Published OnlineFirst January 2, 2013; DOI: 10.1158/1541-7786.

MCR-12-0510

Human Papilloma Virus (HPV) E7-Mediated Attenuation of

Retinoblastoma (Rb) Induces hPygopus2 Expression via Elf-1 in
Cervical Cancer
Youlian R. Tzenov, Phillip G. Andrews, Kim Voisey, et al.
Mol Cancer Res 2013;11:19-30. Published OnlineFirst January 2, 2013.

Updated version
Supplementary
Material

Cited Articles

E-mail alerts
Reprints and
Subscriptions
Permissions

Access the most recent version of this article at:

doi:10.1158/1541-7786.MCR-12-0510
Access the most recent supplemental material at:
http://mcr.aacrjournals.org/content/suppl/2013/06/06/1541-7786.MCR-12-0510.DC1.html

This article cites by 49 articles, 18 of which you can access for free at:
http://mcr.aacrjournals.org/content/11/1/19.full.html#ref-list-1

Sign up to receive free email-alerts related to this article or journal.

To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at
pubs@aacr.org.
To request permission to re-use all or part of this article, contact the AACR Publications Department at
permissions@aacr.org.

Downloaded from mcr.aacrjournals.org on August 11, 2014. 2013 American Association for Cancer Research.

Published OnlineFirst January 2, 2013; DOI: 10.1158/1541-7786.MCR-12-0510

Molecular
Cancer
Research

Cell Cycle and Senescence

Human Papilloma Virus (HPV) E7-Mediated Attenuation of

Retinoblastoma (Rb) Induces hPygopus2 Expression via Elf-1
in Cervical Cancer
Youlian R. Tzenov1, Phillip G. Andrews1, Kim Voisey3, Paul Popadiuk4, Jieying Xiong1, Catherine Popadiuk2,
and Kenneth R. Kao1,3

Abstract
The human papillomavirus (HPV) is the etiologic agent of cervical cancer. In this study, we provide evidence for
the human Pygopus (hPygo)2 gene as a cellular biomarker for HPV-related disease. In a tumor microarray of cervical
cancer progression, hPygo2 levels were greater in high-grade lesions and squamous cell carcinomas than in normal
epithelia. Similarly, hPygo2 mRNA and protein levels were greater in HPV-positive cervical cancer cells relative to
uninfected primary cells. RNA interference (RNAi)-mediated depletion of HPV-E7 increased whereas E74-like
factor (Elf)-1 RNAi decreased association of Retinoblastoma (Rb) tumor suppressor with the hPygo2 promoter in
cervical cancer cell lines. Transfection of dominant-active Rb inhibited Elf-1-dependent activation of hPygo2,
whereas Elf-1 itself increased hPygo2 expression. Chromatin immunoprecipitation assays showed that Rb repressed
hPygo2 by inhibiting Elf-1 at the Ets-binding site in the hPygo2 promoter. These results suggested that abrogation of
Rb by E7 resulted in derepression of Elf-1, which in turn stimulated expression of hPygo2. Thus, initiation of hPygo2
expression by Elf-1 was required for proliferation of cervical cancer cells and its expression therefore may act as a
surrogate marker for dysplasia. Mol Cancer Res; 11(1); 1930.
2012 AACR.

Introduction
The Wnt/b-catenin transcription complex component,
Pygopus, was originally identied as a protein required for
Wnt-dependent transcriptional activation during embryonic development (1). Pygopus is also overexpressed in and is
required for the growth of a number of cancer cell lines of
diverse origin (26). Because of its use by malignant cells,
Pygo is a candidate biomarker both for diagnostic and
potentially therapeutic benet, but our understanding of
the mechanisms that augment its expression in disease is
limited.
An important factor required for hPygo2 expression
in breast and cervical carcinoma cell lines is the E74-like
factor (Elf-1; ref. 7), an E26 transformationspecic
(ETS) family member. ETS factors have well-established
Authors' Afliations: 1Divisions of Biomedical Science, 2Women's Health,
Memorial University; 3Immunohistochemistry Lab, Laboratory Medicine,
Eastern Health, St. John's, Newfoundland; and 4Dalla Lana School of
Public Health, University of Toronto, Toronto, Ontario, Canada
Note: Supplementary data for this article are available at Molecular Cancer
Research Online (http://mcr.aacrjournals.org/).
This research was conducted in partial fulllment of the degree of PhD for
Y.R. Tzenov.
Corresponding Author: Kenneth R. Kao, Division of Biomedical Sciences,
Faculty of Medicine, Memorial University of Newfoundland, 300 Prince
Philip Drive, St. John's, NL, Canada, A1B 3V6. Phone: 709-777-6860; Fax:
709-777-7010; E-mail: kkao@mun.ca
doi: 10.1158/1541-7786.MCR-12-0510

2012 American Association for Cancer Research.

roles in carcinogenesis, regulating oncogenes and tumor

suppressors involved in apoptosis, angiogenesis, invasion,
and metastasis (8). Elf-1 itself activates genes involved in
tumorigenesis (912), is overexpressed in several cancers
(1316), and correlates with poor prognosis (1720). In
normal resting T lymphocytes, the pocket region of the
Retinoblastoma (Rb) tumor suppressor interacts with the
N-terminal LXCXE motif of promoter bound Elf-1 and
blocks its transactivation (21).
Paradigmatic of pathogenically mediated Rb deregulation
is the action of human papillomavirus (HPV) E7 protein, a
key etiologic agent in the initiation of cervical cancer (22). E7
is expressed following infection of proliferation-competent
cervical cells with certain high-risk HPV subtypes. Similar to
other viral proteins such as papovavirus T antigen and
adenovirus E1A (23), E7 is a 98 amino acid nuclear phospho-oncoprotein, which cooperates with the HPV E6 protein to cause immortalization of epithelial cells (24). E7
induces degradation of the Rb protein, resulting in activation
of several transcriptional regulators such as E2F and Elf-1,
which promote cell-cycle progression (25).
We are interested in understanding the mechanism of
expression and requirement of hPygo2 in cancer. In the
initiation of cervical carcinoma, HPV genome integration in
abnormally proliferating cervical cells raises the probability
that they will progress to malignancy because the HPV E2
genomic region is spliced out and the HPV E2 regulatory
protein is no longer produced (26). The loss of E2 causes the
constitutive expression of E6 and E7. E6 recruits the cellular
E3 ubiquitin ligase E6-associated protein and forms a

www.aacrjournals.org

Downloaded from mcr.aacrjournals.org on August 11, 2014. 2013 American Association for Cancer Research.

19

Published OnlineFirst January 2, 2013; DOI: 10.1158/1541-7786.MCR-12-0510

Tzenov et al.

trimeric complex with p53 (27) leading to proteasomal

degradation (28), thereby preventing cell-cycle arrest and/
or inducing apoptosis. The E7 protein, through the LXCXE
motif, binds to the pocket region of Rb and drives its
continuous proteolytic degradation (29), permitting Elf-1
to initiate expression of target genes. We hypothesized that
the increase in hPygo2 expression required for growth of
transformed cervical cells may be the result of Elf-1 activation caused by HPV E7-mediated attenuation of Rb.
In this study, we provide evidence that hPygo2 is overexpressed in cervical cancer and is required for growth of
transformed HPV-infected human endo- and ectocervical
cells. Our data indicated that hPygo2 overexpression in
cervical cancer is augmented by abrogation of Rb function
by E7 leading to derepression of Elf-1, suggesting a mechanistic link between HPV and the cellular response of the
oncogenic hPygo2 transcription factor.
Materials and Methods
Tissue microarray and immunostaining
Cervical tumor microarrays (Cybridi, Lot No. CC10-11004 (192-194) and US Biomax, Inc., (Cat No. CIN481;
Supplementary Table S1) were stained using a Ventana
Benchmark Ultra automated clinical immunostainer optimized for each of the antibodies as provided in Supplementary Table S2 (staining protocols as per manufacturer's
instruction or as indicated in Supplementary Table S2).
Cells were processed and stained for immunouorescence as
previously described (3).
ImageJ (Rasband, W.S., ImageJ, U.S. NIH, Bethesda,
MD; http://imagej.nih.gov/ij/, 19972012) counting software was used to quantify staining intensity in immunohistochemical images. Only cells in which nuclei were visible
were stained and were categorized into negative, weak, and
strong staining (Supplementary Fig. S1). The percentages of
negative, weak, and strong staining cells were calculated for
each core representing 5 diagnoses of disease progression
[normal, CIN 1, CIN 2, CIN 3, and squamous cell carcinoma (SCC)] for 3 observers.
Cell lines
Normal human endocervical (HEN; ref. 30) and ectocervical (HEC; ref. 31) primary cells and their HPV16 or
18 and cigarette smoke condensate transformed subclones,
HEN 16T (32) and HEC 18T (33), were generous gifts
from Dr. A. Pater. Primary cell lines were maintained in
Dulbecco's Modied Eagle's Media (DMEM) supplemented with 2.1 mL/mL bovine pituitary extract (Gibco) and
0.06 mL/mL EGF (Gibco BRL) and cancer cells in DMEM
plus 10% heat-inactivated FBS. Detailed information on cell
lines is provided in Supplementary Table S4 and Supplementary Fig. S4.
RNA extraction, cDNA generation, and quantitative
PCR
Cellular RNA isolation using the Nucleospin RNA II Kit
(Macherey-Nagel) and cDNA generation was previously

20

Mol Cancer Res; 11(1) January 2013

described (34). cDNA samples were subjected to real-time

quantitative PCR using RT2 SYBR green master mix (SA
Biosciences; ref. 7) and analyzed by the relative quantitative
comparative threshold cycle (DDCt) method (3). Primer
sequences are listed in Supplementary Table S5.
Protein extraction, SDS-PAGE, and immunoblotting
Cellular protein extraction and separation on SDS-PAGE
were transferred and detected as previously described (2).
Antibody information is provided in Supplementary Table
S2.
RNA interference and rescue assays
Four siRNA oligonucleotides were synthesized by Dharmacon, including siPy2-X (which targets the 30 -UTR
sequence of hPygo2 of endogenous hPygo2 mRNA),
siPy2-Z (which targets the coding region of hPygo2), siE7
(34), and a nonspecic negative control, siNTC. We have
previously described the Elf-1 targeting siRNA, siElf-1, (7).
Oligonucleotide sequences are listed in Supplementary
Table S5.
A total of 3  105 cells per well were seeded in 6-well plates
and forward transfected with siRNA oligos at nal concentrations of 10 to 25 nmol/L using Lipofectamine RNAiMAX
(Invitrogen) as per the manufacturer's instructions. In rescue
assays, cells were additionally transfected with 1 mg/well of
pCS2 or pCS2 hPygo2 24 hours after the siRNA
transfection using Lipofectamine with Plus Reagent (Invitrogen). Cells were harvested 72 hours after seeding.
Fluorescent-activated cell sorting
Trypsinized and PBS washed cells were xed in 2%
paraformaldehyde and permeabilized in 90% methanol/
1 PBS. After washing, cells were incubated in PBS containing 1 mL of 10 mg/mL propidium iodide and 10 mL of 10
mg/mL RNase at 37 C for 20 minutes in the dark. Samples
were analyzed using a FACSCalibur ow cytometer and
graphically displayed using ModFit analysis software.
Chromatin immunoprecipitation
Chromatin immunoprecipitation (ChIP) assays were conducted as previously described (7). Cross-linked cells were
and sonicated to produce 500 bp genomic DNA fragments.
About 300 mg of precleared chromatin were immunoprecipitated with 2 mg of anti-Rb and anti-Elf-1 antibodies.
Normal rabbit and mouse IgGs were used as negative
immunoprecipitation controls. After washing, reversing
cross-links and purifying DNA, hPygo2, CCNA, and
HCCS1 promoters were amplied by qPCR (sequences
given in Supplementary Table S5). Promoter occupancy
was calculated as a percentage relative to input chromatin
levels.
Plasmids
pECE-PSM-Rb was generously provided by Dr. Brenda
Gallie (35). The pECE-DKPN vector control for pECEPSM-Rb was generated by digesting pECE-PSM-Rb with
KpnI and re-ligating to remove the transcriptional start site.

Molecular Cancer Research

Downloaded from mcr.aacrjournals.org on August 11, 2014. 2013 American Association for Cancer Research.

Published OnlineFirst January 2, 2013; DOI: 10.1158/1541-7786.MCR-12-0510

HPV E7 Induces hPygo2

pCS2 Elf-1 (7) contains the human Elf-1 cDNA. pCS2

was used as the empty vector control for Elf-1 plasmids.
hPygo2 promoter constructs: pGL3-1494 antisense, pGL31494, pGL3-48, and pGL3-basic were used previously (7).
Site-directed mutagenesis to generate the dominant-active
(DA) Rb bindingdecient Elf-1 mutant, pCS2DA-Elf1, using sequences in Wang and colleagues (21), and the Etsbinding site mutant, pGL3-48 mutEBS was conducted
using the QuickChange Kit (Stratagene; primers listed in
Supplementary Table S5). Plasmids were veried by
sequencing.
Transient transfections
Cells were seeded 24 hours before transfection using
Lipofectamine and Plus reagent in 6 well plates (3.0 
105 cells/well) using 1 mg of DNA (per well) or 15-cm plates
(7.5  106 cells) using 5 mg of DNA. Three types of
transfections were conducted: (i) 2 expression plasmids
(pECE-PSM-Rb, pCS2 DA-Elf-1, pCS2 Elf-1) and
their empty plasmid counterparts (0.5 mg of each per well);
(ii) one luciferase reporter plasmid (pGL3-1494 antisense,
pGL3-1494, pGL3-48, pGL3-48 mutEBS, or pGL3-basic)
with a plasmid expressing b-galactosidase, pRSV-bgal (0.5
mg of each per well or 2.5 mg of each per plate); and (iii) 2
expression plasmids (0.25 mg of each per well), one luciferase
reporter plasmid (0.25 mg per well) and pRSV-bgal (0.25 mg
per well). Luciferase assays were conducted and normalized
to b-galactosidase activity as previously described (7).
Image acquisition and analysis
Films were scanned at a resolution of 600 dpi, set to
grayscale, cropped, and used to make gures. No software
adjustments were made. Densitometry was conducted on
scanned lms to quantify relative protein levels using ImageJ
(NIH) software.
Results
hPygo2 protein is overexpressed in CIN3 and SCC
The relative expression of hPygo2 in HPV-infected cervical tissues was assessed using a microarray of tissue core
samples representing various stages of disease progression,
stained by immunohistochemistry, with antibodies against
HPV L1, hPygo2, PCNA, and Elf-1.
HPV protein was detected in normal ectocervical as well as
dysplastic and malignant tissue. The staining intensity using
this antibody was higher in nondiseased ectocervical but not
in endocervical epithelia (Supplementary Fig. S2A; ref. 36).
The proportion of HPV-positive cells did not vary
signicantly.
hPygo2 protein was detected in both normal and pathologic tissues, with the intensity of staining appreciably
higher in cervical intraepithelial neoplasia grade 3 (CIN3)
and SCC cores. The proportion of positively stained cells
increased signicantly from 7% in normal tissues to 93% in
severely dysplastic (CIN3) and malignant tissues. The cellular localization of hPygo2 was either solely cytoplasmic or
both cytoplasmic and nuclear (Supplementary Fig. S2B).

www.aacrjournals.org

Both the expression and staining intensity of the proliferation marker proliferating cell nuclear antigen (PCNA)
(37) signicantly increased with severity of disease, increasing from approximately 20% in normal epithelia and 41.5%
in mild dysplasia (CIN1) to 81% and 76% in CIN3 and
SCCs, respectively. In positive cores, PCNA protein was
primarily localized to the nucleus (Supplementary Fig. S2B).
Elf-1 protein was detected in all tissues and tumors. While
staining intensity clearly decreased with disease severity, the
proportion of positively stained cells remained high in all
epithelia (above 40%) with the exception of normal endocervical cores (24%). As expected, Elf-1 was localized to both
the cytoplasm and nucleus (13).
The relatively higher levels of hPygo2 in CIN3 and SCCs
prompted further examination. We compared the expression
of hPygo2 and 2 established diagnostic markers for cervical
neoplasia (p16 and Ki-67; ref. 38), in a separate cervical
intraepithelial neoplasia tissue array, by immunohistochemistry. The numbers of negative, weak, and strongly staining
cells were determined (Fig. 1 and Supplementary Figs. S1
and S3 and Supplementary Table S3).
The expression of Ki-67 increased progressively with the
highest intensity (P < 0.02), as measured by the percentage of
strongly staining cells, in cores representing SCC (Fig. 1A
and Supplementary Fig. S3). Expression of p16 was absent
from normal and CIN1 cores and slightly increased in CIN2
(Fig. 1B and Supplementary Fig. S3), but staining intensity
was signicantly higher (P < 0.02) in CIN3 and SCC cores.
Using an ordinal logistic random-effects model (Supplementary Table S6), we determined that the intensity of
hPygo2 antibody staining (Fig. 1C and Supplementary Fig.
S3) increased signicantly with disease progression from
CIN2 to CIN3 and SCCs (P < 0.05).
Taken together, these observations suggested that expression of hPygo2 protein was upregulated in many CIN3
lesions and SCC tumors with respect to less serious disease
and normal cervical tissue.
hPygo2 mRNA and protein expression and subcellular
localization in cervical cancer cell lines
The foregoing observations suggested that increased
hPygo2 expression might be a response to HPV integration
and E7 upregulation (occurring in CIN3 and beyond). To
determine the relationship between hPygo2 and E7, we
conducted gene expression analyses in a variety of previously
developed normal and transformed cervical cell lines (30
33). The control cell lines included human primary HPVnegative endo- (HEN) and (HEC) ectocervical cells. The
experimental cancer cell lines included HPV-16 and HPV18 immortalized cells transformed with cigarette smoke
condensate, HEN-16T and HEC-18T, respectively. All 4
human cell lines have been extensively characterized (Supplementary Table S4 and Supplementary Fig. S5).
hPygo2 mRNA expression was slightly higher in HEC
cells relative to HEN cells and signicantly increased (by at
least 9-fold) in HEN 16T and HEC 18T cervical cancer cells
(Fig. 2A). The expression of PCNA and Elf-1 followed the
same trend yet increased in HEC and cancer cell lines. As

Mol Cancer Res; 11(1) January 2013

Downloaded from mcr.aacrjournals.org on August 11, 2014. 2013 American Association for Cancer Research.

21

Published OnlineFirst January 2, 2013; DOI: 10.1158/1541-7786.MCR-12-0510

Tzenov et al.

100
80

Ki67

60

Negative

40

Weak

20

Strong

C
SC

III

II

IN
C

IN

IN
C

or

al

120
100

p16

80
60

Negative

40

Weak

20

Strong

IN

IN
C

SC

III

II

I
IN
C

or

al

100
80

hPygo2

60

Negative

40

Weak

20

Strong
C
SC

III

IN

II
IN

I
IN
C

or

al

Figure 1. Quantication of protein expression by immunohistochemistry.

A cervical intraepithelial neoplasia tissue microarray was stained for (A)
Ki-67, (B) p16, and (C) hPygo2. Proportions of negative, weak, and
strongly positive staining cells in each set of cores representing each
stage of cervical disease progression (horizontal axes) determined by 3
observers. Shown are the means and SDs of each of the calculated values
for each core analyzed. The asterisks over the strongly expressing data
for CIN3 and SCC indicate signicant difference between these
diagnoses and the CIN2 samples.

expected, Rb tumor suppressor levels were lower in the

tumorigenic cell lines. The protein levels of Elf-1 and PCNA
closely paralleled mRNA levels (Fig. 2B). The 80 kDa
(cytoplasmic) and 98 kDa (nuclear) are the predominant
forms of Elf-1. Henceforth, we focus on the nuclear isoform
as it is the only isoform capable of binding to the hPygo2
promoter. hPygo2 is detected as 2 differentially migrating
bands on gel electrophoresis, possibly identifying 2 isoforms,
which were observed using a variety of antibodies (data not
shown). Both forms also paralleled the corresponding
mRNA levels of expression. While not detectable in HEN
cells, hPygo2 was highly localized to nuclei in HEN 16T cells
using immunouorescence (Fig. 2C and Supplementary Fig.
S4). These results suggested that hPygo2 mRNA and protein

22

Mol Cancer Res; 11(1) January 2013

are higher in HPV-transformed cervical cells relative to

primary cells and that hPygo2 expression may be regulated
at the transcriptional level.
hPygo2 and Elf-1 are required for cervical cancer cell
proliferation
The higher levels of hPygo2 observed in HPV-infected cells
and tissues suggested that it may, as in other cancers (26, 39),
have an important function in cervical cancer. HEN16T cells
were therefore transfected with 2 siRNAs directed against
hPygo2, siPy2-X, targeting its 30 -UTR and siPy2-Z targeting
the coding region to determine the requirement of hPygo2.
Either siPy2-X or siPy2-Z signicantly reduced both isoforms
of hPygo2 as compared with the nontarget control siRNA
(siNTC). When transfected with siNTC, the proportion of
cells that accumulated in G1 was 33% but was appreciably
higher at 66% (P < 0.01) after siPy2-X treatment and 55% (P
< 0.01) after siPy2-Z treatment (Fig. 3A). The increase of cells
in G1 was accompanied by a signicant decrease in the
percentage of cells in S-phase. HEC 18T cells redistributed
in a similar manner after hPygo2 siRNA treatment with the
largest proportion of cells arresting in G1 (Fig. 3B).
Associated with G1-dependent cell-cycle arrest is the
activation of the tumor suppressor protein 53 (p53; ref. 40).
Indeed, higher levels of p53 and its target gene cyclindependent inhibitor 1 (p21) coincided with hPygo2 knockdown-mediated, G1 phase cell-cycle arrest in HEN 16T and
HEC 18T cells (Fig. 3C and D). Interestingly, the increase in
p53 levels was reduced to normal levels by transiently cotransfecting cells with hPygo2, suggesting that elevated levels
of hPygo2 protein are a requirement for the proliferation of
HPV-transformed HEN 16T and HEC 18T cervical cells.
Our previous work showed that the oncogenic v-Ets
family transcription factor E74-like factor 1 (Elf-1) bound
to the hPygo2 promoter and could promote its expression in
MCF-7 breast cancer cells (7). We therefore used RNAi to
determine whether Elf-1 was involved in hPygo2-dependent
cervical cancer cell proliferation. siElf-1 increased the proportion of HEN 16T cells in G1 by 27% (P < 0.01) relative
to siNTC (Fig. 3E), which was accompanied by a signicant
decrease in the proportion of cells in S-phase. siElf-1 treatment of HEC 18T cells had a similar effect, increasing the
proportion of cells in G1 by 34% (P < 0.01; Fig. 3F).
In both HEN 16T and HEC 18T cells, the reduction of
nuclear Elf-1 protein levels by siElf-1 paralleled a signicant
decrease in hPygo2 protein expression (Fig. 3G and H).
Similar to our results with the hPygo2 knockdown, there was
an increase in p53 and p21 expression upon depletion of
nuclear Elf-1. The siElf-1induced increase in p53 was
attenuated back to control levels by overexpression of
hPygo2 (Fig. 3G and H), suggesting that Elf-1mediated
activation of hPygo2 expression was required for cell-cycle
progression in these cells.
Rb-dependent regulation of hPygo2 expression via Elf-1
is deactivated by HPV E7 protein
The requirement of Elf-1 and hPygo2 in cervical cancer
suggested that they are activated following HPV integration.

Molecular Cancer Research

Downloaded from mcr.aacrjournals.org on August 11, 2014. 2013 American Association for Cancer Research.

Published OnlineFirst January 2, 2013; DOI: 10.1158/1541-7786.MCR-12-0510

HPV E7 Induces hPygo2

Figure 2. Expression analysis of hPygo2 in cervical cancer cell lines. A, expression of hPygo2, PCNA, Elf-1, and Rb mRNA was analyzed by quantitative PCR in
normal (HEN and HEC) and malignant cervical cell lines (HEN 16T and HEC 18T). mRNA levels were normalized to glyceraldehyde-3-phosphate
dehydrogenase (GAPDH) and set to 2 in HEN cells, adjusted accordingly for the other cell lines and plotted on a logarithmic scale. B, expression levels of
hPygo2, PCNA, Elf-1, and Rb in normal and malignant cervical cell lines were measured using immunoblots. b-Actin was used as a loading control. Molecular
weight is expressed in kDa. C, normal HEN and cervical cancer (HEN 16T) cell lines were processed for immunocytochemical (left column) and
immunouorescent (middle and right columns) detection of hPygo2 (brown and green, respectively). Cells were additionally co-stained with 40 ,6-diamidino-2phenylindole (DAPI) and b-catenin. Full-length blots are presented in Supplementary Fig. S6.

A critical effector for HPV pathogenesis is the deregulated

expression and activity of its E7 viral oncoprotein. As E7
induces degradation of Rb, we predicted that upregulation of
Rb by reducing E7 would be expected to decrease Elf-1
activity, leading to hPygo2 downregulation.
To assess whether E7 was required for increased Elf-1
activity and subsequently hPygo2 expression, we knocked
down E7 protein using RNAi. Treatment of HEN 16T cells
with an E7-specic siRNA siE7 effectively reduced E7
protein levels compared with the nontarget control siRNA
(Fig. 4A). The reduction in E7 was associated with a decrease
in hPygo2, but an increase in Rb protein levels, while not
affecting levels of Elf-1. It is possible, therefore, that Rb
attenuated hPygo2 levels by regulating the activity but not
the expression or stability of Elf-1.
Elf-1 is an Ets-related transcription factor that binds to a
number of target genes including the hPygo2 promoter
whose expression is augmented in cancer (2, 9, 41, 42). In
vitro protein interaction assays showed that Rb represses Elf1 by binding to its transactivation domain in nondividing
cells, attenuating its ability to contribute to the oncogenic
phenotype (21). As Rb is a major degradation target of HPVE7, we determined whether knockdown of E7 affected its
presence, along with Elf-1, at the hPygo2 promoter using
chromatin immunoprecipitation (ChIP) assays. We previously examined a 1,568 bp (1,494 to 74) segment of the
hPygo2 promoter and showed that Elf-1 binds to the 48 to
25 region which contains an Ets-binding site (7). After

www.aacrjournals.org

siNTC treatment of HEN 16T cells, the presence of Elf-1

and Rb was detected at the hPygo2 promoter and at the
promoters of a known Elf-1 target gene, hepatocellular
carcinoma suppressor 1 (HCCS1; ref. 9; Fig. 4B) and an
Rb target gene, cyclin A (CCNA; ref. 43; Fig. 4C). Treating
cells with siE7 did not affect Elf-1 binding to the hPygo2
or HCCS1 promoters. There was, however, a signicant
increase in the presence of Rb at both the hPygo2 (56%, P
0.038) and the CCNA promoters. Thus, E7 reduced the
association of Rb at the hPygo2 promoter while not affecting
Elf-1. This nding suggested that activation of hPygo2 by
HPV in cervical cancer cells was due to derepression of Elf-1
activity by the repressive action of E7 on Rb.
To further analyze the derepression of hPygo2 by HPV
E7, we used a gain-of-function mutant of Rb (PSM-Rb;
ref. 35), a dominant acting Rb-independent mutant of Elf-1
(DA-Elf-1; ref. 21) and wild-type Elf-1 and assessed their
effect on hPygo2 expression. The PSM-Rb construct has 8
phosphorylation sites substituted by alanines (Supplementary Fig. S5A), making it refractory to cyclin/CDK complexes and thus rendering it constitutively repressive (35).
Protein products synthesized using this construct are still
able to bind Elf-1 (Supplementary Fig. S5B). The DA-Elf-1
construct has the 3 crucial amino acids in its LXCXE, which
are required for binding to Rb, substituted with RXRXH
(Supplementary Fig. S5C), rendering it unable to interact
with Rb (Supplementary Fig. S5B). We hypothesized that
overexpression of PSM-Rb would reduce the positive effect

Mol Cancer Res; 11(1) January 2013

Downloaded from mcr.aacrjournals.org on August 11, 2014. 2013 American Association for Cancer Research.

23

Published OnlineFirst January 2, 2013; DOI: 10.1158/1541-7786.MCR-12-0510

Tzenov et al.

Figure 3. Depletion of hPygo2 and

Elf-1 causes activation of p53 and
G1 phase cell-cycle arrest. A and B,
HEN 16T and HEC 18T cells were
treated with either siNTC or
hPygo2 siRNAs. Fluorescentactivated cell sorting (FACS) was
conducted to show the percentage
of cells in each phase of the cell
cycle. C and D, HEN 16T and HEC
18T cells were treated with either
siNTC or hPygo2 siRNA in
combination with either pCS2 or
pCS2 hPygo2 expression
vectors. E and F, HEN 16T and HEC
18T cells were treated with either
siNTC or Elf-1 siRNA (siElf-1) and
subsequently analyzed by FACS. G
and H, HEN 16T and HEC 18T cells
were treated with either siNTC or
siElf-1 in combination with either
pCS2 or pCS2 hPygo2. Fulllength blots are presented in
Supplementary Fig. S6.

of E7 on hPygo2 expression, and in contrast, the DA-Elf-1

mutant would constitutively activate hPygo2 expression.
Our previous work indicated that Elf-1 augmented hPygo2
gene expression by binding to the Ets-binding site (EBS)
present within the rst 48 bp of the hPygo2 promoter (7). To
assess the effect of the Rb and Elf-1 constructs on hPygo2
gene regulation, we overexpressed different combinations of
PSM-Rb, Elf-1, and DA-Elf-1 in conjunction with 2 reporter constructs, both of which included the EBS but differed in
the length of hPygo2 promoter DNA. These included the
minimal hPygo2 promoter containing the rst 48 bp (pGL3

24

Mol Cancer Res; 11(1) January 2013

48), a large segment containing 1,494 bp (pGL3-1494), and

a negative control with no promoter sequence (pGL3 basic;
ref. 7).
Overexpression of PSM-Rb in HEN 16T cells led to 2.9fold (P 0.037) and 7.1-fold (P 0.038) decreases in the
pGL3-48 and pGL3-1494 samples, respectively, compared
with the empty vector control (Fig. 5A). In contrast, transfection with either Elf-1 or DA-Elf-1 signicantly increased
hPygo2 reporter activity compared with the control. Elf-1
overexpression caused a 17.6-fold increase (P 0.017) in the
minimal promoter reporter and a 3.7-fold increase (P

Molecular Cancer Research

Downloaded from mcr.aacrjournals.org on August 11, 2014. 2013 American Association for Cancer Research.

Published OnlineFirst January 2, 2013; DOI: 10.1158/1541-7786.MCR-12-0510

HPV E7 Induces hPygo2

Figure 4. HPV16 E7 knockdown reduces hPygo2 protein levels by increasing the association of Rb with the hPygo2 promoter. A, HEN 16T cells were treated
with either an siNTC or an E7 siRNA (siE7). Immunoblots were conducted to conrm the knockdown of E7 protein and levels of Rb, Elf-1, and hPygo2 were
determined. b-Actin was used as a loading control. B and C, sheared, cross-linked chromatin extracts from HEN 16T cells treated with either siNTC or siE7
were subjected to ChIP assays using antibodies against Rb (B) and Elf-1 (C). A rabbit and mouse IgG mixture was used as a negative control. Promoter regions
of hPygo2, HCCS1, and CCNA were amplied by quantitative PCR. Full-length blots are presented in Supplementary Fig. S6.

0.047) in the large reporter. DA-Elf-1 overexpression

increased the minimal reporter activity by 24.1-fold (P <
0.01) and the large reporter activity by 5.3-fold (P 0.011).
When PSM-Rb and Elf-1 were overexpressed in combination, neither the minimal promoter reporter plasmid nor the
large promoter reporter plasmid activity was signicantly
higher than that of the empty vector control. However, when
PSM-Rb was overexpressed with DA-Elf-1, there was a
signicant increase in both the minimal reporter activity
(7.4-fold, P 0.036) and the large reporter activity (2.3fold, P 0.012) compared with the control.
Overexpression of the Rb and Elf-1 constructs in HEC
18T cells yielded similar results; Rb reduced hPygo2 promoter activity of both reporters whereas Elf-1 increased it
and DA-Elf-1 was able to overcome the repressive effects of
PSM-Rb (Fig. 5B). The similar effects of Rb and Elf-1
overexpression on both the pGL3-48 and pGL3-1494
reporters suggested that Rb and Elf-1 modulated hPygo2
reporter activity through the Ets-binding site in its promoter.
Thus, Rb negatively, whereas Elf-1 positively modulated
hPygo2 reporter activity through its EBS.
To examine whether Rb and Elf-1 could modulate endogenous hPygo2 mRNA and protein levels, we transiently
transfected cells with combinations of PSM-Rb, Elf-1, and
DA-Elf-1 expression vectors. Overexpression of PSM-Rb in
HEN 16T cells resulted in a 5-fold (P < 0.01) reduction in
hPygo2 mRNA levels and a 53% decrease in hPygo2 protein
levels (Fig. 5C). Both Elf-1 and DA-Elf-1 overexpression
increased hPygo2 mRNA and protein levels by 6-fold. When
PSM-Rb and Elf-1 were cotransfected, PSM-Rb inhibited
Elf-1induced hPygo2 expression resulting in no change in
hPygo2 mRNA and protein levels. However, when PSM-Rb
and DA-Elf-1 were cotransfected, DA-Elf-1 overcame the
repressive effects of PSM-Rb resulting in 4.7- and 4.2-fold
increases in hPygo2 mRNA and protein levels, respectively.
Parallel results showing Rb inhibition and Elf-1 induction of
hPygo2 were obtained in HEC 18T cells (Fig. 5D).
The foregoing ndings suggested that Rb and Elf-1
cooperatively modulated hPygo2 expression. ChIP assays
were conducted to determine the presence of Rb and Elf-1 at

www.aacrjournals.org

the hPygo2 promoter. Various combinations of PSM-Rb,

Elf-1, and DA-Elf-1 were overexpressed, and the presence of
Rb and Elf-1 on chromatin located at the hPygo2 promoter
was measured. Interestingly, while overexpression of PSMRb signicantly reduced Elf-1dependent hPygo2 promoter
activity, it did not alter the level of Elf-1 at the hPygo2
promoter compared with the empty vector control (Fig. 5E).
In the remaining cases, in which either Elf-1 or DA-Elf-1 was
overexpressed, there was a consistent increase (3.3- to 4.3fold) in Elf-1 at the promoter. This increase in Elf-1 was not
signicantly affected by the presence of PSM-Rb. Similarly,
Elf-1 at the HCCS1 gene promoter only increased when it
was overexpressed, which was expectedly independent of
PSM-Rb overexpression. These results showed that the
interaction between Elf-1 and Rb is not required for Elf-1
to bind to the hPygo2 promoter.
Given that Rb did not inuence the association of Elf-1 at
the hPygo2 promoter, we hypothesized that the reciprocal
may be true, that its own presence at the hPygo2 promoter is
dependent on Elf-1. While overexpression of PSM-Rb
resulted in a 6.5-fold increase (P 0.017) in Rb (both
endogenous and exogenous) at the promoter relative to the
control (Fig. 5F), the highest levels were achieved when
PSM-Rb was overexpressed in combination with Elf-1 (11fold above control, P 0.024). As expected, when PSM-Rb
was overexpressed with (Rb-independent) DA-Elf-1, its
presence at the hPygo2 promoter was signicantly reduced
compared to when it was co-expressed with wild-type Elf-1
(32% reduction, P 0.015). On the other hand, the
presence of Rb at the CCNA promoter increased binding
whenever PSM-Rb was overexpressed and was not affected
by DA-Elf-1. These results strongly suggested that Rb
associated specically with chromatin at the hPygo2 promoter through its interaction with Elf-1 (Fig. 5G).
We next assessed whether Elf-1 was required for the presence of Rb at the hPygo2 promoter. Treatment of HEN 16T
cells with siElf-1 effectively reduced Elf-1 and hPygo2 protein
levels but without affecting Rb protein levels compared with
the nontarget control siRNA (Fig. 6A). Treatment with the
siNTC control did not alter the presence of Elf-1 and Rb at

Mol Cancer Res; 11(1) January 2013

Downloaded from mcr.aacrjournals.org on August 11, 2014. 2013 American Association for Cancer Research.

25

Published OnlineFirst January 2, 2013; DOI: 10.1158/1541-7786.MCR-12-0510

Tzenov et al.

26

Mol Cancer Res; 11(1) January 2013

Molecular Cancer Research

Downloaded from mcr.aacrjournals.org on August 11, 2014. 2013 American Association for Cancer Research.

Published OnlineFirst January 2, 2013; DOI: 10.1158/1541-7786.MCR-12-0510

HPV E7 Induces hPygo2

Figure 6. Association of Rb with the hPygo2 promoter depends on Elf-1

and an intact ETS-binding site. A, HEN 16T cells were treated with either
an siNTC or an Elf-1 siRNA (siElf-1). Immunoblots were conducted to
conrm the knockdown of Elf-1 protein and levels of Rb and hPygo2 were
determined. b-Actin was used as a loading control. B and C, HEN 16T
cells were processed for ChIP assays and immunoprecipitated with antiElf-1 and anti-Rb antibodies. A rabbit and mouse IgG mixture was used as
a nonspecic control. Promoter regions of hPygo2, HCCS1 and CCNA
were amplied by quantitative PCR. D, HEN 16T cells were transiently
transfected with pGL3-1494, pGL3-48, and pGL3-48 mutEBS luciferase
constructs and a b-galactosidase expression vector. pGL3-basic was
used as an empty luciferase plasmid control. Cells were assayed for
luciferase activity, normalized to b-galactosidase activity, or processed
for ChIP assays (E) and immunoprecipitated with anti-Elf-1 and anti-Rb
antibodies. A rabbit and mouse IgG mixture was used as a nonspecic
control. The 48 to 1 promoter region was amplied by quantitative
PCR. Full-length blots are presented in Supplementary Fig. S6.

the hPygo2 promoter and at the promoters of their target

genes, HCCS1 and CCNA, respectively (Fig. 6B and C).
Treating cells with siElf-1 signicantly reduced Elf-1 binding
to both hPygo2 (2.5-fold, P 0.024) and HCCS1 (2.4-fold,
P 0.031) promoters. The presence of Rb at the hPygo2
promoter was also signicantly reduced (2.1-fold, P 0.033),
but Rb binding to CCNA was not affected. These results
are consistent with aforementioned ndings, which showed
specic Elf-1dependent association of Rb with hPygo2.
Given that the presence of the Rb-Elf-1 complex at the
hPygo2 promoter depended on Elf-1, we hypothesized that
Elf-1 may bind to 1 of the 2 putative EBS, in the promoter.
We used site-directed mutagenesis to change the sequence of
both putative EBSs in the pGL3-48 luciferase reporter
plasmid thereby generating pGL3-48 mutEBS.
First, HEN 16T cells were transfected with pGL3-basic,
pGL3-1494, pGL3-48, or pGL3-48 mutEBS and assayed
for luciferase activity and Rb and Elf-1 promoter occupancy
in the luciferase plasmids. Luciferase reporter activity levels
in the pGL3-1494 and pGL3-48 samples were signicantly
higher (4.5  106 RLU, P < 0.01 and 2  106 RLU, P <
0.01) than in the empty vector control, pGL3-basic, (2.2 
103 RLU; Fig. 6D). The reporter activity level of the pGL348 mutEBS sample was slightly and nonsignicantly higher
(979 RLU, P 0.061) than pGL3-basic.
Second, ChIP assays were used to measure the association
of Rb and Elf-1 to the hPygo2 promoter in each of the
reporter plasmids (Fig. 6E). Elf-1 promoter occupancy was
74.6-fold higher in pGL3-1494 and 70.9-fold higher in
pGL3-48 relative to the empty plasmid control. Rb binding
to the hPygo2 promoters in pGL3-1494 (12.4-fold) and
pGL3-48 (13.4-fold) was signicantly higher than pGL3basic. Mutating both EBS completely abolished the association of Rb and Elf-1 to that of the level of the empty vector
control.
Discussion
The increased expression of hPygo2 in severely dysplastic
lesions and cervical tumors and in cervical cancer cell lines,
along with its requirement for proliferation of these cells,
suggested that hPygo2 may play an important role in cervical
cancer pathogenesis. In our microarray analyses, we used 2
different arrays from 2 separate commercial sources, for
which little pre-analytic information was derived, causing
variability within and between arrays, in immunohistochemical reactivity. While our initial analysis of staining intensity during disease progression suggested that hPygo2 is

Figure 5. hPygo2 protein expression is inhibited by Rb and induced by Elf-1. A and B, HEN 16T and HEC 18T cells were transfected with combinations of PSMRb, Elf-1, and DA-Elf-1 expression vectors in combination with pGL3-48 and pGL3-1494 luciferase constructs and a b-galactosidase expression vector.
Empty pCS2 and pECE expression vectors and the promoter DNA absent pGL3-Basic luciferase construct were used as negative controls. Luciferase
activity was normalized to b-galactosidase activity. Overexpression of Rb and Elf-1 was conrmed by immunoblots. C and D, HEN 16T and HEC 18T cells were
transfected with combinations of PSM-Rb, Elf-1, and DA-Elf-1 expression vectors and their effect on endogenous hPygo2 mRNA and protein levels were
assessed by quantitative PCR and immunoblot, respectively. Overexpression of Rb and Elf-1 was conrmed by immunoblot. E and F, HEN 16T cells were
transfected with combinations of PSM-Rb, Elf-1, and DA-Elf-1 expression vectors and sheared, cross-linked chromatin extracts were subjected to ChIP
assays using antibodies against Rb (E) and Elf-1 (F). A rabbit and mouse IgG mixture was used as a negative control. The promoter regions hPygo2, HCCS1,
and CCNA were amplied by quantitative PCR. G, model explaining the association of Elf-1, DA-Elf-1, and PSM-Rb with the hPygo2 promoter affects hPygo2
gene activation. Full-length blots are presented in Supplementary Fig. S6.

www.aacrjournals.org

Mol Cancer Res; 11(1) January 2013

Downloaded from mcr.aacrjournals.org on August 11, 2014. 2013 American Association for Cancer Research.

27

Published OnlineFirst January 2, 2013; DOI: 10.1158/1541-7786.MCR-12-0510

Tzenov et al.

expressed at higher levels in severe disease, we are presently

optimizing our staining protocol using controlled xation
techniques and array preparation, toward clinical use.
Activation of the Wnt signaling pathway and components
of the pathway have been associated with cervical carcinoma
(44), so it is not unreasonable to suggest that the activity of
hPygo2 in cervical cancer and in cervical dysplasia which we
report in this study, may be linked to, or be a prerequisite for
Wnt signal transduction, given its role as a chromatin
modier in this context (45). Our recent ndings (Andrews
and colleagues, submitted), however, have shown an additional but strong requirement for malignant growthassociated ribosome biogenesis in HeLa cells, a cervical adenocarcinomaderived cell line. While these concurrent results
do not necessarily preclude a role in Wnt signaling in cervical
cancer, the association of hPygo2 with cell-cycle progression
genes we observed in the present study would also support its
role in ribosome biogenesis.
Our ndings are consistent with others showing the
overexpression and requirement of hPygo2 in a broad range
of cancers (2, 3), but they are novel as they showed that the
increase in hPygo2 expression begins at a premalignant stage.
Because not all high-grade (CIN3) lesions progress to cancer,
specic and effective biomarkers are needed to predict when
and whether this progression will occur. The nding that
hPygo2 expression increases specically in the CIN3 precancerous stage indicates that hPygo2 may effective biomarker to predict the transition from high-grade dysplasia to
cancer.
High-grade lesions have an elevated probability for progressing to cancer because HPV genome integration results
in the loss of the E2 regulatory region and subsequent E7
oncoprotein upregulation. Because the increase in hPygo2
expression correlates with the HPV integration event, we
hypothesized that the 2 were causally linked. By knocking
down E7 expression, we observed a decrease in hPygo2
protein levels, which correlated with an increase in Rb
binding to the hPygo2 promoter while not affecting Elf-1
levels or promoter binding. hPygo2 levels were likely
increased in cervical carcinoma, therefore, because E7 upregulation caused the degradation of Rb. Furthermore, the
reduction of p53 by re-expression of hPygo2 in cells depleted
of Elf-1 showed a relationship between these 2 genes in
deregulated growth during oncogenesis. This molecular
interaction may be part of an important signaling axis for
the identication and possible targeting of early stage disease.
Elf-1 protein is continuously expressed and localizes to the
cytoplasm, where it becomes activated through phosphorylation and glycosylation. These modications cause its
translocation to the nucleus, where it binds to target gene

promoters and activates gene expression. Our immunohistochemical analyses of dysplasia core samples revealed nuclear Elf-1 in CIN3, suggesting that Elf-1 is active, consistent
with previous ndings (13). In the nucleus, Rb interacts with
the LXCXE motif of Elf-1, inhibiting its transcriptional
activity. We propose that following HPV integration, E7
displaces Rb from Elf-1 and results in Elf-1mediated
hPygo2 expression.
Initially, Elf-1 was regarded as a lymphoid gene specic
transcription factor. However, it is also regarded as an
important activator or repressor of genes involved in several
processes such as development (41, 46, 47), oncogenesis
(48), and viral gene activation (49). The bulk of research
examining the role of Elf-1 in cancer has focused on the
expression of Elf-1 in relation to other biomarkers, such as
VEGF and PCNA or with clinical correlates. Detailed
studies of Elf-1 gene targets in cancer include Elf-1 activation
of TIE2 (46) and binding to the HCCS1 promoter (13). An
efcient way to potentially identify all Elf-1 gene targets in
these cells would be to conduct a genome-wide chromatin
association analysis and to functionally group these genes to
identify in which processes Elf-1 is involved.
Disclosure of Potential Conicts of Interest
K.R. Kao and C. Popadiuk have Ownership Interest (including patents). No
conicts of interest disclosed. No potential conicts of interest were disclosed by the
other authors.

Authors' Contributions
Conception and design: Y.R. Tzenov, P.G. Andrews, K.R. Kao, C. Popadiuk
Development of methodology: P.G. Andrews, K. Voisey, C. Popadiuk
Acquisition of data (provided animals, acquired and managed patients, provided
facilities, etc.): Y.R. Tzenov, P.G. Andrews, K. Voisey, J. Xiong, K.R. Kao, C.
Popadiuk
Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): Y.R. Tzenov, P. Popadiuk
Writing, review, and/or revision of the manuscript: Y.R. Tzenov, K.R. Kao, C.
Popadiuk
Administrative, technical, or material support (i.e., reporting or organizing data,
constructing databases): Y.R. Tzenov, K.R. Kao, C. Popadiuk
Study supervision: K.R. Kao, C. Popadiuk

Acknowledgments
The authors thank Brenda Gallie for pECE PSM-Rb and Mark Kennedy and
Zhijian He for thoughtful discussion.

Grant Support
Grant support was provided by Canadian Institutes of Health Research, Medical
Research Foundation, Memorial University MOP-14928.
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 with
18 U.S.C. Section 1734 solely to indicate this fact.
Received August 29, 2012; revised October 22, 2012; accepted October 22, 2012;
published OnlineFirst January 2, 2013.

References
1.
2.

28

Jessen S, Gu B, Dai X. Pygopus and the Wnt signaling pathway: a

diverse set of connections. Bioessays 2008;30:44856.
Popadiuk CM, Xiong J, Wells MG, Andrews PG, Dankwa K, Hirasawa
K, et al. Antisense suppression of pygopus2 results in growth arrest of
epithelial ovarian cancer. Clin Cancer Res 2006;12:221623.

Mol Cancer Res; 11(1) January 2013

3.
4.

Andrews PG, Lake BB, Popadiuk C, Kao KR. Requirement of Pygopus

2 in breast cancer. Int J Oncol 2007;30:35763.
Chen J, Luo Q, Yuan Y, Huang X, Cai W, Li C, et al. Pygo2 associates
with MLL2 histone methyltransferase and GCN5 histone acetyltransferase complexes to augment Wnt target gene expression and

Molecular Cancer Research

Downloaded from mcr.aacrjournals.org on August 11, 2014. 2013 American Association for Cancer Research.

Published OnlineFirst January 2, 2013; DOI: 10.1158/1541-7786.MCR-12-0510

HPV E7 Induces hPygo2

5.

6.
7.

8.
9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.
23.

24.

25.

breast cancer stem-like cell expansion. Mol Cell Biol 2010;30:

562135.
Wang ZX, Chen YY, Li BA, Tan GW, Liu XY, Shen SH, et al. Decreased
pygopus 2 expression suppresses glioblastoma U251 cell growth.
J Neurooncol 2010;100:3141.
Chen YY, Li BA, Wang HD, Liu XY, Tan GW, Ma YH, et al. The role of
Pygopus 2 in rat glioma cell growth. Med Oncol 2011;28:63140.
Andrews PG, Kennedy MW, Popadiuk CM, Kao KR. Oncogenic activation of the human Pygopus2 promoter by E74-like factor-1. Mol
Cancer Res 2008;6:25966.
Seth A, Watson DK. ETS transcription factors and their emerging roles
in human cancer. Eur J Cancer 2005;41:246278.
Zhu JD, Fei Q, Wang P, Lan F, Mao dQ, Zhang HY, et al. Transcription of
the putative tumor suppressor gene HCCS1 requires binding of ETS-2
to its consensus near the transcription start site. Cell Res 2006;16:
78096.
Huang X, Brown C, Ni W, Maynard E, Rigby AC, Oettgen P. Critical role
for the Ets transcription factor ELF-1 in the development of tumor
angiogenesis. Blood 2006;107:315360.
Hegen A, Koidl S, Weindel K, Marme D, Augustin HG, Fiedler U.
Expression of angiopoietin-2 in endothelial cells is controlled by
positive and negative regulatory promoter elements. Arterioscler
Thromb Vasc Biol 2004;24:18039.
Scott GK, Chang CH, Erny KM, Xu F, Fredericks WJ, Rauscher FJ III,
et al. Ets regulation of the erbB2 promoter. Oncogene 2000;19:6490
502.
Nicol AF, Pires AR, de Souza SR, Nuovo GJ, Grinsztejn B, Tristao A,
et al. Cell-cycle and suppressor proteins expression in uterine cervix in
HIV/HPV co-infection: comparative study by tissue micro-array (TMA).
BMC Cancer 2008;8:289.
Gavrilov D, Kenzior O, Evans M, Calaluce R, Folk WR. Expression of
urokinase plasminogen activator and receptor in conjunction with the
ets family and AP-1 complex transcription factors in high grade
prostate cancers. Eur J Cancer 2001;37:103340.
Bassuk AG, Barton KP, Anandappa RT, Lu MM, Leiden JM. Expression
pattern of the Ets-related transcription factor Elf-1. Mol Med
1998;4:392401.
Sahin A, Vercamer C, Kaminski A, Fuchs T, Florin A, Hahne JC, et al.
Dominant-negative inhibition of Ets 1 suppresses tumor growth, invasion and migration in rat C6 glioma cells and reveals differentially
expressed Ets 1 target genes. Int J Oncol 2009;34:37789.
Yang DX, Li NE, Ma Y, Han YC, Shi Y. Expression of Elf-1 and survivin in
non-small cell lung cancer and their relationship to intratumoral microvessel density. Chin J Cancer 2010;29:396402.
Yang DX, Han YC, Liu LY, Yu N, Wang X, Shi Y. [Expression and
signicance of Elf-1 and vascular endothelial growth factor in nonsmall cell lung cancer]. Ai Zheng 2009;28:7627.
Takai N, Ueda T, Nishida M, Nasu K, Miyakawa I. The relationship
between oncogene expression and clinical outcome in endometrial
carcinoma. Curr Cancer Drug Targets 2004;4:51120.
Takai N, Miyazaki T, Nishida M, Nasu K, Miyakawa I. The signicance of
Elf-1 expression in epithelial ovarian carcinoma. Int J Mol Med
2003;12:34954.
Wang CY, Petryniak B, Thompson CB, Kaelin WG, Leiden JM. Regulation of the Ets-related transcription factor Elf-1 by binding to the
retinoblastoma protein. Science 1993;260:13305.
Moody CA, Laimins LA. Human papillomavirus oncoproteins: pathways to transformation. Nat Rev Cancer 2010;10:55060.
Pahel G, Aulabaugh A, Short SA, Barnes JA, Painter GR, Ray P, et al.
Structural and functional characterization of the HPV16 E7 protein
expressed in bacteria. J Biol Chem 1993;268:2601825.
Hawley-Nelson P, Vousden KH, Hubbert NL, Lowy DR, Schiller JT.
HPV16 E6 and E7 proteins cooperate to immortalize human foreskin
keratinocytes. EMBO J 1989;8:390510.
Chellappan S, Kraus VB, Kroger B, Munger K, Howley PM,
Phelps WC, et al. Adenovirus E1A, simian virus 40 tumor antigen,
and human papillomavirus E7 protein share the capacity to
disrupt the interaction between transcription factor E2F and the
retinoblastoma gene product. Proc Natl Acad Sci U S A 1992;
89:454953.

www.aacrjournals.org

26. Schwarz E, Freese UK, Gissmann L, Mayer W, Roggenbuck B,

Stremlau A, et al. Structure and transcription of human papillomavirus sequences in cervical carcinoma cells. Nature 1985;314:
1114.
27. Huibregtse JM, Scheffner M, Howley PM. A cellular protein mediates
association of p53 with the E6 oncoprotein of human papillomavirus
types 16 or 18. EMBO J 1991;10:412935.
28. Scheffner M, Werness BA, Huibregtse JM, Levine AJ, Howley PM. The
E6 oncoprotein encoded by human papillomavirus types 16 and 18
promotes the degradation of p53. Cell 1990;63:112936.
29. Munger K, Werness BA, Dyson N, Phelps WC, Harlow E, Howley PM.
Complex formation of human papillomavirus E7 proteins with the
retinoblastoma tumor suppressor gene product. EMBO J 1989;8:
4099105.
30. Tsutsumi K, Belaguli N, Qi S, Michalak TI, Gulliver WP, Pater A, et al.
Human papillomavirus 16 DNA immortalizes two types of normal
human epithelial cells of the uterine cervix. Am J Pathol 1992;140:
25561.
31. Yokoyama M, Tsutsumi K, Pater A, Pater MM. Human papillomavirus
18-immortalized endocervical cells with in vitro cytokeratin expression
characteristics of adenocarcinoma. Obstet Gynecol 1994;83:197
204.
32. Yang X, Nakao Y, Pater MM, Pater A. Identication of two novel cellular
genes associated with multistage carcinogenesis of human endocervical cells by mRNA differential display. Carcinogenesis 1996;17:
5637.
33. Nakao Y, Yang X, Yokoyama M, Pater MM, Pater A. Malignant transformation of human ectocervical cells immortalized by HPV 18: in vitro
model of carcinogenesis by cigarette smoke. Carcinogenesis
1996;17:57783.
34. Chang JT, Kuo TF, Chen YJ, Chiu CC, Lu YC, Li HF, et al. Highly potent
and specic siRNAs against E6 or E7 genes of H. Cancer Gene Ther
2010;17:82736.
35. Hamel PA, Gill RM, Phillips RA, Gallie BL. Regions controlling hyperphosphorylation and conformation of the retinoblastoma gene product
are independent of domains required for transcriptional repression.
Oncogene 1992;7:693701.
36. Balan R, Giusca S, Caruntu ID, Gheorghita V, Neacsu D, Amalinei C.
Immunochemical assessment of p16 and HPV L1 capsid protein in
cervical squamous intraepithelial lesions. Rev Med Chir Soc Med Nat
Iasi 2010;114:111824.
37. Yu CC, Filipe MI. Update on proliferation-associated antibodies applicable to formalin-xed parafn-embedded tissue and their clinical
applications. Histochem J 1993;25:84353.
38. Gupta N, Srinivasan R, Rajwanshi A. Functional biomarkers in cervical
precancer: an overview. Diagn Cytopathol 2010;38:61823.
39. Thompson B, Townsley F, Rosin-Arbesfeld R, Musisi H, Bienz M. A
new nuclear component of the Wnt signalling pathway. Nat Cell Biol
2002;4:36773.
40. Waldman T, Kinzler KW, Vogelstein B. p21 is necessary for the p53mediated G1 arrest in human cancer cells. Cancer Res 1995;55:
518790.
41. Jin E, Liu J, Suehiro J, Yuan L, Okada Y, Nikolova-Krstevski V, et al.
Differential roles for ETS, CREB, and EGR binding sites in mediating VEGF receptor 1 expression in vivo. Blood 2009;114:
555766.
42. Scott GK, Daniel JC, Xiong X, Maki RA, Kabat D, Benz CC. Binding of
an ETS-related protein within the DNase I hypersensitive site of the
HER2/neu promoter in human breast cancer cells. J Biol Chem
1994;269:1984858.
43. Strobeck MW, Knudsen KE, Fribourg AF, DeCristofaro MF, Weissman
BE, Imbalzano AN, et al. BRG-1 is required for RB-mediated cell cycle
arrest. Proc Natl Acad Sci U S A 2000;97:774853.
44. Uren A, Fallen S, Yuan H, Usubutun A, Kucukali T, Schlegel R, et al.
Activation of the canonical Wnt pathway during genital keratinocyte
transformation: a model for cervical cancer progression. Cancer Res
2005;65:6199206.
45. Mosimann C, Hausmann G, Basler K. Beta-catenin hits chromatin:
regulation of Wnt target gene activation. Nat Rev Mol Cell Biol
2009;10:27686.

Mol Cancer Res; 11(1) January 2013

Downloaded from mcr.aacrjournals.org on August 11, 2014. 2013 American Association for Cancer Research.

29

Published OnlineFirst January 2, 2013; DOI: 10.1158/1541-7786.MCR-12-0510

Tzenov et al.

46. Iljin K, Dube A, Kontusaari S, Korhonen J, Lahtinen I, Oettgen P,

et al. Role of ets factors in the activity and endothelial cell
specicity of the mouse Tie gene promoter. FASEB J 1999;13:
37786.
47. Choi HJ, Geng Y, Cho H, Li S, Giri PK, Felio K, et al. Differential
requirements for the Ets transcription factor Elf-1 in the development of
NKT cells and NK cells. Blood 2011;117:18807.

30

Mol Cancer Res; 11(1) January 2013

48. Seth A, Watson DK, Blair DG, Papas TS. c-ets-2 protooncogene has
mitogenic and oncogenic activity. Proc Natl Acad Sci U S A 1989;
86:78337.
49. Leiden JM, Wang CY, Petryniak B, Markovitz DM, Nabel GJ, Thompson CB. A novel Ets-related transcription factor, Elf-1, binds to human
immunodeciency virus type 2 regulatory elements that are required
for inducible trans activation in T cells. J Virol 1992;66:58907.

Molecular Cancer Research

Downloaded from mcr.aacrjournals.org on August 11, 2014. 2013 American Association for Cancer Research.