Professional Documents
Culture Documents
MCR-12-0510
Updated version
Supplementary
Material
Cited Articles
E-mail alerts
Reprints and
Subscriptions
Permissions
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
Downloaded from mcr.aacrjournals.org on August 11, 2014. 2013 American Association for Cancer Research.
Molecular
Cancer
Research
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.
www.aacrjournals.org
Downloaded from mcr.aacrjournals.org on August 11, 2014. 2013 American Association for Cancer Research.
19
Tzenov et al.
20
Downloaded from mcr.aacrjournals.org on August 11, 2014. 2013 American Association for Cancer Research.
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
Downloaded from mcr.aacrjournals.org on August 11, 2014. 2013 American Association for Cancer Research.
21
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
22
Downloaded from mcr.aacrjournals.org on August 11, 2014. 2013 American Association for Cancer Research.
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.
www.aacrjournals.org
Downloaded from mcr.aacrjournals.org on August 11, 2014. 2013 American Association for Cancer Research.
23
Tzenov et al.
24
Downloaded from mcr.aacrjournals.org on August 11, 2014. 2013 American Association for Cancer Research.
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.
www.aacrjournals.org
Downloaded from mcr.aacrjournals.org on August 11, 2014. 2013 American Association for Cancer Research.
25
Tzenov et al.
26
Downloaded from mcr.aacrjournals.org on August 11, 2014. 2013 American Association for Cancer Research.
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
Downloaded from mcr.aacrjournals.org on August 11, 2014. 2013 American Association for Cancer Research.
27
Tzenov et al.
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
3.
4.
Downloaded from mcr.aacrjournals.org on August 11, 2014. 2013 American Association for Cancer Research.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
www.aacrjournals.org
Downloaded from mcr.aacrjournals.org on August 11, 2014. 2013 American Association for Cancer Research.
29
Tzenov et al.
30
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.
Downloaded from mcr.aacrjournals.org on August 11, 2014. 2013 American Association for Cancer Research.