT) and right (for K14E6WT/K14E7WT), the breakdown of total percentage of tumor development by location (top )and the percentage of total area these tumors encompassed (bottom ). In K14E6I128T and K14E6D146-151, the percentage of tumors arising in the vagina was 0%and 4%, respectively (data not shown). In K14E6I128T/K14E7WT and K14E6D146-151/K14E7WT transgenic mice, the tumors that developed in the vagina were 45% and38%, respectively (data not shown). D, quantification of epithelial hyperplasia in the cervix. The average percentage of basal and suprabasal BrdUrd-positive cellswas obtained from eight (!40) microscopic fields per mouse. An average of at least three mice per genotype were used to calculate the percentage.
K14E6WT mice. Thus, the ability of E6 to interact with a-helix both the K14E6I128T/K14E7WT and the K14E6D146-151/K14E7WT partners contributes to both tumor incidence and tumor size.
transgenic mice developed significantly smaller tumors in contrast The interactions of E6 with both A-helix and PDZ partners to the K14E6WT/K14E7WT mice (P = 0.005 and 0.014, respectively; contribute to tumor size and tumor multiplicity in the Table 2). In addition, the total area of tumor invasion of both the reproductive tract in the presence of E7. To understand the K14E6I128T/K14E7WT and the K14E6D146-151/K14E7WT transgenic role of a-helix and PDZ partners of HPV16 E6 in cervical cancer mice were significantly reduced relative to K14E6WT/K14E7WT when E6 is expressed together with HPV16 E7, the same tumors (P = 0.003 and 0.03, respectively; Table 2). Average tumor K14E6mutant lines were crossed onto the K14E7WT background size was also reduced in K14E6I128T/K14E7WT mice (data not and treated with estrogen for 9 months. All K14E7WT mice shown). Differences were also observed in terms of tumor developed cervical cancer in response to estrogen treatment (9).
frequency, with both mouse lines of K14E6mutant/K14E7WT having Therefore, it was not surprising that nearly all treated K14E6WT/ reduced number of tumors compared with K14E6WT/K14E7WT K14E7WT and K14E6mutant/K14E7WT mice also developed cervical mice (Table 2). This reduction was highly significant (P = 0.003) cancer (Table 2). Differences were observed, however, in terms of for the K14E6D146-151/K14E7WT mice, but less so for K14E6I128T/ the size of tumors. Comparing the largest tumor from each mouse, K14E7WT mice (P = 0.187).
Cancer Res 2007; 67: (4). February 15, 2007 E6 tumors develop primarily in the cervix and the cervi- stroma, thereby prohibiting the ability to dissect tumors for covaginal junction. In prior studies, estrogen-treated K14E7WT Western analyses. Hyperplastic reproductive epithelium from transgenic mice efficiently developed tumors in the vagina as well estrogen-treated nontransgenic mice expressed MCM7 only in as in the cervix (9). In contrast, nearly all tumors arising in the basal layer. In contrast, highly dysplastic reproductive K14E6WT and K14E6mutant lines developed in the cervix or at the epithelium from K14E6WT/K14E7WT and K14E7WT mice was junction of the cervix and the vagina. Only 2 of 62 (6%) tumors strongly positive for MCM7 throughout the full thickness of the observed in the three E6 transgenic lines developed in the vagina epithelium, similar to data from our previous studies (26).
proper. In the presence of E7, the percentage of tumors arising in Unexpectedly, approximately one third to two thirds of the the vagina of the K14E6WT/K14E7WT and K14E6mutant/K14E7WT epithelia from all three E6 transgenic lines stained positive for double transgenic mice increased to 38% to 45%. Thus, E6 MCM7, beyond the basal layer of staining in the nontransgenic predisposes animals primarily to tumors of the cervix. In contrast, epithelium. Expression of MCM7 in the epithelia of the three E6 E7 alone or in combination with E6 induces tumors in the vagina transgenic lines was generally uniform. No differences in staining as well as in the cervix, with the largest tumor area predominantly between all doubly transgenic lines were observed, presumably due found in the cervico-vaginal junction (Fig. 2C).
to the strong induction of MCM7 in the E7-expressing tissues. In The ability of E6 to interact with PDZ partners contributes agreement with previous studies (26), MCM7 staining patterns to hyperplasia in the cervical epithelium. K14E6WT transgenic predominantly correlated positively with lesion grade. These mice display epidermal hyperplasia characterized by an induction results indicate that E6 can induce Mcm7, an E2F-responsive gene.
of DNA synthesis within that suprabasal compartment (5). A In tumors, MCM7 expression had a less consistent pattern similar finding was observed in the cervical epithelium (Fig. 2D), compared with the epithelium. Expression of MCM7 in tumors was with significant increases in both basal and suprabasal DNA variable, not correlating with genotype, tumor size, or location.
synthesis (P = 0.02) in K14E6WT transgenic mice (11.6% and 2.9%, Furthermore, levels varied between tumors arising within the same respectively) compared with nontransgenic mice (5.6% and 0.9%, mouse. The sole tumor arising in the nontransgenic mouse had low respectively). Suprabasal DNA synthesis in the K14E6D146-151 MCM7 expression (Fig. 3; Table 3). Nonetheless, all tumors from transgenic mice was reduced compared with K14E6WT transgenic transgenic mice had some level of MCM7 expression. No tumor mice (1.64% versus 2.87%; P = 0.06) and not significantly different was MCM7 negative. Aside from K14E7WT tumors, which were from that observed in nontransgenic mice (P = 0.14). No difference robust for MCM7 staining, all tumors from other genotypes had in suprabasal DNA synthesis was observed between K14E6I128T and low to high MCM7 expression.
K14E6WT transgenic mice (P = 0.26). Thus, in the cervical Cyclin E is another E2F-responsive gene that is also used as a epithelium, the E6 oncogene is able to increase DNA synthesis in biomarker for dysplastic lesions and cervical cancer (27). We both basal and suprabasal layers of the cervical epithelium, evaluated cyclin E expression using immunohistochemistry in the resulting in hyperplasia. This is comparable with results in the epithelia and in tumors for all genotypes. In the absence of E7, skin, where the ability of E6 to increase suprabasal DNA synthesis cyclin E staining was more nuclear and less cytoplasmic. In was mediated at least partially through interactions with PDZ nontransgenic mice, cyclin E expression was restricted mostly to partners (10).
the basal and parabasal layers of cervical epithelia. In both Mcm7 and cyclin E are up-regulated in E6 epithelia and K14E6WT and K14E6mutant cervical epithelium, cyclin E expression tumors of the reproductive tract in the absence of E7. Mcm7 is positively correlated with the level of dysplasia (data not shown).
an E2F-responsive gene and robust biomarker expressed in high- Unlike MCM7, cyclin E expression was not uniform, such that not grade CINs and cervical cancer in humans as well as in K14E6WT/ every cell was positive. The sole nontransgenic tumor had low K14E7WT and K14E7WT mice (26). We evaluated MCM7 expression cyclin E expression. Tumors from the singly E6 transgenic lines had in the epithelia and in tumors of the reproductive tract arising in variable cyclin E expression similar to that observed for MCM7.
nontransgenic, K14E7WT and the three E6 transgenic lines with or Tumors had low to high cyclin E expression with no correlation without E7 (Fig. 3). Analyses of biomarker expression in both the between staining level, tumor size, or location.
epithelium and the tumors were limited to immunohistochemistry.
In the presence of E7, cyclin E diffusely stained both nuclei and Reproductive tract tumors in HPV transgenic mice were often too cytoplasm (Fig. 3; Table 3). K14E7WT reproductive epithelium had small and sessile. The tumors tended to grow inwardly into the at least 50% of cells staining positive for cyclin E. In general, Table 2. Comparison of tumors from the reproductive tract of nontransgenic, K14E7WT, K14E6WT/K14E7WT, andK14E6mutant/K14E7WT transgenic mice Total area of tumor K14E6WT/K14E7WT (n = 6) K14E6I128T/K14E7WT (n = 24) K14E6D146-151/K14E7WT (n = 21) *Compared with K14E6WT/K14E7, P < 0.05, Wilcoxon rank-sum test.
Cancer Res 2007; 67: (4). February 15, 2007 Multiple Properties of E6 Contribute to Cervical Cancer Figure 3. Evaluation of E2F-responsivegene expression in the estrogen-treatedepithelium and tumors from thereproductive tract. Columns 1 and 2,MCM7 staining (brown staining nuclei),which is up-regulated in singly or doubletransgenic reproductive epithelia andtumors, whereas MCM7 expression isrestricted to the basal and parabasal layersof nontransgenic epithelium. Columns 3and 4, cyclin E staining (brown nuclei).
Cyclin E expression, similar to MCM7expression, is also up-regulated in bothsingle and double transgenic mice.
Magnification, !40.
E7-positive tumors had medium to high cyclin E expression. Cyclin was variable, with a range from nil to sporadic highly positive E expression in epithelia and tumors of doubly transgenic mice was (Fig. 4). Tumors from K14E6mutant/K14E7WT transgenic mice (data generally high, with at least 60% of cells staining positive.
not shown) had similar levels of p53 expression relative to Reproductive tract tumors are more likely to express p53 K14E6WT/K14E7WT tumors but were generally less intense. We and in greater intensities in the presence of E7. p53 is generally also noted that vaginal tumors had reduced expression of p53 undetectable in normal tissue unless induced in response to DNA- relative to tumors from the cervix or the cervicovaginal junction.
damaging agents. The reproductive tract is a p53-responsive tissue, p16 expression is inversely correlated with retinoblastoma in which a DNA damage response can be mounted in response to expression in reproductive tumors from HPV transgenic mice.
ionizing radiation if WT p53 is intact. Dominant-acting, missense p16 is up-regulated in several cervical cancer cell lines as well as in mutations in p53, causally associated with tumorigenesis, often human cervical samples (23). This cyclin kinase inhibitor is a lead to the stabilization and accumulation of p53, providing a marker for high-risk HPV infection in human dysplastic lesions and useful indicator of p53 status and the disease state in most tumor cancers of the reproductive tract as well as cancers of the head and types. In HPV-associated cancers, p53 is thought to be inactivated neck (31, 32). We evaluated p16 status in the tumors from our through E6-induced, ubiquitin-mediated protein degradation.
However, p53 mutations have been detected in both premalignantlesions and human cervical tumors at low frequencies. We did p53 Table 3. Summary of biomarker expression in tumors immunohistochemistry to monitor levels of p53 protein in the from the reproductive tract of estrogen-treated mice reproductive tumors of our various HPV transgenic mice (Fig. 4;Table 3). The sole nontransgenic tumor was p53 negative. Tumors from K14E7WT transgenic mice were p53 positive and had low tomedium expression, whereas tumors from K14E6WT transgenic mice were almost completely p53 negative. These observations were consistent with prior studies showing the destabilization of p53 by E6- and E7-induced accumulation of p53 (28–30).
K14E6I128T-expressing epithelium had elevated p53 expression compared with K14E6WT mice, consistent with the reduced ability of K14E6I128T to degrade p53. Tumors from K14E6I128T mice, however, had clearly less intense levels of p53 expression thantumors from K14E7WT mice. In agreement with our predictions, theK14E6D146-151 transgenic tumors were p53 negative. Expression of NOTE: ", negative; F, <5%; +, 5% to 20%; ++, 20% to 50%; +++, >50%.
p53 in the tumors arising in K14E6WT/K14E7WT transgenic mice Cancer Res 2007; 67: (4). February 15, 2007 estrogen-treated mice. Expression of p16 was uniformly diffuse (Fig. 4; Table 3). The sole nontransgenic tumor had low expres- with both nuclear and cytoplasmic expression in all mice. There sion levels of pRb. In agreement with the known ability of E7 to was a general cytoplasmic expression pattern, with nuclear induce the degradation of pRb, E7-expressing tumors had little accumulation localized predominantly to the bottom one third to to no detectable pRb regardless of E6 mutational status (Fig. 4; one half of the epithelium (data not shown). The sole non- Table 3; data not shown). In contrast, tumors from K14E6WT and transgenic tumor had low levels of p16 expression. Tumors from K14E6D146-151 transgenic mice expressed high levels of pRb.
both K14E6WT/K14E7WT and K14E7WT mice displayed high levels of K14E6I128T tumors, however, displayed low levels of pRb relative p16 (Fig. 4; Table 3), similar to the patterns observed in human to K14E6WT and K14E6D146-151 tumors. Hence, HPV E6 is able to cervical samples. Presumably, due to the ability of E7 to induce p16 alter pRb and p16 levels in tumors in a manner distinguishable strongly (33), differences in expression between K14E6mutant/ from HPV E7 and is dependent on its interaction with a-helix K14E7WT tumors were not observed (data not shown). Conversely, partners, as the expression pattern of p16 and pRb no longer expression of p16 in tumors from the K14E6WT and K14E6D146-151 resembles WT E6, but more like E7-expressing tumors.
transgenic mice was either low or negative. This decrease in p16expression was clearly more pronounced in the tumors comparedwith the neighboring epithelium, which was variable (data not shown). In contrast, tumors from K14E6I128T transgenic mice In this study, we dissected the contributions of HPV E6 in both displayed a strong increase in the expression of p16 relative to the presence and the absence of HPV E7 in cervical carcinogenesis K14E6WT and K14E6D146-151 tumors. The intensity of p16 expression by focusing on specific properties of E6 and extending the estrogen in K14E6I128T tumors was generally less robust relative to tumors treatment period. E6, in the absence of E7, induces primarily containing the E7 oncogene.
cervical tumors in the reproductive tract. The abilities of E6 to Because expression of p16 was reduced in K14E6 reproductive interact with both a-helix and PDZ partners contribute to this role tumors and given that pRb contributes to the regulation of p16, we in cervical carcinogenesis. Furthermore, E6 induces a pattern of measured the expression levels of pRb via immunohistochemistry cellular gene expression that is overlapping yet distinct from that Figure 4. p53, p16, and pRb status in thereproductive tract. Images from sectionsstained with antibodies to p53, p16, orpRb (brown ) and counterstained withhematoxylin (blue ). Column 1, p53expression in the cervical epithelium fromvarious genotypes. The nontransgenic(NTG ; top ) sample is from a mouseexposed to 5 Gy of ionizing radiation andis used as a positive control to showp53-positive staining, which is primarilyobserved in the basal and parabasal strata.
The cervical epithelium from unirradiatednontransgenic mice (data not shown) isp53 negative. All other panels are fromunirradiated mice. Columns 2 to 4, thestatus of p53, p16, and retinoblastoma(pRB ) expression in reproductive tumorsfrom various genotypes. Tumorsexpressing the E7 oncogene generallydisplayed variable positivity for p53. Shownin the panel from the E6/E7 tumor is anarea of high sporadic p53 positivity. p16was up-regulated in tumors expressing E7or K14E6I128T. Retinoblastoma expressionwas inversely correlated to p16 in tumorsexpressing either HPV E6 or E7.
Magnification, !40.
Cancer Res 2007; 67: (4). February 15, 2007 Multiple Properties of E6 Contribute to Cervical Cancer tumors of K14E6WT mice. Whether this alternative dysregulation of Table 4. Summary of histopathologic diagnosis in the p16/pRb pathway contributes to the tumorigenesis observed in K14E6WT and K14E6mutant transgenic mice the K14E6I128T transgenic mice is unclear. Regardless, this findingsupports the hypothesis that residual oncogenic activity in the CIN I CIN II CIN III/CIS Cancer K14E6I128T mice reflects a distinct activity of E6 and not a partial retention in the binding capacity to a-helix partners. The absence K14E6I128T (n = 36) of a reduction in the tumorigenic phenotype of K14E6D146-151 K14E6D146-151 (n = 28) compared with K14E6WT mice indicates that PDZ domain partnersare not relevant in the context of these experiments or that theircontribution is modest.
Abbreviations: CIS, carcinoma in situ; NH, normal hyperplasia.
The contribution of E6 to cervical carcinogenesis in the presence of E7 is dependent on interactions with both A-helixand PDZ partners. The above experiments were all carried out inthe absence of E7. Similar studies in the presence of E7 (Table 2) induced by E7. Specifically, E6 leads to a dysregulation of the p16/ revealed that the interaction of E6 with a-helix partners is pRb pathway in a manner different from that of E7 yet led to a important for cervical carcinogenesis. The most obvious difference similar though less robust induction of E2F-responsive genes.
in the tumorigenic phenotypes between K14E6 I128T/K14E7WT and HPV16 E6 has a weaker oncogenic potential than HPV16 E7 K14E6WT/K14E7WT mice was tumor size. In contrast to observa- in the reproductive tract. In our prior studies, K14E6WT tions in the absence of E7, the interactions of E6 with PDZ partners transgenic mice did not develop cervical cancer or even high- also contributed to cervical carcinogenesis in the presence of E7.
grade dysplasias after 6 months of estrogen treatment. K14E7WT Specifically, we observed a reduction in tumor size in the transgenic mice on the hand developed multiple high-grade K14E6D146-151/K14E7WT mice compared with K14E6WT/K14E7WT dysplastic lesions and tumors throughout the entire reproductive mice. Tumors arising in the K14E6mutant/K14E7WT mice were not tract. In this study, we extended the estrogen treatment period to significantly different in size from those arising in the K14E7WT 9 months. A large fraction (41%) of K14E6WT transgenic mice singly transgenic mice. Tumor multiplicity also was reduced in developed cancer. The majority of the remaining K14E6 mice both K14E6mutant/K14E7WT lines relative to K14E6WT/K14E7WT developed at least one high-grade dysplastic lesion (Table 4). This mice. This reduction was only statistically significant for the represented a significant increase in tumorigenesis compared with K14E6D146-151/K14E7WT transgenic mice. Thus, in the presence of nontransgenic mice (6.7%) yet significantly less than that observed E7, E6 contributes to cervical carcinogenesis through at least two in K14E7WT mice (100% tumor incidence). Likewise, tumor distinct mechanisms. This finding is consistent with what we have multiplicity was significantly reduced in K14E6WT mice compared observed previously in the skin, where the ability of E6 to bind both with the K14E7WT mice (1.67 versus 6.67). Furthermore, the a-helix and PDZ domain partners contributed to carcinogenesis K14E6WT transgenic mice did not develop the extensive dysplastic (10, 17, 19).
pathology that occurred throughout the entire reproductive It is unclear which PDZ partner(s) of E6 contributes to both epithelial lining of K14E7WT mice. Thus, HPV16 E6 has a tumor size and tumor multiplicity. The E6 oncogene interacts with demonstrable yet clearly weaker oncogenic activity than HPV16 numerous cellular partners, which contain PDZ motifs, such as Dlg, E7 in the reproductive tract. In contrast, E6 is the more potent Scribble, and Magis (34–37). DLG and/or Scribble are attractive oncogene in the skin, contributing to both the promotion and the candidates given that both are tumor suppressors in Drosophila.
progression stages of skin carcinogenesis and induces primarily Mutations in either of these genes in Drosophila result in the malignant tumors (8). Therefore, the relative potency of the HPV16 development of epithelial hyperplasia, loss of cell-cell contacts E6 and E7 oncogenes differs depending on the tissue evaluated.
(38, 39), and tumorigenesis of the imaginal discs and brain lobes The ability of E6 to bind to A-helix partners contributes to (40). In the human cervix, both hDlg and hScrib are gradually cervical carcinogenesis. K14E6I128T transgenic mice had a altered in cellular localization and expression is lost as low-grade reduction in the incidence and size of reproductive tract tumors lesions progress to invasive cervical carcinomas (41–43). Reduc- compared with K14E6WT transgenic mice. Given the reduced tions in hDlg and hScrib are also seen in human cervical cancer cell tumorigenic phenotype of the K14E6I128T transgenic mice, we lines (19). Because Dlg and Scribble are both expressed in the hypothesize that the inactivation of p53 by E6 contributes to septate junction (44) and seem to have similar functions, both cervical carcinogenesis. Consistent with this hypothesis, slightly genes may contribute to the oncogenic potential of E6 in the elevated levels of p53 protein were seen in tumors arising in reproductive tract. Until analysis of targeted individual PDZ K14E6I128T mice compared with K14E6WT mice.
deletion mutants can be done, the exact E6-PDZ interaction(s) Whereas reduced in their incidence of tumors compared with responsible for the oncogenic potential of E6 remains unclear.
K14E6WT mice, K14E6I128T transgenic mice retained a significant The E6 oncogene induces the E2F-responsive genes, MCM7 increase in their tumorigenic phenotype compared with non- and cyclin E, in reproductive epithelia and tumors in the transgenic mice (Table 1). This increase was evident in tumor absence of E7. All E6 transgenic lines expressed the E2F- multiplicity (1.14 versus 0.07) and average tumor size (1.39 mm2 responsive genes, MCM7 and cyclin E, in both the epithelium versus 0.029 mm2). Such residual oncogenic activity in K14E6I128T and the tumors of the reproductive tract. Expression of these genes transgenic mice could reflect the ability of the I128T mutant was above the levels seen in nontransgenic mice. These results in protein to bind a-helix partners, albeit at 1% to 5% the level of WT the E6 mice were somewhat unexpected given that the E7 E6 protein, or it may reflect an activity of E6 distinct from its ability oncogene was absent in these mice and therefore not available to bind a-helix partners. Tumors arising in K14E6I128T mice also to induce the expression of E2F-responsive genes through pRb displayed a distinct pattern of expression of p16 and pRb relative to inactivation. Hence, the E6 oncogene must be activating the Cancer Res 2007; 67: (4). February 15, 2007 transcription of these E2F-responsive genes by a mechanism How E6 induces levels of CDK4/6 is unknown. Thus, it remains different than E7. In the epithelium of the K14E6WT mice, this up- unclear whether these two hypotheses reflect the same or distinct regulation of E2F-responsive genes correlated with high levels of mechanisms. Regardless, a role of p53 inactivation in mediating p16 and low levels of pRb, as seen in the epithelium and tumors in the dysregulation of p16/pRb pathway by E6 is supported by the the K14E7WT mice. However, there is an inverted pattern of reversed pattern of p16 expression in K14E6I128T tumors, which expression of p16 and pRb in the K14E6WT tumors (Fig. 4; Table 3).
encode a mutant E6 protein defective for inactivating p53 compared Specifically, levels of p16 were low and levels of pRb were high in with K14E6WT tumors (Fig. 4; Table 3).
the reproductive tumors of K14E6WT mice. This result indicates In summary, we report the first in vivo study dissecting the that the alteration of the cell cycle during progression to mechanism of E6 action in cervical carcinogenesis. The E6 inter- malignancy in K14E6WT mice differs from that observed in actions with two groups of cellular partners contributed to cervical K14E7WT mice. Interestingly, the pattern seen in the tumors of carcinogenesis. Additionally, our study revealed that the ability K14E6WT mice is consistent with the low expression levels of p16 of E6 to induce E2F-responsive genes is likely through the dys- and high levels of hyperphosphorylated pRb observed in fibroblast regulation of the p16/pRb pathway by mechanisms distinct from and epithelial cell lines immortalized with the HPV E6 oncogene (33, 45–47). Thus, E6-dependent immortalization in vitro and E6-dependent tumorigenesis in vivo arise through means that lead to asimilar dysregulation of the p16/pRb pathway opposite of that observed in E7-dependent tumorigenesis (this study) or in human Received 9/8/2006; revised 10/30/2006; accepted 12/4/2006.
cervical cancers (48). The inactivation of p53 by E6 and consequent The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance inhibition of p53-induced expression of the cyclin-dependent kinase with 18 U.S.C. Section 1734 solely to indicate this fact.
(CDK) inhibitor p21 might lead to higher CDK activity and thereby We thank Dr. E. Weiss (University Louis Pasteur, Strasbourg, France) for generating and providing the anti-E6 monoclonal antibody 6F4, Drs. Drinkwater and Sugden for increased hyperphosphorylated pRb. Alternatively, E6 could induce the critical reading of the manuscript, and members of the Lambert laboratory for phosphorylation of pRb by up-regulating the levels of CDK4/6 (49).
13. Huibregtse JM, Scheffner M, Beaudenon S, Howley 25. Gulliver GA, Herber RL, Liem A, Lambert PF. Both PM. A family of proteins structurally and functionally conserved region 1 (CR1) and CR2 of the human 1. Sankaranarayanan R, Ferlay J. Worldwide burden of related to the E6-AP ubiquitin-protein ligase. Proc Natl papillomavirus type 16 E7 oncogene are required for gynaecological cancer: the size of the problem. Best Acad Sci U S A 1995;92:5249.
induction of epidermal hyperplasia and tumor forma- Pract Res Clin Obstet Gynaecol 2006;20:207–25.
14. Kishino T, Lalande M, Wagstaff J. UBE3A/E6-AP tion in transgenic mice. J Virol 1997;71:5905–14.
2. Walboomers JM, Jacobs MV, Manos MM, et al. Human mutations cause Angelman syndrome. Nat Genet 1997; 26. Brake T, Connor JP, Petereit DG, Lambert PF.
papillomavirus is a necessary cause of invasive cervical Comparative analysis of cervical cancer in women and cancer worldwide. J Pathol 1999;189:12–9.
15. Nakao M, Sutcliffe JS, Durtschi B, Mutirangura A, in a human papillomavirus-transgenic mouse model: 3. Bosch FX, Lorincz A, Munoz N, Meijer CJ, Shah KV.
Ledbetter DH, Beaudet AL. Imprinting analysis of three identification of minichromosome maintenance protein The causal relation between human papillomavirus and genes in the Prader-Willi/Angelman region: SNRPN, E6- 7 as an informative biomarker for human cervical cervical cancer. J Clin Pathol 2002;55:244–65.
associated protein, and PAR-2 (D15S225E). Hum Mol cancer. Cancer Res 2003;63:8173–80.
4. Fehrmann F, Laimins LA. Human papillomaviruses: 27. Keating JT, Cviko A, Riethdorf S, et al. Ki-67, cyclin E, targeting differentiating epithelial cells for malignant 16. Scheffner M, Huibregtse JM, Vierstra RD, Howley PM.
and p16INK4 are complimentary surrogate biomarkers transformation. Oncogene 2003;22:5201–7.
The HPV-16 E6 and E6-AP complex functions as a for human papilloma virus-related cervical neoplasia.
5. Song S, Pitot HC, Lambert PF. The human ubiquitin-protein ligase in the ubiquitination of p53. Cell Am J Surg Pathol 2001;25:884–91.
papillomavirus type 16 E6 gene alone is sufficient to 28. Scheffner M, Werness BA, Huibregtse JM, Levine AJ, induce carcinomas in transgenic animals. J Virol 1999; 17. Nguyen M, Song S, Liem A, Androphy E, Liu Y, Howley PM. The E6 oncoprotein encoded by human Lambert PF. A mutant of human papillomavirus type 16 papillomavirus types 16 and 18 promotes the degrada- 6. Herber R, Liem A, Pitot H, Lambert PF. Squamous E6 deficient in binding a-helix partners displays reduced tion of p53. Cell 1990;63:1129–36.
epithelial hyperplasia and carcinoma in mice transgenic oncogenic potential in vivo . J Virol 2002;76:13039–48.
29. Jones DL, Thompson DA, Munger K. Destabilization for the human papillomavirus type 16 E7 oncogene.
18. Kiyono T, Foster SA, Koop JI, McDougall JK, Galloway of the RB tumor suppressor protein and stabilization of J Virol 1996;70:1873–81.
DA, Klingelhutz AJ. Both Rb/p16INK4a inactivation and p53 contribute to HPV type 16 E7-induced apoptosis.
7. Riley RR, Duensing S, Brake T, Munger K, Lambert PF, telomerase activity are required to immortalize human Arbeit JM. Dissection of human papillomavirus E6 and epithelial cells. Nature 1998;396:84–8.
30. Eichten A, Westfall M, Pietenpol JA, Munger K.
E7 function in transgenic mouse models of cervical 19. Simonson SJ, Difilippantonio MJ, Lambert PF. Two Stabilization and functional impairment of the tumor carcinogenesis. Cancer Res 2003;63:4862–71.
distinct activities contribute to human papillomavirus suppressor p53 by the human papillomavirus type 16 E7 8. Song S, Liem A, Miller JA, Lambert PF. Human 16 E6's oncogenic potential. Cancer Res 2005;65:8266–73.
oncoprotein. Virology 2002;295:74–85.
papillomavirus types 16 E6 and E7 contribute differently 20. Tachibana KE, Gonzalez MA, Coleman N. Cell-cycle- 31. Sano T, Oyama T, Kashiwabara K, Fukuda T, to carcinogenesis. Virology 2000;267:141–50.
dependent regulation of DNA replication and its Nakajima T. Expression status of p16 protein is 9. Brake T, Lambert PF. Estrogen contributes to the relevance to cancer pathology. J Pathol 2005;205:123–9.
associated with human papillomavirus oncogenic po- onset, persistence, and malignant progression of cervi- 21. Kim Y, Choi EK, Cho NH, et al. Expression of cyclin E tential in cervical and genital lesions. Am J Pathol 1998; cal cancer in a human papillomavirus-transgenic mouse and p27KIP1 in cervical carcinoma. Cancer Lett 2000; model. Proc Natl Acad Sci U S A 2005;102:2490–5.
32. Hafkamp HC, Speel EJ, Haesevoets A, et al. A subset 10. Nguyen ML, Nguyen MM, Lee D, Griep AE, 22. Erlandsson F, Martinsson-Ahlzen HS, Wallin KL, of head and neck squamous cell carcinomas exhibits Lambert PF. The PDZ ligand domain of the human Hellstrom AC, Andersson S, Zetterberg A. Parallel cyclin integration of HPV 16/18 DNA and overexpression of papillomavirus type 16 E6 protein is required for E6's E and cyclin A expression in neoplastic lesions of the p16INK4A and p53 in the absence of mutations in p53 induction of epithelial hyperplasia in vivo . J Virol uterine cervix. Br J Cancer 2006;94:1045–50.
exons 5-8. Int J Cancer 2003;107:394–400.
23. Klaes R, Friedrich T, Spitkovsky D, et al. Over- 33. Khleif SN, DeGregori J, Yee CL, et al. Inhibition of 11. Nguyen MM, Nguyen ML, Caruana G, Bernstein A, expression of p16(INK4A) as a specific marker for cyclin D-CDK4/CDK6 activity is associated with an E2F- Lambert PF, Griep AE. Requirement of PDZ-containing dysplastic and neoplastic epithelial cells of the cervix mediated induction of cyclin kinase inhibitor activity.
proteins for cell cycle regulation and differentiation in the uteri. Int J Cancer 2001;92:276–84.
Proc Natl Acad Sci U S A 1996;93:4350–4.
mouse lens epithelium. Mol Cell Biol 2003;23:8970–81.
24. Giovane C, Trav G, Briones A, Lutz Y, Wasylyk B, 34. Kiyono T, Hiraiwa A, Fujita M, Hayashi Y, Akiyama T, 12. Liu Y, Chen JJ, Gao Q, et al. Multiple functions of Weiss E. Targeting of the N-terminal domain of the Ishibashi M. Binding of high-risk human papillomavirus human papillomavirus type 16 E6 contribute to the human papillomavirus type 16 E6 oncoprotein with E6 oncoproteins to the human homologue of the immortalization of mammary epithelial cells. J Virol monomeric ScFvs blocks the E6-mediated degradation Drosophila discs large tumor suppressor protein. Proc of cellular p53. J Mol Recognit 1999;12:141–52.
Natl Acad Sci U S A 1997;94:11612–6.
Cancer Res 2007; 67: (4). February 15, 2007 Multiple Properties of E6 Contribute to Cervical Cancer 35. Gardiol D, Kuhne C, Glaunsinger B, Lee SS, Javier R, and Lgl in cell polarity, cell proliferation, and cancer.
cence and loss of p16 at immortalization in human Banks L. Oncogenic human papillomavirus E6 proteins papillomavirus 16 E6, but not E7, transformed human target the discs large tumour suppressor for protea- 41. Lin HT, Steller MA, Aish L, Hanada T, Chishti AH.
uroepithelial cells. Cancer Res 1996;56:2886–90.
some-mediated degradation. Oncogene 1999;18:5487–96.
Differential expression of human Dlg in cervical intra- 46. Yamamoto A, Kumakura S, Uchida M, Barrett JC, 36. Nakagawa S, Huibregtse JM. Human scribble (Vartul) epithelial neoplasias. Gynecol Oncol 2004;93:422–8.
Tsutsui T. Immortalization of normal human embryonic is targeted for ubiquitin-mediated degradation by the 42. Cavatorta AL, Fumero G, Chouhy D, et al. Differential fibroblasts by introduction of either the human high-risk papillomavirus E6 proteins and the E6AP expression of the human homologue of drosophila discs papillomavirus type 16 E6 or E7 gene alone. Int J Cancer ubiquitin- protein ligase. Mol Cell Biol 2000;20:8244–53.
large oncosuppressor in histologic samples from human 37. Thomas M, Laura R, Hepner K, et al. Oncogenic papillomavirus-associated lesions as a marker for 47. Tsutsui T, Kumakura S, Yamamoto A, et al. Association human papillomavirus E6 proteins target the MAGI-2 progression to malignancy. Int J Cancer 2004;111:373–80.
of p16(INK4a) and pRb inactivation with immortalization and MAGI-3 proteins for degradation. Oncogene 2002; 43. Nakagawa S, Yano T, Nakagawa K, et al. Analysis of of human cells. Carcinogenesis 2002;23:2111–7.
the expression and localisation of a LAP protein, human 48. Nam EJ, Kim JW, Kim SW, et al. The expressions of 38. Bilder D, Li M, Perrimon N. Cooperative regulation of scribble, in the normal and neoplastic epithelium of the Rb pathway in cervical intraepithelial neoplasia; cell polarity and growth by Drosophila tumor suppres- uterine cervix. Br J Cancer 2004;90:194–9.
predictive and prognostic significance. Gynecol Oncol.
sors. Science 2000;289:113–6.
44. Tepass U, Tanentzapf G, Ward R, Fehon R. Epithelial Epub 2006 Oct 13.
39. Bilder D, Perrimon N. Localization of apical epithelial cell polarity and cell junctions in Drosophila . Annu Rev 49. Malanchi I, Accardi R, Diehl F, et al. Human determinants by the basolateral PDZ protein Scribble.
papillomavirus type 16 E6 promotes retinoblastoma 45. Reznikoff CA, Yeager TR, Belair CD, Savelieva E, protein phosphorylation and cell cycle progression.
40. Humbert P, Russell S, Richardson H. Dlg, Scribble, Puthenveettil JA, Stadler WM. Elevated p16 at senes- J Virol 2004;78:13769–78.
Cancer Res 2007; 67: (4). February 15, 2007

Source: http://beats.datasynth.info/wp-content/uploads/2016/03/Shai_etal_2007.pdf