Content uploaded by Tony Kwok Hung Chung
Author content
All content in this area was uploaded by Tony Kwok Hung Chung on Mar 21, 2014
Content may be subject to copyright.
ORIGINAL ARTICLE
A novel miR-193a-5p-YY1-APC regulatory axis in human
endometrioid endometrial adenocarcinoma
Y Yang
1,2
, L Zhou
1,2
,LLu
2,3
, L Wang
1,2
,XLi
2,4
, P Jiang
2,3
, LKY Chan
1,2
, T Zhang
1,2
,JYu
2,4
, J Kwong
1,2
, TH Cheung
1
, T Chung
1
, K Mak
5
,
H Sun
2,3
and H Wang
1,2
Aberrant expression and altered function of transcription factors (TFs) have vital roles in many aspects of tumor development and
progression. In this study, we investigated the functional significance of a TF, Yin Yang1 (YY1) in tumorigenesis of endometrioid
endometrial carcinoma (EEC). We demonstrated that YY1 is upregulated in EEC cell lines and primary tumors; and its expression
is associated with tumor stages. Depletion of YY1 inhibits EEC cell proliferation and migration both in vitro and in vivo, whereas
overexpression of YY1 promotes EEC cell growth. These results suggest that YY1 functions as an oncogenic factor in EEC.
Transcriptome analysis revealed a significant effect of YY1 on critical aspects of EEC tumorigenesis through inhibition of APC
expression. Further mechanistic investigation uncovered a new epigenetic silencing mode of APC by YY1 through recruitment
of EZH2 and trimethylation of histone 3 lysine 27 on its promoter region. Moreover, YY1 overexpression was found to be a
consequence of miR-193a-5p downregulation through direct miR-193a-5p-YY1 interplay. Our results therefore establish a novel
miR-193a-5p-YY1-APC axis, which contributes to EEC development, and may serve as future intervention target.
Oncogene (2013) 32, 3432–3442; doi:10.1038/onc.2012.360; published online 20 August 2012
Keywords: YY1; EEC; APC; EZH2; H3K27me3; microRNA
INTRODUCTION
Endometrial cancer is the commonest gynecologic malignancy
and ranks fourth in terms of incident cancers in women. About
80–90% of these cancers are endometrioid endometrial carcinoma
(EEC), originating from the single layer of epithelial cells that line
the endometrium and form the endometrial glands. When
diagnosed at an early stage it is often curative, but patients with
advanced stage or recurrent EEC do not respond as well to
therapy. This is to some extent a reflection of an incomplete
understanding of the molecular basis of endometrial carcinogen-
esis.
1
Aberrant expression and function of transcription factors
(TFs) contribute, either directly or indirectly through other
downstream pathways, to some or all the cancer hallmarks,
including insensitivity to antigrowth or apoptotic signals, self-
sufficient growth signals, sustained angiogenesis, limitless
replicative potential and invasive or metastatic capability.
2
Yin Yang 1 (YY1) is a multifunctional TF, which regulates various
processes of development and differentiation.
3
Most processes
mediated by YY1 are cancer-related, which strongly implicates its
importance in cancer development and progression. Indeed,
overexpression of YY1 has been observed in various types of cancers
(reviewed in Sui
4
and Zaravinos and Spandidos
5
). For example, our
group showed that YY1 is upregulated in Rhabdomyosarcoma. Most of
the reports from others and us supported that the overall effect of YY1
is proliferative or oncogenic in cancer development but its oncogenic
mechanisms are not fully explored.
In addition to its DNA-binding activity, YY1 can interact with
numerous proteins, some of which are involved in chromatin
remodeling. As a member of the Polycomb Group (PcG) proteins,
YY1 interacts with Enhancer of Eeste Homolog 2 (EZH2), which
mediates tri-methylation of histone H3-K27, a hallmark of gene
silencing. Previously, we demonstrated that YY1 recruits EZH2-
containing Polycomb repressive complex to promoters/enhancers
of several muscle-specific genes and miRNAs to silence their
expression in the skeletal muscle cells.
6–10
Both YY1 and EZH2 are
overexpressed in multiple cancers but it is unknown whether their
interplay contributes to the aberrant epigenetic status of these
cancers.
Gene activation/inactivation associated with epigenetic
mechanisms represents important events in tumorigenesis.
Hypermethylation of CpG islands of gene promoter is one of the
earliest and most frequent alterations leading to cancer. CpG
island hypermehtylation on several genes, including APC, CDH13,
hMLH1 and p16, has been demonstrated in EEC.[11] As a well-
known tumor suppressor, loss of APC (Adenomatous Polyposis
Coli) function through promoter hypermethylaton was reported
in multiple cancers (reviewed in Aguilera et al.
12
As a critical
component of Wnt/b-catenin signaling pathway, inactivation
of APC allows b-catenin to enter nucleus, where it binds to
member of the TCF/LEF transcriptional effectors to regulate genes
involved in cell proliferation and cell motility. Post-translational
modifications of histones are emerging as the second important
epigenetic mechanism that influences tumorigenesis.
13,14
Specific
histone modification such as HK27me3 has been associated with
epigenetic silencing of tumor suppressor genes in various
cancers.
14
Interestingly, interplays between the two epigenetic
1
Department of Obstetrics and Gynaecology, The Chinese University of Hong Kong, Hong Kong SAR, China;
2
Li Ka Shing Institute of Health Sciences, Prince of Wales Hospital,
The Chinese University of Hong Kong, Hong Kong SAR, China;
3
Department of Chemical Pathology, The Chinese University of Hong Kong, Hong Kong SAR, China;
4
Institute of
Digestive Disease and Department of Medicine and Therapeutics, The Chinese University of Hong Kong, Hong Kong SAR, China and
5
School of Biomedical Sciences, The Chinese
University of Hong Kong, Hong Kong SAR, China. Correspondence: Dr H Wang, 507A Li Ka Shing Institute of Health Sciences, Prince of Wales Hospital, The Chinese University
of Hong Kong, Hong Kong SAR, China.
E-mail: huating.wang@cuhk.edu.hk
Received 19 March 2012; revised 22 June 2012; accepted 30 June 2012; published online 20 August 2012
Oncogene (2013) 32, 3432– 3442
&
2013 Macmillan Publishers Limited All rights reserved 0950-9232/13
www.nature.com/onc
silencing mechanisms were discovered.
13,15
Yet, it is not known
whether histone status contributes to APC silencing in any tumors.
In addition to the transcriptional events regulated by TFs, the
post-transcriptional events in cancer development were recently
investigated owing to the fast-growing field of microRNAs
(miRNAs). miRNAs are 19–25-nt noncoding RNA (ncRNA) tran-
scripts, which modulate mRNA expression post-transcriptionally
by binding to the 30UTRs of target genes.
16
Recent exploration of
miRNA expression profiles in endometrial cancers led to the
identification of a number of dysregulated miRNAs and their
correlation with clinicopathological characteristics.
17–20
However,
their functions and downstream events remain unexplored.
Herein, we show that YY1 has an oncogenic role in EEC
development. Small interefering RNA (siRNA) knockdown of YY1
inhibited tumor growth both in vitro and in vivo. Transcriptome
analysis revealed a significant effect of YY1 on critical aspects of
EEC tumorigenesis. Further mechanistic investigation uncovered
a novel epigenetic silencing mode of APC by YY1 through
H3K27me3. Additionally, YY1 was found under direct regulation by
miR-193a-5p. Our results therefore establish a novel miR-193a-5p-
YY1-APC axis, which contributes to EEC development and
progression.
RESULTS
YY1 is overexpressed in EEC cells and tumors
To probe into the possible involvement of YY1 in EEC develop-
ment, we began by examining the expression of YY1 in five EEC
cell lines, HEC-1-A, HEC-1-B, AN3CA, RL95-2 and KLE using
microdissected normal (N) endometrial tissues as negative control
(NC). Results showed that YY1 mRNA expression was highly
upregulated in all the five EEC cell lines compared with the normal
tissues (Figure 1a and Supplementary Figure S1A). Consistently,
YY1 protein was found to be upregulated in EEC cell lines
(Figure 1b and Supplementary Figure S1B). In addition, immuno-
fluroscence (IF) staining revealed its predominate localization in
nuclei (Figure 1c), in keeping with its role as a TF. To extend this
study from cell lines to patient tumors, YY1 expression was
examined on paraffin-embeded sections of 122 EEC and 35
normal specimens by Immunohistochemistry (IHC) staining.
Results revealed a positive signal at various levels (weak, moderate
or strong) in 490% tumors examined, whereas the majority
(470%) of normal specimens showed negative YY1 signal
(Figure 1d). Consistent with IF staining in cells, YY1 was observed
mostly in the nuclei of glandular epithelium (Figure 1d, inset).
Furthermore, scoring of YY1 staining revealed it is correlated
positively with an ascending FIGO (Federation of International
Gynecologists and Obstetricians) stage: an evident increasing
trend was observed across NE and early stages of EEC (stage I and
II) (Po0.05), but not extended to late stage (III) (Figure 1e).
However, no significant association was found between YY1
staining score and other clinicopathologic features, including
histologic grade, myometrium invasion, lymph node invasion,
peritoneal cytology or recurrence rate within 5 years
(Supplementary Table S1). Collectively, the above results suggest
that YY1 is upregulated in EEC cell lines and primary tumors.
YY1 promotes EEC cell proliferation and migration
The upregulation of YY1 implicated that it could have an
oncogenic role in EEC tumorigenesis. To test this notion, depletion
of YY1 was achieved by transfecting siRNA oligos against YY1
(siYY1) into AN3CA cell, a cell line derived from a lymph node
metastasis of EEC.
21
A near complete loss of YY1 proteins was
detected in siYY1-transfected cells with respect to NC oligo
transfection (siNC) (Figure 2a left) or scrambled (Scr) oligos
(Supplementary Figure S2B). Cell proliferation was then assessed
by MTS assay, which revealed a 50% decrease in siYY1 vs siNC or
Scr cells (Figure 2a, right and Supplementary Figure S2C). This is
strengthened by IF staining of Ki67, which revealed a 40%
decrease of positively stained cells (Figure 2b), as well as colony
formation assay (Figure 2c): The colonies formed in siYY1 cells
were significantly fewer in number and smaller in size than those
in siNC cells. On the contrary, ecotopic expression of YY1 in AN3CA
cells promoted cell proliferation during a 4-day growth course
(Figure 2d), and at a dose-dependent manner (Supplementary
Figure S2A). The above results were largely recapitulated with
similar results in a second cell line, KLE derived from poorly
differentiated primary endometrial carcinoma
22
(Figures 2e–h).
To evaluate YY1 effect on migration ability of EEC cells, we
generated an AN3CA-stable cell line expressing a small hairpin
targeting YY1 (shYY1) or pSuper Vector as NC. Wound-healing
assay revealed (Figure 2i) that YY1 knockdown significantly
suppressed AN3CA cell motility at both 24 (Po0.001) and 48 h
(Po0.0001), resulting in much slower wound closure of the
monolayer. Examination of YY1 effect on EEC cell apoptosis, on
the other hand, did not reveal any significant difference upon YY1
knockdown (Supplementary Figure S2D). Together, the above data
demonstrated that YY1 promotes tumor growth in EEC cells and
warranted further elucidation of its underlying mechanism.
RNA-sequencing reveals YY1-mediated transcriptome change in
EEC cells
To gain insights into the YY1-mediated molecular events in EEC
cells, we conducted a genome-wide analysis to globally char-
acterize YY1-affected transcriptome changes. Poly-A þmRNAs
were extracted from AN3CA cells transfected with siYY1 or siNC
oligos and subjected to mRNA sequencing (RNA-seq). The majority
of reads can be mapped to coding regions (CDS and UTRs, 470%)
and much fewer in introns, intergenic and non-coding regions
(Figure 3a), indicating high specificity for expressed mRNA and
rejection of genomic DNA and unspliced pre-mRNA. A total of
1298 and 1229 genes were found to be up- and down-regulated
in siYY1 cells with respect to siNC cells (Figure 3b and
Supplementary Dataset S1 and Supplementary Dataset S2). For
validation, 13 genes (5 down- and 8 upregulated genes) were
randomly chosen to be examined by qRT–PCR. The results agreed
favorably with the RNA-seq data (Supplementary Figure S3A).
Subsequent Gene Ontology analysis with upregulated list of
genes revealed that the top-ranked lists of enriched Gene
Ontology categories include ‘homophilic cell adhesion’, ‘cell–cell
adhesion’, ‘cell adhesion’, ‘transcription’, ‘negative regulation of
epithelial cell proliferation’ (Figure 3c and Supplementary Dataset
S3). Strikingly, KEGG pathway analysis showed that several
important molecular pathways in ‘endometrial cancer’ were
significantly affected (Supplementary Figure S3B). Expression
levels of key molecules including K-Ras, BRAF, APC, TCF7, TCF7L1
and PI3K from multiple important pathways were changed by YY1
knockdown, strongly arguing that YY1 has a critical role in EEC
development.
YY1 epigenetically silences APC gene through recruiting EZH2 and
inducing H3K27me3 on its promoter region
Among all the potential targets of YY1 regulation, APC caught our
attention owing to its well-known role as a tumor suppressor in
multiple cancers, including EEC. Consistent with RNA-seq data,
APC upregulation upon YY1 knockdown was confirmed by both
semi-quantitative and real-time RT–PCR in AN3CA and KLE cells
(Figures 3d–e). APC has an essential role in inhibiting canonical
Wnt pathway through regulating a transcriptional cofactor,
b-catenin. Expectedly, YY1 depletion downregulated the expres-
sion of several known b-catenin transcriptional targets, Tcf1, Lef1
and c-Myc, in both AN3CA and KLE cells (Figures 3f–g). The effect
of YY1 on b-catenin transcription was further examined by using
Topflash luciferase reporter, which contains three b-catenin
miR-193a-5p-YY1-APC regulatory axis in EEC
Y Yang et al
3433
&2013 Macmillan Publishers Limited Oncogene (2013) 3432– 3442
responsive elements. Stable knockdown of YY1 (shYY1) decreased
the reporter activity; ectopic expression of a b-catenin (b-Cat)
plasmid (wild type) in shYY1 stable cells rescued the reporter
activity, whereas a mutant (Mut) plasmid failed to do so.
Furthermore, knockdown of APC by siRNA or ectopic expression
of YY1 also rescued the reporter activity (Figure 3h), suggesting
that YY1 acts upstream of APC regulating the transactivating
activity of b-catenin. Lastly, knockdown of both YY1 and APC
rescued the inhibitory effect of siYY1 on EEC cell proliferation
(Figure 3i), demonstrating that YY1 functions on EEC cell
proliferation through regulating APC.
Considering the known epigenetic function of YY1 in chromatin
remodeling, we speculated that YY1 might silence APC promoter
through recruitment of Polycomb and subsequent histone
modification. Consistently, three YY1-binding sites (Y1, Y2 & Y3)
were found within 5 kb upstream region of APC transcriptional
start site (Figure 4a). To test whether these sites are competent
for YY1-binding, chromatins from pSuper or shYY1 cells were
subjected to chromatin-immunoprecipitation (ChIP) coupled with
PCR assay using an YY1 antibody. Immunoblotting detection
showed a specific pull down of YY1 protein after ChIP by YY1
but not IgG (Supplementary Figure S4A), demonstrating the
good quality of our antibody. ChIP-PCR results revealed that
all three sites were competent for YY1 binding in pSuper
control cells while the association disappeared (Y1 and Y2)
or decreased (Y3) in shYY1 cells (Figure 4b). Site Y3 displayed
the highest binding affinity for YY1 (B7-fold enrichment).
To strengthen the above findings, we performed in vivo
ChIPs and the binding of YY1 on the above three sites of APC
promoter was also detected in primary tumors but not in normal
tissues (Supplementary Figure S4B). To confirm the above ChIP
results, reporters were generated by fusing regions encompassing
GAPDH
YY1
YY1
GAPDH
Negative
(Normal) Weak
Stron
g
Moderate
***
*
***
***
NE
35
I
45
II
39
III
38
0
2
4
6
8
10
12
KLE
AN3CA
Phase contrast DAPI YY1 Merge
Stage
Cases (No)
N1
N2
N3
HEC-1-A
HEC-1-B
AN3CA
RL95-2
KLE
N1
N2
N3
HEC-1-A
HEC-1-B
AN3CA
RL95-2
KLE
95%CI of staining score
Figure 1. YY1 is overexpressed in EEC cell lines and primary tumors. (a) Expression of YY1 in five EEC cell lines using three microdisected
normal (N) endometrial tissues as controls. (b) The expression of YY1 proteins in the above samples using GAPDH as a loading control. (c)IF
staining for YY1 in AN3CA or KLE cells. Scale bar ¼20 mm. (d) IHC detection of YY1 on paraffin sections of EEC tumors and normal endometrial
specimen. Representative images with various levels of staining (negative from normal tissue, weak, moderate or strong from tumor tissues)
were shown at both 200 and 400 (small inset) magnification. (e) The association of YY1 IHC-staining scores with stages of tumor (I, II or III).
The number of cases are shown below. Data are plotted as mean of 95% confidence interval±s.d. *Po0.05, ***Po0.001. A full colour version of
this figure is available at the Oncogene journal online.
miR-193a-5p-YY1-APC regulatory axis in EEC
Y Yang et al
3434
Oncogene (2013) 3432– 3442 &2013 Macmillan Publishers Limited
each of the three sites to a luciferase gene. As expected,
knock-down of YY1 increased the reporter activities whereas
overexpression of YY1 decreased their activities (Figures 4c and d).
Knowing that YY1 could silence gene expression through
recruiting EZH2, we next examined EZH2 binding on the above
sites. Indeed, EZH2 recruitment was found on all three sites in
pSuper cells (Figure 4e). Subsequent H3K27me3 ChIP indicated
that EZH2 association on these sites led to H3K27 tri-methylation
YY1
Tubulin
Tubulin
YY1
YY1
Tubulin
Tubulin
YY1
Vector
YY1
0
1
2
3
4
Day
Viable cell no.
4321
Ki67
DAPI
siNC siYY1
Ki67
DAPI
siNC siYY1
siNC siYY1
0.0
0.2
0.4
0.6
0.8
%Ki67 positive cells
*
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Fold change
*
0.0
0.2
0.4
0.6
0.8
1.0
*
%Ki67 positive cells
0.0
0.2
0.4
0.6
0.8
*
Absorption
0.0
0.1
0.2
0.3
0.4
0.5
0.6
**
Absorption
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
Fold change
***
siNC siYY1
0h 24h 48h
pSuper
shYY1
Absorption
***
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0
20
40
60
80
100
Wound closure (%)
pSuper
shYY1
48h24h
****
***
siNC
siNC
siNC
siYY1
siYY1
siNC
siYY1
siNC
siYY1
siYY1
siNC
siYY1
siNC
siYY1
siNC
siYY1
Vector
YY1
Vector
Vector
YY1
YY1
Figure 2. YY1 promotes EEC cell proliferation. (a) Left: depletion of YY1 by siRNA in AN3CA cells. Right: measurement of cell proliferation in
siNC- or siYY1-transfected cells by MTS assay. The data are plotted as mean±s.d. from three independent experiments. (b) Left: IF staining of
Ki67 was performed on siNC- or siYY1-transfected cells at 48 h post transfection. Right: Ki67 positively stained cells were shown as percentage
of DAPI positive cells. (c) Left: monolayer colony formation assay was performed in siNC- or siYY1-transfected AN3CA cells. Right: the number
of colonies was counted from three independent experiments. Data are plotted as mean±s.d. (d) Left: ectopic expression of YY1 in AN3CA
cells by transfecting an YY1 expressing or a Vector control plasmid. Right: the number of viable cells was counted at day 1, 2, 3 and 4 post-
transfection. (e–g) Knockdown of YY1 by siRNA oligos decreased KLE cell proliferation as revealed by MTS assay, Ki67 staining and colony
formation assay. (h) Ectopic expression of YY1 increased KLE cell proliferation as measured by MTS assay. (i) Left: AN3CA cells stably expressing
shYY1 or a pSuper vector control plasmid were seeded on six-well plate with 100% confluent monolayer and a ‘wound’ was induced. Phase-
contrast pictures of the wound were taken at 0, 24 and 48 h. Right: the percentage of wound closure was quantified at each indicated time
point. Data are plotted as mean±s.d. from three independent experiments. *Po0.05, **Po0.01, ***Po0.001, ****Po0.0001. A full colour
version of this figure is available at the Oncogene journal online.
miR-193a-5p-YY1-APC regulatory axis in EEC
Y Yang et al
3435
&2013 Macmillan Publishers Limited Oncogene (2013) 3432– 3442
Gene Ontology analysis of up-regulated genes by YY1 knock-down (partial list)
GAPDH
APC
2.60E-02Colorectal cancer hsa05210KEGG_PATHWAY
2.22E-02Endometrial cancer hsa05213 KEGG_PATHWAY
2.05E-02Regulation of actin cytoskeleton hsa04810KEGG_PATHWAY
6.09E-03mTORsignaling pathwayhsa04150KEGG_PATHWAY
7.14E-04Cell adhesionSP_PIR_KEYWORDS
1.36E-04CalciumSP_PIR_KEYWORDS
5.71E-09Alternative splicing SP_PIR_KEYWORDS
9.59E-03Plasma membrane partGO:0044459GOTERM_CC_FAT
6.62E-03Intrinsic to membraneGO:0031224GOTERM_CC_FAT
6.82E-08
Metal ion binding
GO:0046872GOTERM_MF_FAT
2.31E-08Ion bindingGO:0043167 GOTERM_MF_FAT
1.93E-08
Action binding
GO:0043169 GOTERM_MF_FAT
8.59E-03Regulation of transcriptionGO:0045449GOTERM_BP_FAT
8.24E-03Negative regulation of epithelial cell proliferationGO:0050680GOTERM_BP_FAT
3.51E-04TranscriptionGO:0006350GOTERM_BP_FAT
1.46E-04Biological adhesion GO:0022610 GOTERM_BP_FAT
1.41E-04Cell adhesionGO:0007155GOTERM_BP_FAT
1.64E-05Cell-cell adhesionGO:0016337GOTERM_BP_FAT
2.65E-09Homophiliccell adhesionGO:0007156GOTERM_BP_FAT
P-ValueTermGO IDCategory
GAPDH
APC
0
1
2
3
4
Expression fold
siYY1 - +
siYY1
-10
-5
0
10
5
-10 -5 1005
siYY1 (log2)
siNC(log2)
siYY1 - +
0
1
2
3
4
5
6
7
Expression fold
0.0
1.0
2.0
3.0
4.0
5.0
6.0
RLU
shYY1 + + + + +-
-+----
- -+- --
---+--
----+-
β-Cat (WT)
β-Cat (Mut)
siAPC
YY1
**
p=0.109
AN3CA
KLE
1.35%
17.23%
8.95%
11.91%
54.42%
6.14%
55.79%
6.75%
8.61%
8.33%
1.45%
19.07%
CDS
5’UTR
Non-coding
Intron
Intergenic
3’UTR
0.0
0.2
0.4
0.6
0.8
1.0
1.2 siNC
siYY1
Expression fold
0.0
0.2
0.4
0.6
0.8
1.0
1.2 siNC
siYY1
Expression fold
0.0
0.2
0.4
0.6
0.8
1.0
1.2
Absorption
siAPC
siYY1
siNC +
+
+
+
--
-
-
-
0.0
0.2
0.4
0.6
0.8
1.0
Absorption
siAPC
siYY1
siNC +
+
+
+
--
-
-
-
Tcf7
Lef1
c-Myc
Tcf7
Lef1
c-Myc
siNC
Figure 3. Genome-wide analysis of YY1 affected transcriptome changes by RNA-seq. (a) Total RNAs were isolated from AN3CA cells transfected
with siNC (left) or siYY1 (right) oligos, and subjected to high throughput mRNA sequencing (mRNA-seq). The normalized fragment density was
calculated by counting the fragments per kilobase of genomic regions of interests (CDS for coding sequence, 50UTR, 30UTR, non-coding,
intron, intergenic) per million mapped reads. (b) Differentially expressed genes between siNC- and siYY1-transfected AN3CA cells were
determined by RNA-seq. (c) Over-represented Gene Ontology terms by Gene Ontology analysis of down-regulated list of genes. BP: biological
process. MF: molecular function; CC: cellular component; SP_PIR: a database of protein super-family names. KEGG: Kyoto Encyclopedia of
Genes and Genomes. Endometrial cancer pathway is highlighted. (d–e) APC RNA expression was measured in siY Y1 ( þ) or negative control
(NC) oligo (-) transfected AN3CA or KLE cells using both qRT–PCR (top) and semi-quantitative RT–PCR (bottom). (f–g) qRT–PCR measurement
of the expression levels of Tcf1, Lef1 or c-Myc, in siNC- or siYY1-transfected AN3CA or KLE cells. (h–i) AN3CA or KLE cells were transfected with
the indicated siRNA oligos. MTS assays were performed 48 h post transfection to measure cell proliferation. (j) AN3CA cells were transfected
with the indicated plasmids along with Topflash and Lacz reporter plasmid. Luciferase activities were determined 48 h post transfection.
The data represent the average of three independent experiments±s.d. **Po0.01. A full colour version of this figure is available at the
Oncogene journal online.
miR-193a-5p-YY1-APC regulatory axis in EEC
Y Yang et al
3436
Oncogene (2013) 3432– 3442 &2013 Macmillan Publishers Limited
(Figure 4f). Expectedly, knockdown of YY1 (shYY1) decreased both
EZH2 and H3K27me3 enrichment on these sites (Figures 4e and f).
These data suggest that YY1 transcriptionally regulates APC
promoter through recruiting EZH2-containing PRC to multiple
sites to cause H3K27me3. Consistently, knock-down of EZH2
upregulated the APC expression to a comparable degree as YY1
depletion (Figure 4g). Treatment of cells with 3-deazaneplanocin
A, a chemical that depletes EZH2 and H2K27me3 induced APC
expression in a time-dependent manner in multiple EEC cell lines
(Figure 4h and Supplementary Figure S5A). Lastly, expression
levels of several PRC components, including EZH2, EED and RYBP
proteins, were upregulated in all five EEC cell lines compared with
normal endometrial tissues (Figure 4i), supporting possible
involvement of PRC to the aberrant epigenetic status of APC
gene. And EZH2 was also found to localize in the nuclei of EEC
cells where YY1 was found (Supplementary Figure S5B).
Altogether, our findings identify PRC-mediated histone modifica-
tion as a novel epigenetic silencing mechanism on APC promoter.
siEzh2 -+
0.0
1.0
2.0
3.0
4.0
5.0
Expression fold
Enrichment fold
0.0
Y1
1.0
2.0
3.0
4.0
5.0
6.0
7.0
Y2
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
Y3
Y1 Y2 Y3
2
4
6
8
10 IgG
YY1
EZH2
EED
RYBP
GAPDH
pSuper
shYY1
123
Meth
y
lated CpG
Unmethylated CpG
0.0
0.2
0.4
0.6
0.8
1.0
1.2
E1 E2 E3
Vector
YY1
*
RLU
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
E1 E2 E3
pSuper
shYY1
**
RLU
CpG island
-1084
Y3Y2Y1
-2483-4546
Exon 1 APC
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
Enrichment fold
0.5
1.0
1.5
2.0
2.5
3.0
3.5
0.5
1.0
1.5
2.0
2.5
3.0
3.5 IgG
EZH2
siEzh2 -+
0.0
0.5
1.0
1.5
2.0
2.5
Expression fold
Dznep(hr) 0 72
Expression fold
144
0
10
20
30
40
50
60
70
80
90
Dznep(hr) 072
Expression fold
144
0
2
4
6
8
10
12
GTA
-237bp 84bp
pSuper
shYY1
pSuper
shYY1
pSuper
shYY1
pSuper
shYY1
pSuper
shYY1
pSuper
shYY1
IgG
H3K27me3
0.0
1.0
2.0
3.0
4.0
Enrichment fold
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
5
10
15
20
25
Y1 Y2 Y3
pSuper
shYY1
pSuper
shYY1
pSuper
shYY1
TSS
N1
N2
N3
HEC-1-A
HEC-1-B
AN3CA
RL95-2
KLE
Figure 4. YY1 regulates APC through histone modification. (a) Schematic illustration of proximal promoter region of APC-1 A gene. Predicted
YY1-binding sites, Y1, Y2 and Y3, were displayed as diamonds. The genomic location of each site was indicated below. CpG island was shown.
(b) Chromatins were collected from pSuper or shYY1 cells for ChIP assays. Primers amplifying regions encompassing YY1-binding sites, Y1, Y2
or Y3, were used for qRT–PCR analysis. Data are plotted as mean±s.d. (c) Luciferase reporters with enhancers encompassing site Y1, Y2 or Y3,
were transfected into shYY1 or pSuper expressing AN3CA-stable cells together with a LacZ expressing plasmid. Luciferase activities were
measured. (d) The above reporter plasmids were transfected into AN3CA cells together with YY1 expressing or Vector plasmid. Luciferase
activities were determined. (e,f) Chromatin-immunoprecipiation coupled with PCR assays were performed as above to examine the
enrichment of EZH2 and H3K27me3. (g) AN3CA cells were transfected with siRNA oligo against EZH2 ( þ) or negative control (NC) oligos (-)
and examined for the expression of APC 48 h post transfection. (h) Expression of APC in DZnep-treated AN3CA cells. (i) Expression levels of
EZH2, EED and RYBP proteins were examined in five EEC cell lines and three normal (N) endometrial tissues. (j) The cloned bisulfite sequencing
was performed on APC genomic region, -237 to þ84 bp relative to the transcription start site, which includes 23 CpG sites. (k,l) The
experiments were performed in KLE cells as in (g) and (h).
miR-193a-5p-YY1-APC regulatory axis in EEC
Y Yang et al
3437
&2013 Macmillan Publishers Limited Oncogene (2013) 3432– 3442
As APC promoter CpG hypermethylation was previously found
in EEC, we attempted to test if YY1 could have a role in regulating
the hypermethylation. Interestingly, YY1 knockdown did not
significantly change the level of CpG methylation on APC
promoter as measured by bisulfate genomic sequencing
(Figure 4j), suggesting that loss of H3K27me3 alone leads to
APC activation. To strengthen the above argument, we examined
the effect of loss of EZH2/H3K27me3 on APC expression in KLE
cells in which APC promoter was not CpG methylated.
23
Expectedly, knockdown of EZH2 or treatment with 3-
deazaneplanocin A both significantly upregulated APC gene
expression (Figures 4k and l), implying that H3K27me3 status
alone can control APC transcription.
An YY1-APC regulatory axis in vivo
Lastly, we evaluated the oncogenic effect of YY1 in vivo by
establishing AN3CA xenograft tumors in immunocompromised
mice. Control (siNC) oligos or 0.5 nMsiYY1 were directly injected
into tumors. The injections were repeated every other day for 10
days as outlined in Figure 5a. Difference in tumor volume was
observed from the beginning and became more apparent after 5
days and persisted until the experimental endpoint, when siNC-
injected tumors were on average 1.86 times larger than those
injected with siYY1 (n5; P¼0.00156) (Figures 5b and c). Tumors
were resected for further evaluation. RT–PCR and IHC staining
revealed that siYY1 oligos injection effectively decreased YY1
expression at both mRNA and protein levels (Figure 5d, qRT–PCR
0
5
10
15
20
25
30
35
40
0123456789Day
Relative tumor volume
siNC
siYY1
p=0.00156
siNC
siYY1
GAPDH
siNC siYY1
YY1
123 231
siNC siYY1
YY1
Ki67
P-H3
0.0
0.4
0.8
1.2
Fold change
****
0.0
0.4
0.8
1.2
Fold change
****
986420-8
siRNA siRNA siRNA siRNA siRNAImplantation Harvest
-3
Visible tumor
Day
0.0
0.5
1.0
1.5
2.0
2.5
3.0 siNC
siYY1
123
Expression fold
YY1 APC
Case 1
Case 2
c-MycAPC
siNC
siYY1
0.0
0.5
1.0
1.5
231
Expression fold
0
25
50
75
100
Percentage (%)
APC-High (n=15)
APC-Low (n=45)
High
(n=44)
Low
(n=16)
YY1 score
****
siYY1
siNC
siYY1
siNC
Figure 5. YY1-APC regulation in vivo. (a) Schematic illustration of the siYY1 treatment scheme. (b) Tumor volume was recorded daily for 10
days (from day 0–9). Relative tumor volumes are shown with respect to day 0, where the volumes were set to 1. Data are plotted as mean±s.d.
(c) The mice were sacrificed at the end of the treatment and images taken along with the resected tumors from five representative mice. (d)
The expression of YY1 mRNA in the above tumors was measured by semiquantitative RT–PCR. Data from three representative mice were
shown. (e–g) IHC staining of YY1, Ki67 or phospho-Histone3 (P-H3) on the tumor sections. The quantification was done by counting positively
stained cells from 10 randomly chosen fields from a total of six sections per tumor. Data are plotted as mean±s.d. (h,i) The expression of APC
or c-Myc mRNAs in the above tumors. (j) IHC staining of YY1 and APC on sequential sections of EEC patient tumors. The representative images
from two cases were shown at magnification of 400. ****Po0.0001. A full colour version of this figure is available at the Oncogene journal online.
miR-193a-5p-YY1-APC regulatory axis in EEC
Y Yang et al
3438
Oncogene (2013) 3432– 3442 &2013 Macmillan Publishers Limited
and 5E IHC). This was associated with reduced cellular proliferation
as Ki67 and phosphor-histone H3 (p-H3) staining were three- and
fourfold lower, respectively, compared with siNC group (Po0.0001;
Figures 5f and g). Significantly, intratumoral treatment of siYY1
oligos induced APC expression (Figure 5h) and inhibited the
expression of downstream target gene, c-Myc. (Figure 5i). Collec-
tively, these data support that YY1 promotes EEC cell proliferation
through APC. When further examined in 60 patient tumors, APC
expression by IHC staining was found downregulated in EEC and
inversely correlated with YY1 staining score (Po0.0001; Figure 5j).
YY1 is a direct target of miR-193a-5p in EEC cells
After gaining insights into the downstream events of YY1, we
turned our attention to its upstream by asking: what leads to the
YY1 overexpression in EEC cells? Multiple mechanisms have been
demonstrated in controlling YY1 expression, for example, post-
transcriptional control by miRNAs in skeletal muscle cells.
6,7
To
test whether any miRNAs regulate YY1, we examined all the
dysregulated miRNAs in EEC. Downregulation of miR-193a-5p was
found in both tumors and cells by several groups, as well as our
own data (Figures 6a and b). Strikingly, YY1 is a predicted target of
this miRNA with a binding site found in its 30UTR region
(Figure 6c). We thus speculated that YY1 overexpression is a result
of miR-193a-5p downregulation. As expected, restoration of miR-
193a-5p levels in both AN3CA and KLE cells downregulated YY1 at
both RNA and protein levels (Figures 6d and e and Supplementary
Figures S6A and B). This notion was further examined by using
reporters with a fragment of the YY1 30UTR containing the miR-
193a-5p binding site fused downstream of the luciferase gene
(wild type). miR-193a-5p was found to cause significant repression
of luciferase activities (Po0.01; Figure 6f). This regulation
appeared specific to miR-193a-5p binding as the luciferase activity
was restored when miR-193a-5p site was mutated from the YY1
30UTR (Mut). In addition, injection of miR-193a-5p oligos into EEC
Xenagraft tumor in vivo significantly inhibited tumor growth to a
similar extent as siYY1 injection (Supplementary Figures S6C and D
and Figure 6h) and YY1 was decreased in miR-193a-5p injected
tumors compared with tumors injected with NC miRNA oligos
(Figure 6i). These results confirmed the existence of miR-193a-5p
regulation on YY1 in vivo. Lastly, Topflash reporter activity was
found to be significantly decreased upon miR-193a-5p restoration
and rescued upon YY1 overexpression in both AN3CA and KLE
cells (Po0.001 and Po0.001, respectively; Figure 6g), suggesting
that miR-193a-5p-YY1-APC constitutes a linear axis in EEC cells.
DISCUSSION
In this study, we identified YY1 as a promoting factor in EEC
development. Although YY1 dysregulation has been reported in a
variety of cancers,
24
this is the first study to elucidate its
involvement in EEC. The association of YY1 with the tumor stage
demonstrates its predictive value as a molecular marker. As a
potential modulator of EEC carcinogenesis, YY1 may also serve as
an innovative target for EEC therapy. This was nicely illustrated by
our in vivo study using xenograft model (Figure 5). siYY1 oligo
injection successfully delayed tumor growth, pointing to its use as
a potential new therapeutic approach. The underlying mechan-
isms for YY1 oncogenic action are thought to be related to its
regulation of cell cycle and cell death. Key molecular components
controlling these two processes, including cyclin D, p53, c-Myc
and Rb, exhibit the interaction with YY1 either as direct
transcriptional target or interacting partner.
3
Consistent with the
above findings, a marked effect on cell proliferation and cell
migration was found both in vitro and in vivo in EEC. In addition,
our genome-wide survey by RNA-seq uncovered a novel aspect of
YY1 action, which is to regulate cell–cell adhesion. Misregulation
of cell adhesion components can lead to tumorigenesis.
25
YY1 regulation on cell–cell adhesion thus reinforces its critical
function in tumor development.
Inactivation of APC gene through epigenetic mechanism was
demonstrated in several human malignancies. In all the cases,
promoter hypermethylation was thought to be the cause.
However, when examined in 103 EECs, only 46.6% cases showed
APC promoter hypermethylation.
26
No APC mutations were
detected either, suggesting additional inactivating mechanisms.
Our findings, for the first time, implicate histone modification as
an alternative epigenetic silencing mechanism. We demonstrated
that YY1-mediated EZH2 recruitment and subsequent H3K27me3
in the proximal promoter region of APC gene contributes to the
inactivation of APC promoter 1A. This is in line with recent findings
demonstrating H3K27me3 as an important epigenetic regulator
in tumorigenesis.
14
Interestingly, YY1 depletion in AN3CA cells
did not reverse APC promoter hypermethylation despite the
activation of APC gene (Figure 4j), suggesting that CpG
hypermehtylation does not necessarily cause transcriptional
inactivation; loss of H3K27me3 may be enough to promote
chromatin opening, allowing transcription to occur.
In addition to causing H3K27 methylation, there is a possibility
that YY1 could also be involved in CpG hypermehtylation through
recruiting DNMT3A and/or DNMT3B as we found a YY1-binding
site nearby the CpG island of APC promoter. Consistent with this
idea, YY1 was previously shown to recruit DNMTs to CCAAT/
enhancer-binding protein promoter in cervical cancer cells.
27
In this regard, YY1 may function as an important intermediate in
PcG–DNMTs complex-mediated epigenetic regulation of many
tumor suppressor promoters. As APC silencing is a prevalent
phenomenon, our findings suggest that it may be necessary to
examine H3K27me3 status of its promoter region, which may
represent a general silencing mode in many cancers. Furthermore,
EZH2 has been known to be a bonafide oncogene that is essential
for cancer progression and invasion.
28
Both YY1 and EZH2 are
overexpressed in various cancers. Thus, the recruitment of EZH2
by YY1 may contribute to the aberrant epigenetic status of
cancers and constitutes a wide spread mechanism that promotes
tumor progression in many cancers.
It is also interesting to point out that epigenetic silencing of APC
was shown to be an early event in EEC tumor development,
26
concomitant with the appearance of YY1 overexpression at the
very early stage. These suggest that YY1-caused APC silencing
may occur very early, leading to tumor initiation. The inverse
correlation between YY1 and APC expressions in patient tumors
(Figure 5j) further supports the critical involvement of YY1-APC
axis in EEC progression.
Compared with the well-characterized roles of YY1, its regula-
tion is relatively underexplored. Gene duplication has not been
found on human chromosome 14q32 where YY1 locus is located
in EEC,
29
indicating that YY1 overexpression is unlikely to be a
genetic event. The findings in this study demonstrated that miR-
193a-5p directly targets YY1 thus downregulation of miR-193a-5p
is a possible cause of YY1 dysregulation in EEC. This represents the
first report on TF–miRNA interaction in EEC, supporting the notion
that TF-miRNA interplay is a prevalent mechanism in cancers and
the dysregulation of these interacting axes may contribute to
tumor development.
Collectively, our findings establish a novel miR-193a-5p-YY1-APC
axis in EEC. As modeled in Figure 7, YY1 is upregulated in both EEC
cell lines and primary tumors through downregulated miR-193a-5p
expression. YY1 functions to epigenetically silence APC transcrip-
tion by recruiting EZH2 on its promoter region to induce
H3K27me3. Decrease of APC results in b-catenin nuclear transloca-
tion and activation of downstream target genes together with TCF/
LEF. This will lead to increased cell proliferation and migration,
contributing to EEC development and progression. In the future, it
would be interesting to explore whether YY1 has a role in
recruiting DNMT3A and/or DNMT3B to cause hypermethylation of
miR-193a-5p-YY1-APC regulatory axis in EEC
Y Yang et al
3439
&2013 Macmillan Publishers Limited Oncogene (2013) 3432– 3442
the APC promoter. In addition, there is an increasing interest in
exploring mysregulated miRNAs as therapeutic target. Restoration
of the level of a downregulated miRNA by introducing miRNA
mimics was proven to be effective in treating human cancers.
30,31
Our study clearly demonstrated that intramuscular injection of miR-
193a-5p oligos can inhibit EEC xenograft tumor growth, suggesting
that restoring miR-193a-5p levels by miRNA replacement therapy
may serve as an intervention measure for EEC.
MATERIALS AND METHODS
Tissue samples
EEC Specimens were obtained from the tissue bank of the Department of
Obstetrics and Gynecology, Prince of Wales Hospital. Overall, 122 cases of
primary endometrioid endometrial adenocarcinoma and 38 cases of
normal tissues were recruited in this study. Normal endometrial tissue
specimens were collected from women who underwent a hysterectomy or
endometrial curettage for endometrial-unrelated diseases, such as uterus
myoma and uterus prolapse. Histological typing of each tumor was
according to the World Health Organization (WHO) criteria, whereas clinical
staging followed Federation of International Gynecologists and Obste-
tricians (FIGO) standards. All specimens, and their corresponding clinical
information, were obtained under the protocols approved by the Clinical
Research Ethics Committee of the Chinese University of Hong Kong.
Cell
Human endometrial cancer cell lines, AN3CA, KLE, HEC-1-A, HEC-1-B and
RL95-2, were obtained from American Tissue and Cell Culture and cultured
AN3CA
0
0.4
0.8
1.2
1.6
Expression fold
Cell
0
0.4
0.8
1.2
Expression fold
0
0.4
0.8
1.2
1.6
Expression fold
YY1
Tubulin GAPDH
YY1
0.0
0.5
1.0
1.5
2.0
2.5
Expression fold
CGU–UUCUGGGu3’ aguagagcggg
140:5’ aacugaguuaa
5’ hsa-miR-193a-5p
3’ YY1 3’UTR (WT)
GCAGAAGACCCc
CAGACTCTATT (Mut)
miR-193a-5p
AN3CA
RLU
0.0
0.2
0.4
0.6
0.8
1.0
1.2 ****
NC + -
--+
-++ -++YY1
-
KLE
0.2
0.4
0.6
0.8
1.0
1.2 ***
+-
--+
-
NC
miR-193a-5p
WT Mut
0.2
0.4
0.6
0.8
1.0
1.2
**
0.0
0.2
0.4
0.6
0.8
1.0
1.2
WT Mut
RLU
**
AN3CA KLE
0
2
4
6
8
10
12
14
16
18
20
0213 213
456789Day
Relative tumor volume
NC
miR-193a-5p
p = 0.0209
0.0
0.2
0.4
0.6
0.8
1.0
1.2
YY1 expression fold
NC
miR-193a-5p
KLE
Tumor
N1
N1
N2
N2
HEC-1-A
AN3CA
KLE
T1
T2
T3
T4
T5
miR-193a
-5p
NC
miR-193a
-5p
miR-193a
-5p
NC
NC
miR-193a
-5p
NC
Figure 6. YY1 is targeted by miR-193a-5p in EEC. (a,b) The expression of miR-193a-5p in both EEC primary tumors (T) and cell lines was
examined by qRT–PCR using RNU6 as normalization. Normal (N) endometrial tissues were used as negative controls. Expression folds are
shown with respect to N where miR-193a-5p levels were set to a value of 1. Data are plotted as mean±s.d. (c) Predicted miR-193a-5p binding
site in the 30UTR region of YY1. The wild-type (WT) seed sequence was underlined and mutated to the indicated sequence below (Mut) for the
following study. (d) Left: AN3CA cells were transfected with 50 nMmiR-193a-5p or NC oligos. The expression of YY1 RNA was examined in the
transfected cells by qRT–PCR using GAPDH as normalization. Right: the expression of YY1 protein was examined in the above cells by western
blotting using a-tubulin as a loading control. (e) The above experiments in (d) were repeated in KLE cells with comparable results. (f)AWT
luciferase reporter plasmid was generated by fusing miR-193a-5p binding site of the YY1 30UTR downstream of the luciferase reporter gene.
A mutant plasmid was generated by mutating the binding site. W T or mutant reporter constructs were then transfected into AN3CA cells or KLE
cells with NC or miR-193a-5p oligos and a Lacz reporter plasmid. Luciferase activities were determined 48h post transfection and normalized to
b-galactosidase protein. Relative lucifierase units (RLU) are shown with respect to NC, where luciferease activities were set to 1. Data represent
the average of three independent experiments±s.d. (g) Toplash reporter plasmid was transfected into AN3CA cells or KLE cells with indicated
miRNA oligos or plasmid. Luciferase activities were determined as above. (h) NC or miR-19a-5p oligos were injected into AN3CA xenograft
tumor. Tumor volume was recorded daily for 10 days (from day 0–9). Relative tumor volumes are shown with respect to day 0, where the
volumes were set to 1. Data are plotted as mean±s.d. (i) The mice were sacrificed at the end of the treatment and total RNAs were extracted
from the above tumors. The expression of YY1 mRNA was measured by qRT–PCR using GAPDH as NC. Data from three representative mice are
shown. **Po0.01, ***Po0.001, ****Po0.0001.
miR-193a-5p-YY1-APC regulatory axis in EEC
Y Yang et al
3440
Oncogene (2013) 3432 – 3442 &2013 Macmillan Publishers Limited
according to recommendations. To generate an AN3CA cell line with YY1
stably knocked down, pSuper-shYY1-GFP or vector control plasmid was
transfected into AN3CA cells followed by G418 selection for 10 days. The
resistant colonies were pooled together and amplified for further use.
Details for viable cell counting, MTS assay, colony formation assay and
wound-healing assay were described previously.
32
3-deazaneplanocin A treatment
EEC cells were treated with 5 mM3-deazaneplanocin A (Caymen Chemical,
Ann Arbor, MI, USA) for 72 h. Total RNAs were then extracted for RNA
analysis.
Bisulfite genomic sequencing of individual alleles
Genomic DNA was modified by sodium bisulfate as described previously.
32
Plasmid construction
Topflash reporter plasmid, wild type or mutant b-catenin expression
plasmids were gifts from Professor Kingston Mak (CUHK). YY1 expression
plasmid was a gift from Y Shi (Harvard University). To construct a shYY1
expression plasmid, shYY1 oligos were cloned into pSuper-GFP vector
using Bgl II and Hind III sites. To construct the YY1-30UTR-luc wild-type
reporter plasmid, a 224-bp fragment of the YY1 30UTR was amplified by
PCR from human genomic cDNA and cloned into the KpnI and Nhe I site of
the pGL2-control vector (Promega, Madison, WI, USA). The mutant plasmid
was generated by mutating the seed region from 50-GCAGAAGACCC-30to
50-CAGACTCTATT-30. To construct APC-E1, -E2 or -E3 luciferase reporter
plasmids, the enhancer fragment (423, 608 and 253 bp) encompassing
YY1-binding sites Y1, Y2 or Y3 was amplified from human genomic DNA
and cloned into the KpnI and NheI site of the pGL2 vector (Promega)
according to the manufacturer’s instruction. Primers used are listed in
Supplementary Table S2.
Oligonucleotides
Precursor miR-193a-5p or NC oligos were obtained from Life Technologies
(Carlsbad, CA, USA). siRNA against YY1, EZH2, non-targeting control or
scrambled oligos were obtained from Ribobio (Guangzhou, China). In each
case, the concentration used for transient transfections was 50 nM.
Sequences of siRNA oligos are listed in Supplementary Table S3.
RT–PCR and Real-time RT–PCR
Total RNAs from cells were extracted using TRIzol reagent (Life Technologies).
Expression of mature miRNAs was determined using miRNA-specific Taqman
microRNA assay kit (Life Technologies). Expression of mRNA analysis was
performed with SYBR Green Master Mix (Life Technologies) as described.
7
Primers used are listed in Supplementary Table S2.
Luciferase reporter assay
Transfection of different siRNAs or plasmids was performed using
Lipofectamine 2000 Reagent (Life Technologies). Luciferase activity was
measured 48 h after transfection using Dual-Glo Luciferase Assay kit
(Promega).
Immunoblotting and immunostaining
For Western blotting analysis, total cell extracts were prepared and used
as previously described.
23
The following dilutions were used for each
antibody: YY1 (Santa Cruz Biotechnology, Santa Cruz, CA, USA; 1:2000),
EZH2 (Cell Signaling, Boston, MA, USA; 1:1000), RYBP (Millipore, Billerica,
MA, USA; 1:2000), EED (Abcam, Cambridge, MA,USA; 1:2000), a-Tubulin
(Sigma, St Louis, MA, USA; 1:5000) and GAPDH (Santa Cruz Biotechnology;
1:5000). Immunofluorescence of cultured cells was performed using the
following antibodies: YY1 (Santa Cruz Biotechnology; 1:200), EZH2 (Cell
Signaling; 1:100), Ki67 (Santa Cruz Biotechnology, 1:200). For Ki67 and DAPI
quantification, counting was performed on at least 10 randomly chosen
fields using Image-Pro plus6.0 software.
Immunohistochemistry
IHC on paraffin-embedded sections was performed using the following
antibodies: YY1(Abcam,1:100); APC (Abcam, 1:100). IHC on sections from
xenograft tumors was performed using the following antibodies: YY1
(Abcam,1:100), Ki67 (Santa Cruz Biotechnology, 1:100), P-H3 (Cell Signaling,
1:200); for Ki67 and P-H3 quantification, counting was performed on at
least 20 randomly chosen fields from 4-6 sections. For scoring the IHC
staining, positively stained cells were given an intensity score (0 none;
1 weak; 2 intermediate; 3 strong) and a proportion score representing the
estimated proportion of positively stained tumor cells (0, 0%; 1, 1–25%; 2,
26–50%; 3, 51–75%; and 4, 76–100%). A final score ranging from 0–12 was
generated by multiplying proportion score with intensity score. To
examine the correlation between YY1 and APC protein, sequential sections
from 60 EEC tumors were stained for YY1 and APC, respectively. A score of
6 was used to classify low and high levels.
CCAT
EZH2
YY1
CCATCCAT
EZH2
YY1
EZH2
YY1
H3K27me3 H3K27me3 H3K27me3
YY1
YY1
YY1
YY1
miR-193a-5p
Cell proliferation, cell migration EEC development
TCF/LEF
-Catenin
Target genes:
c-Myc, Tcf7, Lef1 etc.
β-Catenin
P
APC
APC
AXIN2
GSK3
β-Catenin β-Catenin
β-Catenin
Figure 7. A model of miR-193a-5p-YY1-APC regulatory axis in EEC tumorogenesis. The model depicts the roles of the miR-193a-5p-YY1-APC
regulatory axis in EEC development. In EEC tumors, the downregulation of miR-193a-5p (k) leads to the overexpression of YY1 (m), which
subsequently silences APC-1A gene by recruiting EZH2-containing PRC to cause H3K27me3 on the proximal promoter region. Inactivation of
APC (k) results in b-catenin activation (m) and nuclear translocation to induce the expression of downstream target genes. The expression of
these genes leads to increased cell proliferation and migration, which contributes to EEC development. Straight line, promoter/enhancer
region of APC with arrow denotes transcriptional start site; CCAT, YY1 binding elements; DNA methylation. A full colour version of this
figure is available at the Oncogene journal online.
miR-193a-5p-YY1-APC regulatory axis in EEC
Y Yang et al
3441
&2013 Macmillan Publishers Limited Oncogene (2013) 3432 – 3442
ChIP assays
ChIP assays were performed as previously described,
6
using 2 mg
of antibodies against YY1 (Santa Cruz Biotechnology), EZH2 (Cell
Signaling), trimethyl-histone H3-K27 (Millipore), or isotype IgG (Santa
Cruz Biotechnology) as a NC. Primers for PCR are listed in (Supplementary
Table S2).
Xenograft mouse model
Studies using female athymic nude mice (3–4-weeks-old) were reviewed
and approved by CUHK Animal Experimentation Ethics Committee. Overall,
210
6
cells were subcutaneously implanted into the left and right flanks
of the mice. Eight days after implantation, siNC or siYY1 oligo were injected
into the left or right tumor, respectively; and the injection was repeated
every other day for five times. Oligos were prepared by pre-incubating
0.5 nMof siRNA oligos with Lipofectamine 2000 (Life Technologies) for
15 min and injections were made in a final volume of 60 ul in OPTI-EM (Life
Technologies). Tumor sizes were measured daily and the tumor volume
was calculated as hlw, with hindicating height, lindicating length
and windicating width. The mice were euthanized at day 16, and the
tumors were excised and snap-frozen for RNA extraction, or paraffin-
embedded for IHC staining. For miRNA injection, 50 nMof NC or miR-193a-5p
oligos were injected into the left and right tumor, respectively.
Sequencing and base calling
Preparation of transcription libraries for sequencing on Illumina (San
Diego, CA, USA) GA2x platform and read mapping were carried out as
described before.
10
Statistical analysis
The staining scores of YY1 IHC in NE and different stages of EEC clinical
samples were compared by trend test and Student’s t-test. Association of
YY1 staining scores with different clinicopathological parameters was
analyzed by one-way analysis of variance. Correlation between YY1 and
APC staining scores in EEC clinical samples was analyzed by Pearson’s
w
2
-test. Data were analyzed using SPSS 17.0 software package (SPSS,
Chicago, IL, USA). Po0.05 was considered statistically significant. For all the
mRNA expression, MTS and luciferase activity data, as well as IHC staining
quantification, the statistical significance was assessed by Student’s t-test.
*Po0.05; **Po0.01; ***Po0.001; ****Po0.0001.
CONFLICT OF INTEREST
The authors declare no conflict of interest.
ACKNOWLEDGEMENTS
The work described in this paper was substantially supported by three General
Research Funds (GRFs) from Research Grants Council (RGC) of the Hong Kong Special
Administrative Region, China (CUHK476309 and CUHK476310 to HW, and
CUHK473211 to HS), and three CUHK direct grants (2041474 to HS and 2041492
and 2041662 to HW).
REFERENCES
1 Bansal N, Yendluri V, Wenham RM. The molecular biology of endometrial cancers
and the implications for pathogenesis, classification, and targeted therapies.
Cancer Control 2009; 16:8–13.
2 Nebert DW. Transcription factors and cancer: an overview. Toxicology 2002; 181-
182: 131–141.
3 Gordon S, Akopyan G, Garban H, Bonavida B. Transcription factor YY1: structure,
function, and therapeutic implications in cancer biology. Oncogene 2006; 25:
1125–1142.
4 Sui G. The regulation of YY1 in tumorigenesis and its targeting potential in cancer
therapy. Mol Cell Pharmacol 2009; 1: 157–176.
5 Zaravinos A, Spandidos DA. Yin yang 1 expression in human tumors. Cell Cycle
2010; 9: 512–522.
6 Lu L, Zhou L, Chen EZ, Sun K, Jiang P, Wang L et al. A novel YY1-miR-1 regulatory
circuit in skeletal myogenesis revealed by genome-wide prediction of yy1-mirna
network. PloS one 2012; 7: e27596.
7 Wang H, Garzon R, Sun H, Ladner KJ, Singh R, Dahlman J et al. NF-kappaB-YY1-
miR-29 regulatory circuitry in skeletal myogenesis and rhabdomyosarcoma.
Cancer Cell 2008; 14: 369–381.
8 Wang H, Hertlein E, Bakkar N, Sun H, Acharyya S, Wang J et al. NF-kappaB reg-
ulation of YY1 inhibits skeletal myogenesis through transcriptional silencing of
myofibrillar genes. Mol Cell Biol 2007; 27: 4374–4387.
9 Wang H, Sun H, Guttridge DC. microRNAs: novel components in a muscle gene
regulatory network. Cell Cycle 2009; 8: 1833–1837.
10 Wang L, Zhou L, Jiang P, Lu L, Chen X, Guttridge DC et al. Loss of miR-29 in
myoblasts contributes to dystrophic muscle pathogenesis. Mol Ther 2012; 20:
1222–1233.
11 Yang HJ, Liu VW, Wang Y, Tsang PC, Ngan HY. Differential DNA methylation
profiles in gynecological cancers and correlation with clinico-pathological data.
BMC Cancer 2006; 6: 212.
12 Aguilera O, Munoz A, Esteller M, Fraga MF. Epigenetic alterations of the Wnt/beta-
catenin pathway in human disease. Endocr Metab Immune Disord Drug Targets
2007; 7: 13–21.
13 Hatziapostolou M, Iliopoulos D. Epigenetic aberrations during oncogenesis. Cell
Mol Life Sci 2011; 68: 1681–1702.
14 Fullgrabe J, Kavanagh E, Joseph B. Histone onco-modifications. Oncogene 2011;
30: 3391–3403.
15 Cedar H, Bergman Y. Linking DNA methylation and histone modification: patterns
and paradigms. Nat Rev Genet 2009; 10: 295–304.
16 Bartel DP. MicroRNAs: genomics, biogenesis, mechanism, and function. Cell 2004;
116: 281–297.
17 Boren T, Xiong Y, Hakam A, Wenham R, Apte S, Wei Z et al. MicroRNAs and their
target messenger RNAs associated with endometrial carcinogenesis. Gynecol
Oncol 2008; 110: 206–215.
18 Chung TK, Cheung TH, Huen NY, Wong KW, Lo KW, Yim SF et al.
Dysregulated microRNAs and their predicted targets associated with endome-
trioid endometrial adenocarcinoma in Hong Kong women. Int J Cancer 2009; 124:
1358–1365.
19 Huang YW, Liu JC, Deatherage DE, Luo J, Mutch DG, Goodfellow PJ et al. Epige-
netic repression of microRNA-129-2 leads to overexpression of SOX4 oncogene in
endometrial cancer. Cancer Res 2009; 69: 9038–9046.
20 Pan Q, Chegini N. MicroRNA signature and regulatory functions in the endome-
trium during normal and disease states. Semin Reprod Med 2008; 26: 479–493.
21 Dawe CJ, Banfield WG, Morgan WD, Slatick MS, Curth HO. Growth in continuous
culture, and in hamsters, of cells from a neoplasm associated with acanthosis
nigricans. J Natl Cancer Inst 1964; 33: 441–456.
22 Richardson GS, Dickersin GR, Atkins L, MacLaughlin DT, Raam S, Merk LP et al. KLE:
a cell line with defective estrogen receptor derived from undifferentiated endo-
metrial cancer. Gynecol Oncol 1984; 17: 213–230.
23 Zysman M, Saka A, Millar A, Knight J, Chapman W, Bapat B. Methylation of ade-
nomatous polyposis coli in endometrial cancer occurs more frequently in tumors
with microsatellite instability phenotype. Cancer Res 2002; 62: 3663–3666.
24 Castellano G, Torrisi E, Ligresti G, Malaponte G, Militello L, Russo AE et al. The
involvement of the transcription factor Yin Yang 1 in cancer development and
progression. Cell Cycle 2009; 8: 1367–1372.
25 Andl CD. The misregulation of cell adhesion components during tumorigenesis:
overview and commentary. J Oncol 2010 pii: 174715.
26 Moreno-Bueno G, Hardisson D, Sanchez C, Sarrio D, Cassia R, Garcia-Rostan G et al.
Abnormalities of the APC/beta-catenin pathway in endometrial cancer. Oncogene
2002; 21: 7981–7990.
27 Ko CY, Hsu HC, Shen MR, Chang WC, Wang JM. Epigenetic silencing of CCAAT/
enhancer-binding protein delta activity by YY1/polycomb group/DNA methyl-
transferase complex. J Biol Chem 2008; 283: 30919–30932.
28 Chase A, Cross NC. Aberrations of EZH2 in cancer. Clin Cancer Res 2011; 17:
2613–2618.
29 Knuutila S, Bjorkqvist AM, Autio K, Tarkkanen M, Wolf M, Monni O et al. DNA copy
number amplifications in human neoplasms: review of comparative genomic
hybridization studies. Am J Pathol 1998; 152: 1107–1123.
30 Takeshita F, Patrawala L, Osaki M, Takahashi RU, Yamamoto Y, Kosaka N et al.
Systemic delivery of synthetic microRNA-16 inhibits the growth of metastatic
prostate tumors via downregulation of multiple cell-cycle genes. Mol Ther 2010;
18: 181–187.
31 Wiggins JF, Ruffino L, Kelnar K, Omotola M, Patrawala L, Brown D et al. Devel-
opment of a lung cancer therapeutic based on the tumor suppressor microRNA-
34. Cancer Res 2010; 70: 5923–5930.
32 Yu J, Cheng YY, Tao Q, Cheung KF, Lam CN, Geng H et al. Methylation of pro-
tocadherin 10, a novel tumor suppressor, is associated with poor prognosis in
patients with gastric cancer. Gastroenterology 2009; 136: 640–651.
Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc)
miR-193a-5p-YY1-APC regulatory axis in EEC
Y Yang et al
3442
Oncogene (2013) 3432 – 3442 &2013 Macmillan Publishers Limited
