Beta-catenin accelerates human papilloma virus type-16 mediated cervical carcinogenesis in transgenic mice.

Abstract

Human papilloma virus (HPV) is the principal etiological agent of cervical cancer in women, and its DNA is present in virtually all of these tumors. However, exposure to the high-risk HPV types alone is insufficient for tumor development. Identifying specific collaborating factors that will lead to cervical cancer remains an unanswered question, especially because millions of women are exposed to HPV. Our earlier work using an in vitro model indicated that activation of the canonical Wnt pathway in HPV-positive epithelial cells was sufficient to induce anchorage independent growth. We therefore hypothesized that constitutive activation of this pathway might function as the "second hit." To address this possibility, we developed two double-transgenic (DT) mouse models, K14-E7/ΔN87βcat and K14-HPV16/ΔN87βcat that express either the proteins encoded by the E7 oncogene or the HPV16 early region along with constitutively active β-catenin, which was expressed by linking it to the keratin-14 (K14) promoter. We initiated tumor formation by treating all groups with estrogen for six months. Invasive cervical cancer was observed in 11% of the K14-ΔN87βcat mice, expressing activated β-catenin and in 50% of the animals expressing the HPV16 E7 oncogene. In double-transgenic mice, coexpression of β-catenin and HPV16 E7 induced invasive cervical cancer at about 7 months in 94% of the cases. We did not observe cervical cancer in any group unless the mice were treated with estrogen. In the second model, K14-HPV16 mice suffered cervical dysplasias, but this phenotype was not augmented in HPV16/ΔN87βcat mice. In summary, the phenotypes of the K14-E7/ΔN87βcat mice support the hypothesis that activation of the Wnt/β-catenin pathway in HPV-associated premalignant lesions plays a functional role in accelerating cervical carcinogenesis.

Full text

Page 1

Beta-Catenin Accelerates Human Papilloma Virus Type-
16 Mediated Cervical Carcinogenesis in Transgenic Mice
Gu ¨lay Bulut1, Shannon Fallen1, Elspeth M. Beauchamp1, Lauren E. Drebing1, Junfeng Sun2, Deborah L.
Berry3, Bhaskar Kallakury4, Christopher P. Crum5, Jeffrey A. Toretsky1, Richard Schlegel4, Aykut U¨ren1*
1Lombardi Comprehensive Cancer Center, Georgetown University Medical Center, Washington, D.C., United States of America, 2Critical Care Medicine Department,
Clinical Center, National Institutes of Health, Bethesda, Maryland, United States of America, 3Histopathology Tissue Shared Resources, Georgetown University Medical
Center, Washington, D.C., United States of America, 4Department of Pathology, Georgetown University Medical Center, Washington, D.C., United States of America,
5Department of Pathology, Harvard University, Boston, Massachusetts, United States of America
Abstract
Human papilloma virus (HPV) is the principal etiological agent of cervical cancer in women, and its DNA is present in
virtually all of these tumors. However, exposure to the high-risk HPV types alone is insufficient for tumor development.
Identifying specific collaborating factors that will lead to cervical cancer remains an unanswered question, especially
because millions of women are exposed to HPV. Our earlier work using an in vitro model indicated that activation of the
canonical Wnt pathway in HPV-positive epithelial cells was sufficient to induce anchorage independent growth. We
therefore hypothesized that constitutive activation of this pathway might function as the ‘‘second hit.’’ To address this
possibility, we developed two double-transgenic (DT) mouse models, K14-E7/DN87bcat and K14-HPV16/DN87bcat that
express either the proteins encoded by the E7 oncogene or the HPV16 early region along with constitutively active b-
catenin, which was expressed by linking it to the keratin-14 (K14) promoter. We initiated tumor formation by treating all
groups with estrogen for six months. Invasive cervical cancer was observed in 11% of the K14-DN87bcat mice, expressing
activated b-catenin and in 50% of the animals expressing the HPV16 E7 oncogene. In double-transgenic mice, coexpression
of b-catenin and HPV16 E7 induced invasive cervical cancer at about 7 months in 94% of the cases. We did not observe
cervical cancer in any group unless the mice were treated with estrogen. In the second model, K14-HPV16 mice suffered
cervical dysplasias, but this phenotype was not augmented in HPV16/DN87bcat mice. In summary, the phenotypes of the
K14-E7/DN87bcat mice support the hypothesis that activation of the Wnt/b-catenin pathway in HPV-associated
premalignant lesions plays a functional role in accelerating cervical carcinogenesis.
Citation: Bulut G, Fallen S, Beauchamp EM, Drebing LE, Sun J, et al. (2011) Beta-Catenin Accelerates Human Papilloma Virus Type-16 Mediated Cervical
Carcinogenesis in Transgenic Mice. PLoS ONE 6(11): e27243. doi:10.1371/journal.pone.0027243
Editor: Hassan Ashktorab, Howard University, United States of America
Received June 15, 2011; Accepted October 12, 2011; Published November 7, 2011
Copyright: ? 2011 Bulut et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This study was supported by grants from the National Institutes of Health (NIH) CA108641 (to AU), Cancer Center Support Grant P30 CA051008 for use
of Shared Resources (H&E and IHC stainings) and NIH ARRA Grant 1G20 RR025828-01 for use of Rodent Barrier Facility Equipment (between 7/20/2009 to 7/19/
2011). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: au26@georgetown.edu
Introduction
Cervical cancer is the second-leading cause of cancer deaths in
women worldwide, and results in approximately 250,000 deaths
each year from this HPV-related disease [1]. Nearly all cervical
cancers are initiated by a subset of high-risk HPVs, predominantly
HPV16 and HPV18 [2]. However, cervical cancers develop only
in a small fraction of these women, typically many years following
initial exposure. Therefore, HPV appears to be required, but insuf-
ficient for developing cervical cancer. Although recent advances in
HPV prevention have been made with the introduction of HPV
vaccines, they will only protect 30% of those subsequently infected
with high risk HPV subtypes, and will most likely have little or no
effect on patients already infected by HPV [1].
Several transgenic mouse models have been developed to
investigate the biology of HPV-mediated tumorigenesis [3,4,5]. In
one clinically relevant mouse model, the keratin-14 (K14)
promoter drives the expression of the HPV E6/E7 oncogenes
[6] specifically in stratified squamous epithelium, including that of
the skin and cervix. These K14-HPV16 mice develop cervical
pathologies only when they are treated with estrogen [7]. The
phenotype appears, at 5 months of age, as increased squamous
hyperplasia, epithelial papillomatosis, dysplasia, and squamous
metaplasia. Microscopic invasive tumors are detected in a small
fraction of the animals [8]. Cervical tumor cells in this transgenic
mouse model exhibit similar biomarker expression patterns to
those of humans [9]. In the second mouse model, the same K14
promoter drives HPV16-E7 oncogene expression [4,9], which
results in development of high-grade cervical dysplasia and
invasive cervical malignancies in 80% of the animals [10]. It
would be reasonable to conclude, therefore, that expression of the
HPV16-E7 oncogene in cervical epithelial cells is sufficient to
initiate oncogenesis.
Members of the Wnt family of secreted growth factors play
important roles during embryogenesis by regulating proliferation,
migration, tissue polarity, and organogenesis, and contribute to the
development of the genitourinary system. In the canonical Wnt
pathway, b-catenin acts as the central component [11]. The
binding of Wnt to its receptor (frizzled) and co-receptor (LRP5/6)
induces accumulation of b-catenin in the cytoplasm and nucleus.
PLoS ONE | www.plosone.org1November 2011 | Volume 6 | Issue 11 | e27243

Page 2

Nuclear b-catenin then binds to members of the T-cell factor
(TCF) transcription factor family, leading to transcription of Wnt
target genes [12]. Dysregulation of the expression of Wnt pathway
components has been implicated in the development of numerous
malignancies, including colon cancer, melanoma, hepatocellular
carcinoma, endometrial carcinoma, ovarian carcinoma, and
prostate cancer [13].
We have investigated the role of Wnt signaling in the
transformation of HPV-positive human keratinocytes [14]. In this
in vitro model, HPV immortalizes primary human keratinocytes but
cannot induce malignant transformation. Therefore, the model
mimics the effect of HPV on the human cervical epithelium.
Activation of the canonical Wnt pathway at multiple levels (plasma
membrane, cytoplasm or nucleus) specifically supports transforma-
tion of HPV-infected primary human keratinocytes. Furthermore,
we and others, have detected cytoplasmic and nuclear expression of
b-catenin, a hallmark of the activated Wnt pathway, in archived
human cervical carcinoma samples [14,15,16]. We hypothesized,
therefore, that activation of canonical Wnt signaling is one of the
potential mechanisms facilitating cervical cancer progression in
HPV-infected cells. To test this hypothesis, we generated two
double-transgenic mouse models to determine whether b-catenin
expression could induce tumors.
Results
Generation of a transgenic mouse model to study the
role of the Wnt signaling in HPV pathogenesis
We used K14-E7 and K14-DN87bcat transgenic animals, in
which the K14 promoter specifically targets expression to the
cervix, to generate K14-E7/DN87bcat double-transgenic mice.
K14-E7 mice express the E7 oncoprotein encoded by the high-risk
HPV type 16 genome [4]. K14-DN87bcat mice express truncated
human DN87bcat mRNA, which directs the synthesis of an
amino-terminal truncated b-catenin molecule that lacks the four
phosphorylation sites required for its degradation [17]. This
mutation results in expression of constitutively active b-catenin.
Under our conditions, the K14-E7 and K14-DN87bcat mice
exhibited phenotypes similar to those reported by others [4,17].
Thus, K14-DN87bcat mice develop benign skin tumors but do
not exhibit histopathologic characteristics of cervical cancer. All
transgenic animals were maintained as heterozygotes. Double-
transgenic mice were generated by crossing K14-DN87bcat mice
with K14-E7 mice. Crossing heterozygote K14-DN87bcat with
heterozygote K14-E7 mice produced offspring at the expected
Mendelian ratio. Only F1 generation female animals were used in
the study. Tumor formation was induced by implanting estrogen
pellets (0.05 mg/60 days) starting at one month of age. Each animal
received 3 pellets (once every 60 days).
K14-E7/DN87bcat double transgenics exhibited a more
severe epidermal phenotype
Animals were analyzed for cervical cancer development in four
experimental groups: wild type, K14-DN87bcat, K14-E7, and
K14-E7/DN87bcat-double transgenic animals. The genotypes of
each group were confirmed by PCR analysis of genomic DNA
(Figure 1A). Expression of DN87bcat protein in target tissues (skin
and cervix) was confirmed by immunoblotting using an anti-b-
catenin antibody (Figure 1B). Heart tissue, which expresses keratin
14 at very low levels, was used as a negative control. All organs
Figure 1. Transgenic genotypes and b-catenin expression. A) PCR genotyping K14-E7/DN87bcat mice. Genomic DNA from tail clips was used
as template for primers specific for K14-E7 and K14-DN87bcat transgenes. DNAs amplified with p53-specific primers served as positive controls.
Samples were electrophoresed through a 2% agarose gel and then imaged. B) Western blot analysis of truncated human b-catenin in K14-E7/
DN87bcat mice. Mouse b-catenin (arrow) and human DN87bcat protein (arrowhead) expression in heart, skin, and cervix tissues of animals was
analyzed by immunoblotting.
doi:10.1371/journal.pone.0027243.g001
Wnt Signaling in Cervical Cancer Progression
PLoS ONE | www.plosone.org2November 2011 | Volume 6 | Issue 11 | e27243

Page 3

analyzed from the four groups showed expression of mouse b-
catenin protein (91 kDa). In contrast, expression of human
DN87bcat was detected only in the skin and cervix of the K14-
DN87bcat and double-transgenic animals (80 kDa) (Figure 1B).
K14-E7/DN87bcat-double transgenic animals were viable and
displayed certain phenotypic features common to K14-E7 and
K14-DN87bcat transgenes (Figure 2) [4,17]. Thus, K14-E7/
DN87bcat-double transgenic exhibited a distinctive phenotype [4]
characterized by wrinkled skin on their torsos, legs, and jowls,
which was apparent early in postnatal life. They developed small
bilateral cataracts in the eyes that were visible at weaning
(Figure 2). As they aged, they exhibited thickened, less translucent
ears, thick ridges around the nose and the ears, and a paucity of
fur around the snout and eyes. Scaliness was apparent on their tail,
ears, and feet. Their coats appeared less dense, greasy, and scruffy,
resembling the K14-E7 phenotype more closely than that of K14-
DN87bcat (Figure 2) [4]. Similar to K14-DN87bcat transgenic
mice, they had grossly enlarged paws (Figure 2) and developed
pilomatricomas, which are benign skin tumors [17]. Due to the
combined effect of both transgenes, they exhibited a more severe
skin phenotype than the K14-DN87bcat and K14-E7 single
transgenics, especially around the snout and the eyelids (Figure 2).
The K14-E7/DN87bcat-double transgenic animals had the lowest
body weights (determined weekly) at the end of the study (Figure 3).
This difference was highly significant when compared with those
in other groups.
Histopathology
Riley et al., observed cervical pathologies after 6 months of
estrogen treatment (7 months of age) in K14-E7 transgenic
Figure 2. Phenotypic features of K14-E7/DN87bcat mice. Photographs of snout, right ear, eye, lips and hind paw are given. K14-E7/DN87bcat
double transgenic animals displayed certain phenotypic features characteristic of both K14-E7 and K14-DN87bcat transgenes. Due to the combined
effect of both transgenes, they had a more severe skin phenotype around the snout, ears, eyelids and lips. They had grossly enlarged paws similar to
those of K14-DN87bcat mice.
doi:10.1371/journal.pone.0027243.g002
Wnt Signaling in Cervical Cancer Progression
PLoS ONE | www.plosone.org3November 2011 | Volume 6 | Issue 11 | e27243

Page 4

animals [10]; therefore, we euthanized all animals at this same age
in the four groups. Proximal vaginal, cervix, and both uterine
horns were dissected en block and analyzed by histopathology.
Proliferating cell nuclear antigen (PCNA) expression was detected
by immunohistochemistry (IHC) to evaluate cell proliferation
within the cervical squamous epithelium (Figure 4). PCNA
expression was more abundant in K14-E7 and K14-E7/
DN87bcat mice than wild type and K14-DN87bcat. In the latter
two groups, PCNA expression was restricted to the basal and
innermost suprabasal layers, whereas in K14-E7 and K14-E7/
DN87bcat transgenics there was an incremental increase in the
labeling index within the multiple layers of the squamous
epithelium. However, PCNA expression in K14-E7/DN87bcat
compared with K14-E7 mice was not significantly different.
Cervical pathologies were analyzed by hematoxylin and eosin
(H&E) staining (Figure 5). Invasive cervical cancer developed in
11% (2/19) and 50% (8/16), respectively, in animals of the K14-
DN87bcat and in the K14-E7 groups, but not in controls (0/15).
Invasive cervical carcinomas developed in 94% (15/16) of the
double transgenics, suggesting that the activation of Wnt pathway
accelerates HPV16-E7-mediated cervical carcinogenesis (p=0.02)
(Figure 5A). A representative case from each genotype is given in
Figure 5B. When estrogen pellets were omitted, we did not observe
cervical cancer in either K14-E7 or K14-E7/DN87bcat (Figure
S1). We performed Periodic Acid Schiff (PAS) staining to evaluate
the basement membranes (Figure 6A) and analyzed CD31
(PECAM) expression by IHC to evaluate the blood vessels
(Figure 6B) in all four groups. The most striking finding was that
invasive tumors lacked a continuous basement membrane, nor did
tumors contain any significant vascular structures.
K14-HPV16/DN87bcat mice did not develop invasive
cervical carcinomas
We also tested progression of cervical pathology in a second
model employing K14-HPV16/DN87bcat transgenic mice. These
animals were generated by crossing K14-HPV16 mice, which
express E6 and E7 oncoproteins of the high risk HPV type 16
[5,10] with K14-DN87bcat mice [17]. K14-HPV16/DN87bcat
animals exhibited skin phenotypes comparable to those of K14-
HPV16 and K14-DN87bcat animals (Figure 7) [5,17]. Similar to
Figure 3. Body weights of K14-E7/DN87bcat mice. A) Whole body images of the four genotypes are given. K14-E7/DN87bcat animals had the
lowest body weight. B) Lifetime weight distributions of animals representing the four genotypes are shown. The differences between the K14-E7/
DN87bcat animals and other three genotypes were highly significant (p,0.0001). Independent t-test was used to determine statistical significance.
doi:10.1371/journal.pone.0027243.g003
Wnt Signaling in Cervical Cancer Progression
PLoS ONE | www.plosone.org4November 2011 | Volume 6 | Issue 11 | e27243

Page 5

the K14-E7/DN87bcat model, body weights of the K14-HPV16/
DN87bcat double transgenic animals were lower than those of
their littermates (Figure 8). In this group, K14-HPV16 mice
showed increased cervical dysplasia compared to wild type
(p,0.0001) and K14-DN87bcat mice (p=0.001), but this
phenotype was not augmented in K14-HPV16/DN87bcat animals
(p=0.59) (Figure 8C). We did not observe any invasive tumors in
K14-HPV16 animals. A representative case from each histopath-
ological group is shown in Figure 8D.
Discussion
Several animal models have been established in attempts to
understand the biology of HPV-mediated tumorigenesis, [3,4,5].
Two of the clinically relevant mouse models, K14-E7 and K14-
HPV16 utilized the K-14 keratin promoter to target transgene
expression to the stratified squamous epithelium, including the
skin and the cervix. In these systems, both types of transgenic mice
develop cervical pathologies following long-term estrogen treat-
ment [8,10]. We took advantage of these mice to test our
hypothesis that activation of the Wnt pathway by b-catenin in the
presence of HPV-E7 oncogene expression can accelerate the
development of cervical cancer. We were able to generate
transgenic mice (designated K14-E7/DN87bcat), which express
the HPV16-E7 oncogene along with constitutively active b-
catenin. We found that in this double-transgenic model, mice
developed invasive cervical cancers with 94% penetrance as early
as seven months of age. In striking contrast, these tumors occurred
in only 50% of K14-E7 mice. The latter result conflicts with the
studies of Herber et al., which reported invasive cervical cancer in
80% of K14-E7 animals by 7 months of age [4]. This difference
may be related to environmental or dietary factors.
The K14-DN87bcat mice used in our study were originally of
the CD1 background and were backcrossed into FVB mice for 15
generations before they were used in our study. Therefore, we do
not consider differences in genetic background to be a significant
difference between our study and the one reported by Herber et al.
Both CD1 and FVB strains are permissive to squamous epithelial
malignancies [5]. Another factor related to this disparity might be
diet, because phytoestrogens in the mouse diet can induce
molecular and physiological changes in the uterus [18]. Phytoes-
trogens in laboratory diets can vary between formulations and
from batch to batch of the same formulation [19]. Phytoestrogens
can also mimic the properties of mammalian estrogens. Since
estrogen dose is an important parameter in the development of
cervical cancer [8], dietary phytoestrogens can influence estrogen-
regulated processes and strongly influence studies on comparative
carcinogenicity. The animals in our study were fed the PMI
Irradiated Rodent 5053 diet from Harlan Laboratories, which
contains 380 ppm of phytoestrogens. Although our present results
differ from other reports regarding tumor penetrance [4], all of our
mice possessed the same genetic background (all F1 littermates),
diet, and housing environment. Therefore, our study was
internally controlled. We feel confident, therefore, in concluding
that our results support our main hypothesis that b-catenin can
accelerate cervical cancer progression.
We observed an intriguing difference in total body weights
between mice in the groups (Figure 3). However, we were unable
to determine the underlying mechanism. Because K14-E7/
DN87bcat mice had the lowest body weights and the highest
tumor incidence, tumor-related cachexia might have been
involved. However, when we compared the weights of K14-E7
animals with tumors (n=8) to ones without (n=8), we did not
detect a statistically significant difference (Figure S2). We could not
perform the same comparison for the other groups because the
number of animals was insufficient for meaningful statistical
analysis. We believe that it is reasonable to conclude that the
weight difference resulted from the progressively worsening
combined skin phenotype. We initially supposed that the double
transgenics would have difficulty in eating or drinking. However,
when we assessed food and water intake by members of each
group for 6 weeks, we did not observe any statistically significant
differences (Figure S3). Therefore, only the combined skin
phenotype remains as a plausible explanation.
Two out of 19 K14-E7/DN87bcat, but none of the K14-
HPV16/DN87b mice (0/10) developed invasive tumors. The
K14-DN87bcat mice that constitute both double transgenic
models are not derived from same genetic background and
therefore cannot be compared with each other. Since only F1
littermates were used in the study, K14-DN87bcat littermates of
K14-HPV16/DN87bcat were derived from one of each parental
heterozygote, namely K14-DN87bcat and K14-HPV16. In
contrast, K14-DN87bcat littermates of the K14-E7/DN87bcat
were derived from the parental heterozygotes K14-DN87bcat and
K14-E7. However, since both K14-HPV16 and K14-E7 possess
the FVB/N background, one would have expected no phenotypic
difference. To evaluate the genetic backgrounds of K14-HPV16
and K14-E7 mice, we performed SNP analyses on three animals
from each group. A 1449 loci medium density SNP linkage
analysis showed that K14-E7 and K14-HPV16 mice, respectively,
were 100% and 99.83% FVB/NTac. These findings suggested
that the alteration in the genetic background of K14-HPV16
animals might have been responsible for the differences between
the two K14-DN87bcat groups. Furthermore, the same genetic
alteration might account for why we observed only dysplasias in
Figure 4. IHC Analysis of PCNA expression in cervical
squamous epithelium. PCNA expression was restricted to the basal
and innermost suprabasal layers in cervical sections of wild type and
K14-DN87bcat transgenics, whereas in K14-E7 and K14-E7/DN87bcat
transgenics, there was an incremental increase in expression within the
multiple layers of the squamous epithelium. Cervical tissues were
harvested at the endpoint of the study, after six months of estrogen
treatment (0.05 mg/60 days, 7 months old). Images are at 680
magnification.
doi:10.1371/journal.pone.0027243.g004
Wnt Signaling in Cervical Cancer Progression
PLoS ONE | www.plosone.org5November 2011 | Volume 6 | Issue 11 | e27243

Page 6

K14-HPV16/DN87bcat mice compared with invasive cancer and
dysplasias no in K14-E7/DN87bcat mice.
We performed IHC to evaluate nuclear b-catenin levels (Figure
S4). We did not observe significant nuclear b-catenin staining
signal intensity in sections prepared from either group group
(Figure S4). The endogenous level of b-catenin is very high in the
mouse cervix (Figure 1B), but IHC was not sensitive enough to
detect protein expressed from transgene. However, our published
data and other IHC studies demonstrate that b-catenin localizes to
the cytoplasm or nucleus in more than two thirds of human
invasive cervical carcinoma samples [14,15,16], suggesting an
active canonical Wnt signaling pathway.
We believe that two mechanisms might account for observed
Wnt pathway activation in cervical cancer. Mutations in the b-
catenin gene are common in colon cancer, whereas b-catenin
mutations in cervical cancer are rare, which suggests activation of
the Wnt pathway upstream of b-catenin in cervical carcinoma
[20]. Epigenetic changes in Wnt pathway regulators in cervical
tumor samples and cell lines, as well as in other cancers have been
reported, which can explain the activation of the canonical Wnt
pathway in the absence of b-catenin mutations [21,22,23,24,
25,26,27]. Alternatively, HPV oncogenes might directly modulate
the Wnt pathway. One recent report provides evidence that
HPV16 E6 is able to augment b-catenin/TCF-dependent
transcription induced by Wnt3a or b-catenin expression. Because
this augmentation is not associated with GSK-3 b activity or major
alterations in the levels, stability, or cellular distribution of b-
catenin, the authors suggest that this relates to activation of
ubiquitin ligase E6AP by E6 [28]. In another study, Rampias et al.
(2010) reported that in oropharyngeal cancer cells, HPV16 E6 and
Figure 5. Histopathology. A) Distribution of cervical pathologies in the K14-E7/DN87bcat transgenics. Histopathological analysis of cervices from
wild type, K14-DN87bcat, K14-E7, and K14-E7/DN87bcat groups is shown. B) Histopathological evaluation of mouse cervical tissues. A representative
case from each genotype is shown. Cervical tissues were harvested at the endpoint of the study, after six months of estrogen treatment (0.05 mg/60
days, 7 months old), sectioned, and stained with H&E. Main panel and inset images are magnified by 640 and 6160, respectively.
doi:10.1371/journal.pone.0027243.g005
Wnt Signaling in Cervical Cancer Progression
PLoS ONE | www.plosone.org6 November 2011 | Volume 6 | Issue 11 | e27243

Page 7

E7 oncogenes are involved in nuclear accumulation of b-catenin
and activation of Wnt signaling mediated by the ubiquitin/
proteosome pathway [29].
Our study provides a potential link between activation of the
Wnt signaling pathway and its contribution to HPV-mediated
cervical cancer. These results indicate that activation of the
canonical Wnt pathway might represent secondary events that are
required for malignant transformation of HPV-infected epithelial
cells. Targeting the canonical Wnt pathway may therefore provide
the basis for developing clinical interventions to prevent disease
progression in populations at risk for HPV infection and to treat
advanced cervical cancers.
Materials and Methods
Materials
Unless stated otherwise, all chemical reagents were purchased
from Sigma (St. Louis, MO).
Ethics Statement
Georgetown University’s Institutional Animal Care and Use
Committee (GUACUC) approved all animal studies, protocol IDs
07-064 and 10-029.
Transgenic Mice
The generation of K14-E7, K14-HPV16 and K14-DN87bcat
transgenic mice has been described [4,5,17]. All three transgenic
mouse models were maintained as heterozygotes. K14-E7 and
K14-HPV16 (FVB) mice were acquired from the Mouse Models of
Human Cancers Consortium (MMHCC) of the National Cancer
Institute. K14-DN87bcat mice were kindly provided by Dr. E.
Fuchs, Rockefeller University. K14-DN87bcat transgenic mice
(CD1 background) were first backcrossed with FVB mice for 15
generations. Double transgenic mice were generated by crossing
K14-DN87bcat mice with either K14-E7 or K14-HPV16 mice.
Only F1 generation female animals were used.
Genotyping and genetic characterization
Offspring were screened for the presence of transgenes by PCR
amplification of genomic DNA isolated from mouse-tails at
weaning. Primers specific for the K14-E7 (F:59-GGCGGA-
TCCTTTTTATGCACCAAAGA-39, R:59-CCCGGATCCTA-
CCTGCAGGATCAGC-39), K14-HPV16 (F:59-AGAACTGCA-
ATGTTTCAGGACCCACAG-39, R:59-TCTGCAACAA GA-
CATACATCGACCGG-39) and K14-DN87bcat (F:59-TCCCAC-
TAATGTCCAGCG-39, R:59-CGCGGCATGCAGGTACC-39)
transgenes were used for genotyping. PCR reactions were initiated
by denaturation at 94uC for 7 min, followed by 1 min at 94uC,
1 min at 55uC, and 1 min at 72uC, for 35 cycles. The reaction was
completed with a final extension step at 72uC for 10 min.
Amplifications with p53-specific primers were used as a positive
control. Samples were run through a 2% agarose gel and then
imaged.
Genetic characterization of the K14-E7 and K14-HPV6 mice
were performed using 1449 dense Single Nucleotide Polymor-
phism (SNP) Marker Panel (GENCON Panel 4, Taconic,
Rensselaer, NY, USA).
DN87bcat Protein Expression
Heart, skin, and cervical tissues were snap frozen in liquid
nitrogen, ground to a powder, lysed with a buffer consisting of
20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 2.5 mM EDTA,
Figure 6. PAS staining and CD31 expression by K14-E7/DN87bcat transgenics. A) In normal cervical epithelium, PAS staining showed a
continuous basement membrane (arrows), whereas tumors lacked this structure of the membrane (arrowhead). Both normal and tumor images are
taken from the same section. B) Blood vessels were evaluated by CD31 (PECAM) expression using IHC. Vascular structures were present in normal
tissues (arrows), but the invasive tumors did not contain any. Both normal and tumor images are taken from the same section. Cervical tissues were
harvested at the study’s end, after six months of estrogen treatment (0.05 mg/60 days, 7 months old). Images are at 6200 magnification.
doi:10.1371/journal.pone.0027243.g006
Wnt Signaling in Cervical Cancer Progression
PLoS ONE | www.plosone.org7 November 2011 | Volume 6 | Issue 11 | e27243

Page 8

10 mM NaF, 10 mM Na2P2O7, 1 mM Na3VO4, 10 mM aproti-
nin, 20 mM leupeptin, and 1 mM PMSF. Total lysates (40 mg for
heart and cervix and 2 mg for skin samples) were resolved on SDS-
PAGE gel. Tissue-specific expression of human DN87bcat was
determined by immunoblot analysis with 250 ng/ml of an anti-b-
catenin antibody (BD Transduction LaboratoriesTM, San Jose,
CA) as described previously [30]. Heart tissue was used as negative
control because K14 promoter activity was not detected in this
tissue.
Hormone Treatment
One-month-old virgin female mice were anesthetized using 5%
isofluorane, and pellets of 17 b-estradiol (Innovative Research of
America, Sarasota, FL) delivering estrogen at a dose of 0.05 mg/
60 days were placed subcutaneously in the dorsal skin, near the
shoulder fat pads. Animals received a new pellet every 60 days.
Histopathological evaluation of mouse cervical tissues
Each organ was sectioned through its entire depth to generate
5 mM thick slices. A representative of 20 sections was stained with
H&E. Histopathology was performed by pathologists B.K. (K14-
E7/DN87bcat mice) and C.P.C. (for K14-HPV16/DN87bcat
mice). Samples were evaluated by a single blind method.
Immunohistochemistry
Five-micron sections from formalin-fixed paraffin embedded
cervical tissues were deparaffinized with xylenes and sequentially
rehydrated using a graded alcohol series. Heat-induced epitope
retrieval was performed by immersing the tissue sections at 98uC
for 20 min in 10 mM citrate buffer (pH 6.0), 0.05% Tween-20.
IHC was performed using the VectaStain Kit from Vector Labs
(Burlingame, CA) according to manufacturer’s instructions.
Briefly, slides were treated with 3% hydrogen peroxide and 10%
Figure 7. Phenotypic features of K14-HPV16/DN87bcat mice. K14-HPV16/DN87bcat double transgenic animals displayed certain phenotypic
features characteristic of both K14-HPV16 and K14-DN87bcat transgenes. Photographs of snout, right ear, eye, lips and hind paw are given.
doi:10.1371/journal.pone.0027243.g007
Wnt Signaling in Cervical Cancer Progression
PLoS ONE | www.plosone.org8 November 2011 | Volume 6 | Issue 11 | e27243

Page 9

normal serum, and incubated with a PCNA antibody (1:600
dilution, Santa Cruz Biotechnology, Inc., Santa Cruz, CA) for 1 h
at room temperature (RT). Slides were treated with the
appropriate species-specific biotin-conjugated secondary antibody
(Vector Labs), Vectastain ABC reagent (Vector Labs), DAB
chromagen (Dako, Carpinteria, CA), and then counterstained with
hematoxylin (Harris Modified Hematoxylin, Fisher Scientific,
Pittsburgh, PA) (1:17 dilution) for 2 min at RT, blued in 1%
ammonium hydroxide for 1 min at RT, dehydrated as described
above, and mounted with Acrymount (StatLab Medical Products,
McKinney, TX). Sections incubated with the secondary antibody
only were used as negative controls. b-catenin (Cell Signaling,
Danvers, MA, USA) and CD31 (PECAM, Santa Cruz) IHCs were
done in a similar manner with both antibody dilutions as 1:200.
All photomicrographs shown here and in all other figures were
taken using a Nikon E600 microscope.
Periodic Acid-Schiff (PAS) Staining
Periodic Acid-Schiff reaction (PAS) was performed as described
by Bancroft and Gamble with minor modifications [31]. Briefly,
five micron sections from formalin fixed paraffin embedded tissues
were de-paraffinized with xylenes and rehydrated through a
graded alcohol series. Sections were treated with 0.5% periodic
acid, microwaved in Schiff’s reagent (Sigma) for 40 seconds,
Figure 8. Body weights of K14-HPV16/DN87bcat mice. A) Whole body images of the four genotypes are given. B) Average body weights at
approximately 5 months of age for the four experimental groups. (*p,0.05, **p,0.001) are shown. Bars represent mean values, and error bars
represent standard deviations. Statistical analysis was done using the independent t-test. C) Distribution of cervical pathologies. Histopathological
analyses of wild type, K14-DN87bcat, K14-HPV16, and K14-HPV16/DN87bcat cervical sections are shown. We did not detect cervical pathology in wild
type (0/16) and K14-DN87bcat (0/10) animals. K14-HPV16 pathology varied as follows: cervical intraepithelial neoplasia-I (CIN-I), (3/15); CIN-II, (6/15);
and CIN-III, (4/15). The frequency of normal cervical histology was (2/15). K14-HPV16/DN87bcat pathologies included CIN-I (4/10), CIN-II (1/10), CIN-III
(2/10), and invasive tumor (1/10). Twenty percent exhibited normal histology (2/10). The average times of estrogen treatment of wild type, K14-
DN87bcat, K14-HPV16, and K14-HPV16/DN87bcat mice were, respectively, 4.6, 4.6, 4.4, and 3.8 months. D) K14-HPV16/DN87bcat cervical
histopathology. H&E stained slides of cervical tissues were analyzed for pathology. A representative case from each histopathological group is shown.
doi:10.1371/journal.pone.0027243.g008
Wnt Signaling in Cervical Cancer Progression
PLoS ONE | www.plosone.org9 November 2011 | Volume 6 | Issue 11 | e27243

Page 10

counterstained with Mayer’s hematoxylin and blued in ammoni-
um hydroxide. Sections were dehydrated, cleared in xylenes and
mounted with Acrymount.
Statistical Analysis
Statistical analyses were performed using Prism version 4.0
(GraphPad Software, La Jolla, CA). Animal weights between
groups were compared using an independent t-test. The Fisher
Exact test was used to compare rates of invasive cervical cancer
and the Wilcoxon rank-sum test was used to compare the
pathology data between experimental groups. Statistical signifi-
cance was defined as p,0.05.
Supporting Information
Figure S1
transgenics that were not treated with estrogen. A)
Cervical tissues were harvested at the study’s end when the
animals were 7 months old. Histopathological analysis of cervical
sections from wild type, K14-DN87bcat, K14-E7 and K14-E7/
DN87bcat mice are shown. B) Histopathological evaluation of
mouse cervical tissues. H&E-stained slides from cervices of mice
not treated with estrogen. A representative case from each
genotype is given. Images are at 6200 magnification.
(TIF)
Cervical phenotypes in K14-E7/DN87bcat
Figure S2
related to the presence of tumors. Average body weights of
Difference in total weight of animals was not
K14-E7 mice with and without tumors were compared. The
difference in between these two groups was non-significant.
Statistical analysis was evaluated using independent t-test.
(TIF)
Figure S3
in panels (A) and (B), respectively. Bars represent mean values, and
error bars represent standard deviations.
(TIF)
Dietary intake. Data for food and water are shown
Figure S4
using IHC in K14-E7/DN87bcat model. We did not observe
significant nuclear b-catenin signal intensity in sections prepared
from mice cervices. Cervical tissues were harvested at the study’s
end, after six months of estrogen treatment (0.05 mg/60 days, 7
months old). Left column is at 6100 and right column is at 6200
magnification.
(TIF)
Evaluation of nuclear bcatenin protein levels
Acknowledgments
We wish to thank Drs. Elaine Fuchs for K14-DN87bcat mice, Jeffrey
Arbeit, and Paul Lambert for helpful discussions.
Author Contributions
Conceived and designed the experiments: GB SF DLB JAT RS AU.
Performed the experiments: GB SF EMB LED DLB. Analyzed the data:
GB JS BK CPC AU. Wrote the paper: GB AU.
References
1. Kahn JA (2009) HPV vaccination for the prevention of cervical intraepithelial
neoplasia. N Engl J Med 361: 271–278.
2. Wolf JK, Franco EL, Arbeit JM, Shroyer KR, Wu TC, et al. (2003) Innovations
in understanding the biology of cervical cancer. Cancer 98: 2064–2069.
3. Song S, Pitot HC, Lambert PF (1999) The human papillomavirus type 16 E6
gene alone is sufficient to induce carcinomas in transgenic animals. J Virol 73:
5887–5893.
4. Herber R, Liem A, Pitot H, Lambert PF (1996) Squamous epithelial hyperplasia
and carcinoma in mice transgenic for the human papillomavirus type 16 E7
oncogene. J Virol 70: 1873–1881.
5. Arbeit JM, Munger K, Howley PM, Hanahan D (1994) Progressive squamous
epithelial neoplasia in K14-human papillomavirus type 16 transgenic mice.
J Virol 68: 4358–4368.
6. Lambert PF, Pan H, Pitot HC, Liem A, Jackson M, et al. (1993) Epidermal
cancer associated with expression of human papillomavirus type 16 E6 and E7
oncogenes in the skin of transgenic mice. Proc Natl Acad Sci U S A 90:
5583–5587.
7. Arbeit JM, Howley PM, Hanahan D (1996) Chronic estrogen-induced cervical
and vaginal squamous carcinogenesis in human papillomavirus type 16
transgenic mice. Proc Natl Acad Sci U S A 93: 2930–2935.
8. Elson DA, Riley RR, Lacey A, Thordarson G, Talamantes FJ, et al. (2000)
Sensitivity of the cervical transformation zone to estrogen-induced squamous
carcinogenesis. Cancer Res 60: 1267–1275.
9. Brake T, Connor JP, Petereit DG, Lambert PF (2003) Comparative analysis of
cervical cancer in women and in a human papillomavirus-transgenic mouse
model: identification of minichromosome maintenance protein 7 as an
informative biomarker for human cervical cancer. Cancer Res 63: 8173–8180.
10. Riley RR, Duensing S, Brake T, Munger K, Lambert PF, et al. (2003) Dissection
of human papillomavirus E6 and E7 function in transgenic mouse models of
cervical carcinogenesis. Cancer Res 63: 4862–4871.
11. Clevers H (2006) Wnt/beta-catenin signaling in development and disease. Cell
127: 469–480.
12. Cadigan KM (2008) Wnt-beta-catenin signaling. Curr Biol 18: R943–947.
13. Polakis P (2000) Wnt signaling and cancer. Genes Dev 14: 1837–1851.
14. Uren A, Fallen S, Yuan H, Usubutun A, Kucukali T, et al. (2005) Activation of
the canonical Wnt pathway during genital keratinocyte transformation: a model
for cervical cancer progression. Cancer Res 65: 6199–6206.
15. Rodriguez-Sastre MA, Gonzalez-Maya L, Delgado R, Lizano M, Tsubaki G,
et al. (2005) Abnormal distribution of E-cadherin and beta-catenin in different
histologic types of cancer of the uterine cervix. Gynecol Oncol 97: 330–336.
16. Shinohara A, Yokoyama Y, Wan X, Takahashi Y, Mori Y, et al. (2001)
Cytoplasmic/nuclear expression without mutation of exon 3 of the beta-catenin
gene is frequent in the development of the neoplasm of the uterine cervix.
Gynecol Oncol 82: 450–455.
17. Gat U, DasGupta R, Degenstein L, Fuchs E (1998) De Novo hair follicle
morphogenesis and hair tumors in mice expressing a truncated beta-catenin in
skin. Cell 95: 605–614.
18. Wang H, Tranguch S, Xie H, Hanley G, Das SK, et al. (2005) Variation in
commercial rodent diets induces disparate molecular and physiological changes
in the mouse uterus. Proc Natl Acad Sci U S A 102: 9960–9965.
19. Jensen MN, Ritskes-Hoitinga M (2007) How isoflavone levels in common rodent
diets can interfere with the value of animal models and with experimental results.
Lab Anim 41: 1–18.
20. Ueda M, Gemmill RM, West J, Winn R, Sugita M, et al. (2001) Mutations of the
beta- and gamma-catenin genes are uncommon in human lung, breast, kidney,
cervical and ovarian carcinomas. Br J Cancer 85: 64–68.
21. Dong SM, Kim HS, Rha SH, Sidransky D (2001) Promoter hypermethylation of
multiple genes in carcinoma of the uterine cervix. Clin Cancer Res 7:
1982–1986.
22. Lee J, Yoon YS, Chung JH (2008) Epigenetic silencing of the WNT antagonist
DICKKOPF-1 in cervical cancer cell lines. Gynecol Oncol 109: 270–274.
23. Lee J, Jeong DJ, Kim J, Lee S, Park JH, et al. (2010) The anti-aging gene
KLOTHO is a novel target for epigenetic silencing in human cervical
carcinoma. Mol Cancer 9: 109.
24. Chung MT, Lai HC, Sytwu HK, Yan MD, Shih YL, et al. (2009) SFRP1 and
SFRP2 suppress the transformation and invasion abilities of cervical cancer cells
through Wnt signal pathway. Gynecol Oncol 112: 646–653.
25. Marsit CJ, McClean MD, Furniss CS, Kelsey KT (2006) Epigenetic inactivation
of the SFRP genes is associated with drinking, smoking and HPV in head and
neck squamous cell carcinoma. Int J Cancer 119: 1761–1766.
26. Marsit CJ, Karagas MR, Andrew A, Liu M, Danaee H, et al. (2005) Epigenetic
inactivation of SFRP genes and TP53 alteration act jointly as markers of invasive
bladder cancer. Cancer Res 65: 7081–7085.
27. Pannone G, Bufo P, Santoro A, Franco R, Aquino G, et al. (2010) WNT
pathway in oral cancer: epigenetic inactivation of WNT-inhibitors. Oncol Rep
24: 1035–1041.
28. Lichtig H, Gilboa DA, Jackman A, Gonen P, Levav-Cohen Y, et al. (2010)
HPV16 E6 augments Wnt signaling in an E6AP-dependent manner. Virology
396: 47–58.
29. Rampias T, Boutati E, Pectasides E, Sasaki C, Kountourakis P, et al. (2010)
Activation of Wnt signaling pathway by human papillomavirus E6 and E7
oncogenes in HPV16-positive oropharyngeal squamous carcinoma cells. Mol
Cancer Res 8: 433–443.
30. Uren A, Wolf V, Sun YF, Azari A, Rubin JS, et al. (2004) Wnt/Frizzled
signaling in Ewing sarcoma. Pediatr Blood Cancer 43: 243–249.
31. Barcroft JD, Gamble M (2002) Theory and Practice of Histological Techniques:
Churchill Livingstone.
Wnt Signaling in Cervical Cancer Progression
PLoS ONE | www.plosone.org10November 2011 | Volume 6 | Issue 11 | e27243